BILLING CODE: 4810-AM-P BUREAU OF CONSUMER FINANCIAL PROTECTION 12 CFR Part 1041 [Docket No. CFPB-2016-0025] RIN 3170–AA40 Payday, Vehicle Title, and Certain High-Cost Installment Loans AGENCY: Bureau of Consumer Financial Protection. ACTION: Final Rule; official interpretations. SUMMARY: The Bureau of Consumer Financial Protection (Bureau or CFPB) is issuing this final rule to establish 12 CFR 1041, which creates consumer protections for certain consumer credit products, and the official interpretations to the rule. First, the rule identifies it as an unfair and abusive practice for a lender to make covered short-term or longer-term balloon-payment loans, including payday and vehicle title loans, without reasonably determining that consumers have the ability to repay the loans according to their terms. The rule exempts certain loans from the underwriting criteria prescribed in the rule if they have specific consumer protections. Second, for the same set of loans along with certain other high-cost longer-term loans, the rule identifies it as an unfair and abusive practice to make attempts to withdraw payment from consumers’ accounts after two consecutive payment attempts have failed, unless the consumer provides a new and specific authorization to do so. Finally, the rule prescribes notices to consumers before attempting to withdraw payments from their account, as well as processes and criteria for registration of information systems, for requirements to furnish and obtain information from them, and for compliance programs and record retention. The rule prohibits 1 evasions and operates as a floor leaving State and local jurisdictions to adopt further regulatory measures (whether a usury limit or other protections) as appropriate to protect consumers. DATES: Effective Date: This regulation is effective [INSERT DATE 21 MONTHS AFTER DATE OF PUBLICATION IN THE FEDERAL REGISTER], except for § 1041.11, which is effective [INSERT DATE 60 DAYS AFTER DATE OF PUBLICATION IN THE FEDERAL REGISTER]. FOR FURTHER INFORMATION CONTACT: Sarita Frattaroli, Counsel; Mark Morelli, Michael G. Silver, Steve Wrone, Senior Counsels; Office of Regulations; Consumer Financial Protection Bureau, at 202-435-7700 or cfpb_reginquiries@cfpb.gov. SUPPLEMENTARY INFORMATION: I. Summary of the Final Rule On June 2, 2016, the Bureau issued proposed consumer protections for payday loans, vehicle title loans, and certain high-cost installment loans. The proposal was published in the Federal Register on July 22, 2016.1 Following a public comment period and review of comments received, the Bureau is now issuing this final rule with consumer protections governing the underwriting of covered short-term and longer-term balloon-payment loans, including payday and vehicle title loans. The rule also contains disclosure and payment withdrawal attempt requirements for covered short-term loans, covered longer-term balloonpayment loans, and certain high-cost covered longer-term loans. Covered short-term loans are typically used by consumers who are living paycheck to paycheck, have little to no access to other credit products, and seek funds to meet recurring or 1 Payday, Vehicle Title, and Certain High-Cost Installment Loans, 81 FR 47864 (July 22, 2016). 2 one-time expenses. The Bureau has conducted extensive research on these products, in addition to several years of outreach and review of the available literature. The Bureau issues these regulations primarily pursuant to its authority under section 1031 of the Dodd-Frank Wall Street Reform and Consumer Protection Act (Dodd-Frank Act) to identify and prevent unfair, deceptive, or abusive acts or practices.2 The Bureau is also using authorities under section 1022 of the Dodd-Frank Act to prescribe rules and make exemptions from such rules as is necessary or appropriate to carry out the purposes and objectives of the Federal consumer financial laws,3 section 1024 of the Dodd-Frank Act to facilitate supervision of certain non-bank financial service providers,4 and section 1032 of the Dodd-Frank Act to require disclosures to convey the costs, benefits, and risks of particular consumer financial products or services.5 The Bureau is not, at this time, finalizing the ability-to-repay determination requirements proposed for certain high-cost installment loans, but it is finalizing those requirements as to covered short-term and longer-term balloon-payment loans. The Bureau is also finalizing certain disclosure, notice, and payment withdrawal attempt requirements as applied to covered shortterm loans, longer-term balloon-payment loans, and high-cost longer-term loans at this time. The Bureau is concerned that lenders that make covered short-term loans have developed business models that deviate substantially from the practices in other credit markets by failing to assess consumers’ ability to repay their loans according to their terms and by engaging in harmful practices in the course of seeking to withdraw payments from consumers’ accounts. The 2 Public Law 111–203, section 1031(b), 124 Stat. 1376 (2010) (hereinafter Dodd-Frank Act). Dodd-Frank Act section 1022(b). 4 Dodd-Frank Act section 1024(b)(7). 5 Dodd-Frank Act section 1032(a). 3 3 Bureau has concluded that there is consumer harm in connection with these practices because many consumers struggle to repay unaffordable loans and in doing so suffer a variety of adverse consequences. In particular, many consumers who take out these loans appear to lack the ability to repay them and face one of three options when an unaffordable loan payment is due: take out additional covered loans (“re-borrow”), default on the covered loan, or make the payment on the covered loan and fail to meet basic living expenses or other major financial obligations. As a result of these dynamics, a substantial population of consumers ends up in extended loan sequences of unaffordable loans. Longer-term balloon-payment loans, which are less common in the marketplace today, raise similar risks. In addition, many lenders may seek to obtain repayment of covered loans directly from consumers’ accounts. The Bureau is concerned that consumers may be subject to multiple fees and other harms when lenders make repeated unsuccessful attempts to withdraw funds from their accounts. In these circumstances, further attempts to withdraw funds from consumers’ accounts are very unlikely to succeed, yet they clearly result in further harms to consumers. A. Scope of the Rule The rule applies to two types of covered loans. First, it applies to short-term loans that have terms of 45 days or less, including typical 14-day and 30-day payday loans, as well as short-term vehicle title loans that are usually made for 30-day terms, and longer-term balloonpayment loans. The underwriting portion of the rule applies to these loans. Second, certain parts of the rule apply to longer-term loans with terms of more than 45 days that have (1) a cost of credit that exceeds 36 percent per annum; and (2) a form of “leveraged payment mechanism” that gives the lender a right to withdraw payments from the consumer’s account. The payments 4 part of the rule applies to both categories of loans. The Bureau had proposed parallel underwriting requirements for high-cost covered longer-term loans. However, at this time, the Bureau is not finalizing the ability-to-repay portions of the rule as to covered longer-term loans other than those with balloon payments. The rule excludes or exempts several types of consumer credit, including: (1) loans extended solely to finance the purchase of a car or other consumer good in which the good secures the loan; (2) home mortgages and other loans secured by real property or a dwelling if recorded or perfected; (3) credit cards; (4) student loans; (5) non-recourse pawn loans; (6) overdraft services and lines of credit; (7) wage advance programs; (8) no-cost advances; (9) alternative loans (similar to loans made under the Payday Alternative Loan program administered by the National Credit Union Administration); and (10) accommodation loans. B. Ability-to-Repay Requirements and Alternative Requirements for Covered Short-Term Loans The rule identifies it as an unfair and abusive practice for a lender to make covered shortterm or longer-term balloon-payment loans without reasonably determining that the consumers will have the ability to repay the loans according to their terms. The rule prescribes requirements to prevent this practice and thus the specific harms to consumers that the Bureau has identified as flowing from the practice, including extended loan sequences for a substantial population of consumers. The first set of requirements addresses the underwriting of these loans. A lender, before making a covered short-term or longer-term balloon-payment loan, must make a reasonable determination that the consumer would be able to make the payments on the loan and be able to 5 meet the consumer’s basic living expenses and other major financial obligations without needing to re-borrow over the ensuing 30 days. Specifically, a lender is required to:  Verify the consumer’s net monthly income using a reliable record of income payment, unless a reliable record is not reasonably available;  Verify the consumer’s monthly debt obligations using a national consumer report and a consumer report from a “registered information system” as described below;  Verify the consumer’s monthly housing costs using a national consumer report if possible, or otherwise rely on the consumer’s written statement of monthly housing expenses;  Forecast a reasonable amount for basic living expenses, other than debt obligations and housing costs; and  Determine the consumer’s ability to repay the loan based on the lender’s projections of the consumer’s residual income or debt-to-income ratio. Furthermore, a lender is prohibited from making a covered short-term loan to a consumer who has already taken out three covered short-term or longer-term balloon-payment loans within 30 days of each other, for 30 days after the third loan is no longer outstanding. Second, and in the alternative, a lender is allowed to make a covered short-term loan without meeting all the specific underwriting criteria set out above, as long as the loan satisfies certain prescribed terms, the lender confirms that the consumer meets specified borrowing history conditions, and the lender provides required disclosures to the consumer. Among other conditions, under this alternative approach, a lender is allowed to make up to three covered 6 short-term loans in short succession, provided that the first loan has a principal amount no larger than $500, the second loan has a principal amount at least one-third smaller than the principal amount on the first loan, and the third loan has a principal amount at least two-thirds smaller than the principal amount on the first loan. In addition, a lender is not allowed to make a covered short-term loan under the alternative requirements if it would result in the consumer having more than six covered short-term loans during a consecutive 12-month period or being in debt for more than 90 days on covered short-term loans during a consecutive 12-month period. A lender is not permitted to take vehicle security in connection with loans that are made according to this alternative approach. C. Payment Practices Rules The rule identifies it as an unfair and abusive practice for a lender to make attempts to withdraw payment from consumers’ accounts in connection with a short-term, longer-term balloon-payment, or high-cost longer-term loan after the lender’s second consecutive attempts to withdraw payments from the accounts from which the prior attempts were made have failed due to a lack of sufficient funds, unless the lender obtains the consumers’ new and specific authorization to make further withdrawals from the accounts. The Bureau found that in these circumstances, further attempted withdrawals are highly unlikely to succeed, but clearly impose harms on consumers who are affected. This prohibition on further withdrawal attempts applies whether the two failed attempts are initiated through a single payment channel or different channels, such as the automated clearinghouse system and the check network. The rule requires that lenders must provide notice to consumers when the prohibition has been triggered and follow certain procedures in obtaining new authorizations. 7 In addition to the requirements related to the prohibition on further payment withdrawal attempts, a lender is required to provide a written notice, depending on means of delivery, a certain number of days before its first attempt to withdraw payment for a covered loan from a consumer’s checking, savings, or prepaid account or before an attempt to withdraw such payment in a different amount than the regularly scheduled payment amount, on a date other than the regularly scheduled payment date, by a different payment channel than the prior payment, or to re-initiate a returned prior transfer. The notice must contain key information about the upcoming payment attempt and, if applicable, alert the consumer to unusual payment attempts. A lender is permitted to provide electronic notices as long as the consumer consents to electronic communications. D. Additional Requirements The rule requires lenders to furnish to provisionally registered and registered information systems certain information concerning covered short-term and longer-term balloon-payment loans at loan consummation, during the period that the loan is an outstanding loan, and when the loan ceases to be an outstanding loan. To be eligible to become a provisionally registered or registered information system, an entity must satisfy the eligibility criteria prescribed in the rule. The rule provides for a registration process that will allow information systems to be registered, and lenders to be ready to furnish required information, at the time the furnishing obligation in the rule takes effect. Consumer reports provided by registered information systems will include a reasonably comprehensive record of a consumer’s recent and current use of covered short-term and longer-term balloon-payment loans. Before making covered short-term and longer-term 8 balloon-payment loans, a lender is required to obtain and consider a consumer report from a registered information system. A lender is required to establish and follow a compliance program and retain certain records. A lender is also required to develop and follow written policies and procedures that are reasonably designed to ensure compliance with the requirements in this rule. Furthermore, a lender is required to retain the loan agreement and documentation obtained for any covered loan or an image thereof, as well as electronic records in tabular format regarding origination calculations and determinations for a short-term or longer-term balloon-payment loan, and regarding loan type and terms. The rule also includes an anti-evasion clause to address the kinds of concerns the Bureau noted in connection with the evasive actions that lenders in this market took in response to the regulations originally adopted on loans made to servicemembers under the Military Lending Act. E. Effective Date The final rule will become effective 21 months after publication of the final rule in the Federal Register. Certain provisions necessary to implement the consumer reporting components of the rule, specifically § 1041.11, will become effective 60 days after publication of the final rule in the Federal Register to facilitate an orderly implementation process. II. Background A. Introduction For most consumers, credit provides a means of purchasing goods or services and spreading the cost of repayment over time. This is true of the three largest consumer credit markets: the market for mortgages ($10.3 trillion in outstanding balances), for student loans 9 ($1.4 trillion), and for auto loans ($1.1 trillion). This is also one way in which certain types of open-end credit—including home equity loans ($0.13 trillion) and lines of credit ($0.472 trillion)—and at least some credit cards and revolving credit ($1.0 trillion)—can be used.6 In addition to the credit markets described above, consumers living paycheck to paycheck and with little to no savings have also used credit as a means of coping with financial shortfalls. These shortfalls may be due to mismatched timing between income and expenses, misaligned cash flows, income volatility, unexpected expenses or income shocks, or expenses that simply exceed income.7 According to a recent survey conducted by the Board of Governors of the Federal Reserve System (Federal Reserve Board), 44 percent of adults reported they would either be unable to cover an emergency expense costing $400 or would have to sell something or borrow money to cover it, and 30 percent reported that they found it “difficult to get by” or were “just getting by” financially.8 Whatever the cause of these financial shortfalls, consumers in these situations sometimes seek what may broadly be termed a “liquidity loan.”9 There are a 6 See Bd. of Governors of the Fed. Reserve Sys., “Mortgage Debt Outstanding (Table 1.54),” (June 2017) (mortgages (one- to four-family)), available at http://www.federalreserve.gov/econresdata/releases/mortoutstand/current.htm; Bd. of Governors of the Fed. Reserve Sys., “Consumer Credit - G.19: July 2017,” (Sept. 8, 2017) (student loans, auto loans, and revolving credit), available at https://www.federalreserve.gov/releases/g19/current/default.htm; Experian-Oliver Wyman, “2017 Q2 Market Intelligence Report: Home Equity Loans Report,” at 16 fig. 21 (2017) and Experian-Oliver Wyman, “2017 Q2 Market Intelligence Report: Home Equity Lines Report,” at 21 fig. 30 (2017) (home equity loans and lines of credit outstanding estimates), available at http://www.marketintelligencereports.com. 7 See generally Rob Levy & Joshua Sledge, “A Complex Portrait: An Examination of Small-Dollar Credit Consumers” (Ctr. for Fin. Servs. Innovation, 2012), available at https://www.fdic.gov/news/conferences/consumersymposium/2012/A%20Complex%20Portrait.pdf. 8 Bd. of Governors of the Fed. Reserve Sys., “Report on the Economic Well-Being of U.S. Households in 2016,” at 2, 8 (May 2017), available at https://www.federalreserve.gov/publications/files/2016-report-economic-well-beingus-households-201705.pdf. 9 If a consumer’s expenses consistently exceed income, a liquidity loan is not likely to be an appropriate solution to the consumer’s needs. 10 variety of loans and products that consumers use for these purposes including credit cards, deposit account overdraft, pawn loans, payday loans, vehicle title loans, and installment loans. Credit cards and deposit account overdraft services are each already subject to specific Federal consumer protection regulations and requirements. The Bureau generally considers these markets to be outside the scope of this rulemaking as discussed further below. The Bureau is also separately engaged in research and evaluation of potential rulemaking actions on deposit account overdraft.10 Another liquidity option—pawn—generally involves non-recourse loans made against the value of whatever item a consumer chooses to give the lender in return for the funds.11 The consumer has the option to either repay the loan or permit the pawnbroker to retain 10 Credit cards and deposit overdraft services would have been excluded from the proposed rule under proposed § 1041.3(e)(3) and (6) as discussed further below. On October 5, 2016, the Bureau released a final rule on prepaid accounts. Among other things, the rule regulates overdraft credit features offered in connection with prepaid accounts, and generally covers under Regulation Z’s credit card rules any such credit feature that is offered by the prepaid account issuer, its affiliate, or its business partner where credit can be accessed in the course of a transaction conducted with a prepaid card. 81 FR 83934 (Nov. 22, 2016). The Bureau later published a final rule delaying the October 1, 2017, effective date of that rule by six months, to April 1, 2018. 82 FR 18975 (Apr. 25, 2017). In preparation for a potential rulemaking regarding possible consumer protection concerns with overdraft programs on checking accounts, the Bureau issued the Notice and Request for Information on the Impacts of Overdraft Programs on Consumers, 77 FR 12031 (Feb. 28, 2012); see Kelly Cochran, “Spring 2017 Rulemaking Agenda,” CFPB Blog (July 20, 2017), available at https://www.consumerfinance.gov/about-us/blog/spring-2017-rulemaking-agenda/. In 2015, banks with over $1 billion in assets reported overdraft and NSF (nonsufficient funds) fee revenue of $11.16 billion. See Gary Stein, “New Insights on Bank Overdraft Fees and 4 Ways to Avoid Them,” CFPB Blog (Feb. 25, 2016), available at https://www.consumerfinance.gov/about-us/blog/new-insights-on-bank-overdraft-fees-and-4ways-to-avoid-them/. The $11.16 billion total does not include credit union overdraft fee revenue and does not separate out overdraft from NSF amounts but overall, overdraft fee revenue accounts for about 72 percent of that amount. Bureau of Consumer Fin. Prot., “Data Point: Checking Account Overdraft,” at 10 (2014), available at http://files.consumerfinance.gov/f/201407_cfpb_report_data-point_overdrafts.pdf. The Federal Reserve Board has adopted a set of regulations of overdraft services. See Electronic Fund Transfers, 75 FR 31665 (June 4, 2010). In addition, the Bureau has published three research reports on overdraft. See Bureau of Consumer Fin. Prot., “Data Point: Frequent Overdrafters” (2017), available at http://files.consumerfinance.gov/f/documents/201708_cfpb_datapoint_frequent-overdrafters.pdf; Bureau of Consumer Fin. Prot., “Data Point: Checking Account Overdraft” (2014), available at http://files.consumerfinance.gov/f/201407_cfpb_report_data-point_overdrafts.pdf; Bureau of Consumer Fin. Prot., “CFPB Study of Overdraft Programs: A White Paper of Initial Data Findings” (2013), available at http://files.consumerfinance.gov/f/201306_cfpb_whitepaper_overdraft-practices.pdf (hereinafter “CFPB Study of Overdraft Programs White Paper”). 11 Pawn lending, also known as pledge lending, has existed for centuries, with references to it in the Old Testament; pawn lending in the U.S. began in the 17th century. See Susan Payne Carter, “Payday Loan and Pawnshop Usage: 11 and sell the pawned property at the end of the loan term, relieving the borrower from any additional financial obligation. This feature distinguishes pawn loans from most other types of liquidity loans. The Bureau is excluding non-recourse possessory pawn loans, as described in proposed § 1041.3(e)(5), from the scope of this rulemaking. This rulemaking is focused on two general categories of liquidity loan products: (1) short-term loans and longer-term balloon-payment loans; and (2) with regard to payment practices, a broader set of liquidity loan products that also includes certain higher-cost longerterm installment loans. The largest category of short-term loans are “payday loans,” which are generally required to be repaid in a lump-sum single-payment on receipt of the borrower’s next income payment, and short-term vehicle title loans, which are also almost always due in a lumpsum single-payment, typically within 30 days after the loan is made. The final rule’s underwriting requirements also apply to depository advance products and other loans of 45 days or less in duration, as well as certain longer-term balloon-payment loans that generally involve a series of small, often interest-only, payments followed by a single final large lump sum payment. The final rule’s payment presentment requirements apply to short-term and longer-term balloonpayment products, as well as to certain higher-cost longer-term installment loans. That latter category includes what are often referred to as “payday installment loans”—that is, loans that are The Impact of Allowing Payday Loan Rollovers,” at 5 (Jan. 15, 2012), available at https://my.vanderbilt.edu/susancarter/files/2011/07/Carter_Susan_JMP_website2.pdf. The two largest pawn firms, EZCORP and FirstCash, account for about 13 percent of approximately 13,000 pawn storefronts. The remaining storefronts are operated by small, independent firms. EZCORP, “Investor Update: Business Transformation Delivering Results,” at 7 (Mar. 7, 2017), available at http://investors.ezcorp.com/download/Investor+Presentation_030717.pdf. FirstCash, Inc., is the company resulting from the September 2016 merger of FirstCash Financial Services and Cash America. FirstCash operates in 26 States. FirstCash, Inc., 2016 Annual Report (Form 10–K), at 1 (Mar. 1, 2017). See generally, John P. Caskey, “Fringe Banking: Cash-Checking Outlets, Pawnshops, and the Poor,” at Chapter 2 (New York: Russell Sage Foundation 1994). 12 repaid in multiple installments with each installment typically due on the borrower’s payday or regularly scheduled income payment and with the lender having the ability to automatically collect payments from an account into which the income payment is deposited. In addition, the latter category includes certain high-cost installment loans made by more traditional finance companies. This rulemaking includes both closed-end loans and open-end lines of credit.12 As described in the section-by-section analysis, the Bureau has been studying these markets for liquidity loans for over five years, gaining insights from a variety of sources. During this time the Bureau has conducted supervisory examinations of a number of payday lenders and enforcement investigations of a number of different types of liquidity lenders, which have given the Bureau insights into the business models and practices of such lenders. Through these processes, and through market monitoring activities, the Bureau also has obtained extensive loan-level data that the Bureau has studied to better understand risks to consumers.13 The Bureau has published five reports based upon these data.14 The Bureau has also carefully reviewed the 12 The Dodd-Frank Act does not define “payday loan,” though it refers to the term in section 1024(a)(1)(E), and the Bureau is not proposing to define it in this rulemaking. The Bureau may do so in a subsequent rulemaking or in another context. In addition, the Bureau notes that various State, local, and Tribal jurisdictions may define “payday loans” in ways that may be more or less coextensive with the coverage of the Bureau’s rule. 13 Information underlying this proposed rule is derived from a variety of sources, including from market monitoring and outreach, third-party studies and data, consumer complaints, the Bureau’s enforcement and supervisory work, and the Bureau’s expertise generally. In publicly discussing information, the Bureau has taken steps not to disclose confidential information inappropriately and to otherwise comply with applicable law and its own rules regarding disclosure of records and information. See 12 CFR 1070.41(c). 14 See Bureau of Consumer Fin. Prot., “Payday Loans and Deposit Advance Products: A White Paper of Initial Data Findings” (2013), available at http://files.consumerfinance.gov/f/201304_cfpb_payday-dap-whitepaper.pdf [hereinafter “CFPB Payday Loans and Deposit Advance Products White Paper”]; Bureau of Consumer Fin. Prot., “CFPB Data Point: Payday Lending” (2014), available at http://files.consumerfinance.gov/f/201403_cfpb_report_payday-lending.pdf [hereinafter “CFPB Data Point: Payday Lending”]; Bureau of Consumer Fin. Prot., “Online Payday Loan Payments” (2016), available at http://files.consumerfinance.gov/f/201604_cfpb_online-payday-loan-payments.pdf [hereinafter CFPB Online Payday 13 published literature with respect to small-dollar liquidity loans and a number of outside researchers have presented their research at seminars for Bureau staff. In addition, over the course of the past five years the Bureau has engaged in extensive outreach with a variety of stakeholders in both formal and informal settings, including several Bureau field hearings across the country specifically focused on the subject of small-dollar lending, meetings with the Bureau’s standing advisory groups, meetings with State and Federal regulators, meetings with consumer advocates, religious groups, and industry trade associations, Tribal consultations, and through a Small Business Review Panel process as described further below. As described in Summary of the Rulemaking Process, the Bureau received and reviewed over one million comments on its proposal, mostly from lenders and borrowers within the respective markets. This Background section provides a brief description of the major components of the markets for short-term loans and longer-term balloon-payment loans, describing the product parameters, industry size and structure, lending practices, and business models of major market segments. The Background section also provides a brief overview of the additional markets for higher-cost longer-term installment loans that are subject to the payment practices components of the final rule. This section also describes recent State and Federal regulatory activity in connection with these various product markets. Market Concerns—Underwriting below, provides a more detailed description of consumer experiences with short-term loans describing Loan Payments]; Bureau of Consumer Fin. Prot., “Single-Payment Vehicle Title Lending” (2016), available at http://files.consumerfinance.gov/f/documents/201605_cfpb_single-payment-vehicle-title-lending.pdf [hereinafter “CFPB Single-Payment Vehicle Title Lending”]; Bureau of Consumer Fin. Prot., “Supplemental Findings on Payday, Payday Installment, and Vehicle Title Loans, and Deposit Advance Products” (2016), available at https://www.consumerfinance.gov/data-research/research-reports/supplemental-findings-payday-paydayinstallment-and-vehicle-title-loans-and-deposit-advance-products/ (hereinafter “CFPB Report on Supplemental Findings”). 14 research about which consumers use the products, why they use the products, and the outcomes they experience as a result of the product structures and industry practices. The Background section also includes an extensive description of the methods by which lenders initiate payments from consumers’ accounts. Market Concerns—Payments, below, describes consumer experiences and concerns with these payment practices. Most of the comments received on the proposal’s Background section agreed in general terms with the descriptions of the markets and products described below, although there may be slight differences in individual lenders’ loan products and business practices. Comments that provided significantly different information are noted below. B. Short-Term, Hybrid, and Balloon-Payment Loans Providing short-term loans for liquidity needs has been a long-term challenge in the consumer financial services market due to the fixed costs associated with loan origination regardless of loan size. At the beginning of the twentieth century, concern arose with respect to companies that were responding to liquidity needs by offering to “purchase” a consumer’s paycheck in advance of it being paid. These companies charged fees that, if calculated as an annualized interest rate, were as high as 400 percent.15 To address these concerns, between 1914 and 1943, 34 States enacted a form of the Uniform Small Loan Law, which was a model law developed by the Russell Sage Foundation. That law provided for lender licensing and permitted interest rates of between 2 and 4 percent per month, or 24 to 48 percent per year. Those rates were substantially higher than pre-existing usury limits (which generally capped interest rates at 15 Salary advances were structured as wage assignments rather than loans to evade much lower State usury caps of about 8 percent per annum or less. John P. Caskey, “Fringe Banking and the Rise of Payday Lending,” at 17, 23 (Patrick Bolton & Howard Rosenthal eds., New York: Russell Sage Foundation, 2005). 15 between 6 and 8 percent per year) but were viewed by proponents as “equitable to both borrower and lender.”16 New forms of short-term small-dollar lending appeared in several States in the 1990s,17 starting with check cashing outlets that would hold a customer’s personal check for a period of time for a fee before cashing it (“check holding” or “deferred presentment”). One of the larger payday lenders began making payday loans in Kansas in 1992, and that same year at least one State regulator issued an administrative interpretation holding that deferred presentment activities were consumer loans subject to that State’s licensing and consumer lending laws.18 One commenter described his role in developing and expanding the deferred presentment lending industry in Tennessee in the early 1990s prior to any regulation in that State, while noting that those same activities required lending licenses in two nearby States. Several market factors converged around the same time that spurred the development of these new forms of short-term small-dollar lending. Consumers were using credit cards more frequently for short-term liquidity lending needs, a trend that continues today. 19 Storefront finance companies, described below in part II.C, that had provided small loans changed their 16 Elisabeth Anderson, “Experts, Ideas, and Policy Change: The Russell Sage Foundation and Small Loan Reform, 1909–-1941,” 37 Theory & Soc’y 271, 276, 283, 285 (2008), available at http://www.jstor.org/stable/40211037 (quoting Arthur Ham, Russell Sage Foundation, Feb. 1911, Quarterly Report, Library of Congress Russell Sage Foundation Archive, Box 55). 17 See Pew Charitable Trusts, “A Short History of Payday Lending Law,” (July 18, 2012), available at http://www.pewtrusts.org/en/research-and-analysis/analysis/2012/07/a-short-history-of-payday-lending-law. 18 QC Holdings, Inc., Registration Statement (Form S–1), at 1 (May 7, 2004);), see, e.g., Laura Udis, Adm’r Colo. Dep’t of Law, Unif. Consumer Credit Code, “Check Cashing Entities Which Provide Funds In Return For A PostDated Check Or Similar Deferred Payment Arrangement And Which Impose A Check Cashing Charge Or Fee May Be Consumer Lenders Subject To The Colorado Uniform Consumer Credit Code,” Administrative Interpretation No. 3.104–9201 (June 23, 1992) (on file). 19 Robert D. Manning, “Credit Card Nation: The Consequences of America’s Addiction to Credit” (Basic Books 2000); Amy Traub, “Debt Disparity: What Drives Credit Card Debt in America,” Demos (2014), available at http://www.demos.org/sites/default/files/publications/DebtDisparity_1.pdf. 16 focus to larger, collateralized products, including vehicle financing and real estate secured loans. At the same time there was substantial consolidation in the storefront installment lending industry. Depository institutions similarly moved away from short-term small-dollar loans. Around the same time, a number of State legislatures amended their usury laws to allow lending by a broader group of both depository and non-depository lenders by increasing maximum allowable State interest rates or eliminating State usury laws, while other States created usury carve-outs or special rules for short-term loans.20 The confluence of these trends has led to the development of markets offering what are commonly referred to as payday loans (also known as cash advance loans, deferred deposit, and deferred presentment loans depending on lender and State law terminology), and short-term vehicle title loans that are much shorter in duration than vehicle-secured loans that have traditionally been offered by storefront installment lenders and depository institutions. Although payday loans initially were distributed through storefront retail outlets, they are now also widely available on the Internet. Vehicle title loans are typically offered exclusively at storefront retail outlets. These markets as they have evolved over the last two decades are not strictly segmented. There is substantial overlap between market products and the borrowers who use them. For example, in a 2015 survey, almost 14.8 percent of U.S. households that had used a payday loan in the prior year had also used a vehicle title loan.21 There is also an established trend away from “monoline” or single-product lending companies. Thus, for example, a number of large payday 20 See Pew Charitable Trusts, “A Short History of Payday Lending Law” (July 18, 2012). This article notes that State legislative changes were in part a response to the ability of Federally- and State-chartered banks to lend without being subject to the usury laws of the borrower’s State. 21 Estimates by the Bureau’s Office of Research are based on data derived from FDIC. Fed. Deposit Ins. Corp., “2015 FDIC National Survey of Unbanked and Underbanked Households” (Oct. 20, 2016), available at https://www.fdic.gov/householdsurvey/2015/2015report.pdf. 17 lenders also offer vehicle title and installment loans.22 The following discussion nonetheless provides a description of major product types. Storefront Payday Loans The market that has received the greatest attention among policy makers, advocates, and researchers is the market for single-payment payday loans. These payday loans are short-term small-dollar loans generally repayable in a single payment due when the consumer is scheduled to receive a paycheck or other inflow of income (e.g., government benefits).23 For most borrowers, the loan is due in a single payment on their payday, although State laws with minimum loan terms—seven days for example—or lender practices may affect the loan duration in individual cases. The Bureau refers to these short-term payday loans available at retail locations as “storefront payday loans,” but the requirements for borrowers taking online payday loans are generally similar, as described below. There are now 35 States that either have created a carve-out from their general usury cap for payday loans or have no usury caps on consumer loans.24 The remaining 15 States and the District of Columbia either ban payday loans or have 22 See, e.g., Advance America, “Title Loan Services,” available at https://www.advanceamerica.net/services/titleloans (last visited Mar. 3, 2016); FirstCash, “Own Your Car? Need Cash Now? Drive Away with Cash in Minutes,” available at http://ww2.firstcash.com/title-loans (last visited May 15, 2017); Check Into Cash, “Auto Title Loans,” available at https://checkintocash.com/commercial/auto-title-loans/ (last visited Sept. 14, 2017); Community Choice Financial/CheckSmart “Get Cash Fast,” available at https://www.ccfi.com/checksmart/ (last visited Mar. 3, 2016); Speedy Cash, “Title Loans,” available at https://www.speedycash.com/title-loans/ (last visited Sept. 14, 2017); PLS Financial Services, “Title Loans,” available at http://pls247.com/il/loans.html (last visited Mar. 3, 2016). Moneytree offers vehicle title and installment loans in Idaho and Nevada. See, e.g., Money Tree Inc., “Title Loans (Idaho),” available at https://www.moneytreeinc.com/loans/idaho/title-loans (last visited Mar. 3, 2016); Money Tree Inc., “Title Loans (Nevada),” available at https://www.moneytreeinc.com/loans/nevada/title-loans (last visited Mar. 3, 2016). 23 For convenience, this discussion refers to the next scheduled inflow of income as the consumer’s next “payday” and the inflow itself as the consumer’s “paycheck” even though these are misnomers for consumers whose income comes from government benefits. 24 See Pew Charitable Trusts, “State Payday Loan Regulation and Usage Rates” (Jan. 14, 2014), available at http://www.pewtrusts.org/en/multimedia/data-visualizations/2014/state-payday-loan-regulation-and-usage-rates 18 fee or interest rate caps that payday lenders apparently find too low to sustain their business models. As discussed further below, several of these States previously had authorized payday lending but subsequently changed their laws. Product definition and regulatory environment. As noted above, payday loans are typically repayable in a single payment on the borrower’s next payday. In order to help ensure repayment, in the storefront environment the lender generally holds the borrower’s personal check made out to the lender—usually post-dated to the loan due date in the amount of the loan’s principal and fees—or the borrower’s authorization to electronically debit the funds from her checking account, commonly known as an automated clearing house (ACH) transaction.25 Payment methods are described in more detail below in part II.D. Payday loan sizes vary depending on State law limits, individual lender credit models, and borrower demand. Many States set a limit on payday loan size; $500 is a common loan limit (for a list of States). Other reports reach slightly different totals of payday authorizing States depending on their categorization methodology. See, e.g., Susanna Montezemolo, “The State of Lending in America & Its Impact on U.S. Households: Payday Lending Abuses and Predatory Practices,” at 32–33 (Ctr. for Responsible Lending 2013), available at http://www.responsiblelending.org/sites/default/files/uploads/10-payday-loans.pdf; Consumer Fed’n of Am., “Legal Status of Payday Loans by State,” available at http://www.paydayloaninfo.org/state-information (last visited Apr. 6, 2016) (lists 32 States as having authorized or allowed payday lending). Since publication of these reports, South Dakota enacted a 36 percent usury cap for consumer loans. Press Release, S.D. Dep’t of Labor and Reg., “Initiated Measure 21 Approved” (Nov. 10, 2016), available at http://dlr.sd.gov/news/releases16/nr111016_initiated_measure_21.pdf. Legislation in New Mexico prohibiting short-term payday and vehicle title loans will go into effect on January 1, 2018. Regulatory Alert, N.M. Reg. and Licensing Dep’t, “Small Loan Reforms,” available at http://www.rld.state.nm.us/uploads/files/HB%20347%20Alert%20Final.pdf. 25 The Bureau is aware from market outreach that at a storefront payday lender’s Tennessee branch, almost 100 percent of customers opted to provide ACH authorization rather than leave a post-dated check for their loans. See also Speedy Cash, “Can Anyone Get a Payday Loan?,” available at https://www.speedycash.com/faqs/payday-loans/ (last visited Feb. 4, 2016) (“If you choose to apply in one of our payday loan locations, you will need to provide a repayment source which can be a personal check or your bank routing information.”); QC Holdings, Inc., 2014 Annual Report (Form 10–K), at 3, 6 (Mar. 12, 2015); FirstCash, Inc., 2016 Annual Report (Form 10–K), at 21. 19 although the limits range from $300 to $1,000.26 In 2013, the Bureau reported that the median loan amount for storefront payday loans was $350, based on supervisory data.27 This finding is broadly consistent with other studies using data from one or more lenders as well as with selfreported information in surveys of payday borrowers28 and State regulatory reports.29 The fee for a payday loan is generally structured as a percentage or dollar amount per $100 borrowed, rather than a periodic interest rate based on the amount of time the loan is outstanding. Many State laws set a maximum amount for these fees, with 15 percent ($15 per 26 At least 19 States cap payday loan amounts between $500 and $600 (Alabama, Alaska, Florida, Hawaii, Iowa, Kansas, Kentucky, Michigan, Mississippi, Missouri, Nebraska, North Dakota, Ohio, Oklahoma, Rhode Island, South Carolina, Tennessee, and Virginia), California limits payday loans to $300 (including the fee), and Delaware caps loans at $1,000. Ala. Code § 5–18A–12(a); Alaska Stat. § 06.50.410; Cal. Fin. Code § 23035(a); Del. Code Ann. tit. 5, § 2227(7); Fla. Stat. § 560.404(5); Haw. Rev. Stat. § 480F–4(c); Iowa Code § 533D.10(1)(b); Kan. Stat. Ann. sec. 16a–2–404(1)(c); Ky. Rev. Stat. Ann. sec. 286.9–100(9); Mich. Comp. Laws § 487.2153(1); Miss. Code Ann. sec. 75–67–519(2); Mo. Rev. Stat. § 408.500(1); Neb. Rev. Stat. § 45–919(1)(b); N.D. Cent. Code § 13–08–12(3); Ohio Rev. Code Ann. § 1321.39(A); Okla. Stat. tit. 59, § 3106(7), R.I. Gen. Laws § 19–14.4–5.1(a); S.C. Code Ann. § 34–39–180(B); Tenn. Code Ann. § 45–17–112(o); Va. Code Ann. § 6.2–1816(5). States that limit the loan amount to the lesser of a percent of the borrower’s income or a fixed-dollar amount include Idaho—25 percent or $1,000, Illinois—25 percent or $1,000, Indiana—20 percent or $550, Washington—30 percent or $700, and Wisconsin—35 percent or $1,500. At least two States cap the maximum payday loan at 25 percent of the borrower’s gross monthly income (Nevada and New Mexico). A few States’ laws are silent as to the maximum loan amount (Utah and Wyoming). Idaho Code Ann. §§ 28–46–413(1), (2); 815 Ill. Comp. Stat. 122/2–5(e); Ind. Code §§24–4.5–7–402, 404; Nev. Rev. Stat. § 604A.425(1)(b); N.M. Stat. Ann. § 58–15–32(A); Utah Code Ann. § 7–23–401; Wash. Rev. Code § 31.45.073(2); Wis. Stat. § 138.14(12)(b); Wyo. Stat. Ann. § 40–14–363. As noted above, the New Mexico statute will be repealed on Jan. 1, 2018. See N.M. H.B. 347, 53d Leg., 1st Sess. (N.M. 2017), available at https://www.nmlegis.gov/Sessions/17%20Regular/final/HB0347.pdf (hereinafter N.M. H.B. 347). 27 CFPB Payday Loans and Deposit Advance Products White Paper, at 15. 28 Leslie Parrish & Uriah King, “Phantom Demand: Short-term Due Date Generates Need for Repeat Payday Loans, Accounting for 76% of Total Volume,” at 21 (Ctr. for Responsible Lending 2009), available at http://www.responsiblelending.org/payday-lending/research-analysis/phantom-demand-final.pdf (reporting $350 as the average loan size); Pew Charitable Trusts, “Payday Lending in America: Who Borrows, Where They Borrow, and Why,” at 9 (Report 3, 2013), available at http://www.pewtrusts.org/~/media/legacy/uploadedfiles/pcs_assets/2012/pewpaydaylendingreportpdf.pdf (reporting $375 as the average). Leslie Parrish & Uriah King, Ctr. 29 See, e.g., Ill. Dep’t. of Fin. & Prof. Reg., “Illinois Trends Report All Consumer Loan Products Through December 2015,” at 15 (Apr. 14, 2016), available at http://www.idfpr.com/DFI/CCD/pdfs/IL_Trends_Report%202015%20FINAL.pdf?ActID=1204&ChapterID=20) ($355.85 is the average for Illinois); Idaho Dep’t. of Fin., “Idaho Credit Code ‘Fast Facts’,” at 5 (Fiscal and Annual Report Data as of January 1, 2016), available at https://www.finance.idaho.gov/ConsumerFinance/Documents/Idaho-Credit-Code-Fast-Facts-With-Fiscal-AnnualReport-Data-01012016.pdf ($350 is the average for Idaho); Wash. State Dep’t. of Fin. Insts., “2015 Payday Lending Report,” at 6 (2015), available at http://www.dfi.wa.gov/sites/default/files/reports/2015-payday-lending-report.pdf ($387.35 is the average for Washington). For example: $355.85 (Illinois average, see Ill. 20 $100 borrowed) being the most common limit.30 The median storefront payday loan fee is $15 per $100; thus for a $350 loan, the borrower must repay $52.50 in finance charges together with the $350 borrowed for a total repayment amount of $402.50.31 The annual percentage rate (APR) on a 14-day loan with these terms is 391 percent.32 For payday borrowers who receive monthly income and thus receive a 30-day or monthly payday loan—many of whom are Social Security recipients 33—a $15 per $100 charge on a $350 loan for a term of 30 days equates to an APR of about 180 percent. The Bureau has found the median loan term for a storefront payday loan to be 14 days, with an average term of 18.3 days. The longer average loan duration is due to State laws that require minimum loan terms that may extend beyond the borrower’s next pay date.34 Fees and loan amounts are higher for online loans, described in more detail below. On the loan’s due date, the terms of the loan obligate the borrower to repay the loan in full. Although the States that created exceptions to their usury limits for payday lending 30 Of the States that expressly authorize payday lending, Rhode Island has the lowest cap at 10 percent of the loan amount. Florida has the same fee amount but also allows a flat $5 verification fee. Oregon’s fees are $10 per $100 capped at $30 plus 36 percent interest. Some States have tiered caps depending on the size of the loan. Generally, in these States the cap declines with loan size. However, in Mississippi, the cap is $20 per $100 for loans under $250 and $21.95 for larger loans (up to the State maximum of $500). Six States do not cap fees on payday loans or are silent on fees: (Delaware, Idaho, Nevada, Texas— (no cap on credit access business fees;), Utah, and Wisconsin—silent on fees.). Depending on State law, the fee may be referred to as a “charge,” “rate,” “interest,”” or other similar term. R.I. Gen. Laws § 19–-14.4–-4(4);), Fla. Stat. § 560.404(6);), Or. Rev. Stat. § 725A.064(1)–()(2);), Miss. Code Ann. § 75–-67–-519(4);), Del. Code Ann. tit. 5, § 2229;, Idaho Code Ann. § 28–-46–-412(3);), Tex. Fin. Code Ann. § 393.602(b);), Utah Code Ann. § 7–-23–-401;, Wis. Stat. § 138.14(10)(a). 31 “CFPB Payday Loans and Deposit Advance Products White Paper,” at 15–17. 32 Throughout part II, APR refers to the annual percentage rate calculated as required by the Truth in Lending Act, 15 U.S.C. 1601 et seq. and Regulation Z, 12 CFR 1026, except where otherwise specified. 33 “CFPB Payday Loans and Deposit Advance Products White Paper,” at 16, 19 (33 percent of payday loans borrowers receive income monthly; 18 percent of payday loan borrowers are public benefits recipients, largely from Social Security including Supplemental Security Income and Social Security Disability, typically paid on a monthly basis). 34 For example, Washington requires the due date to be on or after the borrower’s next pay date but if the pay date is within seven days of taking out the loan, the due date must be on the second pay date after the loan is made. Wash. Rev. Code § 31.45.073(2). A number of States set minimum loan terms, some of which are tied directly to the consumer’s next payday. 21 generally did so on the theory these were short-term loans to which the usual usury rules did not easily apply, in 18 of the States that authorize payday lending the lender is permitted to roll over the loan when it comes due. A rollover occurs when, instead of repaying the loan in full at maturity, the consumer pays only the fees due and the lender agrees to extend the due date.35 By rolling over, the loan repayment of the principal is extended for another period of time, usually equivalent to the original loan term, in return for the consumer’s agreement to pay a new set of fees calculated in the same manner as the initial fees (e.g., 15 percent of the loan principal). The rollover fee is not applied to reduce the loan principal or amortize the loan. As an example, if the consumer borrows $300 with a fee of $45 (calculated as $15 per $100 borrowed), the consumer will owe $345 on the due date, typically 14 days later. On the due date, if the consumer cannot afford to repay the entire $345 due or is otherwise offered the option to roll over the loan, she will pay the lender $45 for another 14 days. On the 28th day, the consumer will owe the original $345 and if she pays the loan in full then, will have paid a total of $90 for the loan. In some States in which rollovers are permitted they are subject to certain limitations such as a cap on the number of rollovers or requirements that the borrower amortize—repay part of the original loan amount—on the rollover. Other States have no restrictions on rollovers. Specially, 17 of the States that authorize single-payment payday lending prohibit lenders from rolling over loans and 11 more States impose some rollover limitations.36 However, in most 35 This rulemaking uses the term “rollover” but this practice is sometimes described under State law or by lenders as a “renewal” or an “extension.” 36 States that prohibit rollovers include California, Florida, Hawaii, Illinois, Indiana, Kentucky, Michigan, Minnesota, Mississippi, Nebraska, New Mexico, Oklahoma, South Carolina, Tennessee, Virginia, Washington, and 22 States where rollovers are prohibited or limited, there is no restriction on the lender immediately making a new loan to the consumer (with new fees) after the consumer has repaid the prior loan. New loans made the same day, or “back-to-back” loans, effectively replicate a rollover because the borrower remains in debt to the lender on the borrower’s next payday.37 Ten States have implemented a cooling-off period before a lender may make a new loan. The most common cooling-off period is one day, although some States have longer periods following a specified number of rollovers or back-to-back loans.38 Wyoming. Cal. Fin. Code sec. 23037(a); Fla. Stat. sec. 560.404(18); Haw. Rev. Stat. sec. 480F-4(d); 815 Ill. Comp. Stat. 122/2-30; Ind. Code sec. 24-4.5-7-402(7); Ky. Rev. Stat. Ann. sec. 286.9-100(14); Mich. Comp. Laws sec. 487.2155(1); Minn. Stat. sec. 47.60(2)(f); Miss. Code Ann. sec. 75-67-519(5); Neb. Rev. Stat. sec. 45-919(1)(f); N.M. Stat. Ann. sec. 58-15-34(A) (to be repealed January 1, 2018 as noted above); Okla. Stat. tit. 59, sec. 3109(A); S.C. Code Ann. sec. 34-39-180(F); Tenn. Code Ann. sec. 45-17-112(q); Va. Code Ann. sec. 6.2-1816(6); Wash. Rev. Code sec. 31.45.073(2); Wyo. Stat. Ann. sec. 40-14-364. Other States such as Iowa and Kansas restrict a loan from being repaid with the proceeds of another loan. Iowa Code sec. 533D.10(1)(e); Kan. Stat. Ann. sec. 16a-2404(6). Other States that permit some degree of rollovers include: Alabama (one); Alaska (two); Delaware (four); Idaho (three); Missouri (six if there is at least 5 percent principal reduction on each rollover); Nevada (may extend loan up to 60 days after the end of the initial loan term); North Dakota (one); Oregon (two); Rhode Island (one); Utah (allowed up to 10 weeks after the execution of the first loan); and Wisconsin (one). Ala. Code sec. 5-18A12(b); Alaska Stat. sec. 06.50.470(b); Del. Code Ann. tit. 5, sec. 2235A(a)(2); Idaho Code Ann. sec. 28-46-413(9); Mo. Rev. Stat. sec. 408.500(6); Nev. Rev. Stat. sec. 604A.480(1); N.D. Cent. Code sec. 13-08-12(12); Or. Rev. Stat. sec. 725A.064(6); R.I. Gen. Laws sec. 19-14.4-5.1(g); Utah Code Ann. sec. 7-23-401(4)(b); Wis. Stat. sec. 138.14 (12)(a). 37 See CFPB Payday Loans and Deposit Advance Products White Paper, at 94; Julie A. Meade, Adm’r of the Colo. Unif. Consumer Credit Code Unit, Colo. Dep’t of Law, “Payday Lending Demographic and Statistical Information: July 2000 through December 2012,” at 24 (Apr. 10, 2014), available at http://www.coloradoattorneygeneral.gov/sites/default/files/contentuploads/cp/ConsumerCreditUnit/UCCC/AnnualR eportComposites/DemoStatsInfo/ddlasummary2000-2012.pdf; Pew Charitable Trusts, “Payday Lending in America: Who Borrows, Where They Borrow, and Why,” at 15 (Report 1, 2012), available at http://www.pewtrusts.org/~/media/legacy/uploadedfiles/pcs_assets/2012/pewpaydaylendingreportpdf.pdf; Leslie Parrish & Uriah King, “Phantom Demand: Short-term Due Date Generates Need for Repeat Payday Loans, Accounting for 76% of Total Volume,” at 7 (Ctr. for Responsible Lending 2009), available at http://www.responsiblelending.org/payday-lending/research-analysis/phantom-demand-final.pdf. 38 States with cooling-off periods include: Alabama (next business day after a rollover is paid in full); Florida (24 hours); Illinois (seven days after a consumer has had payday loans for more than 45 days); Indiana (seven days after five consecutive loans); New Mexico (10 days after completing an extended payment plan) (to be repealed Jan. 1, 2018 as noted above); North Dakota (three business days); Ohio (one day with a two loan limit in 90 days, four per year); Oklahoma (two business days after fifth consecutive loan); Oregon (seven days); South Carolina (one business day between all loans and two business days after seventh loan in a calendar year); Virginia (one day between all loans, 45 days after fifth loan in a 180-day period, and 90 days after completion of an extended payment plan or extended term loan); and Wisconsin (24 hour after renewals). Ala. Code sec. 5-18A-12(b); Fla. Stat. sec. 23 At least 17 States have adopted laws that require payday lenders to offer borrowers the option of taking an extended repayment plan when they encounter difficulty in repaying payday loans.39 Details about the extended repayment plans vary including: borrower eligibility (in some States only prior to the lender instituting collections or litigation); how borrowers may elect to participate in repayment plans; the number and timing of payments; the length of plans; permitted fees for plans; requirements for credit counseling; requirements to report plan payments to a statewide database; cooling-off or “lock-out” periods for new loans after completion of plans; and the consequences of plan defaults. Two States more generally allow lenders the discretion to offer borrowers an extension of time to repay or enter into workout agreements with borrowers having repayment difficulties.40 The effects of these various restrictions are discussed further below in Market Concerns—Underwriting. Industry size and structure. There are various estimates as to the number of consumers who use payday loans on an annual basis. One survey found that 2.5 million households (2 560.404(19); 815 Ill. Comp. Stat. 122/2-5(b); Ind. Code sec. 24-4.5-7-401(2); N.M. Stat. Ann. sec. 58-15-36; N.D. Cent. Code sec. 13-08-12(4); Ohio Rev. Code Ann. sec. 1321.41(E), (N), (R); Okla. Stat. tit. 59, sec. 3110; Or. Rev. Stat. sec. 725A.064(7); S.C. Code Ann. sec. 34–39–270(A), (B); Va. Code Ann. sec. 6.2–1816(6); Wis. Stat. sec. 138.14(12)(a). 39 States with statutory extended repayment plans include: Alabama, Alaska, Florida, Idaho, Illinois, Indiana, Louisiana, Michigan (fee permitted), Nevada, New Mexico (to be repealed Jan. 1, 2018 as noted above), Oklahoma (fee permitted), South Carolina, Utah, Virginia, Washington, Wisconsin, and Wyoming. Florida also requires that as a condition of providing a repayment plan (called a grace period), borrowers make an appointment with a consumer credit counseling agency and complete counseling by the end of the plan. Ala. Code sec. 5-18A-12(c); Alaska Stat. sec. 06.50.550(a); Fla. Stat. sec. 560.404(22)(a); Idaho Code Ann. sec. 28-46-414; 815 Ill. Comp. Stat. 122/2-40; Ind. Code sec. 24-4.5-7-401(3); La. Rev. Stat. Ann. sec. 9:3578.4.1; Mich. Comp. Laws sec. 487.2155(2); Nev. Rev. Stat. sec. 604A.475(1); N.M. Stat. Ann. sec. 58-15-35; Okla. Stat. tit. 59, sec. 3109(D); S.C. Code Ann. sec. 34-39-280; Utah Code Ann. sec. 7-23-403; Va. Code Ann. sec. 6.2-1816(26); Wash. Rev. Code sec. 31.45.084(1); Wis. Stat. sec. 138.14(11)(g); Wyo. Stat. Ann. sec. 40–14–366(a). 40 California (no fees permitted) and Delaware are States that permit payday lenders to extend the time for repayment of payday loans. Cal. Fin. Code sec. 23036(b); Del. Code Ann. tit. 5, sec. 2235A(a)(2). 24 percent of U.S. households) used payday loans in 2015.41 In another survey, 3.4 percent of households reported taking out a payday loan in the past year.42 These surveys referred to payday loans generally, and did not specify whether they were referring to loans made online or at storefront locations. One report estimated the number of individual borrowers, rather than households, was higher at approximately 12 million annually and included both storefront and online loans.43 See Market Concerns—Underwriting for additional information on borrower characteristics. There are several ways to gauge the size of the storefront payday loan industry. Typically, the industry has been measured by counting the total dollar value of each loan made during the course of a year, counting each rollover, back-to-back loan or other re-borrowing as a new loan that is added to the total. By this metric, one industry analyst estimated that from 2009 to 2014, storefront payday lending generated approximately $30 billion in new loans per year and that by 2015 the volume had declined to $23.6 billion,44 although these numbers may include products other than single-payment loans. The analyst’s estimate for combined storefront and online payday loan volume was $45.3 billion in 2014 and $39.5 billion in 2015, down from a peak of about $50 billion in 2007.45 41 Fed. Deposit Ins. Corp., “2015 FDIC National Survey of Unbanked and Underbanked Households,” at 2, 34 (Oct. 20, 2016), available at https://www.fdic.gov/householdsurvey/2015/2015report.pdf. 42 Jesse Bricker, et al., “Changes in U.S. Family Finances from 2013 to 2016: Evidence from the Survey of Consumer Finances,” at 27 (Bd. of Governors of the Fed. Reserve Sys., 103 Fed. Reserve Bulletin No. 3, 2017), available at https://www.federalreserve.gov/publications/files/scf17.pdf. 43 Pew Charitable Trusts, “Payday Lending in America: Who Borrows, Where They Borrow, and Why,” at 4 (Report 3, 2013), available at http://www.pewtrusts.org/~/media/legacy/uploadedfiles/pcs_assets/2012/pewpaydaylendingreportpdf.pdf. 44 John Hecht, “The State of Short-Term Credit Amid Ambiguity, Evolution and Innovation” (2016) (Jefferies LLC, slide presentation) (on file); John Hecht, “The State of Short-Term Credit in a Constantly Changing Environment” at 4 (2015) (Jeffries LLC, slide presentation) (on file). 45 Hecht, “Short-Term Credit Amid Ambiguity.” 25 Alternatively, the industry can be measured by calculating the dollar amount of loan balances outstanding. Given the amount of payday loan re-borrowing, which results in the same funds of the lender being used to finance multiple loan originations to the same borrower, the dollar amount of loan balances outstanding may provide a more nuanced sense of the industry’s scale. Using this metric, the Bureau estimates that in 2012, storefront payday lenders held approximately $2 billion in outstanding single-payment loans.46 In 2015, industry revenue (fees paid on storefront payday loans) was an estimated $3.6 billion, representing 15 percent of loan originations. Combined storefront and online payday revenue was estimated at $8.7 billion in 2014 and $6.7 billion in 2015, down from a peak of over $9 billion in 2012.47 In the last several years, it has become increasingly difficult to identify the largest payday lenders due to firm mergers, diversification by many lenders into a range of products including installment loans and retraction by others into pawn loans, and the lack of available data because most firms are privately held. However, there are at least 10 lenders with approximately 200 or more storefront locations.48 Only a few of these firms are publicly traded companies.49 Most 46 The Bureau’s staff estimate is based on public company financial information, confidential information gathered in the course of statutory functions, and industry analysts’ reports. The estimate is derived from lenders’ singlepayment payday loans gross receivables and gross revenue and industry analysts’ reports on loan volume and revenue. No calculations were done for 2013 to 2016, but that estimate would be less than $2 billion due to changes in the market as the industry has shifted away from single-payment payday loans to products discussed below. 47 Hecht, “Short-Term Credit Amid Ambiguity.” 48 These firms include: ACE Cash Express, Advance America, Amscot Financial, Axcess Financial (CNG Financial, Check ‘n Go, Allied Cash), Check Into Cash, Community Choice Financial (Checksmart), CURO Financial Technologies (Speedy Cash/Rapid Cash), DFC Global Corp (Money Mart), FirstCash, and QC Holdings. See Ace Cash Express, “Store Locator,” available at https://www.acecashexpress.com/locations; Advance America, “Find an Advance America Store Location,” available at https://www.advanceamerica.net/locations/find; Amscot Financial, Inc., “Amscot Locations,” available at https://www.amscot.com/locations.aspx; Check ‘n Go, “State Center,” available at https://www.checkngo.com/resources/state-center; Allied Cash Advance, “Allied Cash Advance Store Directory,” available at https://locations.alliedcash.com/index.html; Check Into Cash, “Payday Loan Information By State,” available at https://checkintocash.com/payday-loan-information-by-state; Community Choice Financial (CheckSmart), “Locations,” available at https://www.ccfi.com/locations/; SpeedyCash, “Speedy 26 large payday lenders are privately held,50 and the remaining payday loan stores are owned by smaller regional or local entities. The Bureau estimates there are about 2,400 storefront payday lenders that are small entities as defined by the Small Business Administration (SBA).51 Several industry commenters, an industry trade association commenter, and a number of payday lenders noted that they offer non-credit products and services at their locations including check cashing, money transmission and bill payments, sale of prepaid cards, and other services, some of which require them to comply with other laws as “money service businesses.” Cash Stores Near Me,” available at https://www.speedycash.com/find-a-store; DFC Global Corp., “Home,” available at http://www.dfcglobalcorp.com/index.html; FirstCash Inc., “Find a Location Near You,” available at http://www.firstcash.com/; QC Holdings, Inc., “Branch Locator,” available at https://www.qcholdings.com/branchlocator.aspx (all sites last visited Jul. 26, 2017). 49 The publicly traded firms are Community Choice Financial Inc./Cash Central/Checksmart (CCFI), EZCORP, Inc. (EZPW), FirstCash Inc. (FCFS), and QC Holdings (QCCO). As noted above, in September 2016, FirstCash Financial Services merged with Cash America, resulting in the company FirstCash Inc. Prior to the merger, in November 2014, Cash America migrated its online loans to a spin-off company, Enova. Cash America International, Inc., 2015 Annual Report (Form 10–K), at 3 (Dec. 14, 2016). Both FCFS and Cash America had been deemphasizing payday lending in the U.S., and shifting towards pawn. In 2016, the new company, FirstCash, had only 45 stand-alone consumer loan locations, in Texas, Ohio, and California, and 326 pawn locations that also offered consumer loans, compared to 1,085 pawn locations. Only 4 percent of its revenue was from non-pawn consumer loans and credit services operations. (Credit services organizations are described below.) FirstCash Inc., 2016 Annual Report (Form 10–K), at 5, 7. In 2015, EZCORP exited payday, installment, and auto title lending, focusing domestically on pawn lending. EZCORP, Inc., 2016 Annual Report (Form 10–K), at 3 (Dec. 14, 2016). QC Holdings delisted from Nasdaq in February 2016 and is traded over-the-counter. QC Holdings, Inc., Suspension of Duty to File Reports Under Sections 13 and 15(d) (Form 15). 50 The larger privately held payday lending firms include Advance America, ACE Cash Express, Axcess Financial (CNG Financial, Check ‘n Go, Allied Cash), Check Into Cash, DFC Global (Money Mart), PLS Financial Services, and Speedy Cash Holdings Corporation. See Susanna Montezemolo, “Payday Lending Abuses and Predatory Practices: The State of Lending in America & Its Impact on U.S. Households” at 9–10 (Ctr. for Responsible Lending, 2013); John Hecht, “Alternative Financial Services: Innovating to Meet Customer Needs in an Evolving Regulatory Framework” (2014) (Stephens, Inc., slide presentation) (on file). 51 Bureau staff estimated the number of storefront payday lenders using licensee information from State financial regulators, firm revenue information from public filings and non-public sources, and, for a small number of States, industry market research relying on telephone directory listings from Steven Graves and Christopher Peterson, available at http://www.csun.edu/~sg4002/research/data/US_pdl_addr.xls. Based on these sources, there are approximately 2,503 storefront payday lenders, including those operating primarily as loan arrangers or brokers, in the United States. Based on the publicly-available revenue information, at least 56 of the firms have revenue above the small entity threshold. Most of the remaining firms operate a very small number of storefronts. Therefore, while some of the firms without publicly available information may have revenue above the small entity threshold, in the interest of being inclusive they are all assumed to be small entities. 27 According to one industry analyst, there were an estimated 16,480 payday loan stores in 2015 in the United States, a decline from 19,000 stores in 2011 and down from the industry’s 2007 peak of 24,043 storefronts.52 The average number of payday loan stores in a county with a payday loan store is 6.32.53 The Bureau has analyzed payday loan store locations in States which maintain lists of licensed lenders and found that half of all stores are less than one-third of a mile from another store, and three-quarters are less than a mile from the nearest store.54 Even the 95th percentile of distances between neighboring stores is only 4.3 miles. Stores tend to be closer together in counties within metropolitan statistical areas (MSA).55 In non-MSA counties the 75th percentile of distance to the nearest store is still less than one mile, but the 95th percentile is 22.9 miles. Research and the Bureau’s own market outreach indicate that payday loan stores tend to be relatively small with, on average, three full-time equivalent employees.56 An analysis of loan data from 29 States found that the average store made 3,541 advances in a year.57 Given rollover 52 Hecht, “Short-Term Credit Amid Ambiguity,” at 7. Although there is no estimate for 2016, the number of storefronts offering payday loans is likely smaller due to the regulatory changes in South Dakota, the exit of EZCORP from payday lending, and the merger of First Cash Financial and Cash America, and its shift away from payday lending. However, it is difficult to precisely measure the number of stores that have shifted from payday to pawn lending, rather than closing. By way of comparison, in 2015 there were 14,259 McDonald’s fast food outlets in the United States. McDonald’s Corp., 2015 Annual Report (Form 10–K), at 23 (Feb. 25, 2016). 53 James R. Barth, et al., “Do State Regulations Affect Payday Lender Concentration?,” at 12 (2015), available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2581622. 54 CFPB Report on Supplemental Findings, at 90. 55 An MSA is a geographic entity delineated by the Office of Management and Budget. An MSA contains a core urban area of 50,000 or more in population. See U.S. Census Bureau, “Metropolitan and Micropolitan,” available at http://www.census.gov/population/metro/ (last visited Apr. 7, 2016). 56 Mark Flannery & Katherine Samolyk, “Payday Lending: Do the Costs Justify the Price?,” (FDIC Ctr. for Fin. Res., Working Paper No. 2005–09, 2005), available at https://www.fdic.gov/bank/analytical/cfr/2005/wp2005/cfrwp_2005-09_flannery_samolyk.pdf. 57 Susanna Montezemolo, “Payday Lending Abuses and Predatory Practices: The State of Lending in America & Its Impact on U.S. Households” at 26 n.2 (Ctr. for Responsible Lending, 2013), available at http://www.responsiblelending.org/state-of-lending/reports/10-Payday-Loans.pdf. 28 and re-borrowing rates, a report estimated that the average store served fewer than 500 customers per year.58 Marketing, underwriting, and collections practices. Payday loans tend to be marketed as a short-term bridge to cover emergency expenses. For example, one lender suggests that, for consumers who have insufficient funds on hand to meet such an expense or to avoid a penalty fee, late fee, or utility shut-off, a payday loan can “come in handy” and “help tide you over until your next payday.”59 Some lenders offer new borrowers their initial loans at no fee (“first loan free”) to encourage consumers to try a payday loan.60 Stores are typically located in high-traffic commuting corridors and near shopping areas where consumers obtain groceries and other staples.61 The evidence of price competition among payday lenders is mixed. In their financial reports, publicly traded payday lenders have reported their key competitive factors to be nonprice related. For instance, they cite location, customer service, and convenience as some of the primary factors on which payday lenders compete with one another, as well as with other financial service providers.62 Academic studies have found that, in States with rate caps, loans 58 Pew Charitable Trusts, “Payday Lending in America: Policy Solutions,” at 18 (Report 3, 2013), available at http://www.pewtrusts.org/~/media/legacy/uploadedfiles/pcs_assets/2013/pewpaydaypolicysolutionsoct2013pdf.pdf. 59 Cash America Int’l Inc., “Cash Advance/Short-term Loans,” available at http://www.cashamerica.com/LoanOptions/CashAdvances.aspx (last visited Apr. 7, 2016). 60 See, e.g., Instant Cash Advance Corp., “Instant PayDay,” available at http://www.instantcashadvancecorp.com/free-loan-offer-VAL312.php (introductory offer of a free (no fee) cash advance of $200) (storefront payday loans); Check N Title Loans, “First Loan Free,” available at http://www.checkntitle.com/ (storefront payday and title loans); AmeriTrust Financial LLC, “1st Advance Free,” available at http://www.americantrustcash.com/payday-loans (storefront payday, title, and installment loans, first loan free on payday loans) (all firm websites last visited on Dec. 21, 2015). 61 See FirstCash, Inc., 2016 Annual Report (Form 10–K), at 9; QC Holdings, Inc., 2014 Annual Report (Form 10– K), at 11; Community Choice Fin. Inc., 2016 Annual Report (Form 10–K), at 6. 62 See QC Holdings, Inc., 2014 Annual Report (Form 10–K), at 12–13. 29 are almost always made at the maximum rate permitted.63 Another study likewise found that in States with rate caps, firms lent at the maximum permitted rate, and that lenders operating in multiple States with varying rate caps raise their fees to those caps rather than charging consistent fees company-wide. The study found, however, that in States with no rate caps, different lenders operating in those States charged different rates. The study reviewed four lenders that operate in Texas64 and observed differences in the cost to borrow $300 per two-week pay period: two lenders charged $61 in fees, one charged $67, and another charged $91, indicating some level of price variation between lenders (ranging from about $20 to $32 per $100 borrowed).65 One industry commenter cited the difference in average loan pricing between storefront (generally lower) and online loans (generally higher), as evidence of price competition but that is more likely due to the fact that state-licensed lenders are generally constrained in the amount they can charge rather than competitive strategies adopted by those lenders. That commenter also notes as evidence of price competition that it sometimes discounts its own loans from its advertised prices; the comment did not address whether such discounts were offered to meet competition. 63 Robert DeYoung & Ronnie Phillips, “Payday Loan Pricing,” at 27–28, (Fed. Reserve Bank of Kan. City, Working Paper No. RWP 09–07, 2009), available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1066761 (studying rates on loans in Colorado between 2000 and 2006); Mark Flannery & Katherine Samolyk, “Payday Lending: Do the Costs Justify the Price?,” at 9–10 (FDIC Ctr. for Fin. Res., Working Paper No. 2005–09, 2005), available at https://www.fdic.gov/bank/analytical/cfr/2005/wp2005/cfrwp_2005-09_flannery_samolyk.pdf. 64 In Texas, these lenders operate as credit services organizations or loan arrangers with no fee caps, described in more detail below. Pew Charitable Trusts, “How State Rate Limits Affect Payday Loan Prices,” (Apr. 2014), available at http://www.pewtrusts.org/~/media/legacy/uploadedfiles/pcs/contentlevel_pages/fact_sheets/stateratelimitsfactsheetpdf.pdf. 65 Pew Charitable Trusts, “How State Rate Limits Affect Payday Loan Prices,” (Apr. 2014), available at http://www.pewtrusts.org/~/media/legacy/uploadedfiles/pcs/contentlevel_pages/fact_sheets/stateratelimitsfactsheetpdf.pdf. 30 The application process for a payday loan is relatively simple. For a storefront payday loan, a borrower must generally provide some verification of income (typically a pay stub) and evidence of a personal deposit account.66 Although a few States impose limited requirements that lenders consider a borrower’s ability to repay,67 storefront payday lenders generally do not consider a borrower’s other financial obligations or require collateral (other than the check or electronic debit authorization) for the loan. Most storefront payday lenders do not consider traditional credit reports or credit scores when determining loan eligibility, nor do they report any information about payday loan borrowing history to the nationwide consumer reporting agencies, TransUnion, Equifax, and Experian.68 From market outreach activities and confidential information gathered in the course of statutory functions, the Bureau is aware that a number of storefront payday lenders obtain data from one or more specialty consumer reporting agencies during the loan application process to check for previous payday loan defaults, identify recent inquiries that suggest an intention to not repay the loan, and perform other due diligence such as identity and deposit account verification. Some storefront payday lenders use analytical 66 See, e.g., Check Into Cash, “Frequently Asked Questions and Policies of Check into Cash,” available at https://checkintocash.com/faqs/in-store-cash-advance/ (last visited Sept. 14, 2017) (process as described by one lender). 67 For example, Utah requires lenders to make an inquiry to determine that the borrower has the ability to repay the loan, which may include rollovers or extended payment plans. This determination may be made through borrower affirmation of ability to repay, proof of income, repayment history at the same lender, or information from a consumer reporting agency. Utah Code sec. 7-23-401. Missouri requires lenders to consider borrower financial ability to reasonably repay under the terms of the loan contract, but does not specify how lenders may satisfy this requirement. Mo. Rev. Stat sec. 408.500(7). Effective July 1, 2017, Nevada lenders must assess borrowers’ reasonable ability to repay by considering, to the extent available, their current or expected income; current employment status based on a pay stub, bank deposit, or other evidence; credit history; original loan amount due, or for installment loans or potential repayment plans, the monthly payment amount; and other evidence relevant to ability to repay including bank statements and borrowers’ written representations. Other States prohibit loans that exceed a certain percentage of the borrower’s gross monthly income (generally between 20 and 35 percent) as a proxy for ability to repay as described above. 68 See, e.g., Neil Bhutta, et al., “Payday Loan Choices and Consequences,” 47 J. of Money, Credit and Banking 223 (2015). 31 models and scoring that attempt to predict likelihood of default.69 Through market outreach and confidential information gathered in the course of statutory functions, the Bureau is aware that many storefront payday lenders only conduct their limited underwriting for first-time borrowers or those returning after an absence. From market outreach, the Bureau is aware that the specialty consumer reporting agencies contractually require any lender that obtains data to also report data to them, although compliance may vary. Reporting usually occurs on a real-time or same-day basis. Separately, 14 States require lenders to check statewide databases before making each loan in order to ensure that their loans comply with various State restrictions.70 These States likewise require lenders to report certain lending activity to the database, generally on a real-time or same-day basis. As discussed in more detail above, these State restrictions may include prohibitions on consumers having more than one payday loan at a time, cooling-off periods, or restrictions on the number of loans consumers may take out per year. Although a consumer is generally required when obtaining a loan to provide a post-dated check or authorization for an electronic debit of the consumer’s account which could be 69 See, e.g., Advance America, “FAQs on Payday Loans/Cash Advances: Is my credit score checked before receiving an in-store Payday Loan?,” available at https://www.advanceamerica.net/questions/payday-loans-cashadvances (last visited May 10, 2017) (the custom scoring model described by one lender). 70 The States with databases are Alabama, Delaware, Florida, Illinois Indiana, Kentucky, Michigan, New Mexico (to be repealed Jan. 1, 2018 as noted above), North Dakota, Oklahoma, South Carolina, Virginia, Washington, and Wisconsin. Illinois also requires use of its database for payday installment loans, vehicle title loans, and some installment loans. Some State laws allow lenders to charge borrowers a fee to access the database that may be set by statute. Ala. Code sec. 5–-18A-13(o); Del. Code Ann. tit. 5, sec. 2235B; Fla. Stat. sec. 560.404(23); 815 Ill. Comp. Stat. 122/2–-15; Ind. Code sec. 24-4.5-7-404(4); Ky. Rev. Stat. Ann. sec. 286.9-100(19)(b); Mich. Comp. Laws sec. 487.2142; N.M. Stat. Ann. sec. 58-15-37(B); N.D. Cent. Code sec. 13-08-12(4); Okla. Stat. tit. 59, sec. 3109(B)(2)(b); S.C. Code Ann. sec. 34-39-175; Va. Code Ann. sec. 6.2-1810; Wash. Rev. Code sec. 31.45.093; Wis. Stat. sec. 138.14(14). 32 presented to the consumer’s bank,71 consumers in practice generally return to the store when the loan is due to “redeem” the check either by repaying the loan or by paying the finance charges and rolling over the loan.72 For example, a major payday lender with a predominantly storefront loan portfolio reported that in 2014, over 90 percent of its payday loan volume was repaid in cash at its branches by consumers either paying in full or by paying the “original loan fee” (finance charges) and rolling over the loan (signing a new promissory note and leaving a new check or payment authorization).73 An industry commenter stated that repayment in cash reflects customers’ preferences. However, borrowers are strongly encouraged and in some cases required by lenders to return to the store when payment is due. Some lenders give borrowers appointment cards with a date and time to encourage them to return with cash. For example, one major storefront payday lender explained that after loan origination “the customer then makes an appointment to return on a specified due date, typically his or her next payday, to repay the cash advance …. Payment is usually made in person, in cash at the center where the cash advance was initiated ….”74 The Bureau is aware, from confidential information gathered in the course of statutory functions and from market outreach, that lenders routinely make reminder calls to borrowers a few days before loan due dates to encourage borrowers to return to the store. One large lender 71 Payments may also be taken from the consumer’s debit card. See, e.g., All American Check Cashing, Inc., Miss. Dep’t of Banking and Consumer Fin., Administrative Order, Cause No. 2016–001, May 11, 2017, available at http://www.dbcf.ms.gov/documents/actions/consumerfin/aa0517.pdf. 72 According to the Bureau’s market outreach, if borrowers provided ACH authorization and return to pay the loan in cash, the authorization may be returned to them or voided. 73 QC Holdings, 2014 Annual Report (Form 10–K), at 7. 74 Advance America, 2011 Annual Report (Form 10–K) at 45 (Mar. 15, 2012). See also Check Into Cash, “Cash Advance Loan FAQs, What is a cash advance?,” available at https://checkintocash.com/faqs/in-store-cash-advance/ (last visited Feb. 4, 2016) (“We hold your check until your next payday, at which time you can come in and pay back the advance.”). 33 reported this practice in a public filing.75 Another storefront payday lender requires its borrowers to return to the store to repay. Its website states: “All payday loans must be repaid with either cash or money order. Upon payment, we will return your original check to you.”76 The Bureau is also aware, from confidential information gathered in the course of statutory functions, that one or more storefront payday lenders have operating policies that specifically state that cash is preferred because only half of their customers’ checks would clear if deposited on the loan due dates. Encouraging or requiring borrowers to return to the store on the due date provides lenders an opportunity to offer borrowers the option to roll over the loan or, where rollovers are prohibited by State law, to re-borrow following repayment or after the expiration of any cooling-off period. Most storefront lenders examined by the Bureau employ monetary incentives that reward employees and store managers for loan volumes, although one industry commenter described the industry’s incentives to employees as rewards for increases in net revenue. Since as discussed below, a majority of loans result from rollovers of existing loans or re-borrowing contemporaneously with or shortly after loans have been repaid, rollovers and re-borrowing contribute substantially to employees’ compensation. From confidential information gathered in the course of statutory functions, the Bureau is aware that rollover and re-borrowing offers are made when consumers log into their accounts online, during “courtesy calls” made to remind borrowers of upcoming due dates, and when borrowers repay in person at storefront locations. In addition, some lenders train their employees to offer rollovers during 75 When Advance America was a publicly traded corporation, it reported: “The day before the due date, we generally call the customer to confirm their payment due date.” Advance America, 2011 Annual Report (Form 10–K), at 11. 76 Instant Cash Advance, “How Cash Advances Work,” available at http://www.instantcashadvancecorp.com/services/payday-loans/ (last visited July 17, 2017). 34 courtesy calls when borrowers notified lenders that they had lost their jobs or suffered pay reductions. Store personnel often encourage borrowers to roll over their loans or to re-borrow, even when consumers have demonstrated an inability to repay their existing loans. In an enforcement action, the Bureau found that one lender maintained training materials that actively directed employees to encourage re-borrowing by struggling borrowers. It further found that if a borrower did not repay or pay to roll over the loan on time, store personnel would initiate collections. Store personnel or collectors would then offer the option to take out a new loan to pay off an existing loan, or refinance or extend the loan as a source of relief from the potentially negative outcomes (e.g., lawsuits, continued collections). This “cycle of debt” was depicted graphically as part of “The Loan Process” in the company’s new hire training manual.77 In Mississippi, another lender employed a companywide practice in which store personnel encouraged borrowers with monthly income or benefits payments to use the proceeds of one loan to pay off another loan, although State law prohibited these renewals or rollovers.78 In addition, though some States require lenders to offer borrowers the option of extended repayment plans and some trade associations have designated provision of such plans as a best 77 Press Release, Bureau of Consumer Fin. Prot., “CFPB Takes Action Against ACE Cash Express for Pushing Payday Borrowers Into Cycle of Debt,” (July 10, 2014), available at http://www.consumerfinance.gov/newsroom/cfpb-takes-action-against-ace-cash-express-for-pushing-paydayborrowers-into-cycle-of-debt/. 78 All American Check Cashing, Inc., Miss. Dep’t of Banking and Consumer Fin., Administrative Order, Cause No. 2016–001, May 11, 2017, available at http://www.dbcf.ms.gov/documents/actions/consumerfin/aa0517.pdf. The lender also failed to refund consumer overpayments. The State regulator ordered revocation of all of the lender’s 75 licenses, consumer refunds, civil penalties of over $1 million, and other relief. All American appealed the order and the matter was settled with terms reducing the penalty to $889,350. Agreed Order of Dismissal with Prejudice, All American Check Cashing Inc. v. Miss. Dep’t of Banking and Consumer Fin., No. G–2017–699 S/2 (Miss. 2017), available at http://www.dbcf.ms.gov/documents/aacc_agreed_060917.pdf. 35 practice, individual lenders may often be reluctant to offer them. In Colorado, for instance, some payday lenders reported, prior to a regulatory change in 2010, that they had implemented practices to restrict borrowers from obtaining the number of loans needed to be eligible for the State-mandated extended payment plan option and that some lenders had banned borrowers who had exercised their rights to elect payment plans from taking new loans.79 The Bureau is also aware, from confidential information gathered in the course of statutory functions, that one or more lenders used training manuals that instructed employees not to mention these plans until after employees first offered rollovers, and then only if borrowers specifically asked about the plans. Indeed, details on implementation of the repayment plans that have been designated by two national trade associations for storefront payday lenders as best practices are unclear, and in some cases place a number of limitations on exactly how and when a borrower must request assistance to qualify for these “off-ramps.” For instance, one trade association representing more than half of all payday loan stores states that as a condition of membership, members must offer an “extended payment plan” but that borrowers must request the plan at least one day prior to the date on which the loan is due, generally in person at the store where the loan was made or otherwise by the same method used to originate the loan.80 Another trade association with over 79 See State of Colo. Dep’t of Law, Off. of Att’y Gen., “2009 Deferred Deposit/Payday Lenders Annual Report,” at 2, available at http://www.coloradoattorneygeneral.gov/sites/default/files/contentuploads/cp/ConsumerCreditUnit/UCCC/AnnualR eportComposites/2009_ddl_composite.pdf. See also Market Concerns— Covered Loans below for additional discussion of lenders’ extended payment plan practices. 80 Community Fin. Servs. Ass’n of America, “About CFSA,” available at http://cfsaa.com/about-cfsa.aspx (last visited Jan. 15, 2016); Community Fin. Servs. Ass’n of America, “CFSA Member Best Practices,” available at http://cfsaa.com/cfsa-member-best-practices.aspx (last visited Sept. 15, 2017); Community Fin. Servs. Ass’n of America, “What Is an Extended Payment Plan?,” available at http://cfsaa.com/cfsa-member-best-practices/what-isan-extended-payment-plan.aspx (last visited Jan. 15, 2016). Association documents direct lenders to display a “counter card” describing the association’s best practices. Plans are to be offered in the absence of State-mandated plans at no charge and payable in four equal payments coinciding with paydays. 36 1,300 members, including both payday lenders and firms that offer non-credit products such as check cashing and money transmission, states that members will provide the option of extended payment plans in the absence of State-mandated plans to customers unable to repay, but details of the plans are not publicly available on its website.81 From confidential information gathered in the course of statutory functions and market outreach, the Bureau is aware that if a borrower fails to return to the store when a loan is due, the lender may attempt to contact the consumer and urge the consumer to make a cash payment before eventually depositing the post-dated check that the consumer had provided at origination or electronically debiting the account. The Bureau is also aware of some situations in which lenders have obtained electronic payments from borrowers’ bank accounts and also accepted cash payments from borrowers at storefronts.82 The Bureau is aware, from confidential information gathered in the course of its statutory functions and from market outreach, that lenders may use various methods to try to ensure that a payment will clear before presenting a check or ACH. These efforts may range from storefront lenders calling the borrower’s bank to ask if a check of a particular size would clear the account to the use of software offered by a 81 Fin. Serv. Ctrs. of America, “Membership,” http://www.fisca.org/AM/Template.cfm?Section=Membership (last visited Sept. 15, 2017); Joseph M. Doyle, “Chairman’s Message,” Fin. Serv. Ctrs. of America, http://www.fisca.org/AM/Template.cfm?Section=Chairman_s_Message&Template=/CM/HTMLDisplay.cfm&Conte ntID=19222 (last visited Jan. 15, 2016); Fin. Serv. Ctrs. of America, “FiSCA Best Practices,” http://www.fisca.org/Content/NavigationMenu/AboutFISCA/CodesofConduct/default.htm (last visited Jan. 15, 2016). 82 See Bureau of Consumer Fin. Prot., “Supervisory Highlights,” at 31–32 (Summer 2017), available at https://s3.amazonaws.com/files.consumerfinance.gov/f/documents/201709_cfpb_Supervisory-Highlights_Issue16.pdf. See also, Press Release, Bureau of Consumer Fin. Prot., “CFPB Takes Action Against Check Cashing and Payday Lending Company for Tricking and Trapping Consumers,” (May 11, 2016), available at https://www.consumerfinance.gov/about-us/newsroom/cfpb-takes-action-against-check-cashing-and-paydaylending-company-tricking-and-trapping-consumers/; All American Check Cashing, Inc., Miss. Dept. of Banking and Consumer Fin., Administrative Order, No. 2016–001 (May 11, 2017), available at http://www.dbcf.ms.gov/documents/actions/consumerfin/aa0517.pdf (for a description of one lender’s alleged failure to refund overpayments resulting from these procedures and an associated State agency’s order against that lender.). 37 number of vendors that attempts to model likelihood of repayment (“predictive ACH”).83 If these attempts are unsuccessful, store personnel at either the storefront level or at a centralized location will then generally engage in collection activity. Collection activity may involve further in-house attempts to collect from the borrower’s bank account.84 If the first attempt fails, the lender may make subsequent attempts at presentment by splitting payments into smaller amounts in hopes of increasing the likelihood of obtaining at least some funds, a practice for which the Bureau recently took enforcement action against a small-dollar lender.85 Or, the lender may attempt to present the payment multiple times, a practice that the Bureau has noted in supervisory examinations.86 A more detailed discussion of payments practices is provided in part D and Markets Concerns—Payments. Eventually, the lender may attempt other means of collection. The Bureau is aware of inhouse debt collections activities, by both storefront employees and employees at centralized collections divisions, including calls, letters, and visits to consumers and their workplaces, 87 as 83 See, e.g., Press Release, Clarity Servs., “ACH Presentment Will Help Lenders Reduce Failed ACH Pulls,” (Aug. 1, 2013), available at https://www.clarityservices.com/clear-warning-ach-presentment-will-help-lenders-reducefailed-ach-pulls/; Factor Trust, “Markets,” http://ws.factortrust.com/products/ (last visited Apr. 8, 2016); Microbilt, “Bank Account Verify. More Predictive. Better Performance. Lower Costs.,” http://www.microbilt.com/bankaccount-verification.aspx (last visited Apr. 8, 2016); DataX. Ltd., “Know Your Customer,” http://www.dataxltd.com/ancillary-services/successful-collections/ (last visited Apr.8, 2016). 84 For example, one payday lender stated in its public documents that it “subsequently collects a large percentage of these bad debts by redepositing the customers’ checks, ACH collections or receiving subsequent cash repayments by the customers.” FirstCash Fin. Servs., 2014 Annual Report (Form 10–K), at 5 (Feb. 12, 2015). As noted above, FirstCash has now largely exited payday lending. 85 Press Release, Bureau of Consumer Fin. Prot., “CFPB Orders EZCORP to Pay $10 Million for Illegal Debt Collection Tactics,” (Dec. 16, 2015), available at http://www.consumerfinance.gov/newsroom/cfpb-orders-ezcorpto-pay-10-million-for-illegal-debt-collection-tactics/. 86 See Bureau of Consumer Fin. Prot., “Supervisory Highlights,” at 20 (Spring 2014), available at http://files.consumerfinance.gov/f/201405_cfpb_supervisory-highlights-spring-2014.pdf. 87 Bureau of Consumer Fin. Prot., “CFPB Compliance Bulletin 2015–07, In-Person Collection of Consumer Debt,” (Dec. 16, 2015), available at http://files.consumerfinance.gov/f/201512_cfpb_compliance-bulletin-in-personcollection-of-consumer-debt.pdf. 38 well as the sale of debt to third-party collectors.88 The Bureau recently conducted a survey of consumer debt collection experiences; 11 percent of consumers contacted about a debt in collection reported the collection activity was related to payday loan debt.89 Further, the Bureau observed in its consumer complaint data that from November 2013 through December 2016, more than 31,000 debt collection complaints had “payday loan” as the underlying debt. In more than 11 percent of the complaints the Bureau handled about debt collection, consumers selected “payday loans” as the underlying debt.90 In addition, in 2016, the Bureau handled approximately 4,400 complaints in which consumers reported “payday loan” as the complaint product and about 26,600 complaints about credit cards.91 As noted above, there are about 12 million payday loan borrowers annually, and approximately 156 million consumers have one or more credit cards.92 Therefore, by way of 88 For example, prior to discontinuing its payday lending operations, EZCorp indicated that it used a tiered structure of collections on defaulted loans (storefront employees, centralized collections, and then third-parties debt sales). EZCORP, Inc., 2014 Annual Report (Form 10–K), at 9 (Nov. 26, 2014). Advance America utilized calls and letters to past-due consumers, as well as attempts to convert the consumer’s check into a cashier’s check, as methods of collection. Advance America, 2011 Annual Report (Form 10–K), at 11. See ACE Cash Express, Inc., Consent Order, CFPB No. 2014–CFPB–0008 (July 10, 2014), available at http://files.consumerfinance.gov/f/201407_cfpb_consent-order_ace-cash-express.pdf; EZCorp Inc., Consent Order, CFPB No. 2015–CFPB–0031 (Dec. 16, 2015), available at http://files.consumerfinance.gov/f/201512_cfpb_ezcorpinc-consent-order.pdf. See also, Bureau of Consumer Fin. Prot., Market Snapshot: Online Debt Sales,” at 5, 7 (Jan. 2017), available at https://www.consumerfinance.gov/data-research/research-reports/market-snapshot-online-debtsales/ (describing a significant share of payday loan portfolios on websites with online debts for sale). 89 Bureau of Consumer Fin. Prot., “Consumer Experiences with Debt Collection: Findings from the CFPB’s Survey of Consumer Views on Debt,” at 19 (Jan. 2017), available at https://www.consumerfinance.gov/documents/2251/201701_cfpb_Debt-Collection-Survey-Report.pdf. 90 Bureau of Consumer Fin. Prot., “Monthly Complaint Report, Vol. 18,” at 12 (Dec. 2016), available at https://www.consumerfinance.gov/data-research/research-reports/monthly-complaint-report-vol-18/. 91 Bureau of Consumer Fin. Prot., “Consumer Response Annual Report, January 1–December 31, 2016,” at 27, 33– 35 (Mar. 2017), available at https://www.consumerfinance.gov/documents/3368/201703_cfpb_Consumer-ResponseAnnual-Report-2016.PDF. 92 The Bureau’s staff estimate is based on finding that 63 percent of American adults hold an open credit card and Census population estimates. Bureau of Consumer Fin. Prot., “The Consumer Credit Card Market Report,” at 36 (Dec. 2015), available at http://files.consumerfinance.gov/f/201512_cfpb_report-the-consumer-credit-cardmarket.pdf; U.S. Census Bureau, “Annual Estimates of Resident Population for Selected Age Groups by Sex for the United States, States, Counties, and Puerto Rico Commonwealth and Municipios: April 1, 2010 to July 1, 2016,” 39 comparison, for every 10,000 payday loan borrowers, the Bureau handled about 3.7 complaints, while for every 10,000 credit card holders, the Bureau handled about 1.7 complaints. Some payday lenders sue borrowers who fail to repay their loans. A study of small claims court cases filed in Utah from 2005 to 2010 found that 38 percent of cases were attributable to payday loans.93 A recent news report found that the majority of non-traffic civil cases filed in 14 Utah justice courts are payday loan collection lawsuits, and in one justice court, the percentage was as high as 98.8 percent.94 In 2013, the Bureau entered into a Consent Order with a large national payday and installment lender based, in part, on the filing of flawed court documents in about 14,000 debt collection lawsuits.95 However, an industry trade association commenter states that many payday lenders do not file lawsuits on defaulted debt. Business model. As previously noted, the storefront payday industry has built a distribution model that involves a large number of small retail outlets, each serving a relatively small number of consumers. That implies that the overhead cost on a per consumer basis is relatively high. (June 2017), available at https://factfinder.census.gov/bkmk/table/1.0/en/PEP/2016/PEPAGESEX. Other estimates of the number of credit card holders have been higher, meaning that 1.7 complaints per 10,000 credit card holders would be a high estimate. The U.S. Census Bureau estimated there were 160 million credit card holders in 2012, and researchers at the Federal Reserve Bank of Boston estimated that 72.1 percent of U.S. consumers held at least one credit card in 2014. U.S. Census Bureau, “Statistical Abstract of the United States: 2012,” at 740 tbl.1188 (Aug. 2011), available at https://www.census.gov/library/publications/2011/compendia/statab/131ed.html; Claire Greene et al., “The 2014 Survey of Consumer Payment Choice: Summary Results,” at 18 (Fed. Reserve Bank of Boston, No. 16–3, 2016), available at https://www.bostonfed.org/-/media/Documents/researchdatareport/pdf/rdr1603.pdf. As noted above in the text, additional complaints related to both credit cards and payday loans are submitted as debt collection complaints with credit card or payday loan listed as the type of debt. 93 Coalition of Religious Communities, ”Payday Lenders and Small Claims Court Cases in Utah,” at 2 (2005–2010), available at http://www.consumerfed.org/pdfs/PDL-UTAH-court-doc.pdf. 94 Lee Davidson, “Payday Lenders Sued 7,927 Utahns Last Year,” The Salt Lake City Tribune, Aug. 2, 2016, http://www.sltrib.com/home/3325528-155/payday-lenders-sued-7927-utahns-last. 95 Press Release, Bureau of Consumer Fin. Prot.,” Consumer Financial Protection Bureau Takes Action Against Payday Lender for Robo-Signing,” (Nov. 20, 2013), available at http://www.consumerfinance.gov/newsroom/consumer-financial-protection-bureau-takes-action-against-paydaylender-for-robo-signing/. 40 Additionally, the loss rates on storefront payday loans—the percentage or amounts of loans that are charged off by the lender as uncollectible—are relatively high. Loss rates on payday loans often are reported on a per-loan basis but, given the frequency of rollovers and renewals, that metric understates the amount of principal lost to borrower defaults. For example, if a lender makes a $100 loan that is rolled over nine times, at which point the consumer defaults, the per-loan default rate would be 10 percent whereas the lender would have in fact lost 100 percent of the amount loaned. In this example, the lender would still have received substantial revenue, as the lender would have collected fees for each rollover prior to default. The Bureau estimates that during the 2011-2012 time frame, charge-offs (i.e., uncollectible loans defaulted on and never repaid) equaled nearly one-half of the average amount of outstanding loans during the year. In other words, for every $1.00 loaned, only $.50 in principal was eventually repaid.96 One academic study found loss rates to be even higher.97 To sustain these significant costs, the payday lending business model is dependent upon a large volume of re-borrowing—that is, rollovers, back-to-back loans, and re-borrowing within a short period of paying off a previous loan—by those borrowers who do not default on their first loan. The Bureau’s research found that over the course of a year, 90 percent of all loan fees comes from consumers who borrowed seven or more times and 75 percent comes from 96 The Bureau’s staff estimate is based on public company financial statements and confidential information gathered in the course of the Bureau’s statutory functions. Ratio of gross charged off loans to average balances, where gross charge-offs represent single-payment loan losses and average balance is the average of beginning and end of year single-payment loan receivables. 97 Mark Flannery & Katherine Samolyk, “Payday Lending: Do the Costs Justify the Price?,” at 16 (FDIC Ctr. for Fin. Res., Working Paper No. 2005–09, 2005), available at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=771624 (estimating annual charge-offs on storefront payday loans at 66.6 percent of outstanding loans). 41 consumers who borrowed 10 or more times.98 Similarly, when the Bureau identified a cohort of borrowers and tracked them over 10 months, the Bureau found that more than two-thirds of all loans were in sequences of at least seven loans, and that over half of all loans were in sequences of 10 or more loans.99 The Bureau defines a sequence as an initial loan plus one or more subsequent loans renewed within 30 days after repayment of the prior loan; a sequence thus captures not only rollovers and back-to-back loans but also re-borrowing that occurs within a short period of time after repayment of a prior loan either at the point at which a State-mandated cooling-off period ends or at the point at which the consumer, having repaid the prior loan, runs out of money.100 A more detailed discussion of sequence length is provided in the section-bysection discussion of §§ 1041.2(a)(14) and 1041.5 and in Market Concerns—Underwriting. Other studies are broadly consistent. For example, a 2013 report based on lender data from Florida, Kentucky, Oklahoma, and South Carolina found that 85 percent of loans were made to borrowers with seven or more loans per year, and 62 percent of loans were made to borrowers with 12 or more loans per year. These four States have restrictions on payday loans such as cooling-off periods and limits on rollovers that are enforced by State-regulated databases, 98 CFPB Payday Loans and Deposit Advance Products White Paper, at 22. CFPB Report on Supplemental Findings, at 129. 100 The Bureau’s Report on Supplemental Findings analyzed payday loan usage patterns with varying definitions of loan sequence length, including 30-days. CFPB Report on Supplemental Findings, at 109-114. Other reports have proposed other definitions of sequence length including 30 days. See Marc Anthony Fusaro & Patricia J. Cirillo, Do Payday Loans Trap Consumers in a Cycle of Debt?, at 12 (2011), available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1960776&download=yes; (sequences based on the borrower’s pay period); nonPrime 101, “Report 7B: Searching for Harm in Storefront Payday Lending, A Critical Analysis of the CFPB’s ‘Debt Trap’ Data,” at 4 n.9 (2016), available at https://www.nonprime101.com/wpcontent/uploads/2016/02/Report-7-B-Searching-for-Harm-in-Storefront-Payday-Lending-nonPrime101.pdf. See Market Concerns—Underwriting below for an additional discussion of these alternative definitions. 99 42 as well as voluntary extended repayment plans. 101 An updated report on Florida payday loan usage derived from the State database noted this trend has continued, with 83 percent of payday loans in 2015 made to borrowers with seven or more loans and 57 percent of payday loans that same year made to borrowers with 12 or more loans.102 In Alabama’s first year of tracking payday loans with a single database, it reported that almost 50 percent of borrowers had seven or more payday loans and almost 37 percent of borrowers had 10 or more payday loans.103 Other reports have found that over 80 percent of total payday loans and loan volume is due to repeat borrowing within 30 days of a prior loan.104 One trade association has acknowledged that “[i]n any large, mature payday loan portfolio, loans to repeat borrowers generally constitute between 70 and 90 percent of the portfolio, and for some lenders, even more.”105 A recent report by a specialty consumer reporting agency confirms that the industry’s business model relies on repeat customers, noting that over half of all loans are made to returning customers and stating “[t]his finding suggests that even though new customers are critical, existing customers are the most 101 Susanna Montezemolo, “The State of Lending in America & Its Impact on U.S. Households: Payday Lending Abuses and Predatory Practices,” at 12 (Ctr. for Responsible Lending 2013), available at http://www.responsiblelending.org/sites/default/files/uploads/10-payday-loans.pdf. For additional information on Florida loan use, see Veritec Solutions, “State of Florida Case Study: Deferred Presentment Program,” (Implemented 2002), available at http://www.veritecs.com/case-studies/floridas-deferred-presentation-databaseand-program-solution/. 102 Brandon Coleman & Delvin Davis, “Perfect Storm: Payday Lenders Harm Florida Consumer Despite State Law,” at 4 (Ctr. for Responsible Lending, 2016), available at http://www.responsiblelending.org/sites/default/files/nodes/files/researchpublication/crl_perfect_storm_florida_mar2016_0.pdf. 103 Veritec Solutions, “State of Alabama Deferred Presentment Services Program, Report on Alabama Deferred Presentment Loan Activity, October 1, 2015 through September 30, 2016,” available at http://www.banking.alabama.gov/pdf/press%20release/InterimRptStatewideDatabase10_1_15to9_30_16.pdf 104 Leslie Parrish & Uriah King, “Phantom Demand: Short-term Due Date Generates Need for Repeat Payday Loans, Accounting for 76% of Total Volume,” at 11–12 (Ctr. for Responsible Lending, 2009), available at http://www.responsiblelending.org/payday-lending/research-analysis/phantom-demand-final.pdf. 105 Letter from Hilary B. Miller, on behalf of Community Fin. Servs. Ass’n. of America to Bureau of Consumer Fin. Prot. (June 20, 2013), available at http://files.consumerfinance.gov/f/201308_cfpb_cfsa-information-quality-actpetition-to-CFPB.pdf (Petition of Community Financial Services Association of America For Retraction of Payday Loans and Deposit Advance Products: A White Paper of Initial Data Findings, at 5.). 43 productive.”106 Market Concerns—Underwriting below discusses the impact of these outcomes for consumers who are unable to repay and either default or re-borrow. Recent regulatory and related industry developments. A number of Federal and State regulatory developments have occurred over the last 15 years as concerns about the effects of payday lending have spread. Regulators have found that the industry has tended to shift to new models and products in response. Since 2000, it has been clear from commentary added to Regulation Z, that payday loans constitute “credit” under the Truth in Lending Act (TILA) and that cost of credit disclosures are required to be provided in payday loan transactions, regardless of how State law characterizes payday loan fees.107 In 2006, Congress enacted the Military Lending Act (MLA) to address concerns that servicemembers and their families were becoming over-indebted in high-cost forms of credit.108 The MLA, as implemented by the Department of Defense’s regulation, imposes two broad classes of requirements applicable to a creditor. First, the creditor may not impose a military annual percentage rate (MAPR)109 greater than 36 percent in connection with an extension of consumer credit to a covered borrower. Second, when extending consumer credit, the creditor must satisfy certain other terms and conditions, such as providing certain information, both 106 Clarity Services, Inc., “2017 Subprime Lending Trends: Insights into Consumers & the Industry,” at 8 (2017), available at https://www.clarityservices.com/wp-content/uploads/2017/03/Subprime-Lending-Report-2017-ClarityServices-3.28.17.pdf. This finding does not distinguish between storefront and online lenders, nor is it expressly limited to single payment loans. 107 12 CFR 1026.2(a)(14)–2. 108 The Military Lending Act, part of the John Warner National Defense Authorization Act for Fiscal Year 2007, was signed into law in October 2006. The interest rate cap took effect October 1, 2007. See 10 U.S.C. 987. 109 The military annual percentage rate is an “all-in” APR that includes a broader range of fees and charges than the APR that must be disclosed under the Truth in Lending Act. See 32 CFR 232.4. 44 orally and in a form the borrower can keep, before or at the time the borrower becomes obligated on the transaction or establishes the account; refraining from requiring the borrower to submit to arbitration in the case of a dispute involving the consumer credit; and refraining from charging a penalty fee if the borrower prepays all or part of the consumer credit. In 2007, the Department of Defense issued its initial regulation under the MLA, limiting the Act’s application to closed-end loans with a term of 91 days or less in which the amount financed did not exceed $2,000; closedend vehicle title loans with a term of 181 days or less; and closed-end tax refund anticipation loans.110 However, the Department found that evasions developed in the market as “the extremely narrow definition of ‘consumer credit’ in the [then-existing rule] permits a creditor to structure its credit products in order to reduce or avoid altogether the obligations of the MLA.”111 As a result, effective October 2015 the Department of Defense expanded its definition of covered credit to include open-end credit and longer-term loans so that the MLA protections generally apply to all credit subject to the requirements of Regulation Z of the Truth in Lending Act, other than certain products excluded by statute.112 In general, creditors must comply with the new regulations for extensions of credit after October 3, 2016; for credit card accounts, creditors are required to comply with the new rule starting October 3, 2017.113 At the State level, the last States to enact legislation authorizing payday lending—Alaska and Michigan—did so in 2005.114 At least 11 States and jurisdictions that previously had authorized payday loans have taken steps to restrict or eliminate payday lending. In 2001, North 110 72 FR 50580 (Aug. 31, 2007). 80 FR 43560, 43567 n.78 (July 22, 2015). 112 80 FR 43560 (July 22, 2015). 113 80 FR 43560 (July 22, 2015). 114 Alaska Stat. secs. 06.50.010–900; Mich. Comp. Laws secs. 487.2121–.2173. 111 45 Carolina became the first State that had previously permitted payday loans to adopt an effective ban by allowing the authorizing statute to expire. In 2004, Georgia also enacted a law banning payday lending. In 2008, the Ohio legislature adopted the Short Term Lender Act with a 28 percent APR cap, including all fees and charges, for short-term loans and repealed the existing Check-Cashing Lender Law that authorized higher rates and fees.115 In a referendum later that year, Ohioans voted against reinstating the Check-Cashing Lender Law, leaving the 28 percent APR cap and the Short Term Lending Act in effect.116 After the vote, some payday lenders began offering vehicle title loans. Other lenders continued to offer payday loans utilizing Ohio’s Credit Service Organization Act117 and the Mortgage Loan Act;118 the latter practice was upheld by the State Supreme Court in 2014.119 Also in 2008, the District of Columbia banned payday lending which had been a permissible activity under the District’s check cashing law, making the loans subject to the District’s 24 percent per annum maximum interest rate cap.120 In 2010, Colorado’s legislature banned short-term single-payment balloon loans in favor of longer-term, six-month loans. Colorado’s regulatory framework is described in more detail in the discussion of payday installment lending below. As of July 1, 2010, Arizona effectively prohibited payday lending after the authorizing statute expired and a statewide referendum that would have continued to permit payday lending 115 Ohio Rev. Code secs. 1321.35 and 1321.40. See generally Ohio Neighborhood Fin., Inc. v. Scott, 139 Ohio St.3d 536, 13 N.E. 3d 1115 (2014). 117 Ohio Rev. Code sec. 4712.01. 118 Ohio Rev. Code sec. 1321.52(C). 119 Scott, 139 Ohio St.3d 536, 13 N.E. 3d 1115 (2014). 120 Payday Loan Consumer Protection Amendment Act of 2007, D.C. Act 17–42 (2007); D.C. Official Code sec. 28–3301(a) (2011). 116 46 failed to pass.121 However, small-dollar lending activity continues in the State. The State financial regulator issued an alert in 2013, in response to complaints about online unlicensed lending, advising consumers and lenders that payday and consumer loans of $1,000 or less are generally subject to a rate of 36 percent per annum and loans in violation of those rates are void.122 In addition, vehicle title loans continue to be made in Arizona as secondary motor vehicle finance transactions.123 The number of licensed vehicle title lenders has increased by about 300 percent since the payday lending law expired and now exceeds the number of payday lenders that were licensed prior to the ban.124 In 2009, Virginia amended its payday lending law. It extended the minimum loan term to the length of two income periods, added a 45-day cooling-off period after substantial time in debt (the fifth loan in a 180-day period) and a 90-day cooling-off period after completing an extended payment plan, and implemented a database to enforce limits on loan amounts and frequency. The payday law applies to closed-end loans. Virginia has no interest rate regulations or licensure requirements for open-end credit.125 After the amendments, a number of lenders that were 121 Ariz. Rev. Stat. sec. 6–1263; Ariz. Sec’y of State, “State of Arizona Official Canvass,” at 15 (2008), available at http://apps.azsos.gov/election/2008/General/Canvass2008GE.pdf; Ariz. Att’y Gen. Off., “Operation Sunset FAQ,” available at https://www.azag.gov/sites/default/files/sites/all/docs/consumer/op-sunset-FAQ.pdf. 122 Ariz. Dep’t of Fin. Insts., to Consumers, Financial Institutions and Enterprises Conducting Business in Arizona, available at http://www.azdfi.gov/LawsRulesPolicy/Forms/FE-AD-PORegulatory_and_Consumer_Alert_CL_CO_13_01%2002-06-2013.pdf. 123 Ariz. Rev. Stat. sec. 44–281 and 44–291; Arizona Dept. of Fin. Insts., “Frequently Asked Questions from Licensees, Question #6 ‘What is a Title Loan’,” http://www.azdfi.gov/Licensing/Licensing_FAQ.html#MVDSFC (last visited Apr. 20, 2016). 124 These include loans “secured” by borrowers’ registrations of encumbered vehicles. Jean Ann Fox et al., “Wrong Way: Wrecked by Debt, Auto Title Lending in America” at 9 (Consumer Fed’n of America, Ctr. for Econ. Integrity, 2016), available at http://consumerfed.org/wp-content/uploads/2016/01/160126_wrongway_report_cfa-cei.pdf. 125 Va. Code Ann. sec. 6.2–312. 47 previously licensed as payday lenders in Virginia, and that offer closed-end payday loans in other States, switched to offering open-end credit in Virginia without State licenses.126 Washington and Delaware have restricted repeat borrowing by imposing limits on the number of payday loans consumers may obtain. In 2009, Washington made several changes to its payday lending law. These changes, effective January 1, 2010, include a cap of eight loans per borrower from all lenders in a rolling 12-month period where there had been no previous limit on the number of total loans, an extended repayment plan for any loan, and a database to which lenders are required to report all payday loans.127 In 2013, Delaware, a State with no fee restrictions for payday loans, implemented a cap of five payday loans, including rollovers, in any 12-month period.128 Delaware defines payday loans as loans due within 60 days for amounts up to $1,000. Some Delaware lenders have shifted from payday loans to longer-term installment loans with interest-only payments followed by a final balloon payment of the principal and an interest fee payment—sometimes called a “flexpay” loan.129 126 See, e.g., CashNetUSA, “What We Offer” https://www.cashnetusa.com/what-we-offer.html (last visited Sept. 16, 2017) (CashNetUSA is part of Enova); Check Into Cash, “Virginia Line of Credit,” https://checkintocash.com/virginia-line-of-credit/ (last visited Sept. 16, 2017); Allied Cash Advance, “Get the Cash You Need Now” https://www.alliedcash.com/ (last visited Sept. 16, 2017) (“VA: Loans made through open-end credit account.”); First Virginia Loans, “Get Cash Fast” https://www.ccfi.com/firstvirginialoans/ (last visited Sept. 16, 2017) (First Virginia is part of Community Choice, see Community Choice Fin. Inc., 2016 Annual Report (Form 10–K), Exhibit 21.1). See also, Commonwealth of Virginia State Corp. Comm’n, “Payday Lender License Surrenders as of January 1, 2012,” available at https://www.scc.virginia.gov/SCCINTERNET/bfi/reg_inst/sur/pay_sur_0112.pdf (for a list of payday lender license surrenders and dates of surrender). 127 Wash. Dep’t of Fin. Insts., “2010 Payday Lending Report,” at 1–3, available at http://www.dfi.wa.gov/sites/default/files/reports/2010-payday-lending-report.pdf. 128 Del. Code Ann. 5 secs. 2227(7), 2235A(a)(1). 129 See, e.g., James v. National Financial, LLC, 132 A.3d 799, 837 (2016) (holding loan agreement unconscionable and invalid). 48 In 2016, South Dakota voters approved a ballot measure instituting a 36 percent APR limit for all consumer loans made by licensed lenders.130 The measure passed with approximately 75 percent of voters supporting it.131 Subsequently, a number of lenders previously licensed to do business in the State either declined to renew their licenses or indicated that they would not originate new loans that would be subject to the cap.132 New Mexico enacted legislation in 2017 that will effectively prohibit single payment payday loans. It requires small-dollar loans to have minimum loan terms of 120 days and be repaid in four or more installments.133 The legislation will take effect on January 1, 2018.134 The legislation also sets a usury limit of 175 percent APR and will apply to short-term vehicle title loans. In 2017, several other States also passed legislation related to payday lending. Arkansas passed a law clarifying that fees charged by credit service organizations are interest under the State’s constitutional usury limit of 17 percent per annum.135 Utah amended its existing law that prohibits rollovers of payday loans for more than 10 weeks by prohibiting lenders from originating new loans for borrowers to repay prior ones.136 130 Press Release, S.D., Dep’t of Labor and Regulation, “Initiated Measure 21 Approved,” (Nov. 10, 2016), available at http://dlr.sd.gov/news/releases16/nr111016_initiated_measure_21.pdf. 131 S.D., Sec’y of State, “South Dakota Official Election Returns and Registration Figures,” at 39 (2016), available at https://sdsos.gov/elections-voting/assets/ElectionReturns2016_Web.pdf. 132 Dana Ferguson, “Payday Lenders Flee South Dakota After Rate Cap,” Argus Leader (Jan. 6, 2017), http://www.argusleader.com/story/news/politics/2017/01/06/payday-lenders-flee-sd-after-rate-cap/96103624/. 133 N.M. H.B. 347. 134 N.M. H.B. 347. 135 2017 Ark. S.B. 658, Arkansas 91st General Assembly, Title: To Create the Credit Repair Services Organizations Act of 2017, and to Repeal the Credit Services Repair Act of 1987. 136 2017 Utah H.B. 40, Utah 62nd Legislature, 2017 General Session, Title: Check Cashing and Deferred Deposit Lending Amendments Sess. 49 At least 41 Texas municipalities have adopted local ordinances setting business regulations on payday lending (and vehicle title lending).137 Some of the ordinances, such as those in Dallas, El Paso, Houston, and San Antonio, include requirements such as limits on loan amounts (no more than 20 percent of the borrower’s gross annual income for payday loans), limits on the number of rollovers, required amortization of the principal loan amount on repeat loans—usually in 25 percent increments, record retention for at least three years, and a registration requirement.138 On a statewide basis, there are no Texas laws specifically governing payday lenders or payday loan terms; credit access businesses that act as loan arrangers or broker payday loans (and vehicle title loans) are regulated and subject to licensing, reporting, and requirements to provide consumers with disclosures about repayment and re-borrowing rates.139 Online Payday Lending With the growth of the Internet, a significant online payday lending industry has developed. Some storefront lenders use the Internet as an additional method of originating payday loans in the States in which they are licensed to do business. In addition, there are now a 137 A description of the municipalities is available at Texas Municipal League. An additional 16 Texas municipalities have adopted land use ordinances on payday or vehicle title lending. Texas Municipal League, “City Regulation of Payday and Auto Title Lenders,” http://www.tml.org/payday-updates (last visited April 26, 2017). 138 Other municipalities have adopted similar ordinances. For example, at least seven Oregon municipalities, including Portland and Eugene, have enacted ordinances that include a 25 percent amortization requirement on rollovers and a requirement that lenders offer a no-cost payment plan after two rollovers. See Portland, Or., Code sec. 7.26.050; Eugene Or., Code sec. 3.556. 139 CABs must include a pictorial disclosure with the percentage of borrowers who will repay the loan on the due date and the percentage who will roll over (called renewals) various times. See Texas Off. of Consumer Credit Commissioner, “Credit Access Businesses” http://occc.texas.gov/industry/cab (last visited Sept. 16, 2017). The CABs, rather than the lenders, maintain storefront locations, and qualify borrowers, service and collect the loans for the lenders. CABs may also guaranty the loans. There is no cap on CAB fees and when these fees are included in the loan finance charges, the disclosed APRs for Texas payday and vehicle title loans are similar to those in other States with deregulated rates. See Ann Baddour, “Why Texas’ Small Dollar Lending Market Matter,” (Fed. Reserve Bank of Dallas, e-Perspectives Issue 2, 2012), available at https://www.fedinprint.org/items/feddep/y2012n2x1.html. In 2004, a Federal appellate court dismissed a putative class action related to these practices. Lovick v. RiteMoney, Ltd., 378 F.3d 433 (5th Cir. 2004). 50 number of lenders offering what are referred to as “hybrid” payday loans, through the Internet. Hybrid payday loans are structured so that rollovers occur automatically unless the consumer takes affirmative action to pay off the loan, thus effectively creating a series of interest-only payments followed by a final balloon payment of the principal amount and an additional fee.140 Hybrid loans structured as single payment loans with automatic rollovers141 and longer-term loans with a final balloon payment142 are covered by the final rule’s Ability-to-Repay requirements as discussed more fully below. Industry size, structure, and products. The size of the online payday market is difficult to measure for a number of reasons. First, many online lenders offer a variety of products beyond single-payment loans (what the Bureau refers to as payday loans) and hybrid loans (which the Bureau views as a form of payday lending and falls within the final rule’s definition of short- 140 nonPrime101, “Report 1: Profiling Internet Small Dollar Lending—Basic Demographics and Loan Characteristics,” at 2–3, (2014), available at https://www.nonprime101.com/wp-content/uploads/2015/02/ProfilingInternet-Small-Dollar-Lending-Final.pdf. The report refers to these automatic rollovers as “renewals.” 141 Examples of hybrid payday loans requiring borrower affirmative action to opt out of automatic rollovers are described in recent litigation by the Bureau and the Federal Trade Commission. Loans by Integrity Advance contained default terms that caused loans to automatically roll over four times with charges added at each rollover before any payments were applied to the principal. See Press Release, Bureau of Consumer Fin. Prot., “CFPB Takes Action Against Online Lender for Deceiving Borrowers,” (Nov.18, 2015), available at https://www.consumerfinance.gov/about-us/newsroom/cfpb-takes-action-against-online-lender-for-deceivingborrowers/. Similarly, OneClickCash was an online lender that offered loans with a TILA disclosure as a single repayment loan, but unless borrowers satisfied certain pre-conditions they were automatically enrolled in a 10 payperiod renewal plan with new finance charges accruing each pay period and no payments applied to the principal balance until the fifth payment. See Order, Fed. Trade Comm’n v. AMG Services, Inc., No. 12–00536 (D. Nev. Mar. 07, 2014), ECF No. 559, available at https://www.ftc.gov/system/files/documents/cases/140319amgorder.pdf. See also, Sierra Lending, “FAQ, How do I repay?,” https://www.sierralending.com/Home/FAQ (last visited July 20, 2017) (consumer must call online payday lender at least three business days prior to due date or lender will automatically withdraw only the finance charge and loan will roll over). 142 The Bureau is aware of a number of examples of storefront and online longer-term loans with final balloon payments. For instance, a loan agreement for a $200 loan from National Financial LLC d/b/a Loan Till Payday LLC required the borrower to pay 26 bi-weekly payments of $60 with a final balloon payment of $260. See, James v. National Financial, LLC, 132 A.3d 799, 837 (2016) (holding loan agreement unconscionable and invalid). Additionally the Bureau is aware of a Texas loan for $365.60, arranged through a credit access business, to be repaid in five payments of $108 with a sixth, final payment of $673.70. 51 term loans), including longer-term installment loans; this poses challenges in sizing the portion of these firms’ business that is attributable to payday and hybrid loans. Second, most online payday lenders are not publicly traded, which means that minimal financial information is available about this market segment. Third, many online payday lenders claim exemption from State lending laws and licensing requirements on the basis that they are located and operated from other jurisdictions. Consequently, these lenders report less information publicly, whether individually or in aggregate compilations, than lenders holding traditional State licenses. Finally, storefront payday lenders who are also using the online channel generally do not separately report their online originations. Bureau staff’s reviews of the largest storefront lenders’ websites indicate an increased focus in recent years on online loan origination. With these caveats, a frequently cited industry analyst has estimated that by 2012, online payday loans had grown to generate nearly an equivalent amount of fee revenue as storefront payday loans on roughly 62 percent of the origination volume, about $19 billion, but originations had then declined somewhat to roughly $15.9 billion by 2015.143 This trend appears consistent with storefront payday loans, as discussed above, and is likely related at least in part to increasing lender migration from short-term into longer-term products. Online payday loan fee revenue has been estimated at $3.1 billion for 2015, or 19 percent of origination volume.144 However, these estimates may be both over- and under-inclusive; they may not differentiate 143 John Hecht, “The State of Short-Term Credit Amid Ambiguity, Evolution and Innovation” (2016) (Jefferies LLC, slide presentation) (on file); John Hecht, “The State of Short-Term Credit in a Constantly Changing Environment” (2015) (Jeffries LLC, slide presentation) (on file); Jessica Silver-Greenberg, “Major Banks Aid in Payday Loans Banned by States,” N.Y. Times, Feb. 23, 2013, http://www.nytimes.com/2013/02/24/business/major-banks-aid-inpayday-loans-banned-by-states.html. 144 John Hecht, “The State of Short-Term Credit Amid Ambiguity, Evolution and Innovation” (2016) (Jefferies LLC, slide presentation) (on file). 52 precisely between online lenders’ short-term and longer-term loans, and they may not account for the online lending activities by storefront payday lenders. Whatever the precise size, the online industry can broadly be divided into two segments: online lenders licensed in the State in which the borrower resides and lenders that are not licensed in the borrower’s State of residence. The first segment consists largely of storefront lenders with an online channel to complement their storefronts as a means of originating loans, as well as a few online-only payday lenders who lend to borrowers in States where they have obtained State lending licenses. Because this segment of online lenders is State-licensed, State administrative payday lending reports include these data but generally do not differentiate loans originated online from those originated in storefronts. Accordingly, this portion of the market is included in the market estimates summarized above, and the lenders consider themselves to be subject to, or generally follow, the relevant State laws discussed above. The second segment consists of lenders that claim exemption from State lending laws. Some of these lenders claim exemption because their loans are made from physical locations outside of the borrower’s State of residence, including from off-shore locations outside of the United States.145 Other lenders claim exemption because they are lending from Tribal lands, with such lenders claiming that they are regulated by the sovereign laws of “federally recognized 145 For example, in 2015 the Bureau filed a lawsuit in Federal district court against NDG Enterprise, NDG Financial Corp., Northway Broker, Ltd., and others alleging that defendants illegally collected online payday loans that were void or that consumers had no obligations to repay, and falsely threatened consumers with lawsuits and imprisonment. Several defendants are Canadian corporations and others are incorporated in Malta. The case is pending. See Press Release, Bureau of Consumer Fin. Prot., “CFPB Sues Offshore Payday Lender” (Aug. 4, 2015), available at http://www.consumerfinance.gov/newsroom/cfpb-sues-offshore-payday-lender/. 53 Indian tribes.”146 These lenders claim immunity from suit to enforce State or Federal consumer protection laws on the basis of their sovereign status.147 A Federal appellate court recently rejected claims of immunity from the Bureau’s civil investigative demands by several Tribalrelated lenders, finding that “Congress did not expressly exclude tribes from the Bureau's enforcement authority.”148 A frequently cited source of data on this segment of the market is a series of reports using data from a specialty consumer reporting agency serving certain online lenders, most of whom are unlicensed.149 These data are not representative of the entire online industry, but nonetheless cover a large enough sample (2.5 million borrowers over a period of four years) to be significant. These reports indicate the following concerning this market segment:  Although the mean and median loan size among the payday borrowers in this dataset are only slightly higher than the information reported above for storefront 146 12 U.S.C. 5481(27). According to a tribal trade association representative, about 30 tribes are involved in the payday lending industry. Julia Harte & Joanna Zuckerman Bernstein, “Payday Nation, When Tribes Team Up with Payday Lenders, Who Profits?,” AlJazeera America, June 17, 2014, http://projects.aljazeera.com/2014/paydaynation/. The Bureau is unaware of other public sources for an estimate of the number of tribal lenders. 147 See First Amended Complaint, Consumer Fin. Prot. Bureau v. CashCall, Inc., No. 13–13167 (D. Mass. Mar. 21, 2014), ECF No. 27, available at http://files.consumerfinance.gov/f/201403_cfpb_amended-complaint_cashcall.pdf; Complaint for Permanent Injunction and Other Relief, Consumer Fin. Prot. Bureau v. Golden Valley Lending Inc., No. 17–3155 (N.D. Ill. Apr. 27, 2017), ECF No. 1, available at http://files.consumerfinance.gov/f/documents/201704_cfpb_Golden-Valley_Silver-Cloud_MajesticLake_complaint.pdf; Order, Fed. Trade Comm’n v. AMG Services, Inc., No. 12–00536 (D. Nev. Mar. 07, 2014), ECF No. 559, available at https://www.ftc.gov/system/files/documents/cases/140319amgorder.pdf; State ex rel. Suthers v. Cash Advance & Preferred Cash Loans, 205 P.3d 389 (Colo. App. 2008), aff’d sub nom; Cash Advance & Preferred Cash Loans v. State, 242 P.3d 1099 (Colo. 2010); California v. Miami Nation Enterprises, 166 Cal.Rptr.3d 800 (2014). 148 CFPB v. Great Plains Lending, LLC, 846 F.3d 1049, 1058 (9th Cir. 2017), reh’g denied (Apr. 5, 2017). 149 nonPrime101, “Report 1: Profiling Internet Small Dollar Lending—Basic Demographics and Loan Characteristics,” at 9, (2014), available at https://www.nonprime101.com/wp-content/uploads/2015/02/ProfilingInternet-Small-Dollar-Lending-Final.pdf. 54 payday loans,150 the online payday lenders charge higher rates than storefront lenders. As noted above, most of the online lenders reporting this data claim exemption from State laws and do not comply with State rate caps. The median loan fee in this dataset is $23.53 per $100 borrowed, compared to $15 per $100 borrowed for storefront payday loans. The mean fee amount is even higher at $26.60 per $100 borrowed.151 Another study based on a similar dataset from three online payday lenders is generally consistent, putting the range of online payday loan fees at between $18 and $25 per $100 borrowed.152  More than half of the payday loans made by these online lenders are hybrid payday loans. As described above, a hybrid loan involves automatic rollovers with payment of the loan fee until a final balloon payment of the principal and fee.153 For the hybrid payday loans, the most frequently reported payment amount 150 The median online payday loan size is $400, compared to a median loan size of $350 for storefront payday loans. nonPrime101, “Report 1: Profiling Internet Small Dollar Lending—Basic Demographics and Loan Characteristics,” at 10, (2014), available at https://www.nonprime101.com/wp-content/uploads/2015/02/Profiling-Internet-SmallDollar-Lending-Final.pdf. 151 nonPrime101, “Report 1: Profiling Internet Small Dollar Lending—Basic Demographics and Loan Characteristics,” at 10, (2014), available at https://www.nonprime101.com/wp-content/uploads/2015/02/ProfilingInternet-Small-Dollar-Lending-Final.pdf. 152 G. Michael Flores, “The State of Online Short-Term Lending, Second Annual Statistical Analysis Report” Bretton-Woods, Inc., at 15 (Feb. 28, 2014), available at http://onlinelendersalliance.org/wpcontent/uploads/2015/07/2015-Bretton-Woods-Online-Lending-Study-FINAL.pdf. 153 nonPrime101, “Report 5 - Loan Product Structures and Pricing in Internet Installment Lending, Similarities to and Differences from Payday Lending and Implications for CFPB Rulemaking,” at 4 (May 15, 2015), available at https://www.nonprime101.com/wp-content/uploads/2015/05/Report-5-Loan-Product-Structures-1.3-5.21.15Final3.pdf. As noted above, these loans may also be called flexpay loans. Such loans would likely be covered longer-term loans under this rule. 55 is 30 percent of principal, implying a finance charge during each pay period of $30 for each $100 borrowed.154  Unlike storefront payday loan borrowers who generally return to the same store to re-borrow, the credit reporting data may suggest that online borrowers tend to move from lender to lender. As discussed further below, however, it is difficult to evaluate whether some of this apparent effect is due to online lenders simply not consistently reporting lending activity.155 Marketing, underwriting, and collection practices. As with most online lenders in other markets, online payday lenders have utilized direct marketing, lead generators, and other forms of advertising for customer acquisition. Lead generators, via websites advertising payday loans usually in the form of banner advertisements or paid search results (the advertisements that appear at the top of an Internet search on Google, Bing, or other search engines) operated by “publishers,” collect consumers’ personal and financial information and electronically offer it to lenders that have expressed interest in consumers meeting certain criteria.156 In July 2016, Google banned ads for loans with APRs over 36 percent or with repayment due in 60 days or 154 nonPrime101, “Report 5 - Loan Product Structures and Pricing in Internet Installment Lending, Similarities to and Differences from Payday Lending and Implications for CFPB Rulemaking,” at 6 (May 15, 2015), available at https://www.nonprime101.com/wp-content/uploads/2015/05/Report-5-Loan-Product-Structures-1.3-5.21.15Final3.pdf. 155 See generally nonPrime101, “Report 7–A - How Persistent is the Borrower-Lender Relationship in Payday Lending,” (Sept. 10, 2015), available at https://www.nonprime101.com/wp-content/uploads/2015/10/Report-7AHow-Persistent-Is-the-Borrow-Lender-Relationship_1023151.pdf. 156 For more information about the use of lead generators in the payday market, see Fed. Trade Comm’n, “Follow the Lead Workshop: Staff Perspective” (Sept. 2016), available at https://www.ftc.gov/system/files/documents/reports/staff-perspective-followlead/staff_perspective_follow_the_lead_workshop.pdf. 56 less.157 From the Bureau’s market outreach activities it is aware that the payday lending industry’s use of lead generators has decreased but that payday lenders may be using other forms of advertising for customer acquisition and retention. Online lenders view fraud (i.e., consumers who mispresent their identity) as a significant risk and also express concerns about “bad faith” borrowing (i.e., consumers with verified identities who borrow without the intent to repay).158 Consequently, online payday and hybrid payday lenders attempt to verify the borrower’s identity and the existence of a bank account in good standing. Several specialty consumer reporting agencies have evolved primarily to serve the online payday lending market. The Bureau is aware from market outreach that online lenders also generally report loan closure information on a real-time or daily basis to the specialty consumer reporting agencies. In addition, some online lenders report to the Bureau that they use nationwide credit report information to evaluate both credit and potential fraud risk associated with first-time borrowers, including recent bankruptcy filings. However, there is evidence that online lenders do not consistently utilize credit report data for every loan, and instead typically 157 Google announced that it was "banning payday loans and some related products from our ads systems," in an attempt to "protect our users from deceptive or harmful financial products." The changes to Google's advertising service, AdWords, went into effect on July 13, 2016, and on its face apply to lenders, lead generators, and others. In the six months following the new policy's introduction, Google reported removing five million payday loan ads from its services. However, some observers have questioned the effectiveness of Google's policy. See David Graff, “An Update to Our AdWords Policy on Lending Products,” Google The Keyword Blog (May 11, 2016), https://blog.google/topics/public-policy/an-update-to-our-adwords-policy-on/; Scott Spencer, “How We Fought Bad Ads, Sites and Scammers in 2016,” Google The Keyword Blog (Jan. 25, 2017), https://blog.google/topics/ads/howwe-fought-bad-ads-sites-and-scammers-2016/; David Dayen, “Google Said It Would Ban All Payday Loan Ads. It Didn’t” The Intercept, Oct. 7, 2016, https://theintercept.com/2016/10/07/google-said-it-would-ban-all-payday-loanads-it-didnt/. 158 For example, Enova states that it uses its own analysis of previous fraud incidences and third party data to determine if applicant information submitted matches other indicators and whether the applicant can authorize transactions from the submitted bank account. In addition, it uses proprietary models to predict fraud. Enova Int’l Inc., 2016 Annual Report (Form 10-K), at 8–9. 57 check and report data only for new borrowers or those returning after an extended absence from the lender’s records.159 Typically, proceeds from online payday loans are disbursed electronically into the consumer’s bank account and the consumer authorizes the lender to electronically debit her account to repay the loan as payments are due. The Bureau is aware from market monitoring that lenders employ various practices to encourage consumers to agree to authorize electronic debits for repayment. Some lenders generally will not disburse electronically if consumers do not agree to ACH repayment, but instead will require the consumer to wait for a paper loan proceeds check to arrive in the mail. 160 Some online payday lenders charge higher interest rates or fees to consumers who do not commit to electronic debits. 161 In addition, some online payday lenders have adopted policies that may delay the crediting of non-ACH payments.162 159 Based on the Bureau’s market outreach with lenders and specialty consumer reporting agencies. See, e.g., Mobiloans, “Line of Credit Terms and Conditions,” www.mobiloans.com/terms-and-conditions (last visited Feb. 5, 2016) (“If you do not authorize electronic payments from your Demand Deposit Account and instead elect to make payments by mail, you will receive your Mobiloans Cash by check in the mail.”). 161 One online payday lender’s website FAQs states: “Q: Am I only able to pay through ACH? A: Paying your cash advance via an electronic funds transfer (EFT) or ACH is certainly the easiest, most efficient, and least expensive method. However, should the need for an alternative payment method arises [sic], we will be happy to discuss that with you.” National Payday, “Frequently Asked Questions,” https://www.nationalpayday.com/faq/ (last visited July 20, 2017). LendUp’s website states there may be a fee to make a MoneyGram payment. LendUp, “Frequently Asked Questions, Paying back your LendUp Loan,” https://www.lendup.com/faq#paying-loan (last visited July 20, 2017). Under the Electronic Fund Transfer Act (EFTA) and its implementing regulation (Regulation E), lenders cannot condition the granting of credit on a consumer’s repayment by preauthorized (recurring) electronic fund transfers, except for credit extended under an overdraft credit plan or extended to maintain a specified minimum balance in the consumer’s account. 12 CFR 1005.10(e). The summary in the text of current lender practices is intended to be purely descriptive. The Bureau is not addressing in this rulemaking the question of whether any of the practices described in text are consistent with EFTA. 162 LendUp’s website states payment by Moneygram or check may involve “processing times” of “1–2 business days” to apply the payment. LendUp, “Frequently Asked Questions, Paying back your LendUp Loan,” https://www.lendup.com/faq#paying-loan (last visited July 20, 2017). LendUp offers both single payment and installment loans, depending on the borrower’s State. 160 58 As noted above, online lenders typically collect payday loans via electronic debits. For a hybrid payday loan the lender seeks to collect the finance charges a pre-set number of times and then eventually collect the principal; for a true payday loan the lender will seek to collect the principal and finance charges when the loan is due. Online payday lenders, like their storefront counterparts, use various models and software, described above, to predict when an electronic debit is most likely to succeed in withdrawing funds from a borrower’s bank account. As discussed further below, the Bureau has observed lenders seeking to collect multiple payments on the same day. This pattern may be driven by a practice of dividing the payment amount in half and presenting two debits at once, presumably to reduce the risk of a larger payment being returned for nonsufficient funds. Indeed, the Bureau found that about one-third of presentments by online payday lenders occur on the same day as another request by the same lender from the same account. The Bureau also found that split presentments almost always result in either payment of all presentments or return of all presentments (in which event the consumer will likely incur multiple nonsufficient funds (NSF) fees from the bank). The Bureau’s study indicates that when an online payday lender’s first attempt to obtain a payment from the consumer’s account is unsuccessful, it will make a second attempt 75 percent of the time and if that attempt fails the lender will make a third attempt 66 percent of the time.163 As discussed further at part II.D, the success rate on these subsequent attempts is relatively low, and the cost to consumers may be correspondingly high.164 163 See generally CFPB Online Payday Loan Payments, at 14. Because these online lenders may offer single-payment payday, hybrid, and installment loans, reviewing the debits does not necessarily distinguish the type of loan involved. Storefront payday lenders were not included. See CFPB Online Payday Loan Payments, at 7, 13. 164 59 There is limited information on the extent to which online payday lenders that are unable to collect payments through electronic debits resort to other collection tactics.165 The available evidence indicates, however, that online lenders sustain higher credit losses and risk of fraud than storefront lenders. One lender with publicly available financial information that originated both storefront and online single-payment loans reported in 2014, a 49 percent and 71 percent charge-off rate, respectively, for these loans.166 Online lenders generally classify as “fraud” both consumers who misrepresented their identity in order to obtain a loan and consumers whose identity is verified but default on the first payment due, which is viewed as reflecting the intent not to repay. Business model. While online lenders tend to have fewer costs relating to operation of physical facilities than do storefront lenders, as discussed above, they face higher costs relating to lead acquisition and marketing, loan origination screening to verify applicant identity, and potentially larger losses due to what they classify as “fraud” than their storefront competitors. Accordingly, it is not surprising that online lenders—like their storefront counterparts— are dependent upon repeated re-borrowing. Indeed, even at a cost of $25 or $30 per $100 borrowed, a typical single online payday loan would generate fee revenue of under $100, which 165 One publicly traded online-only lender that makes single-payment payday loans as well as online installment loans and lines of credit reports that its call center contacts borrowers by phone, email, and in writing after a missed payment and periodically thereafter and that it also may sell uncollectible charged off debt. Enova Int’l Inc., 2016 Annual Report (Form, 10–K), at 9 (Feb. 24, 2017). 166 Net charge-offs over average balance based on data from Cash America and Enova Forms 10–K. See Cash America Int’l, Inc., 2014 Annual Report (Form 10–K), at 102 (Mar. 13, 2015); Enova Int’l Inc., 2014 Annual Report (Form 10–K), at 95 (Mar. 20, 2015). Net charge-offs represent single-payment loan losses less recoveries for the year. Averages balance is the average of beginning and end of year single-payment loan receivables. Prior to November 14, 2014, Enova comprised the e-commerce division of Cash America. Using the 2014 Forms 10–K allows for a better comparison of payday loan activity, than the 2015 Forms 10–K, as Cash America’s payday loan operations declined substantially after 2014. 60 is not sufficient to cover the typical origination costs. Consequently, as discussed above, hybrid loans that roll over automatically in the absence of affirmative action by the consumer account for a substantial percentage of online payday business. These products, while nominally structured as single-payment products, effectively build a number of rollovers into the loan. For example, the Bureau has observed online payday lenders whose loan documents suggest that they are offering a single-payment loan but whose business model is to collect only the finance charges due, roll over the principal, and require consumers to take affirmative steps to notify the lender if consumers want to repay their loans in full rather than allowing them to roll over. The Bureau recently initiated an action against an online lender alleging that it engaged in deceptive practices in connection with such products.167 In a recent survey conducted of online payday borrowers, 31 percent reported that they had experienced loans with automatic renewals.168 As discussed above, a number of online payday lenders claim exemption from State laws and the regulations and limitations established under those laws. As reported by a specialty consumer reporting agency with data from that market, more than half of the payday loans for which information is furnished to it are hybrid payday loans with the most common fee being $30 per $100 borrowed, twice the median amount for storefront payday loans.169 167 Press Release, Bureau of Consumer Fin. Prot., “CFPB Takes Action Against Online Lender for Deceiving Borrowers” (Nov. 18, 2015), available at http://www.consumerfinance.gov/newsroom/cfpb-takes-action-againstonline-lender-for-deceiving-borrowers/. The FTC raised and resolved similar claims against online payday lenders. See Press Release, FTC, FTC Secures $4.4 Million From Online Payday Lenders to Settle Deception Charges (Jan. 5, 2016), available at https://www.ftc.gov/news-events/press-releases/2016/01/ftc-secures-44-million-online-paydaylenders-settle-deception. 168 Pew Charitable Trusts, “Payday Lending in America Fraud and Abuse Online: Harmful Practices in Internet Payday Lending, at 8 (Report 4, 2014), available at http://www.pewtrusts.org/~/media/Assets/2014/10/PaydayLending-Report/Fraud_and_Abuse_Online_Harmful_Practices_in_Internet_Payday_Lending.pdf. 169 nonPrime101, “Report 5 - Loan Product Structures and Pricing in Internet Installment Lending, Similarities to and Differences from Payday Lending and Implications for CFPB Rulemaking,” at 4, 6 (May 15, 2015), available at 61 Similar to associations representing storefront lenders as discussed above, a national trade association representing online lenders includes loan repayment plans as one of its best practices, but does not provide many details in its public material.170 A trade association that represents Tribal online lenders has adopted a set of best practices, but the list does not address repayment plans.171 Vehicle Title Loans, Including Short-Term Loans and Balloon-Payment Products Vehicle title loans—also known as “automobile equity loans”—are another form of liquidity lending permitted in certain States. In a title loan transaction, the borrower must provide identification and usually the title to the vehicle as evidence that the borrower owns the vehicle “free and clear.”172 Unlike payday loans, there is generally no requirement that the borrowers have a bank account, and some lenders do not require a copy of a pay stub or other evidence of income.173 Rather than holding a check or ACH authorization for repayment as with a payday loan, the lender generally retains the vehicle title or some other form of security interest https://www.nonprime101.com/wp-content/uploads/2015/05/Report-5-Loan-Product-Structures-1.3-5.21.15Final3.pdf; CFPB Payday Loans and Deposit Advance Products White Paper, at 16. 170 Online Lenders Alliance, “Best Practices,” at 29 (May 2017), available at http://onlinelendersalliance.org/wpcontent/uploads/2015/01/Best-Practices-2017.pdf. The materials state that its members “shall comply” with any required State repayment plans; otherwise, if a borrower is unable to repay a loan according to the loan agreement, the trade association’s members “should create” repayment plans that “provide flexibility based on the customer’s circumstances.” 171 Native American Fin. Servs. Ass’n, “Best Practices,” http://www.mynafsa.org/best-practices/ (last visited Apr. 20, 2016). 172 Arizona also allows vehicle title loans to be made against as secondary motor vehicle finance transactions. Ariz. Rev. Stat. sec. 44–281, 44–291G; Arizona Dep.t of Fin. Inst., “Frequently Asked Questions from Licensees, Question #6 ‘What is a Title Loan.’” 173 See Fast Cash Title Loans, “FAQ,” http://fastcashvirginia.com/faq/ (last visited Mar. 3, 2016) (“There is no need to have a checking account to get a title loan.”); Title Max, “How Title Loans Work,” https://www.titlemax.com/how-it-works/ (last visited Jan. 15, 2016) (borrowers need a vehicle title and government issued identification plus any additional requirements of State law). 62 that provides it with the right to repossess the vehicle, which may then be sold with the proceeds used for repayment.174 The lender retains the vehicle title or some other form of security interest during the duration of the loan, while the borrower retains physical possession of the vehicle. In some States either the lender files a lien with State officials to record and perfect its interest in the vehicle or charges a fee for non-filing insurance. In a few States, a clear vehicle title is not required, and vehicle title loans may be made as secondary liens against the title or against the borrower’s automobile registration.175 In some States, such as Georgia, vehicle title loans are made under pawnbroker statutes that specifically permit borrowers to pawn vehicle certificates of title.176 Almost all vehicle title lending is conducted at storefront locations, although some title lending does occur online.177 174 See Speedy Cash, “Title Loans FAQs,” https://www.speedycash.com/faqs/title-loans/ (last visited Mar. 29, 2016) (title loans are helpful “when you do not have a checking account to secure your loan….your car serves as collateral for your loan.”). 175 See, for example, the discussion above about Arizona law applicable to vehicle title lending. 176 Ga. Code sec. 44–12–131 (2015). 177 See, e.g., the Bureau’s action involving Wilshire Consumer Credit for illegal collection practices. Consumers primarily applied for Wilshire’s vehicle title loans online. Press Release, Bureau of Consumer Fin. Prot., “CFPB Orders Indirect Auto Finance Company to Provide Consumers $44.1 Million in Relief for Illegal Debt Collection Tactics” (Oct. 1, 2015), available at http://www.consumerfinance.gov/newsroom/cfpb-orders-indirect-auto-financecompany-to-provide-consumers-44-1-million-in-relief-for-illegal-debt-collection-tactics/. See also State actions against Liquidation, LLC d/b/a Sovereign Lending Solutions, LLC and other names, purportedly organized in the Cook Islands, New Zealand. Press Release, Oregon Dep’t of Justice, “AG Rosenblum and DCBS Sue Predatory Title Loan Operator” (Aug. 18, 2015), available at http://www.doj.state.or.us/releases/Pages/2015/rel081815.aspx; Press Release, Michigan Attorney General, “Schuette Stops Collections by High Interest Auto Title Loan Company” (Jan. 26, 2016), available at http://www.michigan.gov/ag/0,4534,7-164-46849-374883--,00.html; Press Release, Pennsylvania Dep’t of Banking and Securities, “Consumers Advised about Illegal Auto Title Loans Following Court Decision” (Feb. 3, 2016), available at http://www.media.pa.gov/pages/banking_details.aspx?newsid=89; Press Release, North Carolina Dep’t of Justice, “Online Car Title Lender Banned from NC for Unlawful Loans, AG Says” (May 2, 2016), available at http://www.ncdoj.gov/Home/SearchResults.aspx?searchtext=Ace%20payday&searchmode=AnyWord&searchscope=SearchCurrentSection&page=82; Final Order: Director’s Consideration, Washington Dep’t of Financial Institutions, Division of Consumer Services v. Auto Loans, LLC a/k/a Car Loan, LLC a/k/a Liquidation, LLC a/k/a Vehicle Liquidation, LLC a/k/a Sovereign Lending Solutions a/k/a Title Loan America, and William McKibbin, Principal, (Apr. 22, 2016), available at http://dfi.wa.gov/sites/default/files/consumer-services/enforcement-actions/C-15-1804-16-FO02.pdf; Press Release, 63 Product definition and regulatory environment. There are three types of vehicle title loans: single-payment loans, installment loans, and in at least one State, balloon payment loans.178 Of the 24 States that permit some form of vehicle title lending, six States permit only single-payment title loans, 13 States allow the loans to be structured as single-payment or installment loans, and five permit only title installment loans.179 (The payment practices of installment title loans are discussed briefly below.) All but three of the States that permit some form of title lending (Arizona, Georgia, and New Hampshire) also permit payday lending. Single-payment vehicle title loans are typically due in 30 days and operate much like payday loans: the consumer is charged a fixed price per $100 borrowed, and when the loan is due the consumer is obligated to repay the full amount of the loan plus the fee but is typically given the opportunity to roll over or re-borrow.180 The Bureau recently studied anonymized data Colo. Dep’t of Law, “AG Coffman Announces Significant Relief for Victims of Illegal Auto Title Loan Scheme” (Nov. 30, 2016), available at https://coag.gov/sites/default/files/contentuploads/cp/ConsumerCreditUnit/PressReleases/UCCC/rsfinancialsoverei gnlending11.30.16.pdf; Press Release, Att’y Gen. of Mass., “AG Obtains Judgment Voiding Hundreds of Illegal Loans to Massachusetts Consumers in Case Against Online Auto Title Lender” (May 25, 2017), available at http://www.mass.gov/ago/news-and-updates/press-releases/2017/2017-05-25-voiding-hundreds-of-illegalloans.html. Consumers applied for the title loans online and sent their vehicle titles to the lender. The lender used local agents for repossession services. 178 The Bureau is aware of Texas vehicle title installment loans structured as longer-term balloon payment loans. One vehicle title loan for $433, arranged through a credit access business, was to be repaid in five payments of $64.91 and a final balloon payment of $519.15. Similarly, another vehicle title loan arranged through a credit access business for $2,471.03 was scheduled to be repaid in five payments for $514.80 with a final balloon payment of $2,985.83. 179 Pew Charitable Trusts, “Auto Title Loans: Market Practices and Borrowers’ Experiences,” at 4 (2015), available at http://www.pewtrusts.org/~/media/Assets/2015/03/AutoTitleLoansReport.pdf?la=en. The report lists 25 States but post-publication, as noted above, South Dakota effectively prohibited vehicle title lending in November 2016 by adopting a 36 percent APR rate cap. And, as of January 1, 2018, New Mexico vehicle title loans will be required to have a 120-day minimum loan term. 180 See Pew Charitable Trusts, “Auto Title Loans: Market Practices and Borrowers’ Experiences,” (2015), available at http://www.pewtrusts.org/~/media/Assets/2015/03/AutoTitleLoansReport.pdf?la=en; see also Idaho Dep’t of Fin., “Idaho Credit Code ‘Fast Facts’”, available at http://www.finance.idaho.gov/ConsumerFinance/Documents/IdahoCredit-Code-Fast-Facts-With-Fiscal-Annual-Report-Data-01012016.pdf; Letter from Greg Gonzales, Comm’r, Tenn. Dep’t of Fin. Insts., to Hon. Bill Haslam, Governor and Hon. Members of the 109th General Assembly, at 4 64 from vehicle title lenders consisting of nearly 3.5 million loans made to over 400,000 borrowers in 20 States. For single-payment vehicle title loans with a typical duration of 30 days, the median loan amount was $694 with a median APR of 317 percent; the average loan amount was $959 and the average APR was 291 percent.181 Two other studies contain similar findings.182 Vehicle title loans are therefore for substantially larger amounts than typical payday loans, but carry similar APRs for similar terms. Some States that authorize vehicle title loans limit the rates lenders may charge to a percentage or dollar amount per $100 borrowed, similar to some State payday lending pricing structures. A common fee limit is 25 percent of the loan amount per month, but roughly half of the authorizing States have no restrictions on rates or fees.183 Some, but not all, States limit the maximum amount that may be borrowed to a fixed dollar amount, a percentage of the borrower’s monthly income (50 percent of the borrower’s gross monthly income in Illinois), or a percentage of the vehicle’s value.184 Some States limit the (Apr. 12, 2016) (Report on the Title Pledge Industry), available at http://www.tennessee.gov/assets/entities/tdfi/attachments/Title_Pledge_Report_2016_Final_Draft_Apr_6_2016.pdf. 181 CFPB Single-Payment Vehicle Title Lending, at 7. 182 Pew Charitable Trusts, “Auto Title Loans: Market Practices and Borrowers’ Experiences,” (2015), available at http://www.pewtrusts.org/~/media/Assets/2015/03/AutoTitleLoansReport.pdf?la=en; Susanna Montezemolo, “The State of Lending in America & Its Impact on U.S. Households: Payday Lending Abuses and Predatory Practices,” at 3 (Ctr. for Responsible Lending 2013), available at http://www.responsiblelending.org/sites/default/files/uploads/10-payday-loans.pdf. 183 States with a 15 percent to 25 percent per month cap include Alabama, Georgia (rate decreases after 90 days), Mississippi, and New Hampshire; Tennessee limits interest rates to 2 percent per month, but also allows for a fee up to 20 percent of the original principal amount. Virginia’s fees are tiered at 22 percent per month for amounts up to $700 and then decrease on larger loans. Ala. Code sec. 5–19A–7(a); Ga. Code Ann. sec. 44–12–131(a)(4); Miss. Code Ann. sec. 75–67–413(1); N.H. Rev. Stat. Ann. sec. 399–A:18(I)(f); Tenn. Code Ann. sec. 45–15–111(a); Va. Code Ann. sec. 6.2–2216(A). 184 For example, some maximum vehicle title loan amounts are $2,500 in Mississippi, New Mexico, and Tennessee, and $5,000 in Missouri. Illinois limits the loan amount to $4,000 or 50 percent of monthly income, Virginia and Wisconsin limit the loan amount to 50 percent of the vehicle’s value and Wisconsin also has a $25,000 maximum loan amount. Examples of States with no limits on loan amounts, limits of the amount of the value of the vehicle, or statutes that are silent about loan amounts include Arizona, Idaho, South Dakota, and Utah. Miss. Code Ann. sec. 75–67–415(f); N.M. Stat. Ann. sec. 58–15–3(A); Tenn. Code Ann. sec. 45–15–115(3); Mo. Rev. Stat. sec. 367.527(2); Ill. Admin. Code tit. 38; sec. 110.370(a); Va. Code Ann. sec. 6.2–2215(1)(d); Wis. Stat. sec. 65 initial loan term to one month but several States authorize rollovers (including, in Idaho and Tennessee, automatic rollovers arranged at the time of the original loan, resembling the hybrid payday structure described above), with a few States requiring mandatory amortization in amounts ranging from five to 20 percent on rollovers.185 Unlike payday loan regulation, few States require cooling-off periods between loans or optional extended repayment plans for borrowers who cannot repay vehicle title loans.186 State vehicle title regulations also sometimes address default, repossession and related fees; any cure periods prior to and after repossession; whether the lender must refund any surplus after the repossession and sale or disposition of the vehicle; and whether the borrower is liable for any deficiency remaining after sale or disposition.187 Of the States that expressly authorize vehicle title lending, nine are “non-recourse” meaning that a lender’s remedy upon the borrower’s default is limited to repossession of the vehicle unless the borrower has impaired the 138.16(1)(c); (2)(a); Ariz. Rev. Stat. Ann. sec. 44–291(A); Idaho Code Ann. sec. 28–46–508(3); S.D. Codified Laws sec. 54–4–44; Utah Code Ann. sec. 7–24–202(3)(c). As noted above, as of January 1, 2018, New Mexico vehicle title loans will be limited to $5,000, with minimum loan terms of 120 days. N.M. H.B. 347. 185 States that permit rollovers include Delaware, Georgia, Idaho, Illinois, Mississippi, Missouri, Nevada, New Hampshire, Tennessee, and Utah. Idaho and Tennessee limit title loans to 30 days but allow automatic rollovers and require a principal reduction of 10 percent and 5 percent respectively, starting with the third rollover. Virginia prohibits rollovers and requires a minimum loan term of at least 120 days. See Del. Code Ann. tit. 5 sec. 2254 (rollovers may not exceed 180 days from date of fund disbursement); Ga. Code Ann. sec. 44–12–138(b)(4); Idaho Code Ann. sec. 28–46–506(1) & (3); Ill. Admin. Code tit. 38; sec. 110.370(b)(1) (allowing refinancing if principal is reduced by 20 percent); Miss. Code Ann. sec. 75–67–413(3); Mo. Rev. Stat. sec. 367.512(4); Nev. Rev. Stat. sec. 604A.445(2); N.H. Rev. Stat. Ann. sec. 399–A:19(II) (maximum of 10 rollovers); Tenn. Code Ann. sec. 45–15– 113(a); Utah Code Ann. sec. 7–24–202(3)(a); Va. Code Ann. sec. 6.2–2216(F). 186 Illinois requires 15 days between title loans. Delaware requires title lenders to offer a workout agreement after default but prior to repossession that repays at least 10 percent of the outstanding balance each month. Delaware does not cap fees on title loans and interest continues to accrue on workout agreements. Ill. Admin. Code tit. 38; sec. 110.370(c); Del. Code Ann. 5 secs. 2255 & 2258 (2015). 187 For example, Georgia allows repossession fees and storage fees. Arizona, Delaware, Idaho, Missouri, South Dakota, Tennessee, Utah, Virginia, and Wisconsin specify that any surplus must be returned to the borrower. Mississippi requires that 85 percent of any surplus be returned. Ga. Code Ann. sec. 44–12–131(a)(4)(C); Ariz. Rev. Stat. Ann. sec. 47–9608(A)(4); Del. Code Ann. tit. 5, sec. 2260; Idaho Code Ann. sec. 28–9–615(d); Mo. Rev. Stat. sec. 408.553; S.D. Codified Laws sec. 54–4–72; Tenn. Code Ann. sec. 45–15–114(b)(2); Utah Code Ann. sec. 7– 24–204(3); Va. Code Ann. sec. 6.2–2217(C); Wis. Stat. sec. 138.16(4)(e); Miss. Code Ann. sec. 75–67–411(5). 66 vehicle by concealment, damage, or fraud.188 Other vehicle title lending statutes are silent and do not directly specify whether a lender has recourse against a borrower for any deficiency balance remaining after repossessing the vehicle. An industry trade association commenter stated that title lenders do not sue borrowers or garnish their wages for deficiency balances. Some States have enacted general requirements that vehicle title lenders consider a borrower’s ability to repay before making a title loan. For example, both South Carolina and Utah require lenders to consider borrower ability to repay, but this may be accomplished through a borrower affirmation that she has provided accurate financial information and has the ability to repay.189 Until July 1, 2017, Nevada required title lenders to generally consider a borrower’s ability to repay and obtain an affirmation of this fact. Effective July 1st, an amendment to Nevada law requires vehicle title lenders (and payday lenders, as noted above) to assess borrowers’ reasonable ability to repay by considering, to the extent available, their current or expected income; current employment status based on a pay stub, bank deposit, or other evidence; credit history; original loan amount due, or for installment loans or potential repayment plans, the monthly payment amount; and other evidence relevant to ability to repay including bank statements and borrowers’ written representations.190 Missouri requires that 188 The non-recourse States include Delaware, Florida (short-term loans), Idaho (short-term loans), Nevada, South Carolina, Tennessee (short-term loans), Utah, Virginia, and Wisconsin. Del. Code 5–22–V sec. 2260; Fla. Stat. sec. 33.537.012 (5) (2016); Idaho Code 28–46–508 (2); NRS 604A.455–2; S.C. Code of Laws sec. 37–2–413(5); Tenn. Code Ann. sec. 45–15–115 (2); Utah Code Ann. sec.7–24–204(1); Va. Code sec. 6.2–2217.A & E; and Wis. Stats. 138.16(4)(f). Kentucky and South Dakota’s title lending laws are also non-recourse but those States also have low rate caps that effectively prohibit title loans. Ky. Rev Stat 286.10–275 (2015); S.D. Codified Laws 54–4–72. In addition, vehicle title loans are sometimes made under State pawn lending laws that may provide that borrowers have no personal liability to repay pawn loans or obligation to redeem pledged items. See, e.g., O.C.G.A. 44–12– 137(a)(7) (2010); La. Rev Stat sec. 37:1803 (2016); Minn. Statutes 325J.08(6) (2016). 189 Utah Code Ann. sec. 7–24–202; S.C. Code Ann. sec. 37–3–413(3). 190 Nev. Rev. Stat. sec. 640A.450(3); A.B. 163, 79th Sess. (Nev. 2017). 67 lenders consider a borrower’s financial ability to reasonably repay the loan under the loan’s contract, but does not specify how lenders may satisfy this requirement.191 Industry size and structure. Information about the vehicle title market is more limited than the storefront payday industry because there are currently no publicly traded monoline vehicle title loan companies, most payday lending companies that offer vehicle title loans are not publicly traded, and less information is generally available from State regulators and other sources.192 One national survey conducted in June 2015 found that 1.7 million households reported obtaining a vehicle title loan over the preceding 12 months.193 Another study extrapolating from State regulatory reports estimated that about two million Americans used vehicle title loans annually.194 In 2014, new vehicle title loan originations were estimated at roughly $2 billion with revenue estimates of $3 to $5.6 billion.195 These estimates may not include the full extent of rollovers, as well as vehicle title loan expansion by payday lenders. There are approximately 8,000 title loan storefront locations in the United States, about half of which also offer payday loans.196 Of those locations that predominately offer vehicle title 191 Mo. Rev. Stat sec. 367.525(4). A trade association representing several larger title lenders, the American Association of Responsible Auto Lenders, does not have a public-facing website but has provided the Bureau with some information about the industry. 193 FDIC, “2016 Unbanked and Underbanked Survey,” at 2, 34. 194 Pew Charitable Trusts, “Auto Title Loans: Market Practices and Borrowers’ Experiences,” at 1 (2015), available at http://www.pewtrusts.org/~/media/Assets/2015/03/AutoTitleLoansReport.pdf?la=en. Pew’s estimate includes borrowers of single-payment and installment vehicle title loans. The FDIC’s survey question did not specify any particular type of title loan. 195 Pew Charitable Trusts, “Auto Title Loans: Market Practices and Borrowers’ Experiences,” at 1 (2015), available at http://www.pewtrusts.org/~/media/Assets/2015/03/AutoTitleLoansReport.pdf?la=en; Jean Ann Fox et al., “Driven to Disaster: Car-Title Lending and Its Impact on Consumers,” at 8 (Ctr. for Responsible Lending, 2013), available at, http://www.responsiblelending.org/other-consumer-loans/car-title-loans/research-analysis/CRL-Car-TitleReport-FINAL.pdf. 196 Pew Charitable Trusts, “Auto Title Loans: Market Practices and Borrowers’ Experiences,” at 1, 33 n.7 (2015), available at http://www.pewtrusts.org/~/media/Assets/2015/03/AutoTitleLoansReport.pdf. 192 68 loans, three privately held firms dominate the market and together account for about 3,000 stores in about 20 States.197 These lenders are concentrated in the southeastern and southwestern regions of the country.198 In addition to the large title lenders, smaller vehicle title lenders are estimated to have about 800 storefront locations,199 and as noted above several companies offer both title loans and payday loans.200 The Bureau understands that for some firms whose core business had been payday loans, the volume of vehicle title loan originations now exceeds payday loan originations. State loan data also show an overall trend of vehicle title loan growth. The number of borrowers in Illinois taking vehicle title loans increased 77 percent from 2009 to 2013, and then declined 14 percent from 2013 to 2015.201 The number of title loans taken out in California increased 183 percent between 2011 and 2016.202 In Virginia, from 2011 to 2013, the number of 197 The largest vehicle title lender is TMX Finance, LLC formerly known as Title Max Holdings, LLC with about 1,200 stores in 17 States. It was publicly traded until 2013 when it was taken private. Its last 10-K reported annual revenue of $656.8 million. TMX Fin. LLC, 2012 Annual Report (Form 10–K), at 21 (Mar. 27, 2013). See TMX Finance, “Careers, We Believe in Creating Opportunity,” https://www.tmxfinancefamily.com/careers/ (last visited Sept. 17, 2017) (for TMX Finance store counts); Community Loans of America “About Us,” https://clacorp.com/about-us (last visited Jun. 19, 2017) (states it has about 1,000 locations across 25 States); Fred Schulte, “Lawmakers protect title loan firms while borrowers pay sky-high interest rates” Public Integrity, (updated Sept. 13, 2016), http://www.publicintegrity.org/2015/12/09/18916/lawmakers-protect-title-loan-firms-whileborrowers-pay-sky-high-interest-rates (Select Management Resources has about 700 stores.). 198 Fred Schulte, “Lawmakers protect title loan firms while borrowers pay sky-high interest rates” Public Integrity, (updated Sept. 13, 2016), http://www.publicintegrity.org/2015/12/09/18916/lawmakers-protect-title-loan-firmswhile-borrowers-pay-sky-high-interest-rates. 199 State reports have been supplemented with estimates from Center for Responsible Lending, revenue information from public filings and from non-public sources. See Jean Ann Fox et al., “Driven to Disaster: Car-Title Lending and Its Impact on Consumers,” at 7 (Ctr. for Responsible Lending, 2013) available at http://www.responsiblelending.org/other-consumer-loans/car-title-loans/research-analysis/CRL-Car-Title-ReportFINAL.pdf. 200 Pew Charitable Trusts, “Auto Title Loans: Market Practices and Borrowers’ Experiences,” at 1 (2015), available at http://www.pewtrusts.org/~/media/Assets/2015/03/AutoTitleLoansReport.pdf?la=en. 201 Ill. Dep’t. of Fin. & Prof. Reg., “Illinois Trends Report All Consumer Loan Products Through December 2015,” at 6 (Apr. 14, 2016), available at http://www.idfpr.com/DFI/CCD/pdfs/IL_Trends_Report%202015%20FINAL.pdf?ActID=1204&ChapterID=20). 202 Compare 38,148 vehicle title loans in CY 2011 to 108,080 in CY 2016. California Dep’t of Corps., “2011 Annual Report Operation of Finance Companies Licensed under the California Finance Lenders Law,” at 12 (2012), 69 motor vehicle title loans made increased by 38 percent from 128,446 to a peak of 177,775, and the number of individual consumers taking title loans increased by 44 percent, from 105,542 to a peak of 152,002. By 2016, the number of title loans in Virginia decreased to 155,996 and the number of individual consumers taking title loans decreased to 114,042. The average number of loans per borrower remained constant at 1.2 from 2011 to 2015; in 2016 the number of loans per borrower increased to 1.4.203 In addition to loans made under Virginia’s vehicle title law, a series of reports noted that some Virginia title lenders offered “consumer finance” installment loans without the corresponding consumer protections of the vehicle title lending law and, accounted for about “a quarter of the money loaned in Virginia using automobile titles as collateral.”204 In Tennessee, the number of licensed vehicle title (title pledge) locations at yearend has been measured yearly since 2006. The number of Tennessee locations peaked in 2014 at 1,071, 52 percent higher than the 2006 levels. In 2015, the number of locations declined to 965. However, in each year from 2013 to 2016, the State regulator has reported more licensed available at http://www.dbo.ca.gov/Licensees/Finance_Lenders/pdf/CFL2011ARC.pdf; California Dep’t of Bus. Oversight, “2016 Annual Report Operation of Finance Companies Licensed Under the California Finance Lenders Law,” at 13 (2017), available at http://www.dbo.ca.gov/Licensees/Finance_Lenders/pdf/2016%20CFLL%20Annual%20Report%20FINAL%207-617.pdf. 203 Va. State Corp. Comm’n, “The 2016 Annual Report of the Bureau of Financial Institutions: Payday Lender Licensees, Check Cashers, Motor Vehicle Title Lender Licensees Operating in Virginia at the Close of Business December 31, 2016,” at 67 (2017), available at https://www.scc.virginia.gov/bfi/annual/ar04-16.pdf; Va. State Corp. Comm’n, “The 2013 Annual Report of the Bureau of Financial Institutions, Payday Lender Licensees, Check Cashers, Motor Vehicle Title Lender Licensees Operating in Virginia at the Close of Business December 31, 2013,” at 80 (2014), available at https://www.scc.virginia.gov/bfi/annual/ar04-13.pdf. Because Virginia vehicle title lenders are authorized by State law to make vehicle title loans to residents of other States, the data reported by licensed Virginia vehicle title lenders may include loans made to out-of-State residents. 204 Michael Pope, “How Virginia Became the Region’s Hub For High-Interest Loans,” WAMU, Oct. 6, 2015, http://wamu.org/news/15/10/06/how_virginia_became_the_regional_leader_for_car_title_loans. 70 locations than existed prior to the State’s title lending regulation, the Tennessee Title Pledge Act.205 Vehicle title loan storefront locations serve a relatively small number of customers. One study estimated that the average vehicle title loan store made 218 loans per year, not including rollovers.206 Another study using data from four States and public filings from the largest vehicle title lender estimated that the average vehicle title loan store serves about 300 unique borrowers per year—or slightly more than one unique borrower per business day.207 The same report estimated that the largest vehicle title lender had 4.2 employees per store.208 But, as mentioned, a number of large payday firms offer both products from the same storefront and may use the same employees to do so. In addition, small vehicle title lenders are likely to have fewer employees per location than do larger title lenders. Marketing, underwriting, and collections practices. Vehicle title loans are marketed to appeal to borrowers with impaired credit who seek immediate funds. The largest vehicle title lender described title loans as a “way for consumers to meet their liquidity needs” and described 205 Letter from Greg Gonzales, Comm’r,Tennessee Dep’t of Fin. Insts., to Hon. Bill Haslam, Governor and Hon. Members of the 108th General Assembly, at 1 (Mar. 31, 2014) (Report on the Title Pledge Industry), available at http://www.tennessee.gov/assets/entities/tdfi/attachments/Title_Pledge_Report_2014.pdf; Letter from Greg Gonzales, Comm’r,Tennessee Dep’t of Fin. Insts., to Hon. Bill Haslam, Governor and Hon. Members of the 109th General Assembly, at 2 (Apr. 12, 2016) (Report on the Title Pledge Industry), available at http://www.tennessee.gov/assets/entities/tdfi/attachments/Title_Pledge_Report_2016_Final_Draft_Apr_6_2016.pdf. 206 Jean Ann Fox et al., “Driven to Disaster: Car-Title Lending and Its Impact on Consumers,” at 7 (Ctr. for Responsible Lending, 2013) available at http://www.responsiblelending.org/other-consumer-loans/car-titleloans/research-analysis/CRL-Car-Title-Report-FINAL.pdf. 207 Pew Charitable Trusts, “Auto Title Loans: Market Practices and Borrowers’ Experiences,” at 5 (2015), available at http://www.pewtrusts.org/~/media/Assets/2015/03/AutoTitleLoansReport.pdf?la=en. The four States were Mississippi, Tennessee, Texas, and Virginia. The public filing was from TMX Finance, the largest lender by store count. Id. at 35 n.37. 208 Pew Charitable Trusts, “Auto Title Loans: Market Practices and Borrowers’ Experiences,” at 22 (2015), available at http://www.pewtrusts.org/~/media/Assets/2015/03/AutoTitleLoansReport.pdf?la=en. The estimate is based on TMX Finance’s total store and employee count reported in its Form 10–K as of the end of 2012 (1,035 stores and 4,335 employees). TMX Fin. LLC, 2012 Annual Report (Form 10–K), at 3, 6. The calculation does not account for employees at centralized non-storefront locations. 71 their customers as those who “often … have a sudden and unexpected need for cash due to common financial challenges.”209 Advertisements for vehicle title loans suggest that title loans can be used “to cover unforeseen costs this month …[if] utilities are a little higher than you expected,” if consumers are “in a bind,” for a “short term cash flow” problem, or for “fast cash to deal with an unexpected expense.”210 Vehicle title lenders advertise quick loan approval “in as little as 15 minutes.”211 Some lenders offer promotional discounts for the initial loan and bonuses for referrals,212 for example, a $100 prepaid card for referring friends for vehicle title loans. 213 The underwriting policies and practices that vehicle title lenders use vary and may depend on such factors as State law requirements and individual lender practices. As noted above, some vehicle title lenders do not require borrowers to provide information about their income and instead rely on the vehicle title and the underlying collateral that may be repossessed and sold in the event the borrower defaults—a practice known as asset-based lending.214 The 209 TMX Fin. LLC, 2012 Annual Report (Form 10–K), at 4, 21. See, e.g., Cash 1, “Get an Instant Title Loan,” https://www.cash1titleloans.com/apply-now/arizona.aspx?stt=cash1titleloans_srch&gclid=Cj0KEQjwoM63BRDK_bf4_MeV3ZEBEiQAuQWqkU6O5gtz6kRjP8T3Al-BvylIbIKksDT-r0NMPjEG4kaAqZe8P8HAQ; Speedy Cash, “Title Loans,” https://www.speedycash.com/title-loans/; Metro Loans, “FAQs,” http://metroloans.com/title-loans-faqs/; Lending Bear, “How it Works,” https://www.lendingbear.com/how-it-works/; Fast Cash Title Loans, “FAQ,” http://fastcashvirginia.com/ (all websites last visited Mar. 24, 2016). 211 Check Smart, “Arizona Vehicle Title Loan,” http://www.checksmartstores.com/arizona/title-loans/ (last visited Jan. 14, 2016); Fred Schulte, “Lawmakers protect title loan firms while borrowers pay sky-high interest rates” Public Integrity, (updated Sept. 13, 2016), http://www.publicintegrity.org/2015/12/09/18916/lawmakers-protecttitle-loan-firms-while-borrowers-pay-sky-high-interest-rates. 212 Ctr. for Responsible Lending, “Car Title Lending: Disregard for Borrowers’ Ability to Repay,” at 1, CRL Policy Brief (May 12, 2014), available at http://www.responsiblelending.org/other-consumer-loans/car-titleloans/research-analysis/Car-Title-Policy-Brief-Abilty-to-Repay-May-12-2014.pdf. 213 Check Smart, “Special Offers,” http://www.checksmartstores.com/arizona/special-offers/ (last visited Mar. 29, 2016). 214 Advance America’s website states “[l]oan amount will be based on the value of your car* (*requirements may vary by state).” Advance America, “Title Loans,” https://www.advanceamerica.net/services/title-loans (last visited Mar. 3, 2016); Pew Charitable Trusts, “Auto Title Loans: Market Practices and Borrowers’ Experiences,” at 1 210 72 largest vehicle title lender stated in 2011 that its underwriting decisions were based entirely on the wholesale value of the vehicle.215 Other title lenders’ websites state that proof of income is required,216 although it is unclear whether employment information is verified or used for underwriting, whether it is used for collections and communication purposes upon default, or for both purposes. The Bureau is aware, from confidential information gathered in the course of its statutory functions, that one or more vehicle title lenders regularly exceed their maximum loan amount guidelines and instruct employees to consider a vehicle’s sentimental or use value to the borrower when assessing the amount of funds they will lend. As in the market for payday loans, there have been some studies about the extent of price competition in the vehicle title lending market. Vehicle title lending is almost exclusively a storefront market, as discussed above. The evidence of price competition is mixed. One large title lender stated that it competes on factors such as location, customer service, and convenience, and also highlights its pricing as a competitive factor.217 An academic study found evidence of price competition in the vehicle title market, citing the abundance of price-related advertising and evidence that in States with rate caps, such as Tennessee, approximately half of the lenders charged the maximum rate allowed by law, while the other half charged lower (2015), available at http://www.pewtrusts.org/~/media/Assets/2015/03/AutoTitleLoansReport.pdf?la=en; Fred Schulte, “Lawmakers protect title loan firms while borrowers pay sky-high interest rates” Public Integrity, (updated Sept. 13, 2016), http://www.publicintegrity.org/2015/12/09/18916/lawmakers-protect-title-loan-firms-whileborrowers-pay-sky-high-interest-rates. 215 TMX Fin. LLC, 2012 Annual Report (Form 10–K), at 5. 216 See, e.g., Check Into Cash, “Unlock The Cash In Your Car,” https://checkintocash.com/title-loans/ (last visited Mar. 3, 2016); Speedy Cash, “Title Loans,” https://www.speedycash.com/title-loans/ (last visited Mar. 3, 2016); ACE Cash Express, “Title Loans,” https://www.acecashexpress.com/title-loans (last visited Mar. 3, 2016); Fast Cash Title Loans, “FAQ,” http://fastcashvirginia.com/faq/ (last visited Mar. 3, 2016). 217 TMX Fin. LLC, 2012 Annual Report (Form 10–K), at 6. 73 rates.218 However, another report found that like payday lenders, title lenders compete primarily on location, speed, and customer service, gaining customers by increasing the number of locations rather than underpricing their competition.219 Loan amounts are typically for less than half the wholesale value of the consumer’s vehicle. Low loan-to-value ratios reduce a lender’s risk. A survey of title lenders in New Mexico found that the lenders typically lend between 25 and 40 percent of a vehicle’s wholesale value.220 At one large title lender, the weighted average loan-to-value ratio was found to be 26 percent of Black Book retail value.221 The same lender has two principal operating divisions; one division requires that vehicles have a minimum appraised value greater than $500, but the lender will lend against vehicles with a lower appraised value through another brand.222 When a borrower defaults on a vehicle title loan, the lender may repossess the vehicle. The Bureau believes, based on market outreach, that the decision whether to repossess a vehicle will depend on factors such as the amount due, the age and resale value of the vehicle, the costs to locate and repossess the vehicle, and State law requirements to refund any surplus amount remaining after the sale proceeds have been applied to the remaining loan balance.223 Available information indicates that lenders are unlikely to repossess vehicles they do not expect to sell. The largest vehicle title lender sold 83 percent of the vehicles it repossessed but did not report 218 Jim Hawkins, “Credit on Wheels: The Law and Business of Auto-Title Lending,” 69 Wash. & Lee L. Rev. 535, 558-559 (2012). 219 Pew Charitable Trusts, “Auto Title Loans: Market Practices and Borrowers’ Experiences,” at 5 (2015), available at http://www.pewtrusts.org/~/media/Assets/2015/03/AutoTitleLoansReport.pdf?la=en. 220 Nathalie Martin & Ozymandias Adams, “Grand Theft Auto Loans: Repossession and Demographic Realities in Title Lending,” 77 Mo. L. Rev. 41 (2012). 221 TMX Fin. LLC, 2011 Annual Report (Form 10–K), at 3 (Mar. 19, 2012). 222 TMX Fin. LLC, 2011 Annual Report (Form 10–K), at 5 (Mar. 19, 2012). 223 See also Pew Charitable Trusts, “Auto Title Loans: Market Practices and Borrowers’ Experiences,” at 13–14 (2015), available at http://www.pewtrusts.org/~/media/Assets/2015/03/AutoTitleLoansReport.pdf?la=en. 74 overall repossession rates.224 In 2012, its firm-wide gross charge-offs equaled 30 percent of its average outstanding title loan balances.225 The Bureau is aware of vehicle title lenders engaging in illegal debt collection activities in order to collect amounts claimed to be due under title loan agreements. These practices include altering caller ID information on outgoing calls to borrowers to make it appear that calls were from other businesses, falsely threatening to refer borrowers for criminal investigation or prosecution, and unlawfully disclosing debt information to borrowers’ employers, friends, and family.226 In addition, approximately 16 percent of consumer complaints handled by the Bureau about vehicle title loans involved consumers reporting concerns about repossession issues.227 Some vehicle title lenders have installed electronic devices on the vehicles, known as starter interrupt devices, automated collection technology, or more colloquially as “kill switches,” that can be programmed to transmit audible sounds in the vehicle before or at the payment due date. The devices may also be programmed to prevent the vehicle from starting when the borrower is in default on the loan, although they may allow a one-time re-start upon the 224 Missouri sales of repossessed vehicles calculated from data linked to St. Louis Post-Dispatch. Walter Moskop, “Title Max is Thriving in Missouri—and Repossessing Thousands of Cars in the Process,” St. Louis Post-Dispatch, Sept. 21, 2015, available at http://www.stltoday.com/business/local/titlemax-is-thriving-in-missouri-andrepossessing-thousands-of-cars/article_d8ea72b3-f687-5be4-8172-9d537ac94123.html. 225 Bureau estimates based on publicly available financial statements by TMX Fin. LLC, 2012 Annual Report (Form 10–K), at 22, 43. 226 Press Release, Bureau of Consumer Fin. Prot., “CFPB Orders Relief for Illegal Debt Collection Tactics,” (Oct. 1, 2015), available at https://www.consumerfinance.gov/about-us/newsroom/cfpb-orders-indirect-auto-financecompany-to-provide-consumers-44-1-million-in-relief-for-illegal-debt-collection-tactics/. In September 2016, the CFPB took action against TMX Finance, alleging that employees made in-person visits to borrowers’ references and places of employment, and disclosed the existence of borrowers’ past due debts to these third-parties. Consent Order, TMX Finance LLC, CFPB No. 2016–CFPB–0022, (Sept. 26, 2016), available at https://www.consumerfinance.gov/documents/1011/092016_cfpb_TitleMaxConsentOrder.pdf. 227 This represents complaints received between November 2013 and December 2016. 75 borrower’s call to obtain a code.228 One of the starter interrupt providers states that “[a]ssuming proper installation, the device will not shut off the vehicle while driving.”229 Due to concerns about consumer harm, a State Attorney General issued a consumer alert about the use of starter interrupt devices specific to vehicle title loans.230 The alert also noted that some title lenders require consumers to provide an extra key to their vehicles. In an attempt to avoid illegal repossessions, Wisconsin’s vehicle title law prohibits lenders from requiring borrowers to provide the lender with an extra key to the vehicle.231 The Bureau has received several complaints about starter interrupt devices. Business model. As noted above, short-term vehicle title lenders appear to have overhead costs relatively similar to those of storefront payday lenders. Default rates on vehicle title loans and lender reliance on re-borrowing activity appear to be even greater than that of storefront payday lenders. Based on data analyzed by the Bureau, the default rate on single-payment vehicle title loans is six percent and the sequence-level default rate is 33 percent, compared with a 20 percent sequence-level default rate for storefront payday loans. One-in-five single-payment vehicle title loan borrowers have their vehicle repossessed by the lender.232 One industry trade association 228 See, e.g., Eric L. Johnson & Corinne Kirkendall, “Starter Interrupt and GPS Devices: Best Practices,” PassTime InTouch, Jan. 14, 2016, available at https://passtimegps.com/starter-interrupt-and-gps-devices-best-practices/. These products may be used in conjunction with GPS devices and are also marketed for subprime automobile financing and insurance. 229 Eric L. Johnson & Corinne Kirkendall, “Starter Interrupt and GPS Devices: Best Practices,” PassTime InTouch, Jan. 14, 2016, available at https://passtimegps.com/starter-interrupt-and-gps-devices-best-practices/. 230 The alert also noted that vehicle title loans are illegal in Michigan. Bill Schuette, Mich. Att’y Gen., “Consumer Alert: Auto Title Loans,” available at http://www.michigan.gov/ag/0,4534,7-164-17337_20942-371738--,00.html. 231 Wis. Stat. sec. 138.16(4)(b). 232 CFPB Single-Payment Vehicle Title Lending, at 23; CFPB Report on Supplemental Findings, at 112. 76 commenter stated that 15 to 25 percent of repossessed vehicles are subsequently redeemed by borrowers after paying off the deficiency balance owed (along with repossession costs). Similarly, the rate of vehicle title re-borrowing appears high. In the Bureau’s data analysis, more than half—56 percent—of single-payment vehicle title loan sequences stretched for at least four loans; over a third—36 percent—were seven or more loans; and 23 percent of loan sequences consisted of 10 or more loans. While other sources on vehicle title lending are more limited than for payday lending, the Tennessee Department of Financial Institutions publishes a biennial report on vehicle title lending. Like the single-payment vehicle title loans the Bureau has analyzed, the vehicle title loans in Tennessee are 30-day single-payment loans. The most recent report shows similar patterns to those the Bureau found in its research, with a substantial number of consumers rolling over their loans multiple times. According to the report, of the total number of loan agreements made in 2014, about 15 percent were paid in full after 30 days without rolling over. Of those loans that are rolled over, about 65 percent were at least in their fourth rollover, about 44 percent were at least in their seventh rollover, and about 29 percent were at least in their tenth, up to a maximum of 22 rollovers.233 The impact of these outcomes for consumers who are unable to repay and either default or re-borrow is discussed in Market Concerns—Underwriting. 233 Letter from Greg Gonzales, Comm’r,Tennessee Dep’t of Fin. Insts., to Hon. Bill Haslam, Governor and Hon. Members of the 109th General Assembly, at 8 (Apr. 12, 2016) (Report on the Title Pledge Industry), available at http://www.tennessee.gov/assets/entities/tdfi/attachments/Title_Pledge_Report_2016_Final_Draft_Apr_6_2016.pdf. See Tenn. Code Ann. sec. 45–17–112(q). 77 Short-Term Lending by Depository Institutions As noted above, within the banking system, consumers with liquidity needs rely primarily on credit cards and overdraft services. Some depository institutions, particularly community banks and credit unions, provide occasional small loans on an accommodation basis to their customers.234 The Bureau’s market monitoring indicates that a number of the banks and credit unions offering these accommodation loans are located in small towns and rural areas and that it is not uncommon for borrowers to be in non-traditional employment or have seasonal or variable income. In addition, some depository institutions have experimented with short-term paydaylike products or partnered with payday lenders, but such experiments have had mixed results and in several cases have prompted prudential regulators to take action discouraging certain types of activity. For a period of time, a handful of banks also offered a deposit advance product as discussed below; that product also prompted prudential regulators to issue guidance that effectively discouraged the offering of the product. National banks, most State-chartered banks, and State credit unions are permitted under existing Federal laws to charge interest on loans at the highest rate allowed by the laws of the State in which the lender is located (lender’s home State).235 The bank or State-chartered credit union may then charge the interest rate of its home State on loans it makes to borrowers in other 234 A trade association representing community banks conducted a survey of its members and found 39 percent of respondents offered short-term personal loans of $1,000 (term of 45 day or less). However, among respondents, personal loan portfolios (including longer-term loans, open-end lines of credit, and deposit advance loans) accounted for less than 3 percent of the dollar volume of their total lending portfolios. Further, the survey noted that these loans are not actively advertised to consumers. Ryan Hadley, “2015 ICBA Community Bank Personal Small Dollar Loan Survey,” at 4 (Oct. 29, 2015) (on file). 235 See generally 12 U.S.C. 85 (governing national banks); 12 U.S.C. 1463 (g) (governing savings associations); 12 U.S.C. 1785 (g) (governing credit unions); 12 U.S.C. 1831d (governing State banks). Alternatively, these lenders may charge a rate that is no more than 1 percent above the 90-day commercial paper rate in effect at the Federal Reserve Bank in the Federal Reserve district in which the lender is located (whichever is higher). Id. 78 States without needing to comply with the usury limits of the States in which it makes the loans (borrower’s home State). Federal credit unions generally must not charge more than an 18 percent interest rate. However, the National Credit Union Administration (NCUA) has taken some steps to encourage federally chartered credit unions to offer “payday alternative loans,” which generally have a longer term than traditional payday products. Federal credit unions are authorized to make these small-dollar loans at rates up to 28 percent interest plus an application fee. This program is discussed in more detail below. Agreements between depository institutions and payday lenders. In 2000, the Office of the Comptroller of the Currency (OCC) issued an advisory letter alerting national banks that the OCC had significant safety and soundness, compliance, and consumer protection concerns with banks entering into contractual arrangements with vendors seeking to avoid certain State lending and consumer protection laws. The OCC noted it had learned of nonbank vendors approaching federally chartered banks urging them to enter into agreements to fund payday and title loans. The OCC also expressed concern about unlimited renewals (what the Bureau refers to as rollovers or re-borrowing), and multiple renewals without principal reduction.236 The agency subsequently took enforcement actions against two national banks for activities relating to payday lending partnerships.237 236 Advisory Letter: AL 2000–10 to Chief Executive Officers of All Nat’l Banks, Dep’t and Div. Heads, and All Examing Personnel from OCC (Nov. 27, 2000) (Payday Lending), available at http://www.occ.gov/static/newsissuances/memos-advisory-letters/2000/advisory-letter-2000-10.pdf. 237 See OCC consent orders involving Peoples National Bank and First National Bank in Brookings. Press Release, OCC Admin of Nat’l Banks, NR 2003–06, “Peoples National Bank to Pay $175,000 Civil Money Penalty And End Payday Lending Relationship with Advance America” (Jan. 31, 2003), available at http://www.occ.gov/static/newsissuances/news-releases/2003/nr-occ-2003-6.pdf; Consent Order, First National Bank in Brookings, OCC No. 2003–1 (Jan. 17, 2003), available at http://www.occ.gov/static/enforcement-actions/ea2003-1.pdf. In December 2016, the OCC released a plan to offer limited special purpose bank charters to fintech companies. In response to 79 The Federal Deposit Insurance Corporation (FDIC) has also expressed concerns with similar agreements between payday lenders and the depositories under its purview. In 2003, the FDIC issued Guidelines for Payday Lending applicable to State-chartered FDIC-insured banks and savings associations; the guidelines were revised in 2005 and most recently in 2015. The guidelines focus on third-party relationships between the chartered institutions and other parties, and specifically address rollover limitations. They also indicate that banks should ensure borrowers exhibit both a willingness and ability to repay when rolling over a loan. Among other things, the guidelines indicate that institutions should: (1) ensure that payday loans are not provided to customers who had payday loans outstanding at any lender for a total of three months during the previous 12 months; (2) establish appropriate cooling-off periods between loans; and (3) provide that no more than one payday loan is outstanding with the bank at a time to any one borrower.238 In 2007, the FDIC issued guidelines encouraging banks to offer affordable small-dollar loan alternatives with APRs of 36 percent or less, reasonable and limited fees, amortizing payments, underwriting focused on a borrower’s ability to repay but allowing flexible documentation, and to avoid excessive renewals.239 criticism that such a charter might enable payday lenders to circumvent some States’ attempts to ban payday lending, the OCC stated it had virtually eliminated abusive payday lending in the federal banking system in the early 2000s, and had “no intention of allowing these practices to return.” Lalita Clozel, “OCC Fintech Charter Opens ‘henhouse’ to Payday Lenders: Consumer Groups,” American Banker, Jan. 13, 2016, available at https://www.americanbanker.com/news/occ-fintech-charter-opens-hen-house-to-payday-lenders-consumer-groups. See “Comptroller’s Licensing Manual Draft Supplement: OCC, Evaluating Charter Application From Financial Technology Companies,” (Mar. 2017), available at https://www.occ.gov/publications/publications-bytype/licensing-manuals/file-pub-lm-fintech-licensing-manual-supplement.pdf. 238 FDIC, “Financial Institution Letters: Guidelines for Payday Lending,” (Revised Nov. 2015), available at https://www.fdic.gov/news/news/financial/2005/fil1405a.html. 239 FDIC, “Financial Institution Letters: Affordable Small-Dollar Loan Products Final Guidelines,” FIL 50–2007 (June 19, 2007), available at https://www.fdic.gov/news/news/financial/2007/fil07050.html. 80 Deposit advance product lending. As the payday lending industry grew, a handful of banks decided to offer their deposit customers a similar product termed a deposit advance product (DAP). While one bank started offering deposit advances in the mid-1990s, the product began to spread more rapidly in the late 2000s and early 2010s. DAP could be structured a number of ways but generally involved a line of credit offered by depository institutions as a feature of an existing consumer deposit account with repayment automatically deducted from the consumer’s next qualifying deposit. Deposit advance products were available to consumers who received recurring electronic deposits if they had an account in good standing and, for some banks, several months of account tenure, such as six months. When an advance was requested, funds were deposited into the consumer’s account. Advances were automatically repaid when the next qualifying electronic deposit, whether recurring or one-time, was made to the consumer’s account rather than on a fixed repayment date. If an outstanding advance was not fully repaid by an incoming electronic deposit within about 35 days, the consumer’s account was debited for the amount due and could result in a negative balance on the account. The Bureau estimates that at the product’s peak from mid-2013 to mid-2014, banks originated roughly $6.5 billion of advances, which represents about 22 percent of the volume of storefront payday loans issued in 2013. The Bureau estimates that at least 1.5 million unique borrowers took out one or more DAP loans during that same period.240 240 CFPB staff analysis based on confidential information gathered in the course of statutory functions. Estimates made by summing aggregated data across a number of DAP-issuing institutions. See John Hecht, “Alternative Financial Services: Innovating to Meet Customer Needs in an Evolving Regulatory Framework,” at 7 (2014) (Stephens, Inc., slide presentation) (on file) (for payday industry size). 81 DAP fees, like payday loan fees, did not vary with the amount of time that the advance was outstanding but rather were set as dollars per amount advanced. A typical fee was $2 per $20 borrowed, the equivalent of $10 per $100. Research undertaken by the Bureau using a supervisory dataset found that the median duration of an episode of DAP usage was 12 days, yielding an effective APR of 304 percent.241 The Bureau further found that while the median draw on a DAP was $180, users typically took more than one draw before the advance was repaid. The multiple draws resulted in a median average daily DAP balance of $343, which is similar to the size of a typical payday loan. With the typical DAP fee of $2 per $20 advanced, the fees for $343 in advances equate to about $34.30. The median DAP user was indebted for 112 days over the course of a year and took advances in seven months. Fourteen percent of borrowers took advances totaling over $9,000 over the course of the year; these borrowers had a median number of days in debt of 254.242 In 2010, the Office of Thrift Supervision (OTS) issued a supervisory directive ordering one bank to terminate its DAP program, which the bank offered in connection with prepaid accounts, after determining the bank engaged in unfair or deceptive acts or practices and violated the OTS’ Advertising Regulation.243 Consequently, in 2011, pursuant to a cease and desist order, the bank agreed to remunerate its DAP consumers nearly $5 million and pay a civil monetary penalty of $400,000.244 241 CFPB Payday Loans and Deposit Advance Products White Paper, at 27–28. CFPB Payday Loans and Deposit Advance Products White Paper, at 33 fig. 11, 37 fig. 14. 243 Meta Fin. Grp., Inc., 2010 Annual Report (Form 10–K), at 59 (Dec. 13, 2010). 244 Meta Fin. Grp., Inc., Quarter Report (Form 10–Q) at 31 (Aug. 5, 2011). The OTS was merged with the OCC effective July 21, 2011. See OCC, “OTS Integration,” http://www.occ.treas.gov/about/who-we-are/occ-foryou/bankers/ots-integration.html (last visited Apr. 27, 2016). 242 82 In November 2013, the FDIC and OCC issued final supervisory guidance on DAP.245 This guidance stated that banks offering DAP should adjust their programs in a number of ways, including applying more scrutiny in underwriting DAP loans and discouraging repetitive borrowing. Specifically, the OCC and FDIC stated that banks should ensure that the customer relationship is of sufficient duration to provide the bank with adequate information regarding the customer’s recurring deposits and expenses, and that the agencies would consider sufficient duration to be no less than six months. In addition, the guidance said that banks should conduct a more stringent financial capacity assessment of a consumer’s ability to repay the DAP advance according to its terms without repeated re-borrowing, while meeting typical recurring and other necessary expenses, as well as outstanding debt obligations. In particular, the guidance stated that banks should analyze a consumer’s account for recurring inflows and outflows at the end, at least, of each of the preceding six months before determining the appropriateness of a DAP advance. Additionally, the guidance noted that in order to avoid re-borrowing, a cooling-off period of at least one monthly statement cycle after the repayment of a DAP advance should be completed before another advance could be extended. Finally, the guidance stated that banks should not increase DAP limits automatically and without a fully underwritten reassessment of a consumer’s ability to repay, and banks should reevaluate a consumer’s eligibility and capacity for DAP at least every six months.246 245 Guidance on Supervisory Concerns and Expectations Regarding Deposit Advance Products, 78 FR 70624 (Nov. 26, 2013); Guidance on Supervisory Concerns and Expectations Regarding Deposit Advance Products, 78 FR 70552 (Nov. 26, 2013). 246 Guidance on Supervisory Concerns and Expectations Regarding Deposit Advance Products, 78 FR 70624 (Nov. 26, 2013); Guidance on Supervisory Concerns and Expectations Regarding Deposit Advance Products, 78 FR 70552 (Nov. 26, 2013). 83 Following the issuance of the FDIC and OCC guidance, banks supervised by the FDIC and OCC ceased offering DAP. Of two DAP-issuing banks supervised by the Federal Reserve Board and therefore not subject to either the FDIC or OCC guidance, one eliminated its DAP program while another continues to offer a modified version of DAP to its existing DAP borrowers.247 Today, with the exception of some short-term lending within the NCUA’s Payday Alternative Loan (PAL) program, described in detail below, relatively few banks or credit unions offer large-scale formal loan programs of this type. Federal credit union payday alternative loans. As noted above, Federal credit unions may not charge more than 18 percent interest. However, in 2010, the NCUA adopted an exception to the interest rate limit under the Federal Credit Union Act that permitted Federal credit unions to make PALs at an interest rate of up to 28 percent plus an application fee, “that reflects the actual costs associated with processing the application” up to $20.248 PALs may be made in amounts of $200 to $1,000 to borrowers who have been members of the credit union for at least one month. PAL terms range from one to six months, PALs may not be rolled over, and borrowers are limited to one PAL at a time and no more than three PALs from the same credit union in a rolling six-month period. PALs must fully amortize and the credit union must establish underwriting guidelines such as verifying borrowers’ employment from at least two recent pay stubs.249 247 The Federal Reserve Board issued a statement to its member banks on DAP. Bd. of Governors of the Fed. Reserve Sys., “Statement on Deposit Advance Products,” (Apr. 25, 2013), available at https://www.federalreserve.gov/supervisionreg/caletters/CA13-07attachment.pdf. 248 12 CFR 701.21(c)(7)(iii). Application fees charged to all applicants for credit are not part of the finance charge that must be disclosed under Regulation Z. See 12 CFR 1026.4(c). 249 12 CFR 701.21(c)(7)(iii). 84 In 2016, about 650 Federal credit unions (nearly 20 percent of all Federal credit unions) offered PALs, with originations at $134.7 million, representing a 9.7 percent increase from 2015.250 In 2015, the average PAL amount was about $700 and carried a median interest rate of 25 percent; in 2016, the average PAL loan amount increased to about $720 with a similar median interest rate of 25 percent.251 C. Longer-Term, High-Cost Loans In addition to short-term loans, certain longer-term, high-cost loans will be covered by the payments protections provisions of this rule. These are longer-term, high-cost loans with a leveraged payment mechanism, as described in more detail in part II.D and Markets Concerns— Payments. The category of longer-term high-cost loans most directly impacted by the payments protections in this rule are payday installment loans. Payday Installment Loans Product definition and regulatory environment. The term “payday installment loan” refers to a high-cost loan repaid in multiple installments, with each installment typically due at the consumer’s payday and with the lender generally having the ability to collect the payment from the consumer’s bank account as money is deposited or directly from the consumer’s paycheck.252 250 Nat’l Credit Union Admin., “5300 Call Report Aggregate Financial Performance Reports (FPRs),” (Dec. 2016), available at https://www.ncua.gov/analysis/Pages/call-report-data/aggregate-financial-performance-reports.aspx. 251 Bureau staff estimates are based on NCUA Call Report data. Nat’l Credit Union Admin., “Credit Union and Corporate Call Report Data,” available at https://www.ncua.gov/analysis/Pages/call-report-data.aspx. 252 Lenders described in part II.C as payday installment lenders may not use this terminology. 85 Two States, Colorado and Illinois, have authorized payday installment loans.253 Through 2010 amendments to its payday loan law, Colorado no longer permits short-term single-payment payday loans. Instead, in order to charge fees in excess of the 36 percent APR cap for most other consumer loans, the minimum loan term must be six months and lenders are permitted to take a series of post-dated checks or payment authorizations to cover each payment under the loan, providing lenders with the same access to borrower’s accounts as a single-payment payday loan.254 In Illinois, lenders have been permitted to make payday installment loans since 2011. These loans must be fully-amortizing for terms of 112 to 180 days and the loan amounts are limited to the lesser of $1,000 or 22.5 percent of gross monthly income.255 A number of other States have adopted usury laws that some payday lenders use to offer payday installment loans in lieu of, or in addition to, more traditional payday loans. Since July 2016, Mississippi lenders can make “credit availability loans”—closed-end fully-amortizing installment loans with loan terms of four to 12 months, whether secured by personal property or unsecured.256 The maximum loan amount on a credit availability loan is limited to $2,500, and lenders may charge a monthly handling fee of up to 25 percent of the outstanding principal balance plus an origination fee of the greater of 1 percent of the amount disbursed or $5.257 253 As noted above, as of January 1, 2018, New Mexico payday loans (and vehicle title loans) must be payable in four substantially equal payments over at least 120 days with an APR of 175 percent or less. 254 Colo. Rev. Stat. sec. 5–3.1–103. Although loans may be structured in multiple installments of substantially equal payments or a single installment, almost all lenders contract for repayment in monthly or bi-weekly installments. 4 Colo. Code Regs. sec. 902–1, Rule 17(B); Adm’r of the Colo. Consumer Credit Unit, “Colorado Payday Lending— July Demographic and Statistical Information: July 2000 through December 2012,” at 15–16, available at https://coag.gov/uccc/info/ar. 255 815 Ill. Comp. Stat. sec. 122/2–5. 256 Miss. Code Ann. sec. 75–67–603(e) (2017). 257 Miss. Code Ann. sec. 75–67–619 (2017) 86 As of 2015, Tennessee lenders may offer “flex loans”—open-end lines of credit that need not have a fixed maturity date and that may be secured by personal property or unsecured.258 The maximum outstanding balance on a flex loan may not exceed $4,000, with an interest rate of up to 24 percent per annum and “customary fees” for underwriting and other purposes not to exceed a daily rate of 0.7 percent of the average daily principal balance.259 At least one lender offering flex loans states that loan payments are “aligned with your payday.”260 Similar legislation has been unsuccessful in other States. For example, in May 2017 the Governor of Oklahoma vetoed legislation that would have authorized high-cost installment loans with interest rates of up to 17 percent per month, or 204 percent APR.261 None of these laws authorizing payday installment loans, credit access loans, or flex loans appear to limit the use of electronic repayment or ACH options for repayment. In addition to States that authorize a specific form of payday installment loan, credit access loan, or flex loan, several other States provide room within their usury laws for high-cost installment products. A recent report found that seven States have no rate or fee limits for closed-end loans of $500 and that 10 States have no rate or fee limits for closed-end loans of $2,000.262 The same report noted that for open-end credit, 13 States do not limit rates for a $500 258 Tenn. Code Ann. sec. 45–12–102(6) (2017). Tenn. Code Ann. sec. 45–12–111(2017). 260 Advance Financial Flex Loan, “Online Tennessee Flex Loans,” https://www.af247.com/tennessee-flex-loans (last visited May 17, 2017). 261 Okla. H.B. 1913, 56th Leg., 1st Sess. (Okla. 2017). http://www.oklegislature.gov/BillInfo.aspx?Bill=HB1913&Session=1700. 262 Nat’l. Consumer Law Ctr., “Predatory Installment Lending in 2017, States Battle to Restrain High-Cost Loans,” at 14 map 1, 15 map 2 (Aug. 2017), available at https://www.nclc.org/images/pdf/pr-reports/installmentloans/report-installment-loans.pdf. Roughly half of the States with no set limits do prohibit unconscionable interest rates. As of January 1, 2008, New Mexico’s status will change from a State with no rate caps for loans of $500 or $2,000 to a State that caps rates at 175 percent APR. 259 87 advance and 15 States do not limit them for a $2,000 advance.263 Another recent study of the websites of five payday lenders that operate both online and at storefront locations found that these five lenders offered payday installment loans in at least 17 States.264 In addition, as discussed above, a substantial segment of the online payday industry operates outside of the constraints of State law, and this segment, too, has migrated towards payday installment loans. For example, a study commissioned by a trade association for online lenders surveyed seven lenders and concluded that, while single-payment loans are still a significant portion of these lenders’ volume, they are on the decline while installment loans are growing. Several of the lenders represented in the report had either eliminated single-payment products or were migrating to installment products while still offering single-payment loans.265 For the practical reasons associated with having no retail locations, online lenders prefer repayment by electronic methods and use various approaches to secure consumers’ authorization for payments electronically through ACH debits. 263 Nat’l. Consumer Law Ctr., “Predatory Installment Lending in 2017, States Battle to Restrain High-Cost Loans,” at 18 map 3, 19 map 4 (Aug. 2017), available at https://www.nclc.org/images/pdf/pr-reports/installmentloans/report-installment-loans.pdf. 264 Diane Standaert, “Payday and Car Title Lenders’ Migration to Unsafe Installment Loans,” at 7 tbl.1 (Ctr. for Responsible Lending, 2015), available at http://www.responsiblelending.org/other-consumer-loans/car-titleloans/research-analysis/crl_brief_cartitle_lenders_migrate_to_installmentloans.pdf. CRL surveyed the websites for: Cash America, Enova International (d/b/a CashNetUSA and d/b/a NetCredit), Axcess Financial (d/b/a Check ‘N Go), and ACE Cash Express. Id. at 10 n.52. 265 G. Michael Flores, “The State of Online Short-Term Lending, Second Annual Statistical Analysis Report,” Bretton-Woods, Inc., at 3 (Feb. 28, 2014), available at http://onlinelendersalliance.org/wpcontent/uploads/2015/07/2015-Bretton-Woods-Online-Lending-Study-FINAL.pdf The report does not address the State licensing status of the study participants but based on its market outreach activities, the Bureau believes that some of the loans included in the study were not made subject to the licensing laws of the borrowers’ States of residence. See also nonPrime101, “Report 1: Profiling Internet Small Dollar Lending—Basic Demographics and Loan Characteristics,” at 9, 11, (2014), available at https://www.nonprime101.com/wpcontent/uploads/2013/10/Clarity-Services-Profiling-Internet-Small-Dollar-Lending.pdf. 88 As with payday loans, and as noted above, as authorized or permitted by some State laws, payday installment lenders often hold borrowers’ checks or obtain their authorization for ACH repayment. Some borrowers may prefer ACH repayment methods for convenience. The Bureau is aware of certain practices used by payday installment lenders to secure repayment through consumers’ accounts including longer waits for distribution of loan proceeds and higher fees for non-electronic payment methods, described above in the Online Payday Loans section, and discussed in more detail in part II.D and Markets Concerns—Payments. To the extent that longer-term payday installment loans meet the cost of credit threshold and include leveraged payment mechanisms, they are subject to this rule’s payments protections.266 Finance Company Installment Loans Product definition and regulatory environment. Before the advent of single-payment payday loans or online lending, and before widespread availability of credit cards, “personal loans” or “personal installment loans” were offered by storefront nonbank installment lenders, often referred to as “finance companies.” Personal loans are typically unsecured loans used for any variety of purposes and distinguished from loans where the lender generally requires the funds be used for a specific intended purpose, such as automobile purchase loans, student loans, and mortgage loans. As discussed below, these finance companies (and their newer online 266 Installment vehicle title loans are title loans that are contracted to be repaid in multiple installments rather than in a single payment. Vehicle title lending almost exclusively occurs at retail storefront locations and consequently, borrowers often repay both in cash at the lender’s location. However, some installment vehicle title lenders allow repayment by ACH from the borrower’s account or by debit card, a practice common to payday installment loans. See, e.g., Auto Loan Store, “Auto Title Loan FAQ,” https://www.autotitlelending.com/faq/ (last visited June 20, 2017); TFC Title Loans, “How Are Title Loans Paid Back?,” TFC Title Loans Blog, https://www.tfctitleloans.com/blog/how-are-title-loans-paid-back/ (last visited Sept. 17, 2017); Presto Title Loans, “You Can Make Payments Online!,” http://prestoautoloans.com/pay-online/!/ (last visited June 20, 2017). To the extent that longer-term installment vehicle title loans meet the cost of credit threshold and the lender obtains a leveraged payment mechanism, the loans are subject to this rule’s protections for payment presentments. 89 counterparts) have a different business model than payday installment lenders. Some of these finance companies limit the APRs on their loans to 36 percent or less, whether required by State law or as a matter of company policy. However, there are other finance companies and installment lenders that offer loans that fall within the rule’s definition of “covered longer-term loan,” as they carry a cost of credit that exceeds 36 percent APR and include repayment through a leveraged payment mechanism—access to the borrower’s account. According to a report from a consulting firm using data derived from a nationwide consumer reporting agency, in 2016 finance companies originated 8.6 million personal loans (unsecured installment loans) totaling $41.7 billion in originations; approximately 6.9 million of these loans worth $25.8 billion, with an average loan size of about $3,727, were made to nonprime consumers (categorized as near prime, subprime, and deep subprime, with VantageScores of 660 and below).267 APRs at storefront locations in States that do not cap rates on installment loans can be 50 to 90 percent for subprime and deep subprime borrowers; APRs in States with rate caps are 24 to 36 percent APR for near prime and subprime borrowers.268 A survey of finance companies conducted in conjunction with a national trade association reported that 80 percent of loans were 267 Experian-Oliver Wyman, “2016 Q4 Market Intelligence Report: Personal Loans Report,” at 11–13 figs. 9, 10, 12 &13 (2017), available at http://www.marketintelligencereports.com; Experian-Oliver Wyman, “2016 Q3 Market Intelligence Report: Personal Loans Report,” at 11-13 figs. 9, 10, 12 & 13 (2016), available at http://www.marketintelligencereports.com; Experian-Oliver Wyman, “2016 Q2 Market Intelligence Report: Personal Loans Report,” at 11–13 figs. 9, 10, 12 & 13 (2016), available at http://www.marketintelligencereports.com; Experian-Oliver Wyman, “2016 Q1 Market Intelligence Report: Personal Loans Report,” at 11–13, figs. 9, 10, 12 & 13 (2016), available at http://www.marketintelligencereports.com. These finance company personal loans are not segmented by cost and likely include some loans with a cost of credit of 36 percent APR or less that would not be covered by the Bureau’s rule. 268 See John Hecht, “Alternative Financial Services: Innovating to Meet Customer Needs in an Evolving Regulatory Framework,” at 11 (2014) (Stephens, Inc., slide presentation) (on file) (for listing of typical rates and credit scores for licensed installment lenders). 90 for $2,000 or less and 85 percent of loans had durations of 24 months or less (60 percent of loans had durations of one year or less).269 The survey did not report an average loan amount. Almost half of the loans had APRs between 49 and 99 percent; 9 percent of loans of $501 or less had APRs between 100 and 199 percent, but there was substantial rate variation among States.270 Except for loans subject to the Military Loan Act described above, APR calculations under Regulation Z include origination fees, but lenders generally are not required to include within the APR costs such as application fees and add-on services such as optional credit insurance and guaranteed automobile protection.271 A wider range and number of such up-front fees and addon products and services appear to be charged by the storefront lenders than by their newer online counterparts. Finance companies typically engage in underwriting that includes a monthly net income and expense budget, a review of the consumer’s credit report, and an assessment of monthly cash flow.272 One trade association representing traditional finance companies has described the 269 Thomas A. Durkin et al., “Findings from the AFSA Member Survey of Installment Lending,” at 24 tbl. 3 (2014), available at http://www.masonlec.org/site/rte_uploads/files/Manne/11.21.14%20JLEP%20Consumer%20Credit%20and%20the %20American%20Economy/Findings%20from%20the%20AFSA%20Member%20Survey%20of%20Installment%20 Lending.pdf. It appears that lenders made loans in at least 27 States, but the majority of loans were from 10 States. Id. at 28 tbl. 9. 270 Thomas A. Durkin et al., “Findings from the AFSA Member Survey of Installment Lending,” at 24 tbl. 3 (2014), available at http://www.masonlec.org/site/rte_uploads/files/Manne/11.21.14%20JLEP%20Consumer%20Credit%20and%20the %20American%20Economy/Findings%20from%20the%20AFSA%20Member%20Survey%20of%20Installment%20 Lending.pdf. It appears that lenders made loans in at least 27 States, but the majority of loans were from 10 States. Id. at 28 tbl. 9 & n.1. 271 12 CFR 1026.4(a) to (d). 272 See American Fin. Servs. Ass’n, “Traditional Installment Loans, Still the Safest and Most Affordable Small Dollar Credit,” available at https://www.afsaonline.org/Portals/0/Federal/White%20Papers/Small%20Dollar%20Credit%20TP.pdf; Sun Loan Company, “Loan FAQs,” http://www.sunloan.com/faq/ (last visited Sept. 23, 2017) (“Yes, we do check your credit report when you complete an application for a Sun Loan Company, but we do not base our approval on your score. 91 underwriting process as evaluating the borrower’s “stability, ability, and willingness” to repay the loan.273 Many finance companies report loan payment history to one or more of the nationwide consumer reporting agencies,274 and the Bureau believes from market outreach that these lenders generally furnish payment information on a monthly basis. With regard to newer online counterparts, the Bureau is aware from its market monitoring activities that some online installment lenders in this market offer products that resemble the types of loans made by finance companies. Many of these online installment lenders engage in highly-automated underwriting that involves substantial use of analytics and technology. The APRs on the loans are over 36 percent and can reach the triple digits.275 Finance companies and online installment lenders offer various methods for consumers to repay their loans. Particularly for online loans, repayment through ACH is common. 276 Some online installment lenders also allow other repayment methods, such as check, debit or credit card, MoneyGram, or Western Union, but may require advance notice for some of these payment methods.277 From its market monitoring functions, the Bureau is aware that finance companies with storefront locations tend to offer a wider array of repayment options. Some finance companies will accept ACH payments in person, set up either during the loan closing process or Your ability, stability and willingness to repay the loan are the most important things we check when making a decision.”). 273 Nat’l Installment Lenders Ass’n, “Best Practices,” http://nilaonline.org/best-practices/ (last visited Apr. 29, 2016). 274 American Fin. Servs. Ass’n, “Traditional Installment Loans, Still the Safest and Most Affordable Small Dollar Credit,” available at https://www.afsaonline.org/Portals/0/Federal/White%20Papers/Small%20Dollar%20Credit%20TP.pdf. 275 APRs on Elevate’s Rise loans can reach 299 percent, APRs on LendUp’s loans can reach about 256 percent, and APRs on Enova’s loans originated through its NetCredit platform can reach 179 percent. Rise, “What it Costs,” https://www.risecredit.com/how-online-loans-work#WhatItCosts (last visited Sept. 17, 2017); LendUp, “Rates & Notices,” https://www.lendup.com/rates-and-notices (last visited Sept. 17, 2017); Enova, “Investor Presentation,” at 7 (May 8, 2017), available at http://ir.enova.com/download/Enova+Investor+Presentation+v5+%28as+of+May+5+2017%29.pdf. 92 at a later date, or by phone.278 Finance companies also traditionally take payments in-store, generally by cash or check, or by mail. Some finance companies charge consumers a fee to use certain payment methods.279 D. Initiating Payment from Consumers’ Accounts As discussed above, payday and payday installment lenders nearly universally obtain at origination one or more authorizations to initiate withdrawal of payment from the consumer’s account. There are a variety of payment options or channels that they use to accomplish this goal, and lenders frequently obtain authorizations for multiple types. Different payment channels are subject to different laws and, in some cases, private network rules, leaving lenders with broad control over the parameters of how a particular payment will be pulled from a consumer’s account, including the date, amount, and payment method. Obtaining Payment Authorization A variety of payment methods enable lenders to use a previously-obtained authorization to initiate a withdrawal from a consumer’s account without further action from the consumer. These methods include paper signature checks, remotely created checks (RCCs) and remotely 276 See, e.g., Elevate, 2017 S–1, at 22; Rise, “Frequently Asked Questions About Rise Loans,” https://www.risecredit.com/frequently-asked-questions (last visited Sept. 23, 2017); Enova, 2016 Annual Report (10–K), at 25. 277 See, e.g., NetCredit, “Frequently Asked Questions: How Can I Repay My Personal Loan,” https://www.netcredit.com/faq (last visited Sept. 17, 2017); Rise, “Frequently Asked Questions About Rise Loans,” https://www.risecredit.com/frequently-asked-questions (last visited Sept. 17, 2017). 278 See Republic Finance, “Payments,” http://republicfinance.com/payment (last visited Sept. 17, 2017). 279 See One lender’s website notes (“Republic Finance has arrangements with a payment processor, PaymentVision, to accept payments from our customers either by phone or online as further described below. By using this service, you contract directly with the payment processor, PaymentVision. If permitted by State law, the payment processor charges a fee for their service. Republic Finance does not receive any portion of that fee.”). Republic Finance, “Payments by Phone (Interactive Voice Response) or Online Payments through Payment Processor,” http://republicfinance.com/payment (last visited Sept. 17, 2017). 93 created payment orders (RCPOs),280 and electronic payments like ACH281 and debit and prepaid card transactions. Payday and payday installment lenders—both online and in storefronts— typically obtain a post-dated check or electronic payment authorization from consumers for repayments of loans.282 For storefront payday loans, lenders typically obtain a post-dated check (or, where payday installment products are authorized, a series of postdated checks) that they can use to initiate a check or ACH transaction from a consumer’s account. For an online loan, a consumer often provides bank account information to receive the loan funds, and the lender often uses that bank account information to obtain payment from the consumer.283 This account 280 A RCC or RCPO is a type of check that is created by the payee—in this case, it would be created by the lender— and processed through the check clearing system. Given that the check is created by the lender, it does not bear the consumer’s signature. See Regulation CC, 12 CFR 229.2(fff) (defining remotely created check); Telemarketing Sales Rule, 16 CFR 310(cc) (defining “remotely created payment order” as a payment instrument that includes remotely created checks). 281 In order to initiate an ACH payment from a consumer’s account, a lender must send a request (also known as an “entry”) through an originating depository financial institution (ODFI). An ODFI is a bank or other financial institution with which the lender or the lender’s payment processor has a relationship. ODFIs aggregate and submit batches of entries for all of their originators to an ACH operator. The ACH operators sort the ACH entries and send them to the receiving depository financial institutions (RDFI) that hold the individual consumer accounts. The RDFI then decides whether to debit the consumer’s account or to send it back unpaid. ACH debit transactions generally clear and settle in one business day after the payment is initiated by the lender. The private operating rules for the ACH network are administered by the National Automated Clearinghouse Association (NACHA), an industry trade organization. 282 See, e.g., QC Holdings, Inc., 2014 Annual Report (Form 10–K), at 6 (Mar. 12, 2015) (“Upon completion of a loan application, the customer signs a promissory note with a maturity of generally two to three weeks. The loan is collateralized by a check (for the principal amount of the loan plus a specified fee), ACH authorization or a debit card.”); Advance America, 2011 Annual Report (Form 10–K) at 45 (Mar. 15, 2012) (“After the required documents presented by the customer have been reviewed for completeness and accuracy, copied for record-keeping purposes, and the cash advance has been approved, the customer enters into an agreement governing the terms of the cash advance. The customer then provides a personal check or an Automated Clearing House (“ACH”) authorization, which enables electronic payment from the customer's account, to cover the amount of the cash advance and charges for applicable fees and interest of the balance due under the agreement.”); ENOVA Int’l, Inc., 2014 Annual Report (Form 10–K), at 6 (Mar. 20, 2015) )(“When a customer takes out a new loan, loan proceeds are promptly deposited in the customer’s bank account or onto a debit card in exchange for a preauthorized debit for repayment of the loan from the customer’s account.”). 283 See, e.g., Great Plains Lending d/b/a Cash Advance Now “Frequently Asked Questions (FAQs),” https://www.cashadvancenow.com/FAQ.aspx (last visited May 16, 2016) (“If we extend credit to a consumer, we will consider the bank account information provided by the consumer as eligible for us to process payments against. In addition, as part of our information collection process, we may detect additional bank accounts under the ownership of the consumer. We will consider these additional accounts to be part of the application process.”). 94 information can be used to initiate an ACH payment from a consumer’s account. Typically, online lenders require consumers to authorize payments from their account as part of their agreement to receive the loan proceeds electronically.284 Some traditional installment lenders also obtain an electronic payment authorization from their customers. Payday and payday installment lenders often take authorization for multiple payment methods, such as taking a post-dated check along with the consumer’s debit card information.285 Consumers usually provide the payment authorization as part of the loan origination process.286 For storefront payday loans, providing a post-dated check is typically a requirement to obtain a loan. Under the Electronic Fund Transfer Act (EFTA) lenders cannot condition credit on obtaining an authorization from the consumer for “preauthorized” (recurring) electronic fund transfers,287 but in practice online payday and payday installment lenders are able to obtain such authorizations from consumers for almost all loans. The EFTA provision concerning 284 See, e.g., Notice of Motion and Motion to Compel Arbitration at exhibit 1, 38, 55, Labajo v. First Int’l Bank & Trust, No. 14–00627 (C.D. Cal. May 23, 2014), ECF No. 26–3. 285 See, e.g., Memorandum of Law in Support of Motion to Dismiss for Failure to State a Claim at exhibit A, Parm v. BMO Harris Bank, N.A., No. 13–03326 (N.D. Ga. Dec. 23, 2013), ECF No. 60–1 (“You may revoke this authorization by contacting us in writing at ach@castlepayday.com or by phone at 1–888–945–2727. You must contact us at least three (3) business days prior to when you wish the authorization to terminate. If you revoke your authorization, you authorize us to make your payments by remotely-created checks as set forth below.”); Declaration re: Motion to Compel Arbitration at exhibit 5, Booth v. BMO Harris Bank, N.A., No. 13–5968 (E.D. Pa. Dec. 13, 2013), ECF No. 41–8 (stating that in the event that the consumer terminates an ACH authorization, the lender would be authorized to initiated payment by remotely created check); Notice of Motion and Motion to Compel Arbitration at exhibit A, Labajo v. First Int’l Bank & Trust, No. 14–00627 (C.D. Cal. May 23, 2014), ECF No. 25–1 (taking ACH and remotely created check authorization). 286 See, e.g., Advance America, 2011 Annual Report (Form 10–K), at 10 (“To obtain a cash advance, a customer typically … enters into an agreement governing the terms of the cash advance, including the customer's agreement to repay the amount advanced in full on or before a specified due date (usually the customer's next payday), and our agreement to defer the presentment or deposit of the customer's check or ACH authorization until the due date.”). 287 EFTA and its implementing regulation, Regulation E, prohibit the conditioning of credit on an authorization for a preauthorized recurring electronic fund transfer. See 12 CFR 1005.10(e)(1) (“No financial institution or other person may condition an extension of credit to a consumer on the consumer's repayment by preauthorized electronic fund transfers, except for credit extended under an overdraft credit plan or extended to maintain a specified minimum balance in the consumer's account.”). 95 compulsory use does not apply to paper checks and one-time electronic fund transfers. Moreover, even for loans subject to the EFTA compulsory use provision, lenders use various methods to obtain electronic authorizations. For example, although some payday and payday installment lenders provide consumers with alternative methods to repay loans, these options may be burdensome and may significantly change the terms of the loan. For example, one lender increases its APR by an additional 61 percent or 260 percent, depending on the length of the loan, if a consumer elects a cash-only payment option for its installment loan product, resulting in a total APR of 462 percent (210 day loan) to 780 percent (140 day loan).288 Other lenders change the origination process if consumers do not immediately provide account access. For example, some online payday lenders require prospective customers to contact them by phone if they do not want to provide a payment authorization and wish to pay by money order or check at a later time. Other lenders delay the disbursement of the loan proceeds if the consumer does not immediately provide a payment authorization.289 Banks and credit unions have additional payment channel options when they lend to consumers who have a deposit account at the same institution. As a condition of certain types of loans, many financial institutions require consumers to have a deposit account at that same institution. 290 The loan contract often authorizes the financial institution to pull payment directly 288 Cash Store, “Installment Loans: Fee Schedule Examples,” https://www.cashstore.com//media/cashstore/files/pdfs/nm%20ins%20552014.pdf (last visited May 16, 2016). 289 See, e.g., Mobiloans, “Line of Credit Terms and Conditions,” www.mobiloans.com/terms-and-conditions (last visited Feb. 5, 2016) (“If you do not authorize electronic payments from your Demand Deposit Account and instead elect to make payments by mail, you will receive your Mobiloans Cash by check in the mail.”). 290 See, e.g., Fifth Third Bank, “Ways to Borrow Money for Your Unique Needs,” https://www.53.com/content/fifththird/en/personal-banking/borrowing-basics/personal-loans.html (last visited May 17, 2016), at 3 (last visited May 17, 2016), available at https://www.53.com/doc/pe/pe-eax-tc.pdf (providing eligibility requirements including that 96 from the consumer’s account. Since these payments can be processed through an internal transfer within the bank or credit union, these institutions do not typically use external payment channels to complete an internal payment transfer. Exercising Payment Authorizations For different types of loans that will be covered under the rule, lenders use their authorizations to collect payment differently. As discussed above, most storefront lenders encourage or require consumers to return to their stores to pay in cash, roll over, or otherwise renew their loans. The lender often will deposit a post-dated check or initiate an electronic fund transfer only where the lender considers the consumer to be in “default” under the contract or where the consumer has not responded to the lender’s communications.291 Bureau examiners have cited one or more payday lenders for threatening to initiate payments from consumer accounts that were contrary to the agreement, and that the lenders did not intend to initiate.292 In contrast, online lenders typically use the authorization to collect all payments, not just those initiated after there has been some indication of distress from the consumer. Moreover, as discussed above, online lenders offering “hybrid” payday loan products structure them so that the lender is authorized to collect a series of interest-only payments—the functional equivalent of paying finance charges to roll over the loan—before full payment or amortizing payments are the consumer “must have a Fifth Third Bank checking deposit account that has been open for the past 90 (ninety) days and is in good standing”). 291 Payday and payday installment lenders may contact consumers a few days before the payment is due to remind them of their upcoming payment. This is a common practice, with many lenders calling the consumer 1 to 3 days before the payment is due, and some providing reminders through text or email. 292 Bureau of Consumer Fin. Prot., “Supervisory Highlights,” at 20 (Spring 2014), available at http://files.consumerfinance.gov/f/201405_cfpb_supervisory-highlights-spring-2014.pdf. 97 due.293 The Bureau also is aware that some online lenders, although structuring their product as nominally a two-week loan, automatically roll over the loan every two weeks unless the consumer takes affirmative action to make full payment.294 The payments processed in such cases are for the cost of the rollover rather than the full balance due. As a result of these distinctions, storefront and online lenders have different success rates in exercising such payment authorizations. Some large storefront lenders report that they initiate payment attempts in less than 10 percent of cases, and that 60 to 80 percent of those attempts are returned for non-sufficient funds.295 Bureau analysis of ACH payments by online payday and payday installment lenders, which typically collect all payments by initiating a transfer from consumers’ accounts, indicates that for any given payment only about 6 percent fail on the first try. However, over an eighteen-month observation period, 50% of online borrowers were found to experience at least one payment attempt that failed or caused an overdraft and one-third of the borrowers experienced more than one such incident. 293 See, e.g., Notice of Charges Seeking Restitution, Digorgement, Other Equitable Relief, and Civil Money Penalties, In the Matter of: Integrity Advance, LLC, No. 2015–CFPB–0029, at 5 (Nov. 18, 2015), available at http://files.consumerfinance.gov/f/201511_cfpb_notice-of-charges-integrity-advance-llc-james-r-carnes.pdf (providing lender contract for loan beginning with four automatic interest-only rollover payments before converting to a series of amortizing payments). 294 See, e.g., Motion to Compel Arbitration, Motion to Stay Litigation at exhibit A, Riley v. BMO Harris Bank, N.A., No. 13–1677 (D.D.C. Jan. 10, 2014), ECF No. 33–2 (interpreting silence from consumer before the payment due date as a request for a loan extension; contract was for a 14-day single-payment loan, loan amount financed was $700 for a total payment due of $875). 295 One major lender with a predominantly storefront loan portfolio, QC Holdings, notes that in 2014, 91.5 percent of its payday and installment loans were repaid or renewed in cash. QC Holdings 2014 Annual Report (Form 10– K), at 7. For the remaining 8.5 percent of loans for which QC Holdings initiated a payment attempt, 78.5 percent were returned due to non-sufficient funds. Id. Advance America, which offers mostly storefront payday and installment loans, initiated check or ACH payments on approximately 6.7 and 6.5 percent, respectively, of its loans in 2011; approximately 63 and 64 percent, respectively, of those attempts failed. Advance America 2011 Annual Report (Form 10–K), at 27. 98 Lenders typically charge fees for these returned payments, sometimes charging both a returned payment fee and a late fee.296 These fees are in addition to fees, such as NSF fees, that may be charged by the financial institution that holds the consumer’s account. The Bureau found that if an electronic payment attempt failed, online lenders try again three-quarters of the time. However, after an initial failure the lender’s likelihood of failure jumps to 70 percent for the second attempt and 73 percent for the third. Of those that succeed, roughly one-third result in an overdraft. Both storefront and online lenders also frequently change the ways in which they attempt to exercise authorizations after one attempt has failed. For example, many typically make additional attempts to collect initial payment due.297 Some lenders attempt to collect the entire payment amount once or twice within a few weeks of the initial failure. The Bureau, however, is aware of online and storefront lenders that use more aggressive and unpredictable payment 296 See Advance America 2011 Annual Report (Form 10–K), at 8 (“We may charge and collect fees for returned checks, late fees, and other fees as permitted by applicable law. Fees for returned checks or electronic debits that are declined for non-sufficient funds (“NSF”) vary by State and range up to $30, and late fees vary by State and range up to $50. For each of the years ended December 31, 2011 and 2010, total NSF fees collected were approximately $2.9 million and total late fees collected were approximately $1 million and $0.9 million, respectively.”); Mypaydayloan.com, “Frequently Asked Questions,” https://www.mypaydayloan.com/faq#loancost (last visited May 17, 2016) (“If your payment is returned due to NSF (or Account Frozen or Account Closed), our collections department will contact you to arrange a second attempt to debit the payment. A return item fee of $25 and a late fee of $50 will also be collected with the next debit.”). 297 See Bureau of Consumer Fin. Prot., “Supervisory Highlights,” at 20 (Spring 2014), available at http://files.consumerfinance.gov/f/201405_cfpb_supervisory-highlights-spring-2014.pdf (“Upon a borrower’s default, payday lenders frequently will initiate one or more preauthorized ACH transactions pursuant to the loan agreement for repayment from the borrower’s checking account.”); FirstCash Fin. Servs., Inc. 2014 Annual Report (Form10–K) at 5 (Feb. 12, 2015) (“Banks return a significant number of ACH transactions and customer checks deposited into the Independent Lender’s account due to insufficient funds in the customers’ accounts. The Company subsequently collects a large percentage of these bad debts by redepositing the customers’ checks, ACH collections or receiving subsequent cash repayments by the customers.”); Advance America, “FAQs on Payday Loans/Cash Advances,” https://www.advanceamerica.net/questions/payday-loans-cash-advances (last visited Sept. 17, 2017) (“Once we present your bank with your ACH authorization for payment, your bank will send the specified amount to CashNetUSA. If the payment is returned because of insufficient funds, CashNetUSA can and will re-present the ACH Authorization to your bank.”). 99 collection practices, including breaking payments into multiple smaller payments and attempting to collect payment multiple times in one day or over a short period of time.298 The cost to lenders to repeatedly attempt payment depends on their contracts with payment processors and commercial banks, but is generally nominal; the Bureau estimates the cost is in a range of 5 to 15 cents for an ACH transaction.299 These practices are discussed in more detail in Market Concerns—Payments. As noted above, banks and credit unions that lend to their account holders can use their internal system to transfer funds from the consumer accounts and do not need to utilize the payment networks. Deposit advance products and their payment structures are discussed further in part II.B. The Bureau believes that many small-dollar loans with depository institutions are paid through internal transfers. Due to the fact that lenders obtain authorizations to use multiple payment channels and benefit from flexibility in the underlying payment systems, lenders generally enjoy broad discretion over the parameters of how a particular payment will be pulled from a consumer’s account, including the date, amount, and payment method. For example, although a check specifies a date, lenders may not present the check on that date. Under UCC section 4-401, merchants can present checks for payment even if the check specifies a later date.300 Lenders 298 See generally CFPB Online Payday Loan Payments. The Bureau reviewed publicly available litigation documents and fee schedules posted online by originating depository institutions to compile these estimates. However, because of the limited availability of private contracts and variability of commercial bank fees, these estimates are tentative. Originators typically also pay their commercial bank or payment processor fees for returned ACH and check payments. These fees appear to range widely, from 5 cents to several dollars. 300 UCC section 4–401(c)(“A bank may charge against the account of a customer a check that is otherwise properly payable from the account, even though payment was made before the date of the check, unless the customer has given notice to the bank of the postdating describing the check with reasonable certainty.”). 299 100 sometimes attempt to collect payment on a different date from the one stated on a check or original authorization. They may shift the attempt date in order to maximize the likelihood that funds will be in the account; some use their own models to determine when to collect, while others use predictive payment products provided by third parties that estimate when funds are most likely to be in the account.301 Moreover, the checks provided by consumers during origination often are not processed as checks. Rather than sending these payments through the check clearing network, lenders often process these payments through the ACH network. They are able to use the consumer account number and routing number on a check to initiate an ACH transaction. When lenders use the ACH network in a first attempt to collect payment, the lender has used the check as a source document and the payment is considered an electronic fund transfer under EFTA and Regulation E,302 which generally provide additional consumer protections—such as error resolution rights—beyond those applicable to checks. However, if a transaction is initially processed through the check system and then processed through the ACH network because the first attempt failed for insufficient funds, the subsequent ACH attempt is not considered an electronic fund transfer under current Regulation E.303 Similarly, consumers may provide their 301 See, e.g., Press Release, Clarity Servs., Inc., “ACH Presentment Will Help Lenders Reduce Failed ACH Pulls” (Aug. 1, 2013), available at https://www.clarityservices.com/clear-warning-ach-presentment-will-help-lendersreduce-failed-ach-pulls/; FactorTrust, “Service Offerings,” http://ws.factortrust.com/products/ (last visited May 4, 2016); Microbilt, “Bank Account Verify,” http://www.microbilt.com/bank-account-verification.aspx (last visited May 4, 2016); DataX, “Credit Risk Mitigation,” http://www.dataxltd.com/ancillary-services/successful-collections/ (last visited May 4, 2016). 302 12 CFR 1005.3(b)(2)(i) (“This part applies where a check, draft, or similar paper instrument is used as a source of information to initiate a one-time electronic fund transfer from a consumer's account. The consumer must authorize the transfer.”). 303 Supplement I, Official Staff Interpretations, 12 CFR 1005.3(c)(1) (“The electronic re-presentment of a returned check is not covered by Regulation E because the transaction originated by check.”). 101 account and routing number to lenders for the purposes of an ACH payment, but the lender may use that information to initiate a remotely created check that is processed through the check system and thus may not receive Regulation E protections.304 Payment System Regulation and Private Network Requirements Different payment mechanisms are subject to different laws and, in some cases, private network rules that affect how lenders can exercise their rights to initiate withdrawals from consumers’ accounts and how consumers may attempt to limit or stop certain withdrawal activity after granting an initial authorization. Because ACH payments and post-dated checks are the most common authorization mechanisms used by payday and payday installment lenders, this section briefly outlines applicable Federal laws and National Automated Clearinghouse Association (NACHA) rules concerning stop-payment rights, prohibitions on unauthorized payments, notices where payment amounts vary, and rules governing failed withdrawal attempts. NACHA recently adopted several changes to the ACH network rules in response to complaints about problematic behavior by payday and payday installment lenders, including a rule that allows it to more closely scrutinize originators who have a high rate of returned payments.305 Issues around monitoring and enforcing those rules and their application to 304 Remotely created checks are particularly risky for consumers because they have been considered to fall outside of protections for electronic fund transfers under Regulation E. Also, unlike signature paper checks, they are created by the entity seeking payment (in this case, the lender)—making such payments particularly difficult to track and reverse in cases of error or fraud. Due to concerns about remotely created checks and remotely created payment orders, the FTC recently banned the use of these payment methods by telemarketers. See FTC Final Amendments to Telemarketing Sales Rule, 80 FR 77520 (Dec. 14, 2015). 305 See NACHA, “ACH Network Risk and Enforcement Topics,” https://www.nacha.org/rules/ach-network-riskand-enforcement-topics (last visited Sept. 23, 2017) (providing an overview of changes to the NACHA Rules); NACHA, “ACH Operations Bulletin #1–2014: Questionable ACH Debit Origination: Roles and Responsibilities of ODFIs and RDFIs” (Sept. 30, 2014), available at https://www.nacha.org/news/ach-operations-bulletin-1-2014questionable-ach-debit-origination-roles-and-responsibilities (“During 2013, the ACH Network and its financial 102 problems in the market for covered loans are discussed in more detail in Market Concerns— Payments. But it should be noted here at the outset that the NACHA rules only apply to payment attempts through ACH and are not enforceable by the Bureau. Stop-payment rights. For preauthorized (recurring) electronic fund transfers,306 EFTA grants consumers a right to stop payment by issuing a stop-payment order through their depository institution.307 The NACHA private rules adopt this EFTA provision along with additional stop-payment rights. In contrast to EFTA, NACHA provides consumers with a stoppayment right for both one-time and preauthorized transfers.308 Specifically, for recurring transfers, NACHA Rules require financial institutions to honor a stop-payment order as long as the consumer notifies the bank at least 3 banking days before the scheduled debit.309 For onetime transfers, NACHA Rules require financial institutions to honor the stop-payment order as long as the notification provides them with a “reasonable opportunity to act upon the order.”310 Consumers may notify the bank or credit union verbally or in writing, but if the consumer does not provide written confirmation the oral stop-payment order may not be binding beyond 14 days. If a consumer wishes to stop all future payments from an originator, NACHA Rules allow institution participants came under scrutiny as a result of the origination practices of certain businesses, such as online payday lenders, in using the ACH Network to debit consumers’ accounts.”). 306 A preauthorized transfer is “an electronic fund transfer authorized in advance to recur at substantially regular intervals. EFTA, 15 U.S.C. 1693a(10); Regulation E, 12 CFR 1005.2(k). 307 “A consumer may stop payment of a preauthorized electronic fund transfer by notifying the financial institution orally or in writing at any time up to three business days preceding the scheduled date of such transfer.” EFTA, 15 U.S.C. 1693e(a); Regulation E, 12 CFR 1005.10(c). 308 See NACHA Rule 3.7.1.2, RDFI Obligation to Stop Payment of Single Entries (“An RDFI must honor a stoppayment order provided by a Receiver, either verbally or in writing, to the RDFI at such time and in such manner as to allow the RDFI a reasonable opportunity to act upon the order prior to acting on an ARC, BOC, POP, or RCK Entry, or a Single Entry IAT, PPD, TEL, or WEB Entry to a Consumer Account.”). 309 NACHA Rule 3.7.1.1. 310 NACHA Rule 3.7.1.2. 103 a bank or credit union to require the consumer to confirm in writing that she has revoked authorization from the originator. Checks are also subject to a stop-payment right under the Uniform Commercial Code (UCC).311 Consumers have a right to stop payment on any check by providing the bank with oral (valid for 14 days) or written (valid for 6 months) notice. To be effective, the stop-payment notice must describe the check “with reasonable certainty” and give the bank enough information to find the check under the technology then existing.312 The stop-payment notice also must be given at a time that affords the bank a reasonable opportunity to act on it before the bank becomes liable for the check under U.C.C. 4-303. Although EFTA, the UCC, and NACHA Rules provide consumers with stop-payment rights, financial institutions typically charge a fee of approximately $32 for consumers to exercise those rights.313 Further, both lenders and financial institutions often impose a variety of requirements that make the process for stopping payments confusing and burdensome for consumers. See the discussion of these requirements in Market Concerns—Payments. Protection from unauthorized payments. Regulation E and NACHA Rules both provide protections with respect to payments by a consumer’s financial institution if the electronic transfer is unauthorized.314 Payments originally authorized by the consumer can become unauthorized under EFTA if the consumer notifies his or her financial institution that the 311 U.C.C. 4–403. U.C.C. 4–403 cmt. 5. 313 Median stop-payment fee for an individual stop-payment order charged by the 50 largest financial institutions in 2015 based on information in the Informa Research Database. See Research Srvs, Inc., “Informa Research Database,” www.informars.com (last visited Mar. 2016). Although information has been obtained from the various financial institutions, the accuracy cannot be guaranteed. 314 NACHA Rule 2.3.1, General Rule, Originator Must Obtain Authorization from Receiver. 312 104 originator’s authorization has been revoked.315 NACHA has a specific threshold for unauthorized returns, which involve transactions that originally collected funds from a consumer’s account but that the consumer is disputing as unauthorized. Under NACHA Rules, originators are required to operate with an unauthorized return rate below 0.5 percent or they risk fines and loss of access to the ACH network.316 Notice of variable amounts. Regulation E and the NACHA Rules both provide that if the debit amount for a preauthorized transfer changes from the previous transfer or from the preauthorized amount, consumers must receive a notice 10 calendar days prior to the debit.317 However, both of these rules have an exception from this requirement if consumers have agreed to a range of debit amounts and the payment does not fall outside that range.318 Based on outreach and market research, the Bureau does not believe that most payday and payday installment lenders making loans that will be covered under the rule are providing a notice of transfers varying in amount. However, the Bureau is aware that many of these lenders take authorizations for a range of amounts. As a result, lenders use these broad authorizations 315 EFTA, 15 U.S.C. 1693a(12) (“The term ‘unauthorized electronic fund transfer’ means an electronic fund transfer from a consumer's account initiated by a person other than the consumer without actual authority to initiate such transfer and from which the consumer receives no benefit, but the term does not include any electronic fund transfer (A) initiated by a person other than the consumer who was furnished with the card, code, or other means of access to such consumer's account by such consumer, unless the consumer has notified the financial institution involved that transfers by such other person are no longer authorized… .”). Regulation E implements this provision at 12 CFR 1005.2(m). 316 NACHA Rule 2.17.2. 317 12 CFR 1005.10(d)(1) (“When a preauthorized electronic fund transfer from the consumer's account will vary in amount from the previous transfer under the same authorization or from the preauthorized amount, the designated payee or the financial institution shall send the consumer written notice of the amount and date of the transfer at least 10 days before the scheduled date of transfer.”); NACHA Rule 2.3.2.6(a). 318 12 CFR 1005.10(d)(2) (“The designated payee or the institution shall inform the consumer of the right to receive notice of all varying transfers, but may give the consumer the option of receiving notice only when a transfer falls outside a specified range of amounts or only when a transfer differs from the most recent transfer by more than an agreed-upon amount.”); NACHA Rule 2.3.2.6(b). 105 rather than fall under the Regulation E requirement to send a notice of transfers varying in amount even when collecting for an irregular amount (for example, by adding fees or a past due amount to a regularly scheduled payment). Some of these contracts provide that the consumer is authorizing the lender to initiate payment for any amount up to the full amount due on the loan.319 Reinitiation Cap. After a payment attempt has failed, NACHA Rules allow an originator—in this case, the lender that is trying to collect payment—to attempt to collect that same payment no more than two additional times through the ACH network.320 NACHA Rules also require the ACH files321 for the two additional attempts to be labeled as “reinitiated” transactions. Because the rule applies on a per-payment basis, for lenders with recurring payment authorizations, the count resets to zero when the next scheduled payment comes due. III. Summary of the Rulemaking Process As described in more detail below, the Bureau has conducted broad outreach with a multitude of stakeholders on a consistent basis over more than five years to learn more about the market for small-dollar loans of various kinds. This outreach has comprised many public events, 319 For example, a 2013 One Click Cash Loan Contract states: The range of ACH debit entries will be from the amount applied to finance charge for the payment due on the payment date as detailed in the repayment schedule in your loan agreement to an amount equal to the entire balance due and payable if you default on your loan agreement, plus a return item fee you may owe as explained in your loan agreement. You further authorize us to vary the amount of any ACH debit entry we may initiate to your account as needed to pay the payment due on the payment date as detailed in the repayment schedule in your loan agreement as modified by any prepayment arrangements you may make, any modifications you and we agree to regarding your loan agreement, or to pay any return item fee you may owe as explained in your loan agreement.”); Notice of Motion and Motion to Compel Arbitration at exhibit 1, 38, 55, Labajo v. First Int’l Bank & Trust, No. 14–00627 (C.D. Cal. May 23, 2014), ECF No. 26–3. (SFS Inc., d/b/a One Click Cash, Authorization to Initiate ACH Debit and Credit Entries). 320 NACHA Rule 2.12.4. 321 ACH transactions are transferred in a standardized electronic file format between financial institutions and ACH network operators. These files contain information about the payment itself along with routing information for the applicable consumer account, originator (or in this case, the lender) account, and financial institution. 106 including field hearings, and hundreds of meetings with both consumer and industry stakeholders on the issues raised by small-dollar lending. In addition to meeting with lenders and other market participants, trade associations, consumer groups, community groups, and others, the Bureau has engaged with individual faith leaders and coalitions of faith leaders from around the country to gain their perspective on how these loans affect their communities and the people they serve. And the Bureau has met frequently with Federal, State, and Tribal officials to consult and share information about these kinds of loans and their consequences for consumers. The Bureau’s understanding of these loans, and how they affect consumers, has also been furthered by its ongoing supervisory activity, which involves exercising its legally mandated authority to conduct formal examinations of companies who make such loans and of debt collectors who collect on such loans. These examinations have canvassed the operations, marketing, underwriting, collections, and compliance management systems at such lenders and continue to do so on an ongoing basis. In addition, the Bureau has investigated and taken enforcement actions against a number of small-dollar lenders, which has provided further insight into various aspects of their operations and the practical effects of their business models on consumers. The Bureau has also undertaken extensive research and analysis over several years to develop the factual foundation for issuance of this final rule. That research and analysis has included multiple white papers and data points on millions of such loans,322 as well as careful 322 Bureau of Consumer Fin. Prot., “Payday Loans, Auto Title Loans, and High-Cost Installment Loans: Highlights from CFPB Research,” (June 2, 2016), available at http://files.consumerfinance.gov/f/documents/Payday_Loans_Highlights_From_CFPB_Research.pdf (summary of the CFPB’s independent research). 107 review of studies and reports prepared by others and the relevant academic literature.323 The Bureau has analyzed its own data on consumer complaints about the issues raised by small-dollar loans and the collections efforts made by lenders and debt collectors on such loans. And the Bureau has consistently engaged in market monitoring activities to gain insights into developing trends in the market for small-dollar loans. All of the input and feedback the Bureau has received from its outreach over the years, its extensive experience of examining and investigating small-dollar lenders, and its research and analysis of the marketplace, have assisted the Bureau in developing and issuing this final rule. The material presented in this section summarizes the Bureau’s work relating to the rule in three categories: • the Bureau’s background and processes in developing the rule; • the key elements of the notice of proposed rulemaking; and • the receipt and consideration of feedback prior to finalizing the rule. A. Bureau Outreach to Stakeholders Birmingham Field Hearing. The Bureau’s formal outreach efforts on this subject began in January 2012, when it held its first public field hearing in Birmingham, Alabama, focused on small-dollar lending. At the field hearing, the Bureau heard testimony and received input from consumers, civil rights groups, consumer advocates, religious leaders, industry and trade association representatives, academics, and elected representatives and other governmental officials about consumers’ experiences with small-dollar loan products. At the same time, the Bureau announced the launch of its program to conduct supervisory examinations of payday 323 See part VII and the Section 1022(b)(2) Analysis for more on the relevant academic literature. 108 lenders pursuant to the Bureau’s authority under section 1024 of the Dodd-Frank Act. As part of this initiative, the Bureau put in place a process to obtain loan-level records from a number of large payday lenders to assist in analyzing the nature and effects of such loans. The Bureau transcribed the field hearing and posted the transcript on its website.324 Concurrently, the Bureau placed a notice in the Federal Register inviting public comment on the issues discussed in the field hearing. The Bureau received 664 public comments in response to that request, which were reviewed and analyzed. Nashville Field Hearing. In March 2014, the Bureau held a field hearing in Nashville, Tennessee to gather further input from a broad range of stakeholders.325 The Bureau heard testimony from consumer groups, industry representatives, academics, and members of the public, including consumers of payday loans. The field hearing was held in conjunction with issuing the second of two research reports on findings by Bureau staff using the supervisory data that it had collected from a number of large payday lenders. In the Director’s opening remarks, he noted three concerns associated with covered loans that had been identified in recent Bureau research: that a significant population of consumers were ending up in extended loan sequences; that some lenders use the electronic payments system in ways that pose risks to consumers; and that a troubling number of companies engage in collection activities that may be unfair or deceptive in one or more ways. While the Bureau was working on these reports and in the 324 Bureau of Consumer Fin. Prot., “In the Matter Of: A Field Hearing on Payday Lending, Hearing Transcript,” (Jan. 19, 2012), available at http://files.consumerfinance.gov/f/201201_cfpb_transcript_payday-lending-fieldhearing-alabama.pdf. 325 Bureau of Consumer Fin. Prot., “Live from Nashville - Field Hearing on Payday Loans,” CFPB Blog (Mar. 25, 2014), available at https://www.consumerfinance.gov/about-us/blog/live-from-nashville/. 109 period following their release, the Bureau held numerous meetings with stakeholders on smalldollar lending in general and to hear their views on potential policy approaches. Richmond Field Hearing. In March 2015, the Bureau held another field hearing in Richmond, Virginia to gather further input from a broad range of stakeholders.326 The focus of this field hearing was the announcement the Bureau simultaneously made of the rulemaking proposals it had under consideration that would require lenders to take steps to make sure consumers can repay their loans and would restrict certain methods of collecting payments from consumers’ bank accounts in ways that lead to substantial penalty fees. The Bureau heard testimony from consumer groups, industry representatives, faith leaders, and members of the public, including consumers of payday loans. In addition to the field hearing, the Bureau held separate roundtable discussions with consumer advocates and with industry members and trade associations to hear feedback on the rulemaking proposals under consideration. A summary of the rulemaking proposals under consideration was released at the time of the Richmond field hearing. This marked the first stage in the process the Bureau is required to follow under the Small Business Regulatory Enforcement and Fairness Act (SBREFA),327 which is discussed in more detail below. The summary was formally known as the Small Business Review Panel Outline. In addition to the discussions that occurred at the time of the Richmond field hearing, the Bureau has met on a number of other occasions with industry members and 326 Bureau of Consumer Fin. Prot., “Field Hearing on Payday Loans in Richmond, VA,” Archive of Past Events (Mar. 26, 2015), available at https://www.consumerfinance.gov/about-us/events/archive-past-events/field-hearingon-payday-lending/. 327 Public Law No. 104–1.21, 110 Stat. 847 (1996). 110 trade associations, including those representing storefront payday lenders, to discuss their feedback on the issues presented in the Outline. Omaha Meeting and Other Events. At the Bureau’s Consumer Advisory Board (CAB) meeting in June 2015 in Omaha, Nebraska, a number of meetings and field events were held about payday, vehicle title, and similar loans. The CAB advises and consults with the Bureau in the exercise of its functions under the Federal consumer financial laws, and provides information on emerging practices in the consumer financial products and services industry, including regional trends, concerns, and other relevant information. The CAB members over several years have included, among others, a payday lending executive and consumer advocates on payday lending. The Omaha events included a visit to a payday loan store to learn more about its operations first-hand and a day-long public session that focused on the Bureau’s proposals in the Small Business Review Panel Outline and trends in payday and vehicle title lending. The CAB also held six subcommittee discussions on the Outline in the spring and summer of 2015, and three more subcommittee discussions on the proposed rule in the summer of 2016. Kansas City Field Hearing. In June 2016, the Bureau held a field hearing in Kansas City, Missouri to gather further input on the issues surrounding potential new Federal regulations of small-dollar lending.328 The focus of this field hearing was the announcement that the Bureau simultaneously made of the release of its Notice of Proposed Rulemaking on Payday, Vehicle Title, and Certain High-Cost Installment Loans. The proposed rule would require lenders to take steps to make a reasonable determination that consumers can afford to repay their loans and 328 Bureau of Consumer Fin. Prot., “Field Hearing on Small Dollar Lending in Kansas City, MO,” Archive of Past Events (June 2, 2016), available at https://www.consumerfinance.gov/about-us/events/archive-past-events/fieldhearing-small-dollar-lending-kansas-city-mo/. 111 would restrict certain methods of collecting payments from consumers’ bank accounts in ways that can lead to substantial penalty fees. The Bureau heard testimony on the proposed rule from consumer groups, industry representatives, and members of the public, including consumers of payday loans. The release of the Notice of Proposed Rulemaking commenced the formal notice-andcomment process under the Administrative Procedure Act. In the Notice, the Bureau stated that comments on the proposed rule would have to be received on or before October 7, 2016 to be considered by the Bureau. The Notice further specified the details of the methods by which comments would be received, which included email, electronic, mail, and hand delivery/courier. The Bureau also noted that all comments submitted would become part of the public record and would be subject to public disclosure. Little Rock Meeting and Other Events. In June 2016, just a week after the field hearing in Kansas City announcing the public release of the proposed rule, the CAB held another public meeting on this topic in Little Rock, Arkansas. Among other things, Bureau officials gave a public briefing on the proposed rule to the CAB members, and the Bureau heard testimony from the general public on the subject. Two of the Bureau’s other advisory bodies have also provided input and feedback on the Bureau’s work to develop appropriate provisions to regulate small-dollar loans. The Community Bank Advisory Council (CBAC) held two subcommittee discussions of the proposals contained in the Small Business Review Panel Outline in March 2015 and November 2015, a Council discussion on the proposed rule in July 2016, and two more subcommittee discussions of the proposed rule in the summer of 2016. In addition, the Bureau’s Credit Union Advisory Council 112 (CUAC) held two subcommittee discussions of the proposals in April 2015 and October 2015, discussed the Outline in its full meeting in March 2016, and held two subcommittee discussions of the proposed rule during the summer of 2016. Faith Leaders. The Bureau has taken part in a large number of meetings with faith leaders, and coalitions of faith leaders, of all denominations to hear their perspective on how small-dollar loans affect their communities and the people they serve. In April 2016, the White House convened a meeting of national faith leaders for this purpose, which included the Bureau’s director. The Bureau has also engaged in outreach to local and national leaders from churches, synagogues, mosques, and temples—both in Washington, D.C. and in many locations around the country. In these sessions, the Bureau has heard from faith leaders about the challenges some of them have faced in seeking to develop alternatives to payday loans that would mitigate what they perceive to be the harms caused to consumers. General Outreach. Various Bureau leaders, including its director, and Bureau staff have participated in and spoken at dozens of events and conferences throughout the country, which have provided further opportunities to gather insight and recommendations from both industry and consumer groups about how to approach the issue of whether and how to regulate smalldollar loans. In addition to gathering information from meetings with lenders and trade associations and through regular supervisory and enforcement activities, Bureau staff made factfinding visits to at least 12 non-depository payday and vehicle title lenders. Inter-Agency Consultation. As discussed in connection with section 1022 of the DoddFrank Act below, the Bureau has consulted with other Federal consumer protection and prudential regulators about these issues and the approaches that the other regulators have taken to 113 small-dollar lending over the years. The Bureau has provided other regulators with information about the proposals under consideration, sought their input, and received feedback that has assisted the Bureau in preparing this final rule. In addition, the Bureau was involved, along with its fellow Federal regulatory agencies, in meetings and other efforts to assist the U.S. Department of Defense as it developed and adopted regulations to implement updates to the Military Lending Act. That statute governs small-dollar loans in addition to various other loan products, and the Bureau developed insights from this work that have been germane to this rulemaking, especially in how to address the potential for lenders to find ways to evade or circumvent its provisions. Consultation with State and Local Officials. The Bureau’s outreach also has included a large number of meetings and calls with State Attorneys General, State financial regulators, and municipal governments, along with the organizations representing the officials charged with enforcing applicable Federal, State, and local laws on small-dollar loans. These discussions have occurred with officials from States that effectively disallow such loans by imposing strict usury caps, as well as with officials from States that allow such loans and regulate them through various frameworks with different substantive approaches. The issues discussed have involved both storefront and online loans. In particular, as the Bureau has worked to develop the proposed registered information system requirements, it has consulted with State agencies from those States that require lenders to provide information about certain small-dollar loans to statewide databases. A group of State Attorneys General submitted a comment claiming that the extent to which the Bureau consulted State and local officials was insufficient. Some other State officials submitted similar comments. Although it is true that the Bureau did not meet with every attorney general or interested official from every State to discuss issues involving the regulation 114 of small-dollar loans, it did meet with many of them, some on multiple occasions. In addition, the Bureau did receive public comments from groups of State Attorneys General and other officials, including both regulators and legislators, and has carefully considered the issues they discussed, which presented many conflicting points of view. Several State Attorneys General requested that the Bureau commit to consulting with State officials before enforcing this regulation. The Bureau will coordinate and consult with State regulators and enforcement officials in the same manner that it does in other enforcement and supervisory matters. Tribal Consultations. The Bureau has engaged in consultation with Indian tribes about this rulemaking. The Bureau’s Policy for Consultation with Tribal Governments provides that the Bureau “is committed to regular and meaningful consultation and collaboration with tribal officials, leading to meaningful dialogue with Indian tribes on Bureau policies that would be expressly directed to tribal governments or tribal members or that would have direct implications for Indian tribes.”329 To date, the Bureau has held three formal consultation sessions related to this rulemaking. The first was held on October 27, 2014, at the National Congress of American Indians 71st Annual Convention and Marketplace in Atlanta, Georgia and before the release of the Bureau’s small-dollar lending SBREFA materials. The timing of the consultation gave Tribal leaders an opportunity to speak directly with the small-dollar lending team about Tribal lender and/or consumer experiences prior to the drafting of proposals that would become the Small Business Review Panel Outline. A second consultation was held on June 15, 2015, at the 329 Bureau of Consumer Fin. Prot., “Consumer Financial Protection Bureau Policy for Consultation with Tribal Governments,” at 1, available at http://files.consumerfinance.gov/f/201304_cfpb_consultations.pdf. 115 Bureau’s headquarters. At that consultation, Tribal leaders responded to the proposals under consideration set forth in the Outline that had recently been released. A third consultation was held on August 17, 2016, at the Sandra Day O’Connor College of Law in Phoenix, Arizona, after the release of the proposed rule. All Federally recognized Indian tribes were invited to attend these consultations, which generated frank and valuable input from Tribal leaders to Bureau senior leadership and staff about the effects such a rulemaking could have on Tribal nations and lenders. In addition, the Bureau has met individually with Tribal leaders, Tribal lenders, and Tribal lending associations in an effort to further inform its small-dollar lending work. A Tribal trade association dealing with financial services issues informed the Bureau that it believed these consultations were inadequate. B. Supervisory and Enforcement Activity In addition to these many channels of outreach, the Bureau has developed a broader understanding of small-dollar lending through its supervisory and enforcement work. This work is part of the foundation of the Bureau’s expertise and experience with this market, which is informed by frequent contact with certain small-dollar lenders and the opportunity to scrutinize their operations and practices up close through supervisory examinations and enforcement investigations. Some illustrative details of this work are related below. The Bureau’s Supervisory Work. The Bureau has been performing supervisory examinations of small-dollar lenders for more than five years. During this time, the Bureau has written and published its guidelines on performing such examinations, which its exam teams 116 have applied and refined further over time.330 All of this work has provided the Bureau with a quite comprehensive vantage point on the operations of payday and other small-dollar lenders and the nature and effects of their loan products for consumers. In its regular published reports known as Supervisory Highlights, the Bureau has summarized, while maintaining confidentiality of supervised entities, the types of issues and concerns that arise in its examinations of non-bank financial companies in general, and of smalldollar lenders in particular. In its Summer 2013 edition, for example, the Bureau emphasized its general finding that “nonbanks are more likely to lack a robust [Compliance Management System] as their consumer compliance-related activities have not been subject to examinations at the federal level for compliance with the Federal consumer financial laws prior to the Bureau’s existence.”331 The Bureau noted that it had identified “one or more instances of nonbanks that lack formal policies and procedures, have not developed a consumer compliance program, or do not conduct independent consumer compliance audits. Lack of an effective CMS has, in a number of instances, resulted in violations of Federal consumer financial laws.”332 In the Spring 2014 edition, the Bureau addressed its supervisory approach to short-term, small-dollar lending in more detail. At that time, the Bureau noted that its exercise of supervisory authority marked the first time any of these lenders had been subject to Federal compliance examinations. The Bureau described a number of shortcomings it had found and 330 See Bureau of Consumer Fin. Prot., “CFPB Examination Procedures, Short-term, Small-Dollar Lending,” available at https://s3.amazonaws.com/files.consumerfinance.gov/f/documents/201309_cfpb_payday_manual_revisions.pdf. 331 Bureau of Consumer Fin. Prot., “Supervisory Highlights,” at 6 (Summer 2013), available at http://files.consumerfinance.gov/f/201308_cfpb_supervisory-highlights_august.pdf. 332 Bureau of Consumer Fin. Prot., “Supervisory Highlights,” at 6 (Summer 2013), available at http://files.consumerfinance.gov/f/201308_cfpb_supervisory-highlights_august.pdf. 117 addressed with the compliance management systems implemented by small-dollar lenders, including lack of oversight, inadequate complaint management, lack of written policies and procedures, failure to train staff adequately, lack of effective compliance audit programs, and more generally a pervasive lack of accountability within the compliance program. It also catalogued many different violations and abuses in the collection methods these lenders used with their customers. Finally, the report noted that Bureau examinations found deceptive practices in the use of preauthorized ACH withdrawals from borrower checking accounts.333 The Summer 2016 edition included a discussion of debt collection issues, which are relevant to many payday lenders, and also included a section explicitly dedicated to small-dollar lending and issues associated with compliance with the Electronic Fund Transfer Act. The Bureau’s examiners found that the “loan agreements of one or more entities failed to set out an acceptable range of amounts to be debited, in lieu of providing individual notice of transfers of varying amounts. These ranges could not be anticipated by the consumer because they contained ambiguous or undefined terms in their descriptions of the upper and lower limits of the range.”334 And the Spring 2017 edition expressed concerns about production incentives relevant to many providers of financial services, noting that “many supervised entities choose to implement incentive programs to achieve business objectives. These production incentives can lead to significant consumer harm if not properly managed.”335 333 Bureau of Consumer Fin. Prot., “Supervisory Highlights,” at 14–20 (Spring 2014), available at http://files.consumerfinance.gov/f/201405_cfpb_supervisory-highlights-spring-2014.pdf. 334 Bureau of Consumer Fin. Prot., “Supervisory Highlights,” at 13 (Summer 2016), available at https://s3.amazonaws.com/files.consumerfinance.gov/f/documents/Supervisory_Highlights_Issue_12.pdf. 335 Bureau of Consumer Fin. Prot., “Supervisory Highlights,” at 27 (Spring 2017), available at https://s3.amazonaws.com/files.consumerfinance.gov/f/documents/201704_cfpb_Supervisory-Highlights_Issue15.pdf 118 In the most recent Summer 2017 edition, the Bureau again described problems that it had addressed with short-term, small-dollar lending, including payday and vehicle title loans. Among them were a variety of collections issues, along with misrepresentations that several lenders had made in the marketing of such loans. Examiners reported that lenders had promised consumers that they could obtain such a loan without a credit check, yet this turned out to be untrue and, in some instances, to lead to loan denials based on the information obtained from the consumers’ credit reports. They also found that certain lenders advertised products and services in their outdoor signage that they did not, in fact, offer. And some lenders advertised their products by making unsubstantiated claims about how they compared with those of competing lenders. These practices were found to be deceptive and changes were ordered to be made.336 The Bureau further found that some lenders misrepresented their processes to apply for a loan online, and others misused references provided by loan applicants on applications for origination purposes by marketing products to the persons listed. Finally, examiners observed that one or more lenders mishandled the payment process by debiting accounts automatically for payments that had already been made, leading to unauthorized charges and overpayments. The entities also failed to implement adequate processes to accurately and promptly identify and refund borrowers who paid more than they owed, who were unable to avoid the injury.337 The Bureau’s Enforcement Work. The Bureau also has developed expertise and experience in this market over time by pursuing public enforcement actions against more than 20 336 Bureau of Consumer Fin. Prot., “Supervisory Highlights,” at 28–30 (Summer 2017), available at https://s3.amazonaws.com/files.consumerfinance.gov/f/documents/201709_cfpb_Supervisory-Highlights_Issue16.pdf. 337 See Bureau of Consumer Fin. Prot., “Supervisory Highlights,” at 31–32 (Summer 2017), available at https://s3.amazonaws.com/files.consumerfinance.gov/f/documents/201709_cfpb_Supervisory-Highlights_Issue16.pdf. 119 small-dollar lenders, including brick-and-mortar storefront lenders, online lenders, and vehicle title lenders (as well as pawn lenders, which are not covered under the rule). A number of these actions have been resolved, but some remain pending in the courts at this time. In every instance, however, before the enforcement action was brought, it was preceded by a thorough investigation of the underlying facts in order to determine whether legal violations had occurred. The issues raised in these actions include engaging in misleading and deceptive marketing practices, making improper disclosures, training employees to hide or obfuscate fees, pushing customers into a cycle of debt by pressuring them to take out additional loans they could not afford, making false statements about whether and how transactions can be canceled or reversed, taking unauthorized and improper electronic withdrawals from customer accounts, and engaging in collections efforts that generate wide-ranging problems.338 The Bureau has determined many of these practices to be violations of the prohibition against unfair, deceptive, or abusive acts or practices. The information and insights that the Bureau has gleaned from these investigations and enforcement actions has further advanced its understanding of this market and of the factual foundations for the policy interventions contained in this final rule. 338 See, e.g., Press Release, Bureau of Consumer Fin. Prot., “CFPB Takes Acton Against Check Cashing and Payday Lending Company for Tricking and Trapping Consumers” (May 11, 2016), available at https://www.consumerfinance.gov/about-us/newsroom/cfpb-takes-action-against-check-cashing-and-paydaylending-company-tricking-and-trapping-consumers/; Press Release, Bureau of Consumer Fin. Prot., “CFPB Fines Titlemax Parent Company $9 Million for Luring Consumers Into More Costly Loans” (Sept. 26, 2016), available at https://www.consumerfinance.gov/about-us/newsroom/cfpb-fines-titlemax-parent-company-9-million-luringconsumers-more-costly-loans/; Press Release, Bureau of Consumer Fin. Prot., “CFPB Sues Five Arizona Title Lenders for Failing to Disclose Loan Annual Percentage Rate to Consumers” (Sept. 21, 2016), available at https://www.consumerfinance.gov/about-us/newsroom/cfpb-sues-five-arizona-title-lenders-failing-disclose-loanannual-percentage-rate-consumers/; Press Release, Bureau of Consumer Fin. Prot., “CFPB Sues Offshore Payday Lender” (Aug. 5, 2015), available at https://www.consumerfinance.gov/about-us/newsroom/cfpb-sues-offshorepayday-lender/; Press Release, Bureau of Consumer Fin. Prot., “Consumer Financial Protection Bureau Takes Action Against Payday Lender for Robo-Signing” (Nov. 20, 2013), available at https://www.consumerfinance.gov/about-us/newsroom/consumer-financial-protection-bureau-takes-action-againstpayday-lender-for-robo-signing/. 120 For example, in 2013 the Bureau resolved a public enforcement action against Cash America, Inc. that arose out of an examination of this large national payday lender. The Bureau cited Cash America for committing three distinct unfair and deceptive practices: robo-signing court documents in debt collection lawsuits; violating the Military Lending Act by overcharging servicemembers and their families; and improperly destroying records in advance of the Bureau’s examination. Cash America was ordered to pay $14 million in refunds to consumers and to pay a civil penalty of $5 million for these violations.339 In 2014, the Bureau filed a public enforcement action against Ace Cash Express that developed out of the Bureau’s prior exam work. The Bureau found through its examination and subsequent investigation that ACE had engaged in unfair, deceptive, and abusive practices by using illegal debt collection tactics to pressure overdue borrowers into taking out additional loans they could not afford. In fact, ACE’s own training manual for its employees had a graphic illustrating this cycle of debt. According to the graphic, consumers begin by applying to ACE for a loan, which ACE approved. Next, if the consumer “exhausts the cash and does not have the ability to pay,” ACE “contacts the customer for payment or offers the option to refinance or extend the loan.” Then, when the consumer “does not make a payment and the account enters collections,” the cycle starts all over again—with the formerly overdue borrower applying for another payday loan.340 339 See Press Release, Bureau of Consumer Fin. Prot., “Consumer Financial Protection Bureau Takes Action Against Payday Lender for Robo-Signing” (Nov. 20, 2013), available at https://www.consumerfinance.gov/aboutus/newsroom/consumer-financial-protection-bureau-takes-action-against-payday-lender-for-robo-signing/. 340 See Press Release, Bureau of Consumer Fin. Prot., “CFPB Takes Action Against Ace Cash Express for Pushing Payday Borrowers Into Cycle of Debt” (July 10, 2014), available at https://www.consumerfinance.gov/aboutus/newsroom/cfpb-takes-action-against-ace-cash-express-for-pushing-payday-borrowers-into-cycle-of-debt/. 121 The Bureau’s examination of ACE was conducted in coordination with the Texas Office of Consumer Credit Commissioner and resulted in an order imposing $5 million in consumer refunds and a $5 million civil penalty. The enforcement action was partially based on ACE’s creation of a false sense of urgency to get delinquent borrowers to take out more payday loans— all while charging new fees each time.341 In September 2015, the Bureau took action against Westlake Services, an indirect auto finance company, and Wilshire Consumer Credit, its auto title lending subsidiary, which offered auto title loans directly to consumers, largely via the Internet, and serviced those loans; Wilshire also purchased and serviced auto title loans made by others. The Bureau concluded that Westlake and Wilshire had committed unfair and deceptive acts or practices by pressuring borrowers through the use of illegal debt collection tactics. The tactics included illegally deceiving consumers by using phony caller ID information (sometimes masquerading as pizza delivery services or flower shops), falsely threatening to refer borrowers for investigation or criminal prosecution, calling under false pretenses, and improperly disclosing information about debts to borrowers’ employers, friends, and family. Wilshire also gave consumers incomplete information about the true cost of the loans it offered. The consent order resolving the matter required the companies to overhaul their debt collection practices and to cease advertising or marketing their products untruthfully. The companies were also ordered to provide consumers with $44.1 million in cash relief and balance reductions, and to pay a civil penalty of $4.25 million. 341 See Press Release, Bureau of Consumer Fin. Prot., “CFPB Takes Action Against Ace Cash Express for Pushing Payday Borrowers Into Cycle of Debt” (July 10, 2014), available at https://www.consumerfinance.gov/aboutus/newsroom/cfpb-takes-action-against-ace-cash-express-for-pushing-payday-borrowers-into-cycle-of-debt/. 122 In December 2015, the Bureau resolved another enforcement action with EZCORP, Inc., a short-term, small-dollar lender. The action was initially generated from a supervisory exam that had exposed significant and illegal debt collection practices. These included in-person collection visits at consumers’ homes or workplaces (which risked disclosing the consumer’s debt to unauthorized third parties), falsely threatening consumers with litigation for not paying their debts, misrepresenting consumers’ rights, and unfairly making multiple electronic withdrawal attempts from consumer accounts that caused mounting bank fees. These practices were found to be unfair and deceptive and to violate the Electronic Fund Transfer Act; as a result, the Bureau ordered EZCORP to refund $7.5 million to 93,000 consumers and pay a $3 million civil penalty, while halting collection of remaining payday and installment loan debts associated with roughly 130,000 consumers. That action also prompted the Bureau to issue an industry-wide warning about potentially unlawful conduct during in-person collections at homes or workplaces.342 In September 2016, the Bureau took action against TitleMax’s parent company TMX Finance, one of the country’s largest auto title lenders, for luring consumers into costly loan renewals by presenting them with misleading information about the terms and costs of the deals. The Bureau’s investigation found that store employees, as part of their sales pitch for the 30-day loans, offered consumers a “monthly option” for making loan payments using a written guide that did not explain the true cost of the loan if the consumer renewed it multiple times, though TMX personnel were well aware of these true costs. In fact, the guide and sales pitch distracted 342 See Press Release, Bureau of Consumer Fin. Prot., “CFPB Orders EZCORP to Pay $10 million for Illegal Debt Collection Tactics,” (Dec. 16, 2015), available at https://www.consumerfinance.gov/about-us/newsroom/cfpborders-ezcorp-to-pay-10-million-for-illegal-debt-collection-tactics/. 123 consumers from the fact that repeatedly renewing the loan, as encouraged by TMX Finance employees, would dramatically increase the loan’s cost, while making it difficult, if not impossible, for a consumer to compare costs for renewing the loan over a given period. The company then followed up with those who failed to repay by making intrusive visits to homes and workplaces that put consumers’ personal information at risk. TMX Finance was ordered to stop its unlawful practices and pay a $9 million penalty.343 Likewise, in December 2016 the Bureau filed a public enforcement action against Moneytree, which offers payday loans and check-cashing services, for misleading consumers with deceptive online advertisements and collections letters. The company was ordered to cease its illegal conduct, refund $255,000 to consumers, and pay a civil penalty of $250,000. In addition to the deceptive advertising, the company was found to have deceptively told consumers that their vehicles could be repossessed when it had no right or ability to do so, and to have improperly withdrawn money from consumers’ accounts without authorization to do so.344 From the Bureau’s experience of carrying out investigations of these kinds of illegal practices and halting them through its enforcement efforts, the Bureau has become much more aware of the nature and likelihood of unfair, deceptive, or abusive practices in this market. And though the Bureau generally has devoted less attention in its supervisory and enforcement programs to issues that it has long intended to address separately, as here, through its rulemaking authority, the Bureau nonetheless has gained valuable experience and expertise from all of this 343 See Press Release, Bureau of Consumer Fin. Prot., “CFPB Fines Titlemax Parent Company $9 Million for Luring Consumers into More Costly Loans,” (Sept. 26, 2016), available at https://www.consumerfinance.gov/aboutus/newsroom/cfpb-fines-titlemax-parent-company-9-million-luring-consumers-more-costly-loans/. 344 See Press Release, Bureau of Consumer Fin. Prot., “CFPB Takes Action Against Moneytree for Deceptive Advertising and Collection Practices,” (Dec. 16, 2016), available at https://www.consumerfinance.gov/aboutus/newsroom/cfpb-takes-action-against-moneytree-deceptive-advertising-and-collection-practices/. 124 work that it now brings to this rulemaking process. Since the inception of its supervision and enforcement program, the Bureau has worked continually to maximize compliance with the Federal consumer financial laws as they apply to payday and other types of small-dollar lenders. Sustained attention to compliance through the Bureau’s supervision and enforcement work is an important adjunct to this rulemaking, but is not a sufficient substitute for it. C. Research and Analysis of Small-Dollar Loans Bureau White Papers. In April 2013, the Bureau issued a white paper on payday loans and deposit advance products, including findings by Bureau staff. For each of these loan products, the Bureau examined loan characteristics, borrower characteristics, intensity of use, and sustained use of the product. These findings were based largely on the data the Bureau had collected from some of the larger payday lenders under its supervisory authority, and covered approximately 15 million loans generated in 33 States and on approximately 15,000 deposit advance product transactions. The report took a snapshot of borrowers at the beginning of the study period and traced their usage of these products over the course of the study period. The report demonstrated that though some consumers use payday loans and deposit advances at relatively low to moderate levels, a sizable share of users conduct transactions on a long-term basis, suggesting they are unable to fully repay the loan and pay other expenses without taking out a new loan shortly thereafter.345 In March 2014, the Bureau issued another white paper on payday lending. This report was based on the supervisory data the Bureau had received from larger payday lenders, truncated somewhat to cover 12-month windows into borrowing patterns. These limitations yielded a 345 CFPB Payday Loans and Deposit Advance Products White Paper. 125 dataset of over 12 million loans in 30 States. Responding to criticisms of the Bureau’s white paper, this report focused on “fresh borrowers,” i.e., those who did not have a payday loan in the first month of the Bureau’s data and whose usage began in the second month. After reviewing this data, the report yielded several key findings. First, of the loans taken out by these borrowers over a period of eleven months over 80 percent are rolled over or followed by another loan within 14 days. Half of all loans are made as part of a sequence that is at least ten loans long, and few borrowers amortize, meaning their principal amounts are not reduced between the first and last loan of a sequence. Monthly borrowers (the majority of whom are receiving government benefits) are disproportionately likely to stay in debt for eleven months or longer. And most borrowing involves multiple renewals following an initial loan, rather than multiple distinct borrowing episodes separated by more than fourteen days.346 Both before and after the release of these white papers, the Bureau held numerous meetings with stakeholders to obtain their perspectives and comments on the methodology and contents of this research. As is also noted below, the Bureau also hosted individual scholars in the field for research presentations Additional Research Reports. In April and May of 2016, the Bureau published two additional research reports on small-dollar loans. In conducting this research, the Bureau used not only the data obtained from the supervisory examinations previously described but also data obtained through orders the Bureau had issued pursuant to section 1022(c)(4) of the Dodd-Frank Act, data obtained through civil investigative demands made by the Bureau pursuant to section 1052 of the Dodd-Frank Act, and data voluntarily supplied to the Bureau by several lenders. 346 See CFPB Data Point: Payday Lending. 126 The first report addressed how online payday and payday installment lenders use access to consumers’ bank accounts to collect loan payments. It found that after a failed ACH payment request made by an online lender, subsequent payment requests to the same account are unlikely to succeed, though lenders often continue to present them, with many online lenders submitting multiple payment requests on the same day. The resulting harm to consumers is shown by the fact that accounts of borrowers who use loans from online lenders and experience a payment that is returned for insufficient funds are more likely to be closed by the end of the sample period than accounts experiencing a returned payment for products other than payday or payday installment loans.347 The other report addressed consumer usage and default patterns on short-term vehicle title loans. Similar to payday loans, the report determined that single-payment vehicle title lenders rely on borrowers who take out repeated loans, with borrowers stuck in debt for seven months or more supplying two-thirds of the title loan business. In over half the instances where the borrower takes out such a loan, they end up taking out four or more consecutive loans, which becomes an unaffordable, long-term debt load for borrowers who are already struggling with their financial situations. In addition to high rates of default, the Bureau found that these loans carried a further adverse consequence for many consumers, as one out of every five loan sequences ends up with the borrower having their vehicle seized by the lender in repossession for failure to repay.348 347 348 See CFPB Online Payday Loan Payments. See CFPB Single-Payment Vehicle Title Lending. 127 In June 2016, the Bureau issued a supplemental report on payday, payday installment, vehicle title loan, and deposit advance products that addressed a wide range of subjects pertinent to the proposed rule. The report studied consumers’ usage and default patterns for title and payday installment loans; analyzed whether deposit advance consumers overdrew accounts or took out payday loans more frequently after banks stopped offering deposit advance products; examined the impact of State laws on payday lending; compared payday re-borrowing rates across States with different renewal and cooling-off period laws; provided findings on payday borrowing and default patterns, using three different loan sequence definitions; and simulated effects of certain lending and collection restrictions on payday and vehicle title loan markets. 349 Consumer Complaint Information. The Bureau also has conducted analysis on its own consumer complaint information. Specifically, the Bureau had received, as of April 1, 2017, approximately 51,000 consumer complaints relating to payday and other small-dollar loan products. Of these complaints, about one-third were submitted by consumers as payday or other small-dollar loan complaints and two-thirds as debt collection complaints where the source of the debt was a payday loan.350 Industry representatives have frequently expressed the view that consumers seem to be satisfied with payday and other covered short-term loan products, as shown by low numbers of complaints and the submission of positive stories about them to the “Tell Your Story” function on the Bureau’s website. Yet, the Bureau has observed from its consumer complaint data that 349 See CFPB Report on Supplemental Findings. The Bureau took a phased approach to accepting complaints from consumers. The Bureau began accepting payday loan complaints in November of 2013, and vehicle title loan complaints in July of 2014, which means that the complaint data it has accumulated on these markets does not cover the same periods as the complaint data it has collected, for example, on the mortgage or credit card markets. 350 128 from November 2013 through December 2016, approximately 31,000 debt collection complaints cited payday loans as the underlying debt, and over 11 percent of the complaints the Bureau has handled about debt collection stemmed directly from payday loans.351 In fact, when complaints about payday loans are normalized in comparison to other credit products, the numbers do not turn out to be low at all. For example, in 2016, the Bureau received about 4,400 complaints in which consumers reported “payday loan” as the complaint product and about 26,600 complaints about credit cards. 352 Yet there are only about 12 million payday loan borrowers annually, and about 156 million consumers have one or more credit cards.353 Therefore, by way of comparison, for every 10,000 payday loan borrowers, the Bureau received about 3.7 complaints, while for every 10,000 credit cardholders, the Bureau received about 1.7 complaints. In addition, the substance of some of the consumer complaints about 351 Bureau of Consumer Fin. Prot., “Monthly Complaint Report, Vol. 9,” at 12 fig. 3 (Mar. 2016), available at http://files.consumerfinance.gov/f/201603_cfpb_monthly-complaint-report-vol-9.pdf. 352 Bureau of Consumer Fin. Prot., “Consumer Response Annual Report, January 1 – December 31, 2016,” at 27, 34 (Mar. 2017), available at https://www.consumerfinance.gov/documents/3368/201703_cfpb_Consumer-ResponseAnnual-Report-2016.pdf. 353 Bureau staff estimate based on finding that 63 percent of American adults hold an open credit card and Census population estimates. See Bureau of Consumer Fin. Prot., “The Consumer Credit Card Market Report,” at 36 (Dec. 2015), available at http://files.consumerfinance.gov/f/201512_cfpb_report-the-consumer-credit-card-market.pdf; U.S. Census Bureau, “Annual Estimates of Resident Population for Selected Age Groups by Sex for the United States, States, Counties, and Puerto Rico Commonwealth and Municipios: April 1, 2010 to July 1, 2016,” (June 2017), available at https://factfinder.census.gov/bkmk/table/1.0/en/PEP/2016/PEPAGESEX. Other estimates of the number of credit card holders have been higher, meaning that 1.7 complaints per 10,000 credit card holders would be a high estimate. The U.S. Census Bureau estimated there were 160 million credit card holders in 2012, U.S. Census Bureau, “Statistical Abstract of the United States: 2012,” at 740 tbl.1188 (Aug. 2011), available at https://www.census.gov/library/publications/2011/compendia/statab/131ed.html, and researchers at the Federal Reserve Bank of Boston estimated that 72.1 percent of U.S. consumers held at least one credit card in 2014, Claire Greene et al., “The 2014 Survey of Consumer Payment Choice: Summary Results,” at 18 (Fed. Reserve Bank of Boston, No. 16–3, 2016), available at https://www.bostonfed.org//media/Documents/researchdatareport/pdf/rdr1603.pdf. As noted above in the text, additional complaints related to both payday loans and credit cards are submitted as debt collection complaints with “payday loan” or “credit card” listed as the type of debt. 129 payday loans as catalogued by the Bureau mirrored many of the concerns that constitute the justification for this rule here.354 Moreover, faith leaders and faith groups of many denominations from around the country collected and submitted comments indicating that many borrowers may direct their personal complaints or dissatisfactions with their experiences elsewhere than to government officials. Market Monitoring. The Bureau has also continuously engaged in market monitoring for the small-dollar loan market, just as it does for the other markets within its jurisdiction. This work involves regular outreach to industry members and trade associations, as well as other stakeholders in this marketplace. It also involves constant attention to news, research, trends, and developments in the market for small-dollar loans, including regulatory changes that may be proposed and adopted by the States and localities around the country. The Bureau has also carefully reviewed the published academic literature on small-dollar liquidity loans, along with research conducted or sponsored by stakeholder groups. In addition, a number of outside researchers have presented their own research at seminars for Bureau staff. D. Small Business Review Panel Small Business Regulatory Enforcement Fairness Act (SBREFA) Process. In April 2015, in accordance with SBREFA, the Bureau convened a Small Business Review Panel with the Chief Counsel for Advocacy of the SBA and the Administrator of the Office of Information and 354 “Consumer confusion relating to repayment terms was frequently expressed. These consumers complained of the lack of clarity about repayment of the loan using automatic withdrawal features on a bank card, on a prepaid card, or by direct deposit. Consumers with multiple advances stated their difficulty managing a short repayment period and more often rolled-over the loan, resulting in an inflated total cost of the loan.” Bureau of Consumer Fin. Prot., “Consumer Response 2016 Annual Report, January 1 – December 31, 2016,” (Mar. 2017), available at https://www.consumerfinance.gov/documents/3368/201703_cfpb_Consumer-Response-Annual-Report-2016.pdf. 130 Regulatory Affairs within the Office of Management and Budget (OMB).355 As part of this process, the Bureau prepared an outline of the proposals then under consideration and the alternatives considered (the Small Business Review Panel Outline), which it posted on its website for review and comment by the general public as well as the small entities participating in the panel process.356 Before formally convening, the Panel took part in teleconferences with small groups of the small entity representatives (SERs) to introduce the Outline and get feedback on the Outline, as well as a series of questions about their business operations and other issues. The Panel gathered information from representatives of 27 small entities, including small payday lenders, vehicle title lenders, installment lenders, banks, and credit unions. The meeting participants represented storefront and online lenders, State-licensed lenders, and lenders affiliated with Indian tribes. The Panel held a full-day meeting on April 29, 2015, to discuss the Small Business Review Panel Outline. The 27 small entities also were invited to submit written feedback, and 24 of them did so. The Panel considered input from the small entities about the potential compliance costs and other impacts on those entities and about impacts on access to credit for small businesses and made recommendations about potential options for addressing those costs and impacts. These recommendations are set forth in the Small Business Review Panel Report, 355 The Small Business Regulatory Enforcement Fairness Act of 1996 (SBREFA), as amended by section 1100G(a) of the Dodd-Frank Act, requires the Bureau to convene a Small Business Review Panel before proposing a rule that may have a substantial economic impact on a significant number of small entities. See Public Law 104–121, tit. II, 110 Stat. 847, 857 (1996) as amended by Public Law 110–28, sec. 8302 (2007), and Public Law 111–203, sec. 1100G (2010). 356 Bureau of Consumer Fin. Prot., “Small Business Advisory Review Panel for Potential Rulemakings for Payday, Vehicle Title, And Similar Loans: Outline of Proposals under Consideration and Alternatives Considered,” (Mar. 26, 2015), available at http://files.consumerfinance.gov/f/201503_cfpb_outline-of-the-proposals-from-smallbusiness-review-panel.pdf. 131 which is made part of the administrative record in this rulemaking.357 The Bureau carefully considered these findings and recommendations in preparing the proposed rule and completing this final rule, as detailed below in the section-by-section analysis of various provisions and in parts VII and VIII. The Bureau also continued its outreach and engagement with stakeholders on all sides since the SBREFA process concluded. Comments Regarding the Bureau’s SBREFA Process. Following the release of the proposed rule, a number of commenters criticized the SBREFA process. Some of these commenters were third parties such as trade associations who were familiar with the SBREFA process. Others were the SERs themselves. Some commenters argued that the Bureau failed to adequately consider the concerns raised and alternatives suggested by the SERs. Some commenters also expressed concerns about the SBREFA procedures. Some commenters objected that in developing the proposed rule the Bureau did not consider policy suggestions made by SERs or recommendations made by the SBREFA Panel. For example, some commenters argued that the Bureau failed to consider whether, as some SERs contended, disclosures could prevent the consumer injury the Bureau is seeking to address in this rulemaking. Some commenters also suggested that the Bureau failed to adequately consider alternative approaches employed by various States. Some commenters criticized the Bureau for ignoring the Panel’s recommendations in developing the proposal, including, for example, the recommendation that the Bureau consider whether the rule should permit loan sequences of more 357 Bureau of Consumer Fin Prot., U.S. Small Bus. Admin., & Office of Mgmt. & Budget, “Final Report of the Small Business Review Panel on CFPB’s Rulemaking on Payday, Vehicle Title, and Similar Loans” (June 25, 2015), available at http://files.consumerfinance.gov/f/documents/3a_-_SBREFA_Panel__CFPB_Payday_Rulemaking_-_Report.pdf (hereinafter Small Business Review Panel Report). 132 than three short-term loans. Other SER commenters argued that the Bureau should adopt the requirements imposed by certain States (like Illinois or Michigan or Utah) or should require lenders to offer off-ramps instead of the requirements herein. Some commenters indicated that they believed the Bureau ultimately ignored or underestimated the rule’s potential impact on small businesses and inadequately considered the rule’s potential impact on rural communities. Some commenters argued that the Bureau did not adequately address issues around the cost of credit to small entities. One commenter noted that some credit unions offer certain short-term loan products and that the Bureau did not consider the impact of the rule on credit union products and small credit unions. The SBA Office of Advocacy submitted comments of its own on the proposed rule and on how it responded to the SBREFA process. Although Advocacy had no complaints about the procedures used or the input received in the process, it did present its views on whether the proposed rule sufficiently reflected the discussions and debates that had occurred during the Panel discussions and the SBREFA process as a whole. To begin with, Advocacy agreed with the Bureau that the proposed rule would have a significant economic impact on small entities, which it found to be a matter of concern and felt had been underestimated by the Bureau. It stated that the ability-to-repay requirements in the proposed rule would be burdensome, and the cooling-off periods in particular would harm small businesses. It encouraged the Bureau to exempt from the rule small businesses that operate in States that currently have payday lending laws and to mitigate its impact on credit unions, Indian tribes, and small communities. Advocacy also commented that the proposed rule would restrict access to credit for consumers 133 and for certain small businesses, and suggested that an exception be made for situations where such a loan may be necessary to address an emergency. The procedural objections to the SBREFA process raised by other commenters included concerns about the make-up of the SBREFA panel and whether it was representative of the small entities who would be most affected by the proposal; the timing of SBREFA meetings; the administration and management of SBREFA-related phone calls; the overall “sufficiency” of the process; and unheeded requests to convene additional Panel sessions or to conduct additional research on specific topics. One trade group commenter incorporated portions of a comment letter from a SER that was sent to the Bureau during the SBREFA process, which raised a number of procedural objections. Another stated the panel excluded open-end lenders. Some expressed concern that the process did not provide them adequate time to realize the full ramifications of the proposed rule and the effects it would have on their business activity. Others suggested that the process was flawed because the Bureau’s analysis allegedly ignored the rule’s potential costs. One commenter also suggested that the SBREFA process was tainted by the Bureau Director’s public comments regarding small-dollar lending in the years preceding the rulemaking. Some commenters noted that the SBREFA process had been effective in considering and responding to certain concerns, including input regarding PAL loans and checking customer borrowing history. Responses to Comments. The Bureau disagrees with commenters arguing that the Bureau did not adequately consider the suggestions of SERs and the Panel. In the proposed rule, the Bureau modified certain aspects of the approach in the Small Business Review Panel Outline in 134 response to feedback from SERs (and others). For example, the Outline included a 60-day cooling-off period after sequences of three short-term loans, but the proposed rule included a 30day cooling-off period, and that change is retained in the final rule. In addition, the Bureau followed the Panel’s recommendation to request comment on numerous specific issues. The feedback received by the Bureau also informed its decision to revise various aspects of the rule. For example, as discussed below, the Bureau revised the ability-to-repay requirements in a number of ways to provide greater flexibility and reduce the compliance burden, such as by not requiring income verification if evidence is not reasonably available. In addition, the rule no longer requires lenders to verify or develop estimates of rental housing expenses based on statistical data. And the Bureau considered all of the alternatives posited by the SERs, as noted where applicable throughout part V and in part VIII. More generally, the Bureau considered and made appropriate modifications to the rule based upon feedback received during the SBREFA process and in response to other feedback provided by the small business community. The Bureau obtained important input through the SBREFA process and all articulated viewpoints were understood—and considered—prior to the promulgation of the final rule. The Bureau disagrees with commenters that it did not consider alternative approaches. For example, in the proposal, the Bureau explained why it believed that disclosures would not be sufficient to address the identified harms and why the approaches of various States also appeared to be insufficient to address those harms. The Bureau likewise explains in this final rule its conclusions about why those approaches would not be sufficient. The Bureau both agrees and disagrees with various comments from Advocacy, and a fuller treatment of these issues is presented below in part VII, which addresses the potential 135 benefits, costs, and impacts of the final rule, including reductions in access to financial products and services and impacts on rural issues, and in part VIII, which addresses among other things the economic impact of the final rule on small entities, including small businesses. But more briefly here, the Bureau would note that it has made many changes in the final rule to reduce the burdens of the specific underwriting criteria in the ability-to-repay requirements; that Advocacy has stated that it appreciates the modification of the 60-day cooling-off period presented in the SBREFA Panel Outline to the 30-day cooling-off period in the proposed rule and now in the final rule; that Advocacy thanked the Bureau for clarifying that the proposed rule (and now the final rule) will not apply to business loans; that adoption of the conditional exemption from the final rule for alternative loans mitigates its impact on credit unions; that the Bureau did engage in another formal Tribal consultation after release of the proposed rule as Advocacy had urged; that the Bureau had consulted further with a range of State officials prior to finalizing the rule; and that the Bureau had extended the implementation period and effective date of the final rule. The Bureau also disagrees with commenters who criticized procedural aspects of the SBREFA process. With respect to the composition of the SERs that participated in the SBREFA process, the Bureau followed legal requirements for categorizing which entities qualified as small entities. The Bureau collaborated with the SBA Office of Advocacy so that the SERs included a variety of different types of lenders that could be affected by the rulemaking, ensuring that participants included a geographically diverse group of storefront payday lenders, online lenders, vehicle title lenders, installment lenders, and banks and credit unions. As noted above, to help ensure that the formal Panel meeting would allow for efficient and effective discussion of 136 substantive issues, the Panel convened several telephone conferences before the formal meeting to provide information about the Outline and to obtain information from the SERs. The Bureau disagrees, further, with the comments raising more specific procedural objections about the teleconferences and the Panel meeting. The Bureau provided agendas in advance of the calls and extended the length of the calls as needed to ensure that SERs were able to participate and provide feedback. While the Bureau appreciates that some SERs may have desired additional time to consider and provide feedback on the Outline, the Bureau notes that the Panel is required by law to report on the SERs’ comments and advice within 60 days after the Panel is convened. The Bureau conducted the process diligently and in accordance with its obligations under the Regulatory Flexibility Act and consistent with prior SBREFA processes. With respect to comments suggesting that the Bureau failed to adequately consider the costs and impact on small businesses and in rural areas, the Bureau notes that the costs and impacts were addressed in the Notice of Proposed Rulemaking, and, for the final rule, are addressed in parts VII and VIII. E. Consumer Testing In developing the disclosures for this rule, the Bureau engaged a third-party vendor, Fors Marsh Group (FMG), to coordinate qualitative consumer testing for the disclosures that were being considered. The Bureau developed several prototype disclosure forms and tested them with participants in one-on-one interviews. Three categories of forms were developed and tested: (1) origination disclosures that informed consumers about limitations on their ability to receive additional short-term loans; (2) upcoming payment notices that alerted consumers about lenders’ future attempts to withdraw money from consumers’ accounts; and (3) expired 137 authorization notices that alerted consumers that lenders would no longer be able to attempt to withdraw money from the consumers’ accounts. Observations and feedback from the testing were incorporated into the model forms developed by the Bureau. Through this testing, the Bureau sought to observe how consumers would interact with and understand prototype forms developed by the Bureau. In late 2015, FMG facilitated two rounds of one-on-one interviews, each lasting 60 minutes. The first round was conducted in September 2015 in New Orleans, Louisiana, and the second round was conducted in October 2015 in Kansas City, Missouri. At the same time the Bureau released the proposed rule, it also made available a report that FMG had prepared on the consumer testing.358 The testing and focus groups were conducted in accordance with OMB Control Number 3170-0022. A total of 28 individuals participated in the interviews. Of these 28 participants, 20 self-identified as having used a small-dollar loan within the past two years. Highlights from Interview Findings. FMG asked participants questions to assess how well they understood the information on the forms. For the origination forms, the questions focused on whether participants understood that their ability to roll this loan over or take out additional loans may be limited. Each participant reviewed one of two different prototype forms: either one for loans that would require an abilityto-repay determination (ATR Form) or one for loans that would be offered under the conditional exemption for covered short-term loans (Alternative Loan Form). During Round 1, many 358 See Fors Marsh Group, “Qualitative Testing of Small Dollar Loan Disclosures, Prepared for the Consumer Financial Protection Bureau,” (Apr. 2016) available at http://files.consumerfinance.gov/f/documents/Disclosure_Testing_Report.pdf (for a detailed discussion of the Bureau’s consumer testing) (hereinafter FMG Report). 138 participants for both form types recognized and valued information about the loan amount and due date; accordingly, that information was moved to the beginning of all the origination forms for Round 2. For the ATR Forms, few participants in Round 1 understood that the “30 days” language was describing a period when future borrowing may be restricted. Instead, several read the language as describing the loan term. In contrast, nearly all participants reviewing the Alternative Loan Form understood that it was attempting to convey that each successive loan they took out after the first in this series had to be smaller than the previous loan, and that after taking out three loans they would not be able to take out another for 30 days. Some participants also reviewed a version of this Alternative Loan Form for when consumers are taking out their third loan in a sequence. The majority of participants who viewed this notice understood it, acknowledging that they would have to wait until 30 days after the third loan was paid off to be considered for another similar loan. During Round 2, participants reviewed two new versions of the ATR Form. One adjusted the “30 days” phrasing and the other completely removed the “30 days” language, replacing it with the phrase “shortly after this one.” The Alternative Loan Form was updated with similar rephrasing of the “30 days” language. In order to simplify the table, the “loan date” column was removed. The results in Round 2 were similar to Round 1. Participants reviewing the ATR forms focused on the language notifying them they should not take out this loan if they are unable to pay the full balance by the due date. Information about restrictions on future loans went largely unnoticed. The edits appeared to have a positive impact on comprehension since no participants interpreted either form as providing information on their loan term. There did not seem to be a 139 difference in comprehension between the group with the “30 days” version and the group with the “shortly” version. As in Round 1, participants who reviewed the Alternative Loan Form noticed and understood the schedule detailing maximum borrowable amounts. These participants understood that the purpose of the Alternative Loan Form was to inform them that any subsequent loans must be smaller. Questions for the payment notices focused on participants’ ability to identify and understand information about the upcoming payment. Participants reviewed one of two payment notices: an Upcoming Withdrawal Notice or an Unusual Withdrawal Notice. Both forms provided details about the upcoming payment attempt and a payment breakdown table. The Unusual Withdrawal Notice also indicated that the withdrawal was unusual because the payment was higher than the previous withdrawal amount. To obtain feedback on participants’ likelihood to open notices delivered in an electronic manner, these notices were presented as a sequence to simulate an email message. In Round 1, all participants, based on seeing the subject line in the e-mail inbox, said that they would open the Upcoming Withdrawal e-mail and read it. Nearly all participants said they would consider the e-mail legitimate. They reported having no concerns about the e-mail because they would have recognized the company name, and because it included details specific to their account along with the lender contact information. When shown the full Upcoming Withdrawal Notice, participants understood that the lender would be withdrawing $40 from their account on a particular date. Several participants also pointed out that the notice described an interest-only payment. Round 1 results were similar for the Unusual Withdrawal Notice; all participants who viewed this notice said they would open the e-mail, and all but one 140 participant—who was deterred due to concerns with the appearance of the link’s URL—would click on the link leading to additional details. The majority of participants indicated that they would want to read the e-mail right away, because the words “alert” and “unusual” would catch their attention, and would make them want to determine what was going on and why a different amount was being withdrawn. For Round 2, the payment amount was increased because some participants found it too low and would not directly answer questions about what they would do if they could not afford payment. The payment breakdown tables were also adjusted to address feedback about distinguishing between principal, finance charges, and loan balance. The results for both the Upcoming Payment and Unusual Payment Notices were similar to Round 1 in that the majority of participants would open the email, thought it was legitimate and from the lender, and understood the purpose. For the consumer rights notice (referred to an “expired authorization notice” in the report), FMG asked questions about participant reactions to the notice, participant understanding of why the notice was being sent, and what participants might do in response to the notice information. As with the payment notices, these notices were presented as a sequence to simulate an email message. In Round 1, participants generally understood that the lender had tried twice to withdraw money from their account and would not be able to make any additional attempts to withdraw payment. Most participants expressed disappointment with themselves for being in a position where they had two failed payments and interpreted the notice to be a reprimand from the lender. 141 For Round 2, the notice was edited to clarify that the lender was prohibited by Federal law from making additional withdrawals. For example, the email subject line was changed from “Willow Lending can no longer withdraw loan payments from your account” to “Willow Lending is no longer permitted to withdraw loan payments from your account.” Instead of simply saying “federal law prohibits us from trying to withdraw payment again,” language was added to both the e-mail message and the full notice saying, “In order to protect your account, federal law prohibits us from trying to withdraw payment again.” More information about consumer rights and the CFPB was also added. Some participants in Round 2 still reacted negatively to this notice and viewed it as reflective of something they did wrong. However, several reacted more positively to this prototype and viewed the notice as protection. To obtain feedback about consumer preferences on receiving notices through text message, participants were also presented with an image of a text of the consumer rights notice and asked how they would feel about getting this notice by text. Overall, the majority of participants in Round 1 (8 of 13) disliked the idea of receiving notices via text. One of the main concerns was privacy; many mentioned that they would be embarrassed if a text about their loan situation displayed on their phone screen while they were in a social setting. In Round 2, the text image was updated to match the new subject line of the consumer rights notice. The majority (10 of the 14) of participants had a negative reaction to the notification delivered via text message. Despite this, the majority of participants said that they would still open the text message and view the link. Most participants (25 out of 28) also listened to a mock voice message of a lender contacting the participant to obtain renewed payment authorization after two payment attempts 142 had failed. In Round 1, most participants reported feeling somewhat intimidated by the voicemail message and were inclined to reauthorize payments or call back based on what they heard. Participants had a similar reaction to the voicemail message in Round 2. F. The Bureau’s Proposal Overview. In June 2016, the Bureau released for public comment a notice of proposed rulemaking on payday, vehicle title, and certain high-cost installment loans, which were referred to as “covered loans.” The proposal was published in the Federal Register in July 2016.359 Pursuant to its authority under the Dodd-Frank Act,360 the Bureau proposed to establish new regulatory provisions to create consumer protections for certain consumer credit products. The proposed rule was primarily grounded on the Bureau’s authority to identify and prevent unfair, deceptive, or abusive acts or practices,361 but also drew on the Bureau’s authority to prescribe rules and make exemptions from such rules as is necessary or appropriate to carry out the purposes and objectives of the Federal consumer financial laws,362 its authority to facilitate supervision of certain non-bank financial service providers (including payday lenders),363 and its authority to require disclosures to convey the costs, benefits, and risks of particular consumer financial products or services.364 In the proposal, the Bureau stated its concern that lenders that make covered loans have developed business models that deviate substantially from the practices in other credit markets by failing to assess consumers’ ability to repay their loans and by engaging in harmful practices 359 81 FR 47864 (July 22, 2016). Public Law 111–203, 124 Stat. 1376 (2010). 361 Dodd-Frank Act section 1031(b). 362 Dodd-Frank Act section 1022(b). 363 Dodd-Frank Act section 1024(b)(7). 364 12 U.S.C. 5532. 360 143 in the course of seeking to withdraw payments from consumers’ accounts. The Bureau preliminarily concluded that there may be a high likelihood of consumer harm in connection with these covered loans because a substantial population of consumers struggles to repay their loans and find themselves ending up in extended loan sequences. In particular, these consumers who take out covered loans appear to lack the ability to repay them and face one of three options when an unaffordable loan payment is due: take out additional covered loans, default on the covered loan, or make the payment on the covered loan and fail to meet other major financial obligations or basic living expenses. Many lenders may seek to obtain repayment of covered loans directly from consumers’ accounts. The Bureau stated its concern that consumers may be subject to multiple fees and other harms when lenders make repeated unsuccessful attempts to withdraw funds from consumers’ accounts. Scope of the Proposed Rule. The Bureau’s proposal would have applied to two types of covered loans. First, it would have applied to short-term loans that have terms of 45 days or less, including typical 14-day and 30-day payday loans, as well as single-payment vehicle title loans that are usually made for 30-day terms. Second, the proposal would have applied to longer-term loans with terms of more than 45 days that have (1) a total cost of credit that exceeds 36 percent; and (2) either a lien or other security interest in the consumer’s vehicle or a form of “leveraged payment mechanism” that gives the lender a right to initiate transfers from the consumer’s account or to obtain payment through a payroll deduction or other direct access to the consumer’s paycheck. Included among covered longer-term loans was a subcategory of loans with a balloon payment, which require the consumer to pay all of the principal in a single payment or make at least one payment that is more than twice as large as any other payment. 144 The Bureau proposed to exclude several types of consumer credit from the scope of the proposal, including: (1) loans extended solely to finance the purchase of a car or other consumer good in which the good secures the loan; (2) home mortgages and other loans secured by real property or a dwelling if recorded or perfected; (3) credit cards; (4) student loans; (5) nonrecourse pawn loans; and (6) overdraft services and lines of credit. Underwriting Requirements for Covered Short-Term Loans. The proposed rule preliminarily identified it as an unfair and abusive practice for a lender to make a covered shortterm loan without reasonably determining that the consumer will have the ability to repay the loan, and would have prescribed requirements to prevent the practice. Before making a covered short-term loan, a lender would first be required to make a reasonable determination that the consumer would be able to make the payments on the loan and be able to meet the consumer’s other major financial obligations and basic living expenses without needing to re-borrow over the ensuing 30 days. Specifically, a lender would have to:  verify the consumer’s net income;  verify the consumer’s debt obligations using a national consumer report and, if available, a consumer report from a “registered information system” as described below;  verify the consumer’s housing costs or use a reliable method of estimating a consumer’s housing expense based on the housing expenses of similarly situated consumers;  estimate a reasonable amount of basic living expenses for the consumer— expenditures (other than debt obligations and housing costs) necessary for a 145 consumer to maintain the consumer’s health, welfare, and ability to produce income;  project the amount and timing of the consumer’s net income, debt obligations, and housing costs for a period of time based on the term of the loan; and  determine the consumer’s ability to repay the loan and continue paying other obligations and basic living expenses for a period of thirty days thereafter based on the lender’s projections of the consumer’s income, debt obligations, and housing costs and estimate of basic living expenses for the consumer. Under certain circumstances, a lender would be required to make further assumptions or presumptions when evaluating a consumer’s ability to repay a covered short-term loan. The proposal specified certain assumptions for determining the consumer’s ability to repay a line of credit that is a covered short-term loan. In addition, if a consumer were to seek a covered shortterm loan within 30 days of a covered short-term or longer-term balloon-payment loan, a lender generally would be required to presume that the consumer is not able to afford the new loan. A lender could overcome the presumption of unaffordability for a new covered short-term loan only if it could document a sufficient improvement in the consumer’s financial capacity. Furthermore, a lender would have been prohibited for a period of 30 days from making a covered short-term loan to a consumer who has already taken out three covered short-term loans within 30 days of each other. Under the proposal, a lender would also have been allowed to make a covered short-term loan without complying with all the underwriting criteria just specified, as long as the conditionally exempt loan satisfied certain prescribed terms to prevent and mitigate the risks and 146 harms of unaffordable loans leading to extended loan sequences, and the lender confirmed that the consumer met specified borrowing history conditions and provided required disclosures to the consumer. Among other conditions, a lender would have been allowed to make up to three covered short-term loans in short succession, provided that the first loan had a principal amount no larger than $500, the second loan had a principal amount at least one-third smaller than the principal amount on the first loan, and the third loan had a principal amount at least two-thirds smaller than the principal amount on the first loan. In addition, a lender would not have been allowed to make a covered short-term loan under the alternative requirements if it would result in the consumer having more than six covered short-term loans during a consecutive 12-month period or being in debt for more than 90 days on covered short-term loans during a consecutive 12-month period. Under the proposal, a lender would not be permitted to take vehicle security in connection with these loans. Underwriting Requirements for Covered Longer-Term Loans. The proposed rule would have identified it as an unfair and abusive practice for a lender to make certain covered longerterm loans without reasonably determining that the consumer will have the ability to repay the loan. The coverage would have been limited to high-cost loans of this type and for which the lender took a leveraged payment mechanism, including vehicle security. The proposed rule would have prescribed requirements to prevent the practice for these loans, subject to certain exemptions and conditions. Before making a covered longer-term loan, a lender would have had to make a reasonable determination that the consumer has the ability to make all required payments as scheduled. This determination was to be made by focusing on the month in which the payments under the loan would be the highest. The proposed ability-to-repay requirements 147 for covered longer-term loans closely tracked the proposed requirements for covered short-term loans with an added requirement that the lender, in assessing the consumer’s ability to repay a longer-term loan, must reasonably account for the possibility of volatility in the consumer’s income, obligations, or basic living expenses during the term of the loan. The Bureau has determined not to finalize this aspect of the proposal at this time (other than for covered longer-term balloon-payment loans), and will take any appropriate further action on this subject after the issuance of this final rule. Payments Practices Related to Small-Dollar Loans. The proposed rule would have identified it as an unfair and abusive practice for a lender to attempt to withdraw payment from a consumer’s account in connection with a covered loan after the lender’s second consecutive attempt to withdraw payment from the account has failed due to a lack of sufficient funds, unless the lender obtains from the consumer a new and specific authorization to make further withdrawals from the account. This prohibition on further withdrawal attempts would have applied whether the two failed attempts are initiated through a single payment channel or different channels, such as the automated clearinghouse system and the check network. The proposed rule would have required that lenders provide notice to consumers when the prohibition has been triggered and follow certain procedures in obtaining new authorizations. In addition to the requirements related to the prohibition on further payment withdrawal attempts, the proposed rule would require a lender to provide a written notice at least three business days before each attempt to withdraw payment for a covered loan from a consumer’s checking, savings, or prepaid account. The notice would have contained key information about the upcoming payment attempt, and, if applicable, alerted the consumer to unusual payment 148 attempts. A lender could provide electronic notices as long as the consumer consented to electronic communications. Additional Requirements. The Bureau also proposed to require lenders to furnish to provisionally registered and registered information systems certain information concerning covered loans at loan consummation, any updates to that information over the life of the loan, and certain information when the loan ceases to be outstanding. To be eligible to become a provisionally registered or registered information system, an entity would have to satisfy the eligibility criteria prescribed in the proposed rule. The Bureau proposed a sequential process to allow information systems to be registered and lenders to be ready to furnish at the time the furnishing obligation in the proposed rule would take effect. For most covered loans, registered information systems would provide a reasonably comprehensive record of a consumer’s recent and current borrowing. Before making most covered loans, a lender would have been required to obtain and consider a consumer report from a registered information system. The proposal would require a lender to establish and follow a compliance program and retain certain records, which included developing and following written policies and procedures that are reasonably designed to ensure compliance with the proposed requirements. A lender would also be required to retain the loan agreement and documentation obtained for a covered loan, and electronic records in tabular format regarding origination calculations and determinations for a covered loan, for a consumer who qualifies for an exception to or overcomes a presumption of unaffordability for a covered loan, and regarding loan type, terms, payment history, and loan performance. The proposed rule also included an anti-evasion clause and a severability clause. 149 Effective Date. The Bureau proposed that, in general, the final rule would become effective 15 months after publication of the final rule in the Federal Register. It also proposed that certain provisions necessary to implement the consumer reporting components of the proposal would become effective 60 days after publication of the final rule in the Federal Register to facilitate an orderly implementation process. G. Public Comments on the Proposed Rule Overview. Reflecting the broad public interest in this subject, the Bureau received more than 1.4 million comments on the proposed rulemaking. This is the largest comment volume associated with any rulemaking in the Bureau’s history. Comments were received from consumers and consumer advocacy groups, national and regional industry trade associations, industry participants, banks, credit unions, nonpartisan research and advocacy organizations, members of Congress, program managers, payment networks, payment processors, fintech companies, Tribal leaders, faith leaders and coalitions of faith leaders, and State and local government officials and agencies. The Bureau received well over 1 million comments from individuals regarding the proposed rule, often describing their own circumstances or those of others known to them in order to illustrate their views, including their perceptions of how the proposed rule might affect their personal financial situations. Some individuals submitted multiple separate comments. The Bureau has not attempted to tabulate precise results for how to tally the comments on both sides of the rule. Nor would it be easy to do so in any practical way, and of course some of the comments did not appear to take a side in advocating for or against the rule, though only a small number would fall in this category. Nonetheless, it was possible to achieve a rough 150 approximation that broke down the universe of comments in this manner and the Bureau made some effort to do so. As an approximation, of the total comments submitted, more than 300,000 comments generally approved of the Bureau’s proposal or suggested that the Bureau should adopt a rule that is more restrictive of these kinds of loans in some way or other. Over one million comments generally opposed the proposed rule and took the view that its provisions would be too restrictive of these kinds of loans. The Bureau received numerous submissions generated through mass mail campaigns and other organized efforts, including signatures on a petition or multiple letters, postcards, emails, or web comments. These campaigns were conducted by opponents and supporters of the proposed rule. The Bureau also received stand-alone comments submitted by a single commenter, individual, or organization. Of the approximately 1.4 million comments submitted, a substantial majority were generated by mass-mail campaigns or other organized efforts. In many cases, these submissions contained the same or similar wording. Of those 1.4 million comments, approximately 300,000 were handwritten and often had either the same or similar content or advanced substantially similar themes and arguments. These comments were posted as attachments to the electronic docket at www.regulations.gov. For many of the comments that were submitted as part of mass mail campaigns or other organized efforts, a sample comment was posted to the electronic docket at www.regulations.gov, with the total number of such comments received reflected in the docket entries. Accordingly, these comments, whose content is represented on the electronic docket via 151 the sample comment, were not individually posted to the electronic docket at www.regulations.gov. In addition, the 1.4 million comments included more than 100,000 signatures or comments contained on petitions, with some petitions containing tens of thousands of signatures. These petitions were posted as attachments to the electronic docket at www.regulations.gov. Whenever relevant to the rulemaking, these submissions and comments were considered in the development of the final rule. Form of Submission. As detailed in the proposed rule,365 the Bureau accepted comments through four methods: email, electronic,366 regular mail, and hand delivery or courier (including delivery services like FedEx). Approximately 800,000 comments, or roughly 60% of the total, were paper comments received by mail or couriers, while approximately 600,000 (or about 40%) were submitted electronically, either directly to the electronic docket at www.regulations.gov or by email. The electronic submissions included approximately 100,000 scanned paper comments sent as PDF attachments to thousands of emails. In addition, the Bureau also processed and considered comments that were received after the comment period had closed, as well as more than 50 ex parte submissions. The ex parte materials were generally presentations and summary memoranda relevant to the rulemaking that were provided to Bureau personnel in the normal course of their work, but outside the procedures for submitting written comments to the rulemaking docket referenced above. They were 365 366 See 81 FR 47863 (July 22, 2016). Electronic submissions were made via http://www.regulations.gov. 152 considered in accordance with the Bureau’s established rulemaking procedures governing ex parte materials. Materials on the record, including ex parte submissions and summaries of ex parte meetings and telephone conferences, are publicly available at www.regulations.gov. Other relevant information is discussed below as appropriate. In the end, the Bureau considered all of the comments it received about the proposed rule prior to finalizing the rule. Stand-Alone Comments. Tens of thousands appear to have been “stand-alone” comments – comments that did not appear to have been submitted as part of a mass mail campaign or other organized effort. Nevertheless, many of these stand-alone comments contained language and phrasing that were highly similar to other comments. In addition, pre-printed postcards or other form comments with identical language submitted as part of an organized effort sometimes also included additional notations, such as “we need this product” or “don’t take this away.” Some comment submissions also attached material, including copies of news articles, loan applications, loan advertisements, and even personal financial documents. Many of the comments from lenders, trade associations, consumer advocacy groups, research and advocacy organizations, and government officials included specific discussion about particular provisions of the proposed rule, and the substantive issues raised in those comments are discussed in connection with those provisions. However, as noted above, a high volume of comments were received from individuals, rather than from such entities (or their official representatives). Many of these individual comments focused on personal experiences rather than legal or financial analysis of the details of the provisions of the proposed rule. The discussion below summarizes what the commenters—more than a million in total—had to say to 153 the Bureau about the proposed rule. The comments can be broken into three general categories: (1) individual comments made about the rule that were more factual in nature regarding the uses and benefits of covered short-term loans; (2) individual comments stating or explaining the grounds on which the commenters opposed the rule, both generally and in more specific respects; and (3) individual comments stating or explaining the grounds on which the commenters supported the rule, again both generally and in more specific respects. The individual comments as so categorized are set forth below, and they have helped inform the Bureau’s consideration of the issues involved in deciding whether and how to finalize various aspects of the proposed rule. Comments Not Specifically Supporting or Opposing the Rule. Many commenters noted, as a factual matter, the uses they make of covered short-term loans. These uses include: rent, childcare, food, vacation, school supplies, car payments, power/utility bills, cell phone bills, credit card bills, groceries, medical bills, insurance premiums, student educational costs, daily living costs, gaps between paychecks, money to send back to a home country, necessary credit, to “make ends meet,” “hard times,” and “bills.” In considering these types of comments, the Bureau generally interpreted them as critical of the rule for going too far to regulate covered short-term loans. Some individual commenters talked about how they would cover various costs and expenses if the rule caused previously available payday loans to become less available or unavailable. Among the alternatives they cited were credit cards, borrowing from family or friends, incurring NSF or overdraft charges, or seeking bank loans. 154 The comments included many suggestions about the consumer financial marketplace that reached beyond the scope of the proposed rule. Some of these comments suggested that the Bureau should regulate interest rates or limit the amounts that could be charged for such loans by imposing a nationwide usury cap. Comments Opposing the Proposed Rule. The nature of criticism varied substantially. Some commenters were broadly opposed to the rule without further explanation, while others objected to the government’s participation in regulating the activity affected by the rule. Some objected to the means by which the rule was being considered or enacted while others objected to various substantive aspects of the rule. Some commenters combined these various types of criticisms. Unexplained opposition included some very brief comments like “No” or “Are you crazy?” Others based their opposition on general anti-government sentiments. Some objected simply to the fact of the rulemaking. These objections included comments like “I’m against Washington stopping me from getting a loan.” More specific comments stated that the government should not be in the business of limiting how much people can borrow and that consumers can manage their own funds. Others contended that similar regulatory efforts in other countries had been unsuccessful. Some were opposed on the ground that the proposed rule was too complicated, with a few objecting simply to its length and complexity or its reliance on dated evidence. A considerable number of commenters, including some State and local governmental officials, opined that existing State laws and regulations adequately addressed any regulatory need in this area. Some suggested that any regulation of covered short-term loans should be left 155 to the States or that the Bureau should “work with state governments.” Some suggested that the Bureau had not adequately consulted with State officials before proposing the rule. And though the specific intent of the comments was not always made clear, some suggested that, either in promulgating or implementing the rule, the Bureau should consult State law and compare different rates and requirements in different States. Some comments were implicitly critical of the proposal, even if not expressly so, when they proposed alternative approaches like the suggestion that the Bureau “should follow the Florida Model.” Many comments were from individuals who indicated they were users of payday loans, were able to reliably pay them back, and objected to new restrictions. Some of those comments came with notations that they had been specifically asked by loan providers to submit such comments. Many opposed the rule in whole or in part. Some supported some parts of the rule and opposed other parts. Hundreds of thousands of individuals submitted comments generally supporting the availability of small-dollar loans that would have been covered by the proposed rule. Many but not all were submitted by consumers of these loans, who mentioned their need for access to small loans to address financial issues they faced with paying bills or dealing with unexpected expenses. Certain consumers stated that they could not access other forms of credit and favored the convenience and simplicity of these loans. Many expressed their opposition to caps or limits on the number of times they would be able to borrow money on such loans. As noted above, many commenters simply indicated that they like and use payday loans. The Bureau generally understood these comments as expressions of concerns that the proposed rule might or would restrict their access to covered loans. In contending for greater availability 156 of such loans, commenters specifically noted their use of payday loans for a substantial range of financial needs and reasons. They explained that these loans are used to cover, among other financial needs, overdraft fees, the last piece of tuition rather than losing enrollment, a portion of rent so as not to incur a rent penalty, various bills so as to avoid incurring late fees, utilities so they would not be turned off, college student necessities not covered by student loans, and funds to cover a gap in available resources before the next paycheck. Several commenters specifically noted that payday loan costs were cheaper than bank overdraft fees that would otherwise be incurred. Some indicated they had no alternative to payday products because they lacked credit for credit cards and could not borrow from family or friends or relatives. Some commenters focused on the favorable environment they experienced in using payday loans, often in juxtaposition to their less welcoming experience with banks. A number of loan providers commented that low-income, non-English speaking immigrants are treated well by those who make these loans to them. Various borrowers related that they have been treated well at payday storefronts and that employees are helpful with their loan applications. Others indicated that local communities support local payday lenders and the loans they provide and these lenders in turn are leading small businesspersons in their communities. Others noted that payday lenders often provide other services like check cashing, bill paying, and loading of pre-paid cards, sometimes with no fees. Still others echoed that payday lenders do more than other lenders to help their individual customers, and are all about “finding a solution” for the customer. Some commented that payday lenders do not pressure customers to take out loans whereas banks do. 157 One commenter noted that even with substantial income, payday loans still provided convenience due to a favorable ongoing relationship with the lender. Others commented more generally that the loans are convenient because they require no application and no credit check, they are easy to get and easy to renew, and they are provided at locations where it is convenient to get a check cashed. One expressly noted that despite the recognized expense of such loans, their availability and convenience made them worth it. Various commenters noted that small loans were difficult or impossible to obtain from banks. Others objected that banks require too much personal information when lending funds, like credit checks and references. Some noted that they had a poor credit history or insufficient credit history and therefore could not get loans from banks or credit cards. Some indicated that small-dollar loans may be necessary for assuring available cash flow at some small businesses. These commenters indicated that payday loans are often critical when bank loans have been denied, the business is awaiting customer payments, and funds are needed to make payroll. Some said that alternatives were unsafe or unable to meet their needs. Others claimed that pawn shops have a bad reputation, that loan sharks might be an available option but for the possible “outcome,” and foreign and “underground” lenders were not viable options. Some merely signed their name to the contents of printed text. Others sometimes added related messages in filling out such forms. Other forms provided space for and encouraged individualized messages and explanations rather than simply presenting uniform prepared text. Some comments opposing the proposed rule were submitted by lender employees, and those comments also ranged fairly widely in the extent of their individualized content; some referred to their fears of losing their jobs if the proposed rule were to become effective in its current form. 158 Some of these commenters indicated that payday loan proceeds were used to pay bills for which non-payment would result in penalties or late fees or suspension of vital services; many of them expressed, or seemed implicitly to suggest, concern that the rule would restrict their access to funds for meeting these needs. Some commenters discussed general or specific concerns about their understanding of the effect the rule would have without expressly indicating support for or opposition to the rule, though a fair reading of their comments showed them to be expressing concern that the proposed rule would, or might restrict their access to covered loans and thus appeared to be critical of the proposed rule. For example, specific concerns about the perceived negative effects of the rule included its potential effect on the cost of covered loans, including fees and interest rates, restrictions on product availability because of re-borrowing limits, and lack of clarity about what products would replace those made unavailable by the rule. A number of comments expressed concern or confusion about the alternative lending options they would have following the enactment of the rule, and whether these alternatives would be acceptable options. Some had very specific concerns about the potential effects of the rule, including a potential lack of liquidity in the market, and expressed a general concern that the rule might lead to increased consumer fraud. Others were concerned about the security of the personal financial information they would have to provide to get a loan. Some expressed concern that the new requirements would lead to loan denials that would hurt their credit scores. Many employees of the lenders affected by the proposed rule were concerned about their continued employment status if the rule were to be adopted. 159 Some commenters proposed exclusions from the effects of the rule, either directly or indirectly, indicating, for example, the auto title or credit union loans should be unaffected by the final rule. It was also suggested that there should be a safe harbor if lenders do their own underwriting or engage in income verification. Others suggested that various types of lenders should be excluded from the rule. These included credit unions, on the ground that they make “responsible” loans that use the ability to repay as an eligibility screen already, and “flex loans” because they are like lines of credit. At least one commenter suggested that the Bureau should exempt FDIC-regulated banks from any coverage under the rule. In addition to more general criticisms of the rule, individual commenters also offered objections and concerns about the substantive provisions of the proposed rule. Some were general, like the suggestion that repayment should be more flexible. Others were more focused on specific features of the rule, including claims that the proposed rule would violate existing laws in unspecified ways. Many commenters were concerned about the burdens and length of the “30-day waiting period” or cooling-off period, noting that they would be unable to access such loans during those periods even if they had an urgent need for funds. Others similarly commented that the various requirements and restrictions would result in loan denials and impede their ability to access needed funds easily and quickly. Many specifically noted the need for funds for unexpected emergencies, like car repairs. Some simply declared these limits “unwarranted,” saying that they understood the risks associated with these loans and appreciated their availability nonetheless. Some commenters focused on the procedural difficulties of obtaining covered loans under the rule. They objected to the length and detail of the loan application process when funds 160 were needed quickly and easily to cope with emergencies, with car repairs cited frequently. They stated that the process for getting a small-dollar loan should be short and easy and that otherwise it was not worth the effort. Others felt that the proposed rule would require them to disclose too much information about their income and expenses, which would invade their privacy. Some stated that credit checks should not be required for small-dollar loans. Still others expressed concern that the government should not be able to demand such information or require that borrowers provide it. A few commenters noted that it would be hard for lenders to comply with the rule, which would impose additional compliance costs. A few specifically suggested that the Bureau should consider having lenders use the State databases that lenders must currently use rather than the approach laid out in the proposed rule. Finally, though the vast majority of critical comments opposed the proposed rule and the restrictions it would impose, a substantial number of individual commenters were critical because they did not believe the rule went far enough or imposed enough restrictions. These included views that allowing consumers to receive as many as six loans a year or more would sink them into further debt, that “big banks” would benefit from the rule, or that the rule should “go after big banks” rather than smaller payday lenders. Many critics of the proposed rule stated that it should more directly impose a cap on interest rates, as many States have done and as has proved effective in limiting the making of these kinds of loans. Others suggested that the proposed rule could have “unintended consequences,” though without clearly explaining what those consequences might be, and that more should be done to prevent them. 161 Comments Supporting the Proposed Rule. Many individuals submitted comments that either supported the thrust of the proposed rule or argued that it needed to be strengthened in particular ways to accomplish its purposes. Some were submitted by consumers of these loans, and others were submitted through groups such as nonprofit organizations or coalitions of faith leaders who organized the presentation of their individual stories. Many were submitted as part of campaigns organized by consumer advocacy groups and a variety of nonprofit organizations concerned about the dangers they perceived to flow from these types of loans. These comments tended to dwell on the risks and financial harms that many consumers incur from small-dollar loans. These accounts consistently centered on those borrowers who find themselves ending up in extended loan sequences and bearing the negative collateral consequences of re-borrowing, delinquency, and default, especially the inability to keep up with their other major financial obligations and the loss of control over their budgetary decisions. Many of these commenters cited the special risks posed by loans that are extended without a reasonable determination of the consumer’s ability to repay the loan without re-borrowing. Some went further and urged that such loans be outlawed altogether based on their predatory nature and the extremely high costs to consumers of most of these loan products. Some of these comments described their first-hand experiences with extended loan sequences and the financial harms that had resulted either to themselves or to friends or family members. Some colored their accounts with considerable anger and frustration about these experiences, how they were treated, and the effects that these loans had in undermining or ruining their financial situations. 162 Many comments were generated or collected by faith leaders and faith groups, with individuals often presenting their views in terms of moral considerations, as well as financial effects. Some of these comments cited scripture and offered religiously based objections to covered loan activity, with particular opposition to the high interest rates associated with covered loans. Others, without necessarily grounding their concerns in a specific religious orientation, noted that current covered loans harm certain financially vulnerable populations, including the elderly, low-income consumers, and single mothers. They also recounted efforts they and others had made to develop so-called “rescue” products to extricate members of their congregations from the cumulative harms of extended loan sequences. Some employees of lenders, especially credit unions, offered views in favor of the proposed rule based on what they had seen of the negative experiences that their customers had encountered with these types of loans. Many commenters who favored the proposed rule dwelled on their concerns about the risks posed by the types of covered loans that are currently available to consumers. Overall, these comments tended to focus on the risks and financial harms that many consumers incur when using short-term small-dollar loans. They expressed concerns about borrowers who find themselves in extended loan sequences and bearing increasingly negative effects as a result. Commenters often stressed that these situations left consumers unable to keep up with other major financial obligations and that they lost control over their personal budgetary decisions. Like the favorable comments regarding current payday loan activity – which the Bureau understood to be critical of the proposed rule – critics of current covered loan practices did not always specify their views about the proposed rule. Nonetheless, absent specific indications to 163 the contrary, comments that were critical of current payday lending activity were understood to be supportive of the proposed rule as an effective potential response to those concerns. Some comments simply indicated a general policy view that there was a need to “stop the debt trap” or that rollover loans were “out of hand.” Others objected to the perception that covered loans are “geared to people with fixed incomes.” Many opposed what they viewed as the common situation that these loans were unaffordable and put people in a position in which they are unable to pay off the principal and must roll over the loans to avoid default. Some comments focused on the specific consumer protective nature of the proposed rule, indicating that the rule was needed because current lenders do not care about people’s ability to repay the loans, knowing that they can profit from continuing re-borrowing. A handful of comments from current or former employees of such lenders said they supported the proposed rule because of the negative experiences they had seen their customers encounter with these types of loans. One commenter opined that even NSF fees were less damaging to consumers than the cumulative effects of these loans, with the fees they imposed and frequency with which they landed many consumers in continued debt traps. Many others commenting on these types of loans indicated that their “debt trap” nature was reinforced in the context of vehicle loans, since repossession of a vehicle could dramatically deepen the downward debt spiral. Still, one commenter argued that even the repossession of the borrower’s vehicle might not be as bad as the continuing predicament of self-perpetuating loan sequences with their escalating fees and loan balances. Some indicated that other loans were better alternatives to payday loans, sometimes citing PAL loans in this regard. And some were concerned about the character of the lenders 164 associated with covered loans, with one comment relating that a recent payday lender had been indicted for illegal conduct associated with payday lending. Some individual commenters indicated that they were representatives of or otherwise affiliated with national consumer organizations, and other national organizations, and were supportive of the rule. Some commenters noted that they were current payday loan borrowers working to pay off their loans and were supportive of the rule. Others supported the rule based on their own generally negative personal experiences with covered loans, with some specifying that they only supported the rule as applied to lenders that made loans without determining whether borrowers had the ability to repay them. Many individual commenters indicated support for time limits on these loans and the proposed “cooling-off period” because they believed it would ultimately help consumers better manage their funds. Some thought that the rule would have the effect of lowering interest rates. Some individual commenters who identified themselves as State officials, including individual legislators, commented that the rule would favorably supplement existing statutes that dealt with covered loans in their respective States. Individuals affiliated with some industry groups indicated their general support for the rule, but expressed concern that, in unexplained ways, the rule may go “too far.” In contrast, others recommended that the standards in the proposed rule should be applied in the context of all consumer lending rather than just in this market. The Bureau’s Consideration of Individual Comments. Although the specific treatment of discrete issues is addressed more fully in part V below, which presents the section-by-section analysis explaining the components of the final rule, it may be useful here to provide some of the 165 uses that the Bureau made of the individual comments. First, it is a notable and commendable fact that over a million individual commenters would take the time and effort to respond to the Bureau with their thoughts and reactions, both pro and con, to this proposed rule. Public comments are not just an obligatory part of the rulemaking process required by the Administrative Procedure Act, they are welcome as a means of providing insight and perspective in fashioning such rules. Perhaps needless to say, that inviting solicitation was put to the test here. As noted earlier, many of the individual comments turned out to be duplicative and redundant of one another. In part, that was because both the industry groups, on the one side, and the consumer and community groups, on the other side, employed campaigns to solicit large numbers of individual comments. The Bureau does not view any of those efforts as improper or illegitimate, and it has not discounted any comments on their merits as a result of their apparent origins. It did create challenges, however, for figuring out how to manage this large volume of comments – how to receive and process them, how to handle and organize them, and how to review and consider them. In the end, the Bureau proceeded as laid out in its earlier discussion in this section, and though the process took many months and considerable effort, it was eventually completed in a satisfactory way. The Bureau also does not view the repetition and redundancy among many of the comments as being immaterial. The Bureau considered not only what views the public has, but how intensely they are felt and maintained. The Bureau has frequently noted, in its handling of consumer complaints, that when the same concern arises more frequently, it may reflect an emerging pattern and be worthy of more attention than if the same concern arises only once or 166 twice and thus appears to reflect a more isolated set of circumstances. The same may be true here, with the caveat that, depending on the circumstances, comments generated primarily through campaigns may or may not truly reflect any widespread or deeply felt convictions, depending on the level of the individual’s actual involvement. Having said that, the processes that Congress has created for Federal administrative rulemaking, both in the Administrative Procedure Act generally and here in the Dodd-Frank Act in particular, were not designed or intended to be governed by some rough assessment of majority vote or even majority sentiment. While rough estimates of pro and con submissions are provided above, the Bureau has simply sought to understand the consumer experiences reported in these comments and address the substance of these comments on their merits. As a general matter, the individual comments have helped inform the Bureau’s understanding of factual matters surrounding the circumstances and use of covered loans. In the sections on Market Concerns—Underwriting and Market Concerns—Payments, they helped add depth and content to the Bureau’s description of issues such as borrower characteristics, the circumstances of borrowing, their expectations of and experience with extended loan sequences, including harms they have suffered as a consequence of delinquency, default, and loss of control over budgeting. Many of these concerns were already known at the outset of the rule-writing process, as a result of extensive outreach and feedback the Bureau has received on the subject, as well as through the research that the Bureau and others have performed on millions of covered loans, all of which is discussed above. Nonetheless, the Bureau’s review of large numbers of individual comments has reinforced certain points and prompted further consideration of others. For example, many 167 individuals stated great concern that the proposed rule would make the underwriting process for small-dollar loans too burdensome and complex. They commented positively on the speed and convenience of obtaining such loans, and were concerned that the process described in the proposed rule would lead to fewer such loans being offered or made. This has influenced the Bureau’s consideration of the details of the underwriting process addressed in § 1041.5 of the final rule and contributed to the Bureau’s decision to modify various aspects of that process. At the same time, many other individual commenters had much to say about the perils of extended loan sequences and how they had harmed either themselves or others, which helped underscore the need for the Bureau to finalize a framework that would be sufficiently protective of consumers. In particular, many commenters supported the general requirement that lenders must reasonably assess the borrower’s ability to repay before making a loan according to specific underwriting criteria, and that limited exceptions to those criteria would be made only where other conditions applied to ensure that lenders would not end up in extended loan sequences. There are also many other places in the Bureau’s discussion and explanation of the final rule where individual comments played a role in the Bureau’s analysis. Further Inter-Agency Consultation. In addition to the inter-agency consultation that the Bureau engaged in prior to issuing the Notice of Proposed Rulemaking, pursuant to section 1022(b)(2) of the Dodd-Frank Act, the Bureau has consulted further with the appropriate prudential regulators and the FTC during the comment process. As a result of these consultations, the Bureau has made a number of changes to the rule and has provided additional explanation for various determinations it has made about the provisions of the rule, which have been discussed with the other regulators and agencies during the consultation process. 168 Ex Parte Submissions. In addition, the Bureau considered the comments it received after the comment period had closed, as well as other input from more than 50 ex parte submissions, meetings, and telephone conferences.367 All such materials in the record are available to the public at http://www.regulations.gov. Relevant information received is discussed below in the section-by-section analysis and subsequent parts of this notice, as applicable. The Bureau considered all the comments it received about the proposal, made certain modifications, and is adopting the final rule as described more fully in part V below. IV. Legal Authority The Bureau is issuing this final rule pursuant to its authority under the Dodd-Frank Act. The rule relies on rulemaking and other authorities specifically granted to the Bureau by the Dodd-Frank Act, as discussed below. A. Section 1031 of the Dodd-Frank Act Section 1031(b)—The Bureau’s Authority to Identify and Prevent UDAAPs Section 1031(b) of the Dodd-Frank Act provides the Bureau with authority to prescribe rules to identify and prevent unfair, deceptive, or abusive acts or practices, or UDAAPs. Specifically, section 1031(b) of the Act authorizes the Bureau to prescribe rules “applicable to a covered person or service provider identifying as unlawful unfair, deceptive, or abusive acts or practices in connection with any transaction with a consumer for a consumer financial product or service, or the offering of a consumer financial product or service.” Section 1031(b) of the Act 367 See also Bureau of Consumer Fin. Prot., “CFPB Bulletin 11–3, CFPB Policy on Ex Parte Presentations in Rulemaking Proceedings,” (Aug. 16, 2011), available at http://files.consumerfinance.gov/f/2011/08/Bulletin_20110819_ExPartePresentationsRulemakingProceedings.pdf, updated and revised, Policy on Ex Parte Presentations in Rulemaking Proceedings, 82 FR 18687 (Apr. 21, 2017). 169 further provides that, “Rules under this section may include requirements for the purpose of preventing such acts or practice.” There are notable similarities between the Dodd-Frank Act and the Federal Trade Commission Act (FTC Act) provisions relating to unfair and deceptive acts or practices. Accordingly, these FTC Act provisions, and case law and Federal agency rulemakings relying on them, inform the scope and meaning of the Bureau’s rulemaking authority with respect to unfair and deceptive acts or practices under section 1031(b) of the Dodd-Frank Act.368 Courts evaluating exercise of agency rulemaking authority under the unfairness and deception standards of the FTC Act have held that there must be a “reasonable relation” between the act or practice identified as unlawful and the remedy chosen by the agency.369 The Bureau agrees with this approach and therefore maintains it is reasonable to interpret section 1031(b) of the Dodd-Frank Act to permit the imposition of requirements to prevent acts or practices that are identified by the Bureau as unfair or deceptive, as long as the preventive requirements being imposed by the Bureau have a reasonable relation to the identified acts or practices. The Bureau likewise maintains that it is reasonable to interpret section 1031(b) of the Dodd-Frank Act to provide that same degree of discretion to the Bureau with respect to the imposition of requirements to prevent acts or practices that are identified by the Bureau as abusive. Throughout this rulemaking process, the Bureau has relied on and applied this 368 Section 18 of the FTC Act similarly authorizes the FTC to prescribe “rules which define with specificity acts or practices which are unfair or deceptive acts or practices in or affecting commerce” and provides that such rules “may include requirements prescribed for the purpose of preventing such acts or practices.” 15 U.S.C. 57a(a)(1)(B). As discussed below, the Dodd-Frank Act, unlike the FTC Act, also permits the Bureau to prescribe rules identifying and preventing “abusive” acts or practices. 369 See Am. Fin. Servs. Ass’n v. FTC, 767 F.2d 957, 988 (D.C. Cir. 1985) (AFSA) (holding that the FTC “has wide latitude for judgment and the courts will not interfere except where the remedy selected has no reasonable relation to the unlawful practices found to exist” (citing Jacob Siegel Co. v. FTC, 327 U.S. 608, 612–13 (1946)). 170 interpretation in formulating and designing requirements to prevent acts or practices identified as unfair or abusive. Section 1031(c)—Unfair Acts or Practices Section 1031(c)(1) of the Dodd-Frank Act provides that the Bureau “shall have no authority under this section to declare an act or practice in connection with a transaction with a consumer for a consumer financial product or service, or the offering of a consumer financial product or service, to be unlawful on the grounds that such act or practice is unfair,” unless the Bureau “has a reasonable basis” to conclude that: “(A) the act or practice causes or is likely to cause substantial injury to consumers which is not reasonably avoidable by consumers; and (B) such substantial injury is not outweighed by countervailing benefits to consumers or to competition.”370 Section 1031(c)(2) of the Act provides that, “[i]n determining whether an act or practice is unfair, the Bureau may consider established public policies as evidence to be considered with all other evidence. Such public policy considerations may not serve as a primary basis for such determination.”371 In sum, the unfairness standard under section 1031(c) of the Dodd-Frank Act requires primary consideration of three elements: the presence of a substantial injury, the absence of consumers’ ability to reasonably avoid the injury, and the countervailing benefits to consumers or to competition associated with the act or practice. The Dodd-Frank Act also permits secondary consideration of public policy objectives. As noted above, the unfairness provisions of the Dodd-Frank Act are similar to the 370 371 12 U.S.C. 5531(c)(1). 12 U.S.C. 5531(c)(2). 171 unfairness standard under the FTC Act.372 That standard was developed, in part, when in 1994, Section 5(n) of the FTC Act was amended to incorporate the principles set forth in the FTC’s December 17, 1980 “Commission Statement of Policy on the Scope of Consumer Unfairness Jurisdiction” (the FTC Policy Statement on Unfairness).373 Due to the similarities between unfairness provisions in the Dodd-Frank and FTC Acts, the scope and meaning of the Bureau’s authority under section 1031(b) of the Dodd-Frank Act to issue rules that identify and prevent acts or practices that the Bureau determines are unfair pursuant to section 1031(c) of the Dodd-Frank Act are naturally informed by the FTC Act unfairness standard, the FTC Policy Statement on Unfairness, FTC and other Federal agency rulemakings,374 and related case law. The Bureau believes it is reasonable to interpret section 1031 of the Dodd-Frank Act consistent with the specific positions discussed in this section on Legal Authority. The Bureau’s interpretations are based on its expertise with consumer financial products, services, and markets, and its experience with implementing this provision in 372 Section 5(n) of the FTC Act, as amended in 1994, provides that, “The [FTC] shall have no authority . . . to declare unlawful an act or practice on the grounds that such act or practice is unfair unless the act or practice causes or is likely to cause substantial injury to consumers which is not reasonably avoidable by consumers themselves and not outweighed by countervailing benefits to consumers or to competition. In determining whether an act or practice is unfair, the [FTC] may consider established public policies as evidence to be considered with all other evidence. Such public policy considerations may not serve as a primary basis for such determination.” 15 U.S.C. 45(n). 373 Letter from the FTC to Hon. Wendell Ford and Hon. John Danforth, Committee on Commerce, Science and Transportation, United States Senate, Commission Statement of Policy on the Scope of Consumer Unfairness Jurisdiction (December 17, 1980), reprinted in In re Int’l Harvester Co., 104 F.T.C. 949, 1070 (1984) (Int’l Harvester). See also S. Rept. 103–130, at 12–13 (1993) (legislative history to FTC Act amendments indicating congressional intent to codify the principles of the FTC Policy Statement on Unfairness). 374 In addition to the FTC’s rulemakings under unfairness authority, certain Federal prudential regulators have prescribed rules prohibiting unfair practices under section 18(f)(1) of the FTC Act and, in doing so, they applied the statutory elements consistent with the standards articulated by the FTC. The Federal Reserve Board, FDIC, and the OCC also issued guidance generally adopting these standards for purposes of enforcing the FTC Act’s prohibition on unfair and deceptive acts or practices. See 74 FR 5498, 5502 (Jan. 29, 2009) (background discussion of legal authority for interagency Subprime Credit Card Practices rule). 172 supervisory and enforcement actions. The Bureau also generally finds persuasive the reasons provided by the authorities supporting these positions as discussed in this section. Substantial Injury The first element required for a determination of unfairness under section 1031(c)(1) of the Dodd- Frank Act is that the act or practice causes, or is likely to cause, substantial consumer injury. As noted above, Bureau rulemaking regarding the meaning of the elements of this unfairness standard is informed by the FTC Act unfairness standard, the FTC Policy Statement on Unfairness, FTC and other Federal agency rulemakings, and related case law. The FTC noted in the FTC Policy Statement on Unfairness that substantial injury ordinarily involves monetary harm, and that trivial or speculative harms are not cognizable under the test for substantial injury.375 The FTC also noted that an injury is “sufficiently substantial” if it consists of a small amount of harm to a large number of individuals or if it raises a significant risk of harm.376 In addition, the FTC has also found that substantial injury may involve a large amount of harm experienced by a small number of individuals.377 And while the FTC has said that emotional impact and other more subjective types of harm ordinarily will not constitute substantial injury,378 the D.C. Circuit held that psychological harm can form part of the 375 See FTC Policy Statement on Unfairness, Int’l Harvester, 104 F.T.C. 949, 1073 (1984). For example, in the Higher-Priced Mortgage Loan (HPML) Rule, the Federal Reserve Board concluded that a borrower who cannot afford to make the loan payments as well as payments for property taxes and homeowners insurance because the lender did not adequately assess the borrower’s ability to repay suffers substantial injury, due to the various costs associated with missing mortgage payments (e.g., large late fees, impairment of credit records, foreclosure related costs). See 73 FR 44522, 44541–42 (July 30, 2008). 376 See FTC Policy Statement on Unfairness, Int’l Harvester, 104 F.T.C. at 1073 n.12. 377 See Int’l Harvester, 104 F.T.C. at 1064. 378 See FTC Policy Statement on Unfairness, Int’l Harvester, 104 F.T.C. at 1073. 173 substantial injury along with financial harm.379 Not Reasonably Avoidable The second element required for a determination of unfairness under section 1031(c)(1) of the Dodd-Frank Act is that the substantial injury is not reasonably avoidable by consumers. Again, the FTC Act unfairness standard, the FTC Policy Statement on Unfairness, FTC and other Federal agency rulemakings, and related case law inform the meaning of this element of the unfairness standard. The FTC has noted that knowing the steps for avoiding injury is not enough for the injury to be reasonably avoidable; rather, the consumer must also understand the necessity of taking those steps.380 As the FTC explained in its Policy Statement on Unfairness, most unfairness matters are brought to “halt some form of seller behavior that unreasonably creates or takes advantage of an obstacle to the free exercise of consumer decision making.”381 The D.C. Circuit held that such behavior can create a “market failure” and the agency “may be required to take corrective action.”382 Reasonable avoidability also takes into account the costs of making a choice other than the one made and the availability of alternatives in the marketplace. 383 379 See AFSA, 767 F.2d at 973–74, n.20 (1985) (discussing the potential psychological harm resulting from lenders’ taking of non-possessory security interests in household goods and associated threats of seizure, which was part of the FTC’s rationale for intervention in the Credit Practices Rule). 380 See Int’l Harvester, 104 F.T.C. at 1066. 381 FTC Policy Statement on Unfairness, Int’l Harvester, 104 F.T.C. at 1074. 382 AFSA, 767 F.2d at 976. The D.C. Circuit noted that Congress intended for the FTC to develop and refine the criteria for unfairness on a “progressive, incremental” basis. Id. at 978. The court upheld the FTC’s Credit Practices Rule by reasoning in part that “the fact that the [FTC’s] analysis applies predominantly to certain creditors dealing with a certain class of consumers (lower-income, higher-risk borrowers) does not, as the dissent suggests, undercut its validity. [There is] a market failure with respect to a particular category of credit transactions which is being exploited by the creditors involved to the detriment of the consumers involved.” Id. at 982 n.29. 383 See FTC Policy Statement on Unfairness, Int’l Harvester, 104 F.T.C. at 1074 n.19 (“In some senses any injury 174 Countervailing Benefits to Consumers or Competition The third element required for a determination of unfairness under section 1031(c)(1) of the Dodd- Frank Act is that the act or practice’s countervailing benefits to consumers or to competition do not outweigh the substantial consumer injury. Once again, the FTC Act unfairness standard, the FTC Policy Statement on Unfairness, FTC and other Federal agency rulemakings, and related case law inform the meaning of this element of the unfairness standard. In applying the FTC Act’s unfairness standard, the FTC has stated that it is important to consider both the costs of imposing a remedy and any benefits that consumers enjoy as a result of the practice.384 Authorities addressing the FTC Act’s unfairness standard indicate that the countervailing benefits test does not require a precise quantitative analysis of benefits and costs, because such an analysis may be unnecessary or, in some cases, impossible. Rather, the agency is expected to gather and consider reasonably available evidence.385 can be avoided—for example, by hiring independent experts to test all products in advance, or by private legal actions for damages—but these courses may be too expensive to be practicable for individual consumers to pursue.”); AFSA, 767 F.2d at 976–77 (reasoning that because of factors such as substantial similarity of contracts, “consumers have little ability or incentive to shop for a better contract”). 384 See FTC Policy Statement on Unfairness, Int’l Harvester, 104 F.T.C. at 1073–74 (noting that an unfair practice must be “injurious in its net effects” and that “[t]he Commission also takes account of the various costs that a remedy would entail. These include not only the costs to the parties directly before the agency, but also the burdens on society in general in the form of increased paperwork, increased regulatory burdens on the flow of information, reduced incentives to innovation and capital formation, and similar matters.”). 385 See S. Rept. 103–130, at 13 (1994) (legislative history for the 1994 amendments to the FTC Act noting that, “In determining whether a substantial consumer injury is outweighed by the countervailing benefits of a practice, the Committee does not intend that the FTC quantify the detrimental and beneficial effects of the practice in every case. In many instances, such a numerical benefit-cost analysis would be unnecessary; in other cases, it may be impossible. This section would require, however, that the FTC carefully evaluate the benefits and costs of each exercise of its unfairness authority, gathering and considering reasonably available evidence.”); Pennsylvania Funeral Directors Ass’n v. FTC, 41 F.3d 81, 91 (3d Cir. 1994) (in upholding the FTC’s amendments to the Funeral Industry Practices Rule, the Third Circuit noted that “much of a cost-benefit analysis requires predictions and speculation”); Int’l Harvester, 104 F.T.C. at 1065 n.59 (“In making these calculations we do not strive for an unrealistic degree of precision . . . . We assess the matter in a more general way, giving consumers the benefit of the doubt in close issues . . . . What is important . . . is that we retain an overall sense of the relationship between costs 175 Public Policy As noted above, section 1031(c)(2) of the Dodd-Frank Act provides that, “[i]n determining whether an act or practice is unfair, the Bureau may consider established public policies as evidence to be considered with all other evidence. Such public policy considerations may not serve as a primary basis for such determination.”386 Section 1031(d)—Abusive Acts or Practices The Dodd-Frank Act, in section 1031(b), authorizes the Bureau to identify and prevent abusive acts and practices. The Bureau believes that Congress intended for the statutory phrase “abusive acts or practices” to encompass conduct by covered persons that is beyond what would be prohibited as unfair or deceptive acts or practices, although such conduct could overlap and thus satisfy the elements for more than one of the standards.387 Under section 1031(d) of the Dodd-Frank Act, the Bureau “shall have no authority . . . to declare an act or practice abusive in connection with the provision of a consumer financial product or service” unless the act or practice meets at least one of several enumerated conditions. For example, under section 1031(d)(2)(A) of the Act, an act or practice might “take[] unreasonable advantage of” a consumer’s “lack of understanding . . . of the material risks, costs, or conditions of the [consumer financial] product or service” (i.e., the lack of understanding and benefits. We would not want to impose compliance costs of millions of dollars in order to prevent a bruised elbow.”). 386 12 U.S.C. 5531(c)(2). 387 See, e.g., S. Rept. No. 111–176, at 172 (Apr. 30, 2010) (“Current law prohibits unfair or deceptive acts or practices. The addition of ‘abusive’ will ensure that the Bureau is empowered to cover practices where providers unreasonably take advantage of consumers.”); Public Law No. 111–203 (listing, in the preamble to the Dodd- Frank Act, one of the purposes of the Act as “protect[ing] consumers from abusive financial services practices”). 176 prong).388 Under section 1031(d)(2)(B) of the Act, an act or practice might “take[] unreasonable advantage of” the “inability of the consumer to protect the interests of the consumer in selecting or using a consumer financial product or service” (i.e., the inability to protect prong).389 The Dodd-Frank Act does not further elaborate on the meaning of these terms, leaving it to the Bureau to interpret and apply these standards. Although the legislative history on the meaning of the Dodd-Frank Act’s abusiveness standard is fairly limited, it suggests that Congress was particularly concerned about the widespread practice of lenders making unaffordable loans to consumers. A primary focus was on unaffordable home mortgages and mortgages made without adequate or responsible underwriting.390 However, there is some indication that Congress also intended the Bureau to use the authority under section 1031(d) of the Dodd-Frank Act to address payday lending through the Bureau’s rulemaking, supervisory, and enforcement authorities. For example, the Senate Committee on Banking, Housing, and Urban Affairs report on the Senate version of the legislation listed payday loans as one of several categories of consumer financial products and services, other than mortgages, where “consumers have long faced problems” because they lack “adequate Federal rules and enforcement,” noting further that “[a]busive lending, high and 388 12 U.S.C. 5531(d)(2)(A). 12 U.S.C. 5531(d)(2)(B). The Dodd-Frank Act’s abusiveness standard also permits the Bureau to intervene under section 1031(d)(1) if the Bureau determines that an act or practice “materially interferes with the ability of a consumer to understand a term or condition of a consumer financial product or service,” 12 U.S.C. 5531(d)(1), and under section 1031(d)(2)(C) if an act or practice “takes unreasonable advantage of” the consumer’s “reasonable reliance” on the covered person to act in the consumer’s interests, 12 U.S.C. 5531(d)(2)(C). 390 While Congress sometimes described other products as abusive, it frequently applied the term to unaffordable mortgages and mortgages made without adequate or responsible underwriting. See, e.g., S. Rept. No. 111–176, at 11 (noting that the “financial crisis was precipitated by the proliferation of poorly underwritten mortgages with abusive terms”). 389 177 hidden fees, unfair and deceptive practices, confusing disclosures, and other anti-consumer practices have been a widespread feature in commonly available consumer financial products such as credit cards.”391 The same section of the Senate committee report included a description of the basic features of payday loans and the problems associated with them, specifically noting that many consumers are unable to repay the loans while meeting their other obligations and that many of these borrowers re-borrow, which results in a “perpetual debt treadmill.”392 These portions of the legislative history reinforce other indications in the Dodd-Frank Act that Congress consciously intended to confer direct authority upon the Bureau to address issues concerning payday loans. 393 B. Section 1032 of the Dodd-Frank Act Section 1032(a) of the Dodd-Frank Act provides that the Bureau may prescribe rules to ensure that the features of any consumer financial product or service, “both initially and over the term of the product or service,” are “fully, accurately, and effectively disclosed to consumers in a manner that permits consumers to understand the costs, benefits, and risks associated with the 391 See S. Rept. 111–176, at 17. In addition to credit cards, the Senate committee report listed overdraft, debt collection, payday loans, and auto dealer lending as the consumer financial products and services warranting concern. Id. at 17–23. 392 See S. Rept. 111–176, 20–21; see also 155 Cong. Rec. 31250 (Dec. 10, 2009) (during a colloquy on the House floor with the one of the authors of the Dodd-Frank Act, Representative Barney Frank, Representative Henry Waxman stated that the “authority to pursue abusive practices helps ensure that the agency can address payday lending and other practices that can result in pyramiding debt for low income families.”). 393 Section 1024(a)(1)(E) of the Dodd-Frank Act also expressly confers authority upon the Bureau to take specific acts concerning “any covered person who . . . offers or provides to a consumer a payday loan.” These include the use of supervisory authority to “conduct examinations” for the purpose of “assessing compliance with the requirements of Federal consumer financial law,” to exercise “exclusive” authority to “enforce Federal consumer financial law,” and to exercise “exclusive” authority to “issue regulations” for the purpose of “assuring compliance with Federal consumer financial law.” Congress conferred this authority only for a defined and limited universe of consumer financial products—payday loans, mortgage loans, and student loans—and in certain other specified instances, thus indicating its intent to empower the Bureau to consider and carry out broad regulatory and oversight activity with respect to the market for payday loans, in particular. 178 product or service, in light of the facts and circumstances.”394 This authority is broad, and empowers the Bureau to prescribe rules regarding the disclosure of the “features” of consumer financial products and services generally. Accordingly, the Bureau may prescribe rules containing disclosure requirements even if other Federal consumer financial laws do not specifically require disclosure of such features. Section 1032(c) of the Dodd-Frank Act provides that, in prescribing rules pursuant to section 1032 of the Act, the Bureau “shall consider available evidence about consumer awareness, understanding of, and responses to disclosures or communications about the risks, costs, and benefits of consumer financial products or services.”395 Section 1032(b)(1) of the Dodd-Frank Act provides that “any final rule prescribed by the Bureau under this section requiring disclosures may include a model form that may be used at the option of the covered person for provision of the required disclosures.”396 Section 1032(b)(2) of the Act provides that such a model form “shall contain a clear and conspicuous disclosure that, at a minimum—(A) uses plain language comprehensible to consumers; (B) contains a clear format and design, such as an easily readable type font; and (C) succinctly explains the information that must be communicated to the consumer.”397 Section 1032(b)(3) of the Dodd-Frank Act provides that any such model form “shall be validated through consumer testing.”398 And section 1032(d) of the Act provides that, “Any covered person that uses a model form included with a rule issued under this section shall be 394 12 U.S.C. 5532(a). 12 U.S.C. 5532(c). 396 12 U.S.C. 5532(b)(1). 397 12 U.S.C. 5532(b)(2). 398 12 U.S.C. 5532(b)(3). 395 179 deemed to be in compliance with the disclosure requirements of this section with respect to such model form.”399 C. Other Authorities Under the Dodd-Frank Act Section 1022(b)(1) of the Dodd-Frank Act provides that the Bureau’s director “may prescribe rules and issue orders and guidance, as may be necessary or appropriate to enable the Bureau to administer and carry out the purposes and objectives of the Federal consumer financial laws, and to prevent evasions thereof.”400 “Federal consumer financial law” includes rules prescribed under Title X of the Dodd-Frank Act,401 including sections 1031(b) to (d) and 1032. Section 1022(b)(2) of the Dodd-Frank Act prescribes certain standards for rulemaking that the Bureau must follow in exercising its authority under section 1022(b)(1) of the Act. 402 For a discussion of the Bureau’s standards for rulemaking under section 1022(b)(2) of the Act, see part VII below. Section 1022(b)(3)(A) of the Dodd-Frank Act authorizes the Bureau, by rule, to “conditionally or unconditionally exempt any class of covered persons, service providers, or consumer financial products or services” from any provision of Title X or from any rule issued under Title X as the Bureau determines “necessary or appropriate to carry out the purposes and objectives” of Title X. In doing so, the Bureau must, “tak[e] into consideration the factors” set forth in section 1022(b)(3)(B) of the Act,403 which specifies three factors that the Bureau shall, as 399 12 U.S.C. 5532(d). 12 U.S.C. 5512(b)(1). 401 12 U.S.C. 5481(14). 402 12 U.S.C. 5512(b)(2). 403 12 U.S.C. 5512(b)(3)(A). 400 180 appropriate, take into consideration in issuing such an exemption.404 Furthermore, §§ 1041.10 and 1041.11 of the final rule are authorized by other DoddFrank Act authorities, such as sections 1021(c)(3),405 1022(c)(7),406 1024(b)(1),407 and 1024(b)(7) of the Act.408 A more complete description of the Dodd-Frank Act authorities on which the Bureau is relying for §§ 1041.10 and 1041.11 of the final rule is contained in the section-by-section analysis of those provisions. D. Section 1041 of the Dodd-Frank Act and Preemption Section 1041(a)(1) of the Dodd-Frank Act provides that Title X of the Act, other than sections 1044 through 1048, “may not be construed as annulling, altering, or affecting, or exempting any person subject to the provisions of [Title X] from complying with,” the statutes, regulations, orders, or interpretations in effect in any State (sometimes hereinafter, State laws), “except to the extent that any such provision of law is inconsistent with the provisions of [Title X], and then only to the extent of the inconsistency.” 409 Section 1041(a)(2) of the Act provides that, for purposes of section 1041, “a statute, regulation, order, or interpretation in effect in any State is not inconsistent with” the Title X provisions “if the protection that such statute, regulation, order, or interpretation affords to consumers is greater than the protection provided” 404 12 U.S.C. 5512(b)(3)(B) (“(B) Factors—In issuing an exemption, as permitted under subparagraph (A), the Bureau shall, as appropriate, take into consideration—(i) the total assets of the class of covered persons; (ii) the volume of transactions involving consumer financial products or services in which the class of covered persons engages; and (iii) existing provisions of law which are applicable to the consumer financial product or service and the extent to which such provisions provide consumers with adequate protections.”). 405 12 U.S.C. 5511(c)(3). 406 12 U.S.C. 5512(c)(7). 407 12 U.S.C. 5514(b)(1). 408 12 U.S.C. 5514(b)(7). 409 12 U.S.C. 5551(a)(1). Section 1002(27) of the Dodd-Frank Act defines “State” to include any “Federally recognized Indian Tribe.” See 12 U.S.C. 5481(27). 181 under Title X.410 This section further provides that “[a] determination regarding whether a statute, regulation, order, or interpretation in effect in any State is inconsistent with the provisions of [Title X] may be made by the Bureau on its own motion or in response to a nonfrivolous petition initiated by any interested person.”411 The requirements of the final rule set minimum Federal standards for the regulation of covered loans. They thus accord with the common preemption principle that Federal law provides a floor and not a ceiling on consumer financial protection,412 as provided in section 1041(a)(2) of the Dodd-Frank Act. The requirements of this rule will thus coexist with State laws that pertain to the making of loans that the rule treats as covered loans (hereinafter, “applicable State laws”). Consequently, any person subject to the final rule will be required to comply with both the requirements of this rule and all applicable State laws, except to the extent that the applicable State laws are inconsistent with the requirements of the rule.413 This approach reflects the established framework of cooperative federalism between Federal and State laws in many other substantive areas. Accordingly, the arguments advanced by some commenters that the payday rule would “occupy the field” are incorrect. Where Federal law occupies an entire field, “even complementary State regulation is impermissible” because field preemption 410 12 U.S.C. 5551(a)(2). 12 U.S.C. 5551(a)(2). 412 The Bureau received a comment from a group of State Attorneys General asking the Bureau to codify the statement that this is a floor and not a ceiling. The Bureau does not believe this is necessary, and that it would conflict with the regulatory scheme of the rule, which is primarily aimed at obligations on the part of lenders. This section should suffice for purposes of communicating the Bureau’s intent with regard to preemption. 413 The requirements of the final rule will also coexist with applicable laws in cities and other localities, and the Bureau does not intend the rule to annul, alter, affect, or exempt any person from complying with the regulatory frameworks of cities and other localities to the extent those frameworks provide greater consumer protections than the requirements of this rule. 411 182 “foreclose[s] any State regulation in the area, even if it is parallel to Federal standards.”414 This rule would not have that effect. As noted above, section 1041(a)(2) of the Dodd-Frank Act specifies that State laws which afford greater consumer protection than is provided under Title X are not inconsistent with the provisions of Title X. Specifically, as discussed in part II, different States have taken different approaches to regulating loans that are treated as covered loans under the final rule, with many States electing to permit the making of such loans according to varying conditions, and other States choosing not to do so by imposing usury caps that effectively render it impractical to make such loans in those States. Particularly in the States where fixed usury caps effectively prohibit these types of loans, nothing in this rule is intended or should be construed to undermine or cast doubt on whether those provisions are sound public policy. Because Title X does not confer authority on the Bureau to establish usury limits,415 its policy interventions, as embodied in the final rule, are entirely distinct from such measures as are beyond its statutory authority. Therefore, nothing in this rule should be construed as annulling or even as inconsistent with a regulatory or policy approach to such loans based on usury caps, which are wholly within the prerogative of the States to lawfully impose. Indeed, as described in part II, South Dakota became the most recent State to impose a usury cap on payday loans after conducting a ballot initiative in 2016 in which the public voted to approve the measure by a substantial margin. 414 Arizona v. United States, 567 U.S. 387 (2012). Section 1027(o) of the Dodd-Frank Act provides that “No provision of this title shall be construed as conferring authority on the Bureau to establish a usury limit applicable to an extension of credit offered or made by a covered person to a consumer, unless explicitly authorized by law.” 12 U.S.C. 5517(o). 415 183 The requirements of the final rule will coexist with different approaches and frameworks for the regulation of such covered loans as reflected in applicable State laws.416 The Bureau is aware of certain applicable State laws that may afford greater protections to consumers than do the requirements of this rule. For example, as described in part II and just discussed above, certain States have fee or interest rate caps (i.e., usury limits) that payday lenders may find are set too low to sustain their business models. The Bureau regards the fee and interest rate caps in these States as providing greater consumer protections than, and thus as not inconsistent with, the requirements of the final rule. Aside from those provisions of State law just discussed, the Bureau declines to determine definitively in this rulemaking whether any other individual statute, regulation, order, or interpretation in effect in any State is inconsistent with the rule. Comments on the proposal and internal analysis have led the Bureau to conclude that specific questions of preemption should be decided upon application, and the Bureau will respond to nonfrivolous petitions initiated by interested persons in accordance with section 1041(a)(2) of the Dodd-Frank Act. The Bureau believes that in most cases entities can apply the principles articulated above in a straightforward manner to determine their rights and obligations under both the rule and State law. Moreover, in light of the variety of relevant State law provisions and the range of practices that may be covered by those laws, it is impossible for the Bureau to provide a definitive description of all interactions or to anticipate all areas of potential concern. 416 Some State officials expressed concern that the identification of unfair and abusive acts or practices in this rulemaking may be construed to affect or limit provisions in State statutes or State case law. The Bureau has identified unfair and abusive acts or practices under the statutory definitions in section 1031(c) and (d) of the DoddFrank Act. The final rule is not intended to limit the further development of State laws protecting consumers from unfair or deceptive acts or practices as defined under State laws, or from similar conduct prohibited by State laws, consistent with the principles set forth in the Dodd-Frank Act as discussed further above. 184 Some commenters argued that because section 1041 of the Dodd-Frank Act includes only the term “this title,” and not “any rule or order prescribed by the Bureau under this title,” Congress contemplated only statutory and not regulatory preemption of State law. The Bureau disagrees and believes section 1041 is best interpreted to apply to Title X and rules prescribed by the Bureau under that Title. Section 1041 was modeled in large part on similar provisions from certain enumerated consumer laws. Consistent with longstanding case law holding that State laws can be pre-empted by Federal regulations promulgated in the exercise of delegated authority,417 those provisions were definitively interpreted to apply to requirements imposed by implementing regulations, even where the statutory provisions include explicit reference only to the statutes themselves.418 Congress is presumed to have been aware of those applications in enacting Title X, and section 1041 is best interpreted similarly. Moreover, the Bureau’s interpretation furthers principles of consistency, uniformity, and manageability in interpreting Title X and legislative rules with the force and effect of law implementing that statute. Finally, while section 1041 of the Act instructs preemption analyses, any actual pre-emptive force derives from the substantive provisions of Title X and its implementing rules, not from section 1041 itself. A reading that section 1041 would apply only to Title X itself could lead to the conclusion that rules prescribed by the Bureau under Title X have broader preemptive effect than does Title X itself. The better interpretation is that the preemptive effect of regulations exercised under delegated authority should be guided by the provisions of section 1041. Lastly, the Bureau intends this rule to interact in the same manner with laws or 417 See, e.g., Hillsborough County v. Automated Med. Laboratories, 471 U.S. 707, 713 (1985). See, e.g., 15 U.S.C. 1610(a)(1) & 12 CFR 1026.28 (TILA & Regulation Z); 15 U.S.C. 1691d(f) & 12 CFR 1002.11 (ECOA & Regulation B). 418 185 regulations at other government levels, like city or locality laws or regulations. E. General Comments on the Bureau’s Legal Authority In addition to setting out the Bureau’s legal authority for this rulemaking and responding to comments directed to specific sources of authority, it is necessary to address several more general comments that challenged or criticized certain aspects of the Bureau’s ability to proceed to finalize this rule. They will be addressed here. Some industry commenters and State Attorneys General have contended that the Bureau lacks the legal authority to adopt this rule because the Bureau itself or its statutory authority is unconstitutional on various grounds, including separation-of-powers, the non-delegation doctrine, and the 10th Amendment. No court has ever held that the Bureau is unable to issue regulations on the basis that it is unconstitutional, and in fact the Bureau has issued dozens of regulations to date, including many major rules that have profoundly affected key consumer markets such as mortgages, prepaid accounts, remittance transfers, and others—a number of which were mandated by Congress. In addition, longstanding precedent has established that a government agency lacks the authority to decide the constitutionality of congressional enactments.419 One commenter argued that the timing of the proposed rule prevented the Bureau from using data gathered in Treasury Department Financial Empowerment Studies on small dollar loans conducted under Title XII of the Dodd-Frank Act, and that the combination of Title XII and section 1022 of the Dodd-Frank Act evidence Congress’s intent to not grant the Bureau 419 See, e.g., Johnson v. Robison, 415 U.S. 361, 368 (1974); Public Utils. Comm’n v. United States, 355 U.S. 534, 539 (1958). 186 authority to issue a rule that reduces the availability of payday loans. There is nothing in either the plain language or structure of the Dodd-Frank Act to suggest that Congress intended the Bureau to postpone any regulation of unfair and abusive payday lending practices until after Treasury had established the multiyear grant program that Congress authorized Treasury to establish. Indeed, it is noteworthy that Title XII does not mandate that Treasury create such programs—it merely authorizes Treasury to do so. Moreover, contrary to the commenter’s assertions, the final rule will not end payday lending and it will not undermine the rationale for the grants for which Congress provided in Title XII. There is no basis to conclude that the Bureau is under any obligation to wait for such grant programs to play out to prevent UDAAPs. Some industry commenters have made the claim that the Bureau had impermissibly prejudged the evidence about whether and how to proceed with this rule and failed to comply with its own ex parte policy by engaging in improper communications with special interest groups prior to the publication of the Notice of Proposed Rulemaking. The Bureau does not agree with these claims for several reasons. First, part III of the final rule, which summarizes in detail the Bureau’s rulemaking process, shows that these claims are without basis. That discussion reflects the Bureau’s considerable experience with these issues and with this market for over five years of steady work. It also includes a description of the Bureau’s approach to handling the great volume of public comments received on the proposed rule, as well as a number of ex parte communications, which have been documented and incorporated into the administrative record and are available to the public at www.regulations.gov. Second, both the proposed rule and the final rule are based on the Bureau’s careful review of the relevant evidence, including evidence generated by the Bureau’s own studies, as well as evidence 187 submitted by a broad range of stakeholders, including industry stakeholders. Finally, the numerous changes made in the final rule in response to stakeholder comments, including industry stakeholders, is further evidence that the Bureau has not prejudged any issues. A number of industry commenters have argued that the rule conflicts with the Bureau’s statutory purpose under section 1021(b)(4) of the Dodd-Frank Act, which is to enforce the law consistently for all persons, regardless of their status as depository institutions, because it addresses covered loans but does not address other types of financial products, such as overdraft services or credit card accounts. The Bureau notes in response that each of these products has its own features, characteristics, historical background, and prior regulatory treatment, as discussed further in the section-by-section analysis of § 1041.3(d). Just as it has not been judged to be impermissibly inconsistent for Federal and State authorities (including the Congress) to treat these distinct products differently as a matter of statutory law and regulation, despite certain similarities of product features and uses, even so it is not inconsistent for the Bureau to do so for the purposes of this rule. Further, while it may be true that more nonbanks will be impacted by this rule than banks by virtue of the products that banks and nonbanks are currently providing, that does not mean that this rule conflicts with section 1021(b)(4), but simply reflects the current makeup of this marketplace. Finally, and more narrowly, some Tribal and industry commenters have averred that the Bureau lacks authority to adopt regulations pursuant to section 1031 of the Dodd-Frank Act that apply to Indian tribes or to any of the entities to which they have delegated Tribal authority. These arguments raised on behalf of Tribal lenders have also been raised in Tribal consultations that the Bureau has held with federally recognized Indian Tribes, as discussed in part III, and in 188 various court cases to date. They rest on what the Bureau believes is a misreading of the Act and of Federal law and precedents governing the scope of Tribal immunity, positions that the Bureau has briefed extensively to the Federal courts in some key cases testing these issues.420 V. Section-by-Section Analysis Subpart A—General Section 1041.1 Authority and Purpose Proposed § 1041.1 provided that the rule is being issued pursuant to Title X of the DoddFrank Wall Street Reform and Consumer Protection Act.421 It also provided that the purpose of this part is to identify certain unfair and abusive acts or practices in connection with certain consumer credit transactions; to set forth requirements for preventing such acts or practices; and to prescribe requirements to ensure that the features of those consumer credit transactions are fully, accurately, and effectively disclosed to consumers. It also noted that this part prescribes processes and criteria for registration of information systems. The Bureau did not receive any comments on proposed § 1041.1 and is finalizing this provision as proposed. Section 1041.2 Definitions 420 See CFPB v. Great Plains Lending, 846 F.3d 1049 (9th Cir. 2017), reh’g denied (Apr. 5, 2017) (Court of Appeals affirmed District Court ruling upholding and enforcing the Bureau’s authority to issue civil investigative demands to payday lenders claiming Tribal affiliation and rejecting their claim of “tribal sovereign immunity”; a petition for certiorari to the Supreme Court is now pending); see also Otoe-Missouria Tribe of Indians v. New York State Dep’t of Fin. Servs., 769 F.3d 105, 107 (2d Cir. 2014) (upholding the State’s claim to be able to be able to pursue an enforcement action against payday lenders claiming Tribal affiliation that “provide short-term loans over the Internet, all of which have triple-digit interest rates that far exceed the ceiling set by New York law;” the Bureau filed an amicus curiae brief in support of the State’s position). 421 Public Law 111–203, 124 Stat. 1376, 1955 (2010). 189 Proposed § 1041.2 set forth definitions for certain terms relevant to the proposal. Additional definitions were set forth in proposed §§ 1041.3, 1041.5, 1041.9, 1041.14, and 1041.17 for further terms used in those respective sections. To the extent those definitions are used in the final rule and have not been moved into § 1041.2, as discussed below, they are addressed in the context of those particular sections (some of which have been renumbered in the final rule). In general, the Bureau proposed to incorporate a number of defined terms under the Dodd-Frank Act and under other statutes or regulations and related commentary, particularly Regulation Z and Regulation E as they implement the Truth in Lending Act (TILA)422 and the Electronic Fund Transfer Act (EFTA),423 respectively. The Bureau believed that basing the proposal’s definitions on previously defined terms may minimize regulatory uncertainty and facilitate compliance, especially where the other regulations are likely to apply to the same transactions in their own right. However, as discussed further below, the Bureau proposed, in certain definitions, to expand or modify the existing definitions or the concepts enshrined in such definitions for purposes of the proposal to ensure that the rule had its intended scope of effect, particularly as industry practices may evolve. The Bureau received numerous comments about these proposed terms and their definitions, as well as some suggestions to define additional concepts left undefined in the proposal. The Bureau is finalizing § 1041.2 with some revisions and deletions from the proposal, as discussed further below, including the addition of a rule of construction as § 422 423 Public Law 90–321, 82 Stat. 146 (1968). Public Law 95–630, 92 Stat. 3641 (1978). 190 1041.2(b) to provide general guidance concerning the incorporation of terms from other statutes and regulations in the context of part 1041. 2(a) Definitions 2(a)(1) Account Proposed § 1041.2(a)(1) would have defined account by cross-referencing to the definition of that same term in Regulation E, 12 CFR part 1005. Regulation E generally defines account to include demand deposit (checking), savings, or other consumer asset accounts (other than an occasional or incidental credit balance in a credit plan) held directly or indirectly by a financial institution and established primarily for personal, family, or household purposes.424 The term account was also used in proposed § 1041.3(c), which would provide that a loan is a covered loan if, among other requirements, the lender or service provider obtains repayment directly from a consumer’s account. This term was also used in proposed § 1041.14, which would impose certain requirements when a lender seeks to obtain repayment for a covered loan directly from a consumer’s account, and in proposed § 1041.15, which would require lenders to provide notices to consumers before attempting to withdraw payments from consumers’ accounts. The Bureau stated that defining this term consistently with an existing regulation would reduce the risk of confusion among consumers, industry, and regulators. The Bureau considered the Regulation E definition to be appropriate because that definition is broad enough to capture the types of transactions that may implicate the concerns addressed by this part. 424 Regulation E also specifically includes payroll card accounts and certain government benefit card accounts. As specifically noted in the proposal here, 81 FR 47864, 47904 n.416 (July 22, 2016), the Bureau was considering in a separate rulemaking whether to provide comprehensive consumer protections under Regulation E to a broader category of prepaid accounts. The Bureau later finalized that proposed rule. See 81 FR 83934 (Nov. 22, 2016). 191 Proposed comment 2(a)(1)-1 also made clear that institutions may rely on 12 CFR 1005.2(b) and its related commentary in determining the meaning of account. One commenter stated that the definition of account should be expanded to include general-use prepaid cards, regardless of whether they are labeled and marketed as a gift card, as defined in 12 CFR 1005.20(a)(3). The Bureau recently finalized a separate rule creating comprehensive consumer protections for prepaid accounts, and in the process amended the definition of account in 12 CFR 1005.2(b) to include “a prepaid account,” so the thrust of the comment is already effectively addressed.425 The definition of “prepaid account” in that rulemaking only excludes gift cards that are both labeled and marketed as a gift card, which are subject to separate rules under Regulation E.426 The Bureau does not believe that such products are likely to be tendered as a form of leveraged payment mechanism, but will monitor the market for this issue and take appropriate action if it appears that lenders are using such products to evade coverage under the rule. The Bureau did not receive any other comments on this portion of the proposal and is finalizing this definition as proposed. Proposed comment 2(a)(1)-1 has now been incorporated into comment 2(b)(1)-1 to illustrate the broader rule of construction discussed in § 1041.2(b). 2(a)(2) Affiliate Proposed § 1041.2(a)(2) would have defined affiliate by cross-referencing to the definition of that same term in the Dodd-Frank Act, 12 U.S.C. 5481(1). The Dodd-Frank Act defines affiliate as any person that controls, is controlled by, or is under common control with 425 See 81 FR 83934, 83965–83978, 84325–84326 (Nov. 22, 2016). See 81 FR 83934, 83976–83978 (Nov. 22, 2016) (discussing § 1005.2(b)(3)(ii)(D) and comment 2(b)(3)(ii)–3 of the final prepaid rule.). 426 192 another person. Proposed §§ 1041.6 and 1041.10 would have imposed certain limitations on lenders making loans to consumers who have outstanding covered loans with an affiliate of the lender, and the Bureau’s analyses of those proposed sections discussed in more detail the particular requirements related to affiliates. The Bureau stated that defining this term in the proposal consistently with the Dodd-Frank Act would reduce the risk of confusion among consumers, industry, and regulators. Although the limitations in proposed §§ 1041.6 and 1041.10 are not being finalized, the final rule includes a number of other provisions in which the term affiliate is used, including the conditional exemption in § 1041.3(f). The Bureau did not receive any comments on this portion of the proposal and is finalizing this definition as proposed. 2(a)(3) Closed-end credit Proposed § 1041.2(a)(3) would have defined closed-end credit as an extension of credit to a consumer that is not open-end credit under proposed § 1041.2(a)(14). This term is used in various parts of the rule where the Bureau proposed to tailor provisions specifically for closedend and open-end credit in light of their different structures and durations. Most notably, proposed § 1041.2(a)(18) prescribed slightly different methods of calculating the total cost of credit for closed-end and open-end credit. Proposed § 1041.16(c) also required lenders to furnish information about whether a covered loan is closed-end or open-end credit to registered information systems. Proposed comment 2(a)(3)-1 also made clear that institutions may rely on 12 CFR 1026.2(a)(10) and its related commentary in determining the meaning of closed-end credit, but without regard to whether the credit is consumer credit or is extended to a consumer, as those terms are defined in 12 CFR 1026.2(a). 193 The Bureau did not receive any comments on the definition of closed-end credit contained in the proposal and is finalizing the definition and commentary as proposed. The Bureau did, however, receive a number of comments on the definition of open-end credit contained in the proposal and made some changes to that definition in light of the comments received, all as discussed below. Because the term closed-end credit is defined in contradistinction to the term open-end credit, the changes made to the latter definition will affect the parameters of this definition as well. 2(a)(4) Consumer Proposed § 1041.2(a)(4) would have defined consumer by cross-referencing the definition of that term in the Dodd-Frank Act, which defines consumer as an individual or an agent, trustee, or representative acting on behalf of an individual.427 The term is used in numerous provisions across proposed part 1041 to refer to applicants for and borrowers of covered loans. The Bureau stated that this definition, rather than the arguably narrower Regulation Z definition of consumer—which defines consumer as “a cardholder or natural person to whom consumer credit is offered or extended”—is appropriate to capture the types of transactions that may implicate the concerns addressed by the proposed rule. In particular, the definition of this term found in the Dodd-Frank Act expressly includes agents and representatives of individuals, rather than just individuals themselves. The Bureau believed this definition might more comprehensively foreclose possible evasion of the specific consumer protections imposed by proposed part 1041 than would the definition found in Regulation Z. 427 12 U.S.C. 5481(4). 194 The Bureau did not receive any comments on this portion of the proposal and is finalizing this definition as proposed. 2(a)(5) Consummation Proposed § 1041.2(a)(5) would have defined consummation as the time that a consumer becomes contractually obligated on a new loan, which is consistent with the definition of the term in Regulation Z § 1026.2(a)(13), or the time that a consumer becomes contractually obligated on a modification of an existing loan that increases the amount of the loan. The proposal used the term both in defining certain categories of covered loans and in defining the timing of certain proposed requirements. The time of consummation was important both in applying certain proposed definitions for purposes of coverage and in applying certain proposed substantive requirements. For example, under proposed § 1041.3(b)(1), whether a loan is a covered short-term loan would depend on whether the consumer is required to repay substantially all of the loan within 45 days of consummation. Under proposed § 1041.3(b)(2)(i), the determination of whether a loan is subject to a total cost of credit exceeding 36 percent per annum would be made at the time of consummation. Pursuant to proposed §§ 1041.6 and 1041.10, certain limitations would potentially apply to lenders making covered loans based on the consummation dates of those loans. Pursuant to proposed § 1041.15(b), lenders would have to furnish certain disclosures before a loan subject to the requirements of that section is consummated. In the proposal, the Bureau stated that defining this term consistently with Regulation Z with respect to new loans would reduce the risk of confusion among consumers, industry, and regulators. Proposed comment 2(a)(5)-1 also made clear that the question of when a consumer 195 would become contractually obligated with regard to a new loan is a matter to be determined under applicable law; for example, a contractual commitment agreement that binds the consumer to the loan would be a consummation. However, the comment stated that consummation does not occur merely because the consumer has made some financial investment in the transaction (for example, by paying a non-refundable fee), unless applicable law holds otherwise. The Bureau also provided guidance as to consummation with respect to particular loan modifications, so as to further the intent of proposed §§ 1041.3(b)(1), 1041.3(b)(2), 1041.5(b), and 1041.9(b), all of which would impose requirements on lenders as of the time that the loan amount increases on an existing loan. The Bureau concluded that defining these increases in loan amounts as consummations would improve clarity for consumers, industry, and regulators. The abovereferenced sections, as proposed, would impose no duties or limitations on lenders when a loan modification decreases the amount of the loan. Accordingly, in addition to incorporating Regulation Z commentary as to the general definition of consummation for new loans, proposed comment 2(a)(5)-2 explained the time at which certain modifications of existing loans would be considered to be a consummation for purposes of the rule. Proposed comment 2(a)(5)-2 explained that a modification would be considered a consummation if the modification increases the amount of the loan. Proposed comment 2(a)(5)-2 also explained that a cost-free repayment plan, or “off-ramp” as it is commonly known in the market, would not result in a consummation under proposed § 1041.2(a)(5). In the proposal, the Bureau stated that it considered expressly defining a new loan in order to clarify when lenders would need to make the ability-to-repay determinations prescribed in proposed §§ 1041.5 and 1041.9. The definition that the Bureau considered would have 196 defined a new loan as a consumer-purpose loan made to a consumer that (a) is made to a consumer who is not indebted on an outstanding loan, (b) replaces an outstanding loan, or (c) modifies an outstanding loan, except when a repayment plan, or “off-ramp” extends the term of the loan and imposes no additional fees. Although some commenters requested more guidance to distinguish a loan modification from an instance of re-borrowing or a loan refinancing, the Bureau has concluded that the examples provided in the commentary sufficiently address all of the relevant scenarios where ambiguity could arise about whether consummation occurs. No other comments were received on any other aspect of this portion of the proposal. The Bureau has reworded parts of comment 2(a)(5)-2 for clarity in describing what types of loan modifications trigger substantive requirements under part 1041, but otherwise is finalizing this definition and the commentary as proposed. 2(a)(6) Cost of credit Proposed § 1041.2(a)(18) set forth the method for lenders to calculate the total cost of credit to determine whether a longer-term loan would be covered under proposed § 1041.3(b)(2). Proposed § 1041.2(a)(18) generally would have defined the total cost of credit as the total amount of charges associated with a loan expressed as a per annum rate, including various charges that do not meet the definition of finance charge under Regulation Z. The charges would be included even if they were paid to a party other than the lender. The Bureau proposed to adopt this approach to defining loan costs from the Military Lending Act, and also to have adopted the MLA’s 36 percent threshold in defining what covered longer-term loans were subject to part 1041. The effect would have been that a loan with a term of longer than 45 days 197 must have a total cost of credit exceeding a rate of 36 percent per annum in order to be a covered loan. The Bureau thus proposed using an all-in measure of the total cost of credit rather than the definition of annual percentage rate (APR) under Regulation Z because it was concerned that lenders might otherwise shift their fee structures to fall outside traditional Regulation Z concepts. This in turn would lead them to fall outside the proposed underwriting criteria for covered longer-term loans, which they could do, for example, by imposing charges in connection with a loan that are not included in the calculation of APR under Regulation Z. The Bureau acknowledged that lenders were less familiar with the approach involving the MLA calculations than they are with the more traditional APR approach and calculations under Regulation Z. Therefore, the Bureau specifically sought comment on the compliance burdens of the proposed approach and whether to use the more traditional APR approach instead. The Bureau received many comments on the definition of the total cost of credit, which reflected its functional position in the proposed rule as the trigger for the additional underwriting criteria applicable to covered longer-term loans. A number of comments addressed what kinds of fees and charges should be included or excluded from the total cost of credit and demanded more technical guidance, which reflected the increased complexity of using this method. One lender noted a specific loan program that would only be included in the rule because of the inclusion of participation fees in the proposed definition. Various commenters noted the greater simplicity of the APR calculation in Regulation Z, and contended that greater burdens would be imposed and less clarity achieved by applying the proposed definition of total cost of credit. The latter, they suggested, would confuse consumers who are accustomed to Regulation Z’s APR definition, would be difficult to administer properly, and would be likely to have unintended 198 consequences, such as causing many lenders to choose not to offer optional ancillary products like credit life and disability insurance, to the detriment of borrowers. Consumer groups, by contrast, generally preferred the proposed definition of total cost of credit, though they offered suggestions to tighten and clarify it in several respects. As noted earlier, the Bureau is not finalizing the portions of the proposed rule governing underwriting criteria for covered longer-term loans at this time. Given that covered longer-term loans are only subject to the payment requirements in subpart C, and in view of the comments received, the Bureau concludes that the advantages of simplicity and consistency militate in favor of adopting an APR threshold as the measure of the cost of credit, which is widely accepted and built into many State laws, and which is the cost that will be disclosed to consumers under Regulation Z. Moreover, the Bureau believes that the other changes in the rule mean that the basis for concern that lenders would shift their fee structures to fall outside traditional Regulation Z definitions has been reduced. Instead, the cost-of-credit threshold is now relevant only to determine whether the portions of the final rule governing payments apply to longer-term loans, which the Bureau has concluded are much less likely to prompt lenders to seek to modify their fee structures simply to avoid the application of those provisions. The Bureau notes that in determining here that the Regulation Z definition of cost of credit would be simpler and easier to use for the limited purpose of defining the application of the payment provisions of subpart C of this rule, the Bureau does not intend to decide or endorse this measure of the cost of credit—as contrasted with the total cost of credit adopted under the MLA—for any subsequent rule governing the underwriting of covered longer-term loans without balloons. The stricter and more encompassing measure used for the MLA rule may well be more 199 protective of consumers,428 and the Bureau will consider the applicability of that measure as it considers how to address longer-term loans in a subsequent rule. To effectuate this change, the Bureau has adopted as the final rule’s defined term “cost of credit,” which is an APR threshold rather than a threshold based on the total cost of credit as defined in the proposed rule. The cost of credit is defined to be consistent with Regulation Z and thus includes finance charges associated with the credit as stated in Regulation Z, 12 CFR 1026.4. As discussed further below in connection with § 1041.3(b)(3), for closed-end credit, the total cost of credit must be calculated at consummation and according to the requirements of Regulation Z, 12 CFR 1026.22, but would not have to be recalculated at some future time, even if a leveraged payment mechanism is not obtained until later. For open-end credit, the total cost of credit must be calculated at consummation and, if it does not cross the 36 percent threshold at that time, at the end of each billing cycle thereafter according to the rules for calculating the effective annual percentage rate for a billing cycle as stated in Regulation Z, 12 CFR 1026.14(c) and (d). This is a change from the proposal in order to determine coverage in situations in which there may not be an immediate draw, which was not expressly addressed in the proposal. The Bureau has concluded that defining the term cost of credit consistently with Regulation Z would reduce the risk of confusion among consumers, industry, and regulators. It also reduces burden and avoids undue complexities, especially now that the Bureau is not finalizing the underwriting criteria that were proposed for covered longer-term loans at this time. 428 In particular, the Bureau notes the statement that the Department of Defense made in the MLA rule that “unqualified exclusions from the MAPR [military annual percentage rate] for certain fees, or all non-periodic fees, could be exploited by a creditor who would be allowed to preserve a high-cost, open-end credit product by offering a relatively lower periodic rate coupled with an application fee, participation fee, or other fee,” in declining to adopt any such exclusions, which indicates the more protective nature of a “total cost of credit” definition when coupled with such further measures as necessary to protect consumers. 80 FR 43563. 200 For these reasons, the Bureau is finalizing the definition of cost of credit in a manner consistent with the discussion above, as renumbered, and with some minor additional wording revisions from the proposed rule for clarity and consistency. The proposed commentary associated with the term total cost of credit is no longer relevant and has been omitted from the final rule. 2(a)(7) Covered longer-term balloon-payment loan Proposed § 1041.2(a)(7) would have defined a covered longer-term balloon-payment loan as a covered longer-term loan described in proposed § 1041.3(b)(2)—as further specified in the next definition below—where the consumer is required to repay the loan in a single payment or through at least one payment that is more than twice as large as any other payment(s) under the loan. Proposed § 1041.9(b)(2) contained certain rules that lenders would have to follow when determining whether a consumer has the ability to repay a covered longer-term balloon-payment loan. Moreover, some of the restrictions imposed in proposed § 1041.10 would apply to covered longer-term balloon-payment loans in certain situations. The term covered longer-term balloon-payment loan would include loans that are repayable in a single payment notwithstanding the fact that a loan with a “balloon” payment is often understood in other contexts to mean a loan repayable in multiple payments with one payment substantially larger than the other payments. In the proposal, the Bureau found as a preliminary matter that both structures pose similar risks to consumers, and proposed to treat both types of loans the same way for the purposes of proposed §§ 1041.9 and 1041.10. Accordingly, the Bureau proposed to use a single defined term for both loan types to improve the proposal’s readability. 201 Apart from including single-payment loans within the definition of covered longer-term balloon-payment loans, the proposed term substantially tracked the definition of balloon payment contained in Regulation Z § 1026.32(d)(1), with one additional modification. The Regulation Z definition requires the larger loan payment to be compared to other regular periodic payments, whereas proposed § 1041.2(a)(7) required the larger loan payment to be compared to any other payment(s) under the loan, regardless of whether the payment is a regular periodic payment. Proposed comments 2(a)(7)-2 and 2(a)(7)-3 explained that payment in this context means a payment of principal or interest, and excludes certain charges such as late fees and payments that are accelerated upon the consumer’s default. Proposed comment 2(a)(7)-1 would have specified that a loan described in proposed § 1041.3(b)(2) is considered to be a covered longer-term balloon-payment loan if the consumer must repay the entire amount of the loan in a single payment. A coalition of consumer advocacy groups commented that this proposed definition is under-inclusive because it fails to include other loans that create risk that consumers will need to re-borrow because larger payments inflict payment shock on the borrowers. The commenter suggested that a more appropriate definition would be the one found in the North Carolina Retail Installment Sales Act, which defines a balloon payment as a payment that is more than 10 percent greater than other payments, except for the final payment, which is a balloon payment if it is more than 25 percent greater than other payments. In light of this comparison, the commenter recommended that any payment that is 10 percent greater than any other payment should be considered a balloon payment. 202 The Bureau recognizes these concerns, but notes that the proposed definition is generally consistent with how balloon-payment loans are defined and treated under Regulation Z, and therefore believes that adopting that definition for purposes of this rule would promote consistency and reduce the risk of confusion among consumers, industry, and regulators. The Bureau will be alert to the risk that smaller irregular payments that are not as large as twice the amount of the other payments could still cause expense shock for some consumers and lead to the kinds of problems addressed here, and thus could trigger a finding of unfairness or abusiveness in particular circumstances. In addition, the Bureau has experience with the rules adopted to implement the Military Lending Act, where loan products and lending practices adopted by some lenders in this industry evolved to circumvent the provisions of those rules. In particular, as noted in the proposal, lenders began offering payday loans greater than 91 days in duration and vehicle title loans greater than 181 days in duration, along with open-end products, in a direct response intended to evade the MLA rules—a development that prompted further Congressional and regulatory intervention. If problems begin to appear in this market from practices that are intended to circumvent the provisions of this rule, the Bureau and other regulators would be able to address any unfair or abusive practices with respect to such loan products through supervision or enforcement authority, or by amending this rule to broaden the definition. Some industry commenters contended that the Bureau’s concerns about re-borrowing for covered longer-term loans were most applicable to loans with balloon-payment structures, and they therefore argued that any ability-to-repay restrictions and underwriting criteria should be limited to longer-term balloon-payment loans. The Bureau agrees that many of its concerns 203 about covered longer-term balloon-payment loans are similar to its concerns about covered shortterm loans. Yet the Bureau also has considerable concerns about certain lending practices with respect to other covered longer-term loans, and will continue to scrutinize those practices under its supervision and enforcement authority and in a future rulemaking. At this time, however, as described more fully below in the section on Market Concerns—Underwriting, the Bureau has observed longer-term loans involving balloon payments where the lender does not reasonably assess the borrower’s ability to repay before making the loan, and in those circumstances it has observed many of the same types of consumer harms that it has observed when lenders fail to reasonably assess the borrower’s ability to repay before making covered short-term loans. As noted in part I, for a number of reasons the Bureau has decided not to address the underwriting of all covered longer-term loans at this time. Nonetheless, as just mentioned and as discussed more fully below in Market Concerns—Underwriting, the Bureau is concerned that if subpart B is not applied to covered longer-term balloon-payment loans, then lenders would simply extend the terms of their current short-term products beyond 45 days, without changing the payment structures of those loans or their current inadequate underwriting practices, as a way to circumvent the underwriting criteria for covered short-term loans. As stated above, the balloon-payment structure of these loans tend to pose very similar risks and harms to consumers as for covered short-term loans, including likely poses similar forecasting problems for consumers in repaying such loans. Therefore, in § 1041.5 of the final rule, the specific underwriting criteria that apply to covered short-term loans are made applicable to covered longer-term balloon-payment loans also. The Bureau has also modified the definition of covered longer-term balloon-payment loan so that it applies to all loans with the payment structures 204 described in the proposal. This represents an expansion in scope as compared to the proposal, as longer-term balloon-payment loans are now being covered without regard to the cost of credit or whether the lender has taken a leveraged payment mechanism in connection with the loan. In the proposal, the Bureau specifically sought comment on this potential modification, and the reasons for it are set out more extensively below in Market Concerns—Underwriting. And along with other covered longer-term loans, these particular loans remain covered by the sections of the final rule on payments as well. In light of the decision to treat covered longer-term balloon-payment loans differently from other covered longer-term loans, the Bureau decided to shift the primary description of the requirements for covered longer-term balloon-payment loans to § 1041.3(b)(2). Accordingly, the language of § 1041.2(a)(7) of the final rule has been revised to mirror the language of § 1041.2(a)(8) and (a)(10), which simply cross-reference the descriptions of the various types of covered loans specified in proposed § 1041.3(b). As a housekeeping matter, therefore, the substantive definition for longer-term balloon-payment loans is now omitted from this definition and is addressed instead in a comprehensive manner in § 1041.3(b)(2) of this final rule, where it has been expanded to address in more detail various loan structures that constitute covered longer-term balloon-payment loans. For the same reason, proposed comments 2(a)(7)-1 to 2(a)(7)-3 are omitted from the final rule and those matters are addressed in comments 3(b)(2)-1 to 3(b)(2)-4 of the final rule, as discussed below. The term covered longer-term balloon-payment loan is therefore defined in the final rule as a loan described in § 1041.3(b)(2). 2(a)(8) Covered longer-term loan 205 Proposed § 1041.2(a)(8) would have defined a covered longer-term loan to be a loan described in proposed § 1041.3(b)(2). That proposed section, in turn, described a covered loan as one made to a consumer primarily for personal, family, or household purposes that is not subject to any exclusions or exemptions, and which can be either: (1) closed-end credit that does not provide for multiple advances to consumers, where the consumer is not required to repay substantially the entire amount due under the loan within 45 days of consummation; or (2) all other loans (whether open-end credit or closed-end credit), where the consumer is not required to repay substantially the entire amount of the advance within 45 days of the advance under the loan and, in either case, two other conditions are satisfied—the total cost of credit for the loan exceeds an annual rate of 36 percent, as measured at specified times; and the lender or service provider obtains a leveraged payment mechanism, including but not limited to vehicle security, at specified times. Some restrictions in proposed part 1041 would have applied only to covered longer-term loans described in proposed § 1041.3(b)(2). For example, proposed § 1041.9 would have prescribed the ability-to-repay determination that lenders are required to perform when making covered longer-term loans. Proposed § 1041.10 would have imposed limitations on lenders making covered longer-term loans to consumers in certain circumstances that may indicate the consumer lacks the ability to repay. The Bureau proposed to use a defined term for the loans described in proposed § 1041.3(b)(2) for clarity. The Bureau received many comments on this definition that focused primarily on whether the definition was appropriate for purposes of the proposed underwriting requirements or for inclusion in the rulemaking generally, rather than with regard to the payment interventions 206 in particular. A law firm representing a traditional installment lending client commented that the definition of covered longer-term loan in the proposed rule would include traditional installment loans to a greater extent than the Bureau anticipated, with a correspondingly larger impact on credit availability as installment lenders would be forced to replace their proven underwriting techniques with burdensome and untried approaches. Others contended that the Bureau had presented no evidence indicating that the practices associated with traditional installment loans are unfair or abusive. Several commenters noted that a number of traditional installment loan products may exceed a total cost of credit of 36 percent, and some may even exceed a 36 percent annual percentage rate under TILA as well. A trade association said that such a stringent all-in annual percentage rate could encompass many bank loan products. More broadly, some commenters criticized the use of any form of interest rate threshold to determine the legal status of any loans as potentially violating the prohibition in section 1027(o) of the Dodd-Frank Act against imposing usury limits on extensions of consumer credit. Many commenters offered their views on the prong of the definition that focused on the taking of a leveraged payment mechanism or vehicle security, again often in the context of application of the underwriting requirements rather than the payment requirements. Those concerns have largely been addressed or mooted by the Bureau’s decisions to apply only the payment requirements to covered longer-term loans and to narrow the definition of such loans to focus only on those types of leveraged payment mechanisms that involve the ability to pull money from consumers’ accounts, rather than vehicle security. Comments focusing on that 207 narrower definition of leveraged payment mechanism are addressed in more depth in connection with § 1041.3(c) below. Therefore, in light of these comments and the considerations discussed above and in connection with § 1041.3(b)(3) below, the Bureau is finalizing the definition of covered longerterm loan in § 1041.2(a)(8) as discussed, with the cross-reference to proposed § 1041.3(b)(2) now edited and renumbered as § 1041.3(b)(3). As for the latter section now referenced in this definition, it too has been edited to clarify that covered longer-term loans no longer encompass covered longer-term balloon-payment loans, which are now treated separately, as the former are no longer subject to specific underwriting criteria whereas the latter are subject to the same specific underwriting criteria as covered short-term loans, which are set out in § 1041.5 of the final rule. The term covered longer-term loan is therefore defined in the final rule, as described in § 1041.3(b)(3), as one made to a consumer primarily for personal, family, or household purposes that is not subject to any exclusions or exemptions, and which can be neither a covered shortterm loan nor a covered longer-term balloon-payment loan—and thus constitutes a covered longer-term loan without a balloon-payment structure—and which meets both of the following conditions: the cost of credit for the loan exceeds a rate of 36 percent per annum; and the lender or service provider obtains a leveraged payment mechanism as defined in § 1041.3(c) of the final rule. The details of that description, and how it varies from the original proposed description of a covered longer-term loan, are provided and explained more fully in the section-by-section analysis of § 1041.3(b)(3) of the final rule. 208 2(a)(9) Covered person The Bureau has decided to include in the final rule a definition of the term covered person, which the final rule defines by cross-referencing the definition of that same term in the Dodd-Frank Act, 12 U.S.C. 5481(6). In general, the Dodd-Frank Act defines covered person as any person that engages in offering or providing a consumer financial product or service and any affiliate of such person if the affiliate acts as a service provider to such person. The Bureau concludes that defining the term covered person consistently with the Dodd-Frank Act is a mere clarification that reduces the risk of confusion among consumers, industry, and regulators, since this term is used throughout the final rule. The Bureau therefore is including this definition in the final rule as § 1041.2(a)(9). 2(a)(10) Covered short-term loan Proposed § 1041.2(a)(6) would have defined a covered short-term loan to be a loan described in proposed § 1041.3(b)(1). That proposed section, in turn, described a covered loan as one made to a consumer primarily for personal, family, or household purposes that is not subject to any exclusions or exemptions, and which can be either: closed-end credit that does not provide for multiple advances to consumers, where the consumer is required to repay substantially the entire amount due under the loan within 45 days of consummation, or all other loans (whether open-end credit or closed-end credit), where the consumer is required to repay substantially the entire amount of the advance within 45 days of the advance under the loan. Some provisions in proposed part 1041 would apply only to covered short-term loans as described in proposed § 1041.3(b)(1). For example, proposed § 1041.5 would prescribe the ability-to-repay determination that lenders are required to perform when making covered short209 term loans. Proposed § 1041.6 would impose limitations on lenders making sequential covered short-term loans to consumers. And proposed § 1041.16 would impose the payment provisions on covered short-term loans as well. The Bureau proposed to use a defined term for the loans described in § 1041.3(b)(1) for clarity. Various commenters stated that this definition is extraordinarily broad and sweeps in many different types of short-term loans, and institutions and trade associations both argued for exempting the types of loans they or their members commonly make. For example, one credit union commenter argued that the Bureau should exclude loans with total cost of credit under 36 percent. Consumer advocates argued, to the contrary, that broad coverage under the proposed rule is necessary to capture the relevant market, which can differ legally and functionally from one State to another. The Bureau finds that covered short-term loans pose substantial risks and harms for consumers, as it has detailed more thoroughly below in Market Concerns— Underwriting and the section-by-section analysis for § 1041.4 of the final rule. At the same time, the Bureau is adopting various exclusions and exemptions from coverage under the rule in § 1041.3(d), (e), and (f) below, and has discussed commenters’ requests for exclusions of various categories of loans and lenders in connection with those provisions. The Bureau has expanded the alternative loan exclusion, which now triggers off of cost of credit as defined under Regulation Z, and thus, it appears likely that the products of the credit union noted above are excluded. In light of the aggregate effect of this broad definition coupled with those exclusions and exemptions, the Bureau concludes that its definition of covered short-term loan is specific, yet necessarily broad in its coverage, in order to effectuate protections for consumers against practices that the Bureau has found to be unfair and abusive in the market for these loans. The 210 Bureau is finalizing as proposed other than renumbering. Likewise, the provision referenced in this definition—proposed § 1041.3(b)(1)—is being finalized with only non-substantive language changes, though additional commentary on that provision has been added in the final rule and will be addressed below in the discussion of that portion of the rule. 2(a)(11) Credit Proposed § 1041.2(a)(9) would have defined credit by cross-referencing the definition of credit in Regulation Z, 12 CFR part 1026. Regulation Z defines credit as the right to defer payment of debt or to incur debt and defer its payment. This term was used in numerous places throughout the proposal to refer generically to the types of consumer financial products that would be subject to the requirements of proposed part 1041. The Bureau stated that defining this term consistently with an existing regulation would reduce the risk of confusion among consumers, industry, and regulators. The Bureau also stated that the definition in Regulation Z is appropriately broad so as to capture the various types of transaction structures that implicate the concerns addressed by proposed part 1041. Proposed comment 2(a)(9) further made clear that institutions may rely on 12 CFR 1026.2(a)(14) and its related commentary in determining the meaning of credit. One consumer group commented that the definition of credit did not include a definition of loan and that these commonly related terms should be clarified to avoid the potential for confusion—a point that is addressed in §§ 1041.2(a)(13) and 1041.3(a) of the final rule. The Bureau did not receive any other comments on this portion of the proposal and is finalizing this definition and the commentary as proposed. 2(a)(12) Electronic fund transfer 211 Proposed § 1041.2(a)(10) would have defined electronic fund transfer by crossreferencing the definition of that same term in Regulation E, 12 CFR part 1005. Proposed § 1041.3(c) would provide that a loan may be a covered longer-term loan if the lender or service provider obtains a leveraged payment mechanism, which can include the ability to withdraw payments from a consumer’s account through an electronic fund transfer. Proposed § 1041.14 would impose limitations on how lenders use various payment methods, including electronic fund transfers. Proposed comment 2(a)(10)-1 also made clear that institutions may rely on 12 CFR 1005.3(b) and its related commentary in determining the meaning of electronic fund transfer. The Bureau stated that defining this term consistently with an existing regulation would reduce the risk of confusion among consumers, industry, and regulators. The Bureau did not receive any comments on this portion of the proposal and is finalizing this definition as renumbered and the commentary as proposed. 2(a)(13) Lender Proposed § 1041.2(a)(11) would have defined lender as a person who regularly makes loans to consumers primarily for personal, family, or household purposes. This term was used throughout the proposal to refer to parties that are subject to the requirements of proposed part 1041. This proposed definition is broader than the general definition of creditor under Regulation Z in that, under this proposed definition, the credit that the lender extends need not be subject to a finance charge as that term is defined by Regulation Z, nor must it be payable by written agreement in more than four installments. The Bureau proposed a broader definition than in Regulation Z for many of the same reasons that it proposed using the total cost of credit as a threshold for covering longer-term 212 loans rather than the traditional definition of annual percentage rate as defined by Regulation Z, which was discussed in the analyses of §§ 1041.2(a)(11) and 1041.3(b)(2)(i) of the proposed rule. In both instances, the Bureau was concerned that lenders might otherwise shift their fee structures to fall outside of traditional Regulation Z concepts and thus outside the coverage of proposed part 1041. For example, the Bureau stated that some loans that otherwise would meet the requirements for coverage under proposed § 1041.3(b) could potentially be made without being subject to a finance charge as that term is defined by Regulation Z. If the Bureau adopted that particular Regulation Z requirement in the definition of lender, a person who regularly extended closed-end credit subject only to an application fee, or open-end credit subject only to a participation fee, would not be deemed to have imposed a finance charge. In addition, many of the loans that would be subject to coverage under proposed § 1041.3(b)(1) are repayable in a single payment, so those same lenders might also fall outside the Regulation Z trigger for loans payable in fewer than four installments. Thus, the Bureau proposed to use a definition that is broader than the one contained in Regulation Z to ensure that the provisions proposed in part 1041 would apply as intended. The Bureau proposed to carry over from the Regulation Z definition of creditor the requirement that a person “regularly” makes loans to a consumer primarily for personal, family, or household purposes in order to be considered a lender under proposed part 1041. Proposed comment 2(a)(11)-1 explained that the test for determining whether a person regularly makes loans is the same as in Regulation Z, as explained in 12 CFR 1026.2(a)(17)(v), and depends on the overall number of loans made to a consumer for personal, family, or household purposes, not just covered loans. The Bureau stated in the proposal that it would be appropriate to exclude 213 from the definition of lender those persons who make loans for personal, family, or household purposes on an infrequent basis so that persons who only occasionally make loans would not be subject to the requirements of proposed part 1041. Such persons could include charitable, religious, or other community institutions that make loans very infrequently or individuals who occasionally make loans to family members. Consumer groups noted in commenting on the definition of lender that the proposed rule did not explicitly define what a loan is and urged the Bureau to include a definition of this term as well, as it is used frequently throughout the rule. They also commented that the definition of lender should be broadened to encompass service providers as well. For the reasons explained above in the section-by-section analysis of § 1041.2(a)(6), with respect to the definition of the term cost of credit, the Bureau has now narrowed the coverage of longer-term loans by using a threshold that is based on finance charges under Regulation Z rather than the broader range of items included in the proposed definition of total cost of credit. At the same time, it has decided to maintain the broader definition of lender, which includes parties that extend credit even if it is not subject to a finance charge as defined in Regulation Z, nor payable by written agreement in more than four installments. With regard to covered short-term and longer-term balloon-payment loans, the Bureau has concluded that it is important to maintain broad coverage over such products, even if the companies that provide them may try to structure them so as to avoid qualifying as a “creditor” under Regulation Z. The reasons for revising the definition of cost of credit, again as explained further below, were driven in large part by the Bureau’s decision not to address the underwriting of other covered longer-term loans in this rule at this time, given the benefits of alignment with Regulation Z and greater simplicity. The 214 broader definition of lender remains germane, however, to the types of loans that are subject to the underwriting provisions of the final rule. In addition, the Bureau does not find it necessary to supplement these definitions further by adding a new definition of loan in addition to the modified definitions of credit and lender. Instead, the Bureau is addressing the commenters’ point by modifying the definition of lender in § 1041.2(a)(13) to refer to a person who regularly “extends credit” rather than making loans, and has revised § 1041.3(a) to refer to a lender who “extends credit by making covered loans.” The loans covered by the final rule are credit as defined in the rule and are made by lenders as defined in the rule. In addition, key subsets of the broader universe of loans—including covered short-term loans, covered longer-term loans, and covered longer-term balloon-payment loans— are also defined explicitly in the final rule. And these definitions are premised in turn on the explication of what is a covered loan in proposed § 1041.3(b). As for the relationship between the terms lender and service provider, the Bureau is satisfied that these relationships and their effects are addressed in a satisfactory manner by defining lender as set forth here and by including separate definitions of covered person and service provider in conformity to the DoddFrank Act, as discussed in § 1041.2(a)(9) and (a)(18) of the final rule. The relationship between lender and service provider is discussed further below in the section-by-section analysis of § 1041.2(a)(18), which concerns the definition of service provider. One other segment of commenters sought to be excluded or exempted from coverage under this rule, raising many of the same points that they had raised during Bureau outreach prior to release of the proposal. As stated in the proposal, some stakeholders had suggested to the Bureau that the 215 definition of lender should be narrowed so as to exempt financial institutions that predominantly make loans that would not be covered loans under the proposed rule. They stated that some financial institutions only make loans that would be covered loans as an accommodation to existing customers, and that providing such loans is such a small part of the overall business that it would not be practical for the institutions to develop the required procedures for making covered loans. The Bureau solicited comment on whether it should narrow the definition of lender based on the quantity of covered loans an entity offers, and, if so, how to define such a de minimis test. Similarly, during the comment period many commenters, including but not limited to smaller depository institutions, presented their views that this kind of accommodation lending is longstanding and widespread and so should not be subject to coverage under the rule. At the same time, stakeholders had urged and the Bureau recognized at the time it issued the proposed rule that some newly formed companies are providing services that, in effect, allow consumers to draw on money they have earned but not yet been paid. Certain of these services do not require the consumer to pay any fees or finance charges, relying instead on voluntary “tips” to sustain the business, while others are compensated through electronic fund transfers from the consumer’s account. Some current or future services may use other business models. The Bureau also noted the existence of some newly formed companies providing financial management services to low- and moderate-income consumers that include features to smooth income. The Bureau solicited comment on whether such entities should be considered lenders under the regulation. During the public comment period, a coalition of consumer groups, some “fintech” firms, and others expressed concern about how the definition of lender would apply to new businesses 216 that are creating services to consumers to access earned income for a fee—thereby jeopardizing certain promising innovations by making them subject to the constraining provisions of this rule—and others offered views on that set of issues as well. Commenters also offered their thoughts on other innovative income-smoothing and financial-management initiatives. The Bureau has decided to address the issues raised by commenters that were seeking an exclusion or exemption from this rule not by altering the definition of lender but instead by fashioning specific exclusions and conditional exemptions as addressed below in § 1041.3(d), (e), and (f) of the final rule. Therefore in light of the comments and responses, the Bureau is finalizing this definition as renumbered and the commentary as proposed, with the one modification—use of the phrase “extends credit”—as discussed above. 2(a)(14) Loan sequence or sequence Proposed § 1041.2(a)(12) generally would have defined a loan sequence or sequence as a series of consecutive or concurrent covered short-term loans in which each of the loans (other than the first loan) is made while the consumer currently has an outstanding covered short-term loan or within 30 days thereafter. It would define both loan sequence and sequence the same way because the terms are used interchangeably in various places throughout the proposal. Furthermore, it also specified how to determine a given loan’s place within a sequence (for example, whether a loan constitutes the first, second, or third loan in a sequence), which would implicate other provisions of the proposed rule. The Bureau’s rationale for proposing to define loan sequence in this manner was discussed in more detail in the section-by-section analysis of proposed §§ 1041.4 and 1041.6. 217 The Bureau also sought comment on whether alternative definitions of loan sequence may better address its concerns about how a consumer’s inability to repay a covered loan may cause the need for a successive covered loan. Some consumer advocates commented that this definition would be clarified by including language from local ordinances or State laws that have the same effective meaning so as to avoid any confusion in compliance and enforcement. Consumer groups commented that the rule should treat a loan made within 60 days of another loan, rather than 30 days, as part of the same loan sequence in order to better effectuate its purpose of addressing the flipping of both shortterm and longer-term loans and to include late fees as rollover fees. Some industry commenters argued for a shorter period. The Bureau has considered a number of ways to specify and clarify the definition of loan sequences in order to minimize or avoid evasions of the final rule. Adopting local or State definitions would not appear to clarify the issues, as they are inconsistent from one jurisdiction to another. However, as discussed in greater detail below in Market Concerns—Underwriting and in §§ 1041.4 and 1041.5(d) of the final rule, the Bureau has decided to incorporate covered longer-term balloon-payment loans into this definition, reflecting concerns about the harms that can occur to consumers who take out a series of covered longer-term balloon-payment loans in quick succession as well as the Bureau’s concerns about potential evasions of the underwriting criteria. As discussed in the proposal, the Bureau also has considered various time frames for the definition of loan sequence, including 14 days as well as 30 days and 60 days, and decided in finalizing the rule to adhere to 30 days as a reasonable and appropriate frequency for use in this 218 definition, to align with consumer expense cycles, which often involve recurring expenses that are typically a month in length. This is designed to account for the fact that where repaying a loan causes a shortfall, the consumer may seek to return during the same expense cycle to get funds to cover downstream expenses. In addition, a number of consumers receive income on a monthly basis. The various considerations involved in resolving these issues are discussed more fully in the section-by-section analysis of § 1041.5(d) of the final rule. In light of the discussion above, the Bureau otherwise is finalizing this renumbered definition as modified. In addition, wherever the proposed definition had referred to a covered short-term loan, the definition in the final rule refers instead to a covered short-term loan or a covered longer-term balloon-payment loan—or, where pluralized, the definition in the final rule refers instead to covered short-term loans or covered longer-term balloon-payment loans, or a combination thereof. 2(a)(15) Motor vehicle In connection with proposing to subject certain longer-term loans with vehicle security to part 1041, in proposed § 1041.3(d) the Bureau would have defined vehicle security to refer to the term motor vehicle as defined in section 1029(f)(1) of the Dodd-Frank Act. That definition encompasses not only vehicles primarily used for on-road transportation, but also recreational boats, motor homes, and other categories. As described below, the Bureau has now decided to narrow the definition of covered-longer term loan to focus only on loans that meet a certain rate threshold and involve the taking of a leveraged payment mechanism as defined in § 1041.3(c) of the final rule, without regard to whether vehicle security is taken on the loan. However, the definitions of vehicle security and motor vehicle are still relevant to § 1041.6(b)(3), which 219 prohibits lenders from making covered short-term loans under § 1041.6 if they take vehicle security in connection with such a loan, for the reasons explained in the section-by-section analysis of that provision. Upon further consideration in light of this context and its experience from other related rulemakings, the Bureau has decided to narrow the definition of motor vehicle in the final rule to focus on any self-propelled vehicle primarily used for on-road transportation, but not including motor homes, recreational vehicles, golf carts, and motor scooters. Some commenters did suggest that vehicle title loans should encompass boats, motorcycles, and manufactured homes. Nonetheless, the Bureau has concluded that it is more appropriate to use a narrower definition because the term motor vehicle is germane to the vehicle title loans addressed in the final rule, which involve the prospect of repossession of the vehicle for failing to repay the loan. The impact to consumers from default or repossession likely operates differently for basic on-road transportation used to get to work or manage everyday affairs, thus creating different pressures to repay loans based on these kinds of vehicles as compared to loans based on other forms of transportation. Moreover, from the Bureau’s prior experience of writing rules with respect to vehicles, most notably in the Bureau’s larger participant rule authorizing its supervision authority over certain entities in the market for auto loans, it is aware that treatment of this category of items requires clarification in light of what can be some difficult and unexpected boundary issues. The definition included here in § 1041.2(a)(15) of the final rule is thus similar to the language used in 220 the Bureau’s larger participant rule for the auto loan market,429 which generally encompasses the kinds of vehicles—specifically cars and trucks and motorcycles—that consumers primarily use for on-road transportation rather than for housing or recreation. The Bureau also notes that it had proposed to exclude loans secured by manufactured homes under § 1041.3(e)(2), and has finalized that provision in § 1041.3(d)(2) as discussed below. 2(a)(16) Open-end credit Proposed § 1041.2(a)(14) would have defined open-end credit by cross-referencing the definition of that same term in Regulation Z, 12 CFR part 1026, but without regard to whether the credit is consumer credit, as that term is defined in Regulation Z § 1026.2(a)(12), is extended by a creditor, as that term is defined in Regulation Z § 1026.2(a)(17), or is extended to a consumer, as that term is defined in Regulation Z § 1026.2(a)(11). In general, Regulation Z § 1026.2(a)(20) provides that open-end credit is consumer credit in which the creditor reasonably contemplates repeated transactions, the creditor may impose a finance charge from time to time on an outstanding unpaid balance, and the amount of credit that may be extended to the consumer during the term of the plan (up to any limit set by the creditor) is generally made available to the extent that any outstanding balance is repaid. For the purposes of defining openend credit under proposed part 1041, the term credit, as defined in proposed § 1041.2(a)(9), was substituted for the term consumer credit in the Regulation Z definition of open-end credit; the term lender, as defined in proposed § 1041.2(a)(11), was substituted for the term creditor in the same Regulation Z definition; and the term consumer, as defined in proposed § 1041.2(a)(4), was substituted for the term consumer in the Regulation Z definition of open-end credit. 429 80 FR 37496 (June 30, 2015). 221 The term open-end credit was used in various parts of the proposal where the Bureau tailored requirements separately for closed-end and open-end credit in light of their different structures and durations. Most notably, proposed § 1041.2(a)(18) would require lenders to employ slightly different methods when calculating the total cost of credit of closed-end versus open-end loans. Proposed § 1041.16(c) also would require lenders to report whether a covered loan is a closed-end or open-end loan. In the proposal, the Bureau stated that generally defining this term consistently across regulations would reduce the risk of confusion among consumers, industry, and regulators. With regard to the definition of consumer, however, the Bureau proposed that, for the reasons discussed in connection with proposed § 1041.2(a)(4), it would be more appropriate to incorporate the definition from the Dodd-Frank Act rather than the definition from Regulation Z, which is arguably narrower. Similarly, the Bureau indicated that it would be more appropriate to use the broader definition of lender contained in proposed § 1041.2(a)(11) than the Regulation Z definition of creditor. One commenter recommended that the Bureau defer action on lines of credit entirely (not just overdraft lines of credit as would be excluded in proposed § 1041.3) and address these loan products in a future rulemaking. A number of commenters stated that the underwriting criteria for such products should be aligned with the provisions of the Credit CARD Act and the Bureau’s rule on prepaid accounts, and raised questions about the timing calculations on line-ofcredit payments. In response, the Bureau continues to judge it to be important to address open-end lines of credit in this rule in order to achieve more comprehensive coverage, outside of those lines of 222 credit that are excluded under final § 1041.3(d)(6) as discussed below. In response to many comments, including those urging closer alignment with other standards for assessing ability to repay under other statutory schemes, the Bureau has also modified the underwriting criteria in § 1041.5 of the final rule in a number of respects, as explained further below. The Bureau is therefore finalizing § 1041.2(a)(16) largely as proposed, with one substantive clarification that credit products that otherwise meet the definition of open-end credit under Regulation Z should not be excluded from the definition of open-end credit under § 1041.2(a)(16) because they do not involve a finance charge. This change will assure that products are appropriately classified as open-end credit under part 1041, rather than as closedend credit. The Bureau has also revised comment 2(a)(16)-1 to reflect this change and to streamline guidance clarifying that for the purposes of defining open-end credit under part 1041, the term credit, as defined in § 1041.2(a)(11), is substituted for the term consumer credit, as defined in 12 CFR 1026.2(a)(12); the term lender, as defined in § 1041.2(a)(13), is substituted for the term creditor, as defined in 12 CFR 1026.2(a)(17); and the term consumer, as defined in § 1041.2(a)(4), is substituted for the term consumer, as defined in 12 CFR 1026.2(a)(11). For all the reasons discussed above, the Bureau is finalizing this definition and the commentary as renumbered and revised. 2(a)(17) Outstanding loan Proposed § 1041.2(a)(15) would have generally defined outstanding loan as a loan that the consumer is legally obligated to repay, except that a loan ceases to be outstanding if the consumer has not made any payments on the loan within the previous 180 days. Under this definition, a loan is an outstanding loan regardless of whether the loan is delinquent or subject to 223 a repayment plan or other workout arrangement if the other elements of the definition are met. Under proposed § 1041.2(a)(12), a covered short-term loan would be considered to be within the same loan sequence as a previous such loan if it is made within 30 days of the consumer having the previous outstanding loan. Proposed §§ 1041.6 and 1041.7 would impose certain limitations on lenders making covered short-term loans within loan sequences, including a prohibition on making additional covered short-term loans for 30 days after the third loan in a sequence. In the proposal, the Bureau stated that if the consumer has not made any payment on the loan for an extended period of time, it may be appropriate to stop considering the loan to be an outstanding loan for the purposes of various provisions of the proposed rule. Because outstanding loans are counted as major financial obligations for purposes of underwriting and because treating a loan as outstanding would trigger certain restrictions on further borrowing by the consumer under the proposed rule, the Bureau attempted to balance several considerations in crafting the proposed definition. One is whether it would be appropriate for very stale and effectively inactive debt to prevent the consumer from accessing credit, even if so much time has passed that it seems relatively unlikely that the new loan is a direct consequence of the unaffordability of the previous loan. Another is how to define such stale and inactive debt for purposes of any cut-off, and to account for the risk that collections might later be revived or that lenders would intentionally exploit a cut-off in an attempt to encourage new borrowing by consumers. The Bureau proposed a 180-day threshold as striking an appropriate balance, and noted that this approach would generally align with the policy position taken by the Federal Financial Institutions Examination Council (FFIEC), which generally requires depository institutions to 224 charge off open-end credit at 180 days of delinquency. Although that policy also requires that closed-end loans be charged off after 120 days, the Bureau found as a preliminary matter that a uniform 180-day rule for both closed-end and open-end loans may be more appropriate, given the underlying policy considerations discussed above, as well as for simplicity. Proposed comment 2(a)(15)-1 would clarify that the status of a loan that otherwise meets the definition of outstanding loan does not change based on whether the consumer is required to pay a lender, affiliate, or service provider or whether the lender sells the loan or servicing rights to a third party. Proposed comment 2(a)(15)-2 would clarify that a loan ceases to be an outstanding loan as of the earliest of the date the consumer repays the loan in full, the date the consumer is released from the legal obligation to repay, the date the loan is otherwise legally discharged, or the date that is 180 days following the last payment that the consumer made on the loan. Additionally, proposed comment 2(a)(15)-2 would explain that any payment the consumer makes restarts the 180-day period, regardless of whether the payment is a scheduled payment or in a scheduled amount. Proposed comment 2(a)(15)-2 would further clarify that once a loan is no longer an outstanding loan, subsequent events cannot make the loan an outstanding loan. The Bureau proposed this one-way valve to ease compliance burden on lenders and to reduce the risk of consumer confusion. One consumer group commented that, with respect to loans that could include more than one payment, it would be helpful for the definition to refer to an installment in order to ensure its alignment with terms used in State and local laws. Other consumer groups suggested various other changes to clarify details of timing addressed in this definition, as well as urging that the 180-day period should be changed to 365 days so that more loans would be considered as 225 outstanding. Several commented that the definition should be changed so that the 180-day period should run from either the date of the last payment by the consumer or from the date of the last debt collection activity by the collector, in order to more accurately determine what is truly stale debt and to broaden the scope of what loans are outstanding to ensure that older loans are not being used by lenders to encourage consumers to re-borrow. To support compliance under the modified definition, they also urged that lenders be required to report collection activity to the registered information systems. The Bureau has concluded that language in final comment 2(a)(17)-2 emphasizing that any payment restarts the 180-day clock is sufficient to address the commenter’s concern without having to incorporate new terminology to align the term with its use in State and local laws. With respect to the comments about the time frame, and 365 days in particular, the Bureau was not persuaded of the reasoning or need to broaden the scope of outstanding loans to this extent. The Bureau’s proposed 180-day period was already aligned to the longer end of the FFIEC treatment of these issues, by adopting the 180 days that the FFIEC has applied to open-end credit rather than the 120 days that it has applied to closed-end credit. In addition, the Bureau’s experience with these markets suggests that these types of lenders typically write off their debts even sooner than 180 days. The Bureau concludes that the various suggested changes that were offered to tighten the proposed standard are not necessary to be adopted at this time, though such matters could be revisited over time as supervision and enforcement of the final rule proceed in the future. In particular, the comment that lenders should be required to report collection activity to the registered information systems would have broadened the requirements of the rule and the 226 burdens imposed in significant and unexpected ways that did not seem warranted at this juncture. The Bureau also carefully considered the comments made about extending the period of an outstanding loan, which suggested that it should run not just 180 days from the date of the last payment made on the loan but also 180 days from the date of the last debt collection activity on the loan. The Bureau declines to adopt this proposed change, for several reasons. It would add a great deal of complexity that would encumber the rule, not only in terms of ensuring compliance but in terms of carrying out supervision and enforcement responsibilities as well. For example, this modification would appear not to be operational unless debt collection activities were reported to the registered information systems, which as noted above would add significant and unexpected burdens to the existing framework. Moreover, timing the cooling-off period to any debt collection activity could greatly extend how long a consumer would have to wait to reborrow after walking away from a debt, thereby disrupting the balance the Bureau was seeking to strike in the proposal between these competing objectives. The Bureau also judged that if the comment was aimed at addressing and discouraging certain types of debt collection activities, it would be better addressed in the rulemaking process that the Bureau has initiated separately to govern debt collection issues. Finally, this suggestion seems inconsistent with the Bureau’s experience, which indicates that lenders in this market typically cease their own collection efforts within 180 days. For these reasons, the Bureau is finalizing this definition as renumbered and the commentary as proposed with minor changes for clarity. The Bureau has also added a sentence to comment 2(a)(17)-2 to expressly state that a loan is outstanding for 180 days after consummation if the consumer does not make any payments on it, the consumer is not otherwise 227 released from the legal obligation to pay, and the loan is not otherwise legally discharged. 2(a)(18) Service provider Proposed § 1041.2(a)(17) would have defined service provider by cross-referencing the definition of that same term in the Dodd-Frank Act, 12 U.S.C. 5481(26). In general, the DoddFrank Act defines service provider as any person that provides a material service to a covered person in connection with the offering or provision of a consumer financial product or service, including one that participates in designing, operating, or maintaining the consumer financial product or service or one that processes transactions relating to the consumer financial product or service. Moreover, the Act specifies that the Bureau’s authority to identify and prevent unfair, deceptive, or abusive acts or practices through its rulemaking authority applies not only to covered persons, but also to service providers.430 Proposed § 1041.3(c) and (d) would provide that a loan is covered under proposed part 1041 if a service provider obtains a leveraged payment mechanism or vehicle title and the other coverage criteria are otherwise met. The definition of service provider and the provisions in proposed § 1041.3(c) and (d) were designed to reflect the fact that in some States, covered loans are extended to consumers through a multi-party transaction. In these transactions, one entity will fund the loan, while a separate entity, often called a credit access business or a credit services organization, will interact directly with, and obtain a fee or fees from, the consumer. This separate entity will often service the loan and guarantee the loan’s performance to the party funding the loan. The credit access business or credit services organization, and not the party funding the loan, will in many cases obtain the leveraged payment mechanism or vehicle security. In these cases, the credit access 430 12 U.S.C. 5531(a) and (b). 228 business or credit services organization is performing the responsibilities normally performed by a party funding the loan in jurisdictions where this particular business arrangement is not used. Despite the formal division of functions between the nominal lender and the credit access business, the loans produced by such arrangement are functionally the same as those covered loans issued by a single entity and appear to present the same set of consumer protection concerns. Accordingly, the Bureau stated in the proposal that it is appropriate to bring loans made under these arrangements within the scope of coverage of proposed part 1041. Proposed comment 2(a)(17)-1 further made clear that persons who provide a material service to lenders in connection with the lenders’ offering or provision of covered loans during the course of obtaining for consumers, or assisting consumers in obtaining, loans from lenders are service providers, subject to the specific limitations in section 1002(26) of the Dodd-Frank Act. The Bureau stated that defining the term service provider consistently with the DoddFrank Act reduces the risk of confusion among consumers, industry, and regulators. Consumer groups commented that the rule should apply to service providers, including credit service organizations and their affiliates, whenever it applies to lenders and their affiliates. The Bureau concludes that the definitions of and references to lender and service provider, including incorporation of the statutory definitions of covered person and service provider into the regulatory definitions, throughout the regulation text and commentary are sufficiently well articulated to make these points clear as to the applicability and scope of coverage of part 1041. Both section 1031(a) and section 1036(a) of the Dodd-Frank Act specify that a service provider can be held liable on the same terms as a covered person—which includes a lender as defined by § 1041.2(13)—to the extent that a service provider engages in conduct that violates this rule on 229 behalf of a lender, or entities such as credit access businesses and credit service organizations that provide a material service to a lender in making these kinds of covered loans.431 The Bureau did not receive any other comments on this portion of the proposal and is finalizing this definition and the commentary as just discussed and as renumbered. 2(a)(19) Vehicle security The Bureau has decided to make “vehicle security” a defined term, incorporating language that described the practice of taking vehicle security from proposed § 1041.3(d). Its role is now more limited, however, due to other changes in the rule, which no longer governs the underwriting of covered longer-term loans (other than balloon-payment loans), which instead are now subject only to the payment provisions. Nonetheless, the Bureau is preserving the language explaining vehicle security and moving it here for purposes of defining the exclusion of vehicle title loans from coverage under § 1041.6 of the final rule, which provides for conditionally exempted loans. As to the definition itself, the proposal would have stated that for purposes of defining a covered loan, a lender or service provider obtains vehicle security if it obtains an interest in a consumer’s motor vehicle (as defined in section 1029(f)(1) of the Dodd-Frank Act) as a condition of the credit, regardless of how the transaction is characterized by State law, including: (1) any security interest in the motor vehicle, motor vehicle title, or motor vehicle registration whether or not the security interest is perfected or recorded; or (2) a pawn transaction in which 431 See 12 U.S.C. 5531(a) (“The Bureau may take any action authorized under subtitle E [Enforcement powers] to prevent a covered person or service provider from committing or engaging in an unfair, deceptive, or abusive act or practice under Federal law in connection with any transaction with a consumer for a consumer financial product or service, or the offering of a consumer financial product or service.”); 12 U.S.C. 5536(a) (equating covered persons and service providers for purposes of prohibited acts in violation of Federal consumer financial law, including liability for violations for engaging in “any unfair, deceptive, or abusive act or practice”). 230 the consumer’s motor vehicle is the pledged good and the consumer retains use of the motor vehicle during the period of the pawn agreement. Under the proposal, the lender or service provider would obtain vehicle security if the consumer is required, under the terms of an agreement with the lender or service provider, to grant an interest in the consumer’s vehicle to the lender in the event that the consumer does not repay the loan. As noted in the proposal, because of exclusions contained in proposed § 1041.3(e)(1) and (e)(5), the term vehicle security would have excluded loans made solely and expressly for the purpose of financing a consumer’s initial purchase of a motor vehicle in which the lender takes a security interest as a condition of the credit, as well as non-recourse pawn loans in which the lender has sole physical possession and use of the property for the entire term of the loan. Proposed comment 3(d)(1)-1 also would have clarified that mechanic liens and other situations in which a party obtains a security interest in a consumer’s motor vehicle for a reason that is unrelated to an extension of credit do not trigger coverage. The Bureau proposed that the security interest would not need to be perfected or recorded in order to trigger coverage under proposed § 1041.3(d)(1). The Bureau reasoned that consumers may not be aware that the security interest is not perfected or recorded, nor would it matter in many cases. Perfection or recordation protects the lender’s interest in the vehicle against claims asserted by other creditors, but does not necessarily affect whether the consumer’s interest in the vehicle is at risk if the consumer does not have the ability to repay the loan. Even if the lender or service provider does not perfect or record its security interest, the security interest can still change a lender’s incentives to determine the consumer’s ability to repay the loan and exacerbate the harms the consumer experiences if the consumer does not have the ability to repay the loan. 231 The Bureau received many comments on the prong of the definition that focused on the taking of a leveraged payment mechanism or vehicle security, again often in the context of application of the underwriting requirements rather than the payment requirements. Those concerns have largely been addressed or mooted by the Bureau’s decisions to apply only the payment requirements to covered longer-term loans and to narrow the definition of such loans to focus only on those types of leveraged payment mechanisms that involve the ability to pull money from consumers’ accounts, rather than vehicle security. Comments focusing on that narrower definition of leveraged payment mechanism are addressed in more depth in connection with § 1041.3(c) below. Importantly, the term vehicle security as defined in proposed § 1041.3(d) was further limited in its effect by the provisions of proposed § 1041.3(b)(3)(ii), which had stated that a lender or service provider did not become subject to the proposed underwriting criteria merely by obtaining vehicle security at any time, but instead had to obtain vehicle security before, at the same time as, or within 72 hours after the consumer receives the entire amount of funds that the consumer is entitled to receive under the loan. Many commenters criticized the 72-hour requirement as undermining consumer protections and fostering evasion of the rule. Because of various changes that have occurred in revising the coverage of the underwriting criteria and reordering certain provisions in the final rule, this limitation is no longer necessary to effectuate any of those purposes of the rule. The definition of vehicle security remains relevant to the provisions of § 1041.6 of the final rule, but it is unclear how a 72-hour limitation is germane to establishing the scope of coverage under that section, and so it has been eliminated from the final rule. 232 One consumer group suggested that a vehicle title loan should be covered under the rule regardless of whether the title was a condition of the loan. The Bureau does not find it necessary to alter the definition in this manner in order to accomplish the purpose of covering vehicle title loans, particularly in light of the language in comment 2(a)(19)-1, which indicates that vehicle security will attach to the vehicle for reasons that are related to the extension of credit. With respect to comments on the details of the definition of vehicle security, one commenter had suggested that the final rule should make clear that the proposed restrictions on this form of security interest do not interfere with or prohibit any statutory liens that have been authorized by Congress. Because nothing in the language of the final rule purports to create any such interference or prohibition, the Bureau does not find it necessary to modify its definition of vehicle security in this regard. Other commenters made various points about the meaning and coverage of the term motor vehicle in the Bureau’s treatment of the term vehicle security. Those comments are addressed separately in the discussion of the definition of motor vehicle in § 1041.2(a)(15) of the final rule. The Bureau has moved the discussion of vehicle security from proposed § 1041.3(d) to § 1041.2(a)(19) in the general definitions section, and has narrowed the definition of motor vehicle contained in section 1029(f)(1) of the Dodd-Frank Act, replacing it with the somewhat narrower definition of motor vehicle contained in § 1041.2(a)(15) of the final rule as described above. The definition of vehicle security still includes the other elements of the proposal, as slightly rewritten for clarity to focus on this term itself rather than on the actions of a lender or service provider. Accordingly, the term vehicle security is defined in the final rule as an interest in a 233 consumer’s motor vehicle obtained by the lender or service provider as a condition of the credit, regardless of how the transaction is characterized by State law, including: (1) any security interest in the motor vehicle, motor vehicle title, or motor vehicle registration whether or not the security interest is perfected or recorded; or (2) a pawn transaction in which the consumer’s motor vehicle is the pledged good and the consumer retains use of the motor vehicle during the period of the pawn agreement. This definition also carries with it proposed comment 3(d)(1)-1, now finalized as comment 2(a)(19)-1, which explains that an interest in a consumer’s motor vehicle is a condition of credit only to the extent the security interest is obtained in connection with the credit, and not for a reason that is unrelated to an extension of credit, such as the attachment of a mechanic’s lien. This comment is finalized with the language unchanged.432 2(b) Rule of construction After reserving this provision in the proposal, the Bureau has determined to add a rule of construction for purposes of part 1041, which states that where definitions are incorporated from other statutes or regulations, the terms have the meaning and incorporate the embedded definitions, appendices, and commentary from those other laws except to the extent that part 1041 provides a different definition for a parallel term. The Bureau had included versions of this basic principle in the regulation text and commentary for certain individual provisions of the proposed rule, but has concluded that it would be helpful to memorialize it as a general rule of construction. Accordingly, the Bureau moved certain proposed commentary for individual 432 Two definitions in the proposal are no longer operative and so have been omitted from the final rule. First, proposed § 1041.2(a)(13) would have defined the term non-covered bridge loan. Second, proposed § 1041.2(a)(16) would have defined the term prepayment penalty. Because the Bureau is not finalizing the portions of the proposed rule on underwriting of covered longer-term loans at this time, along with other changes made in §§ 1041.5 and 1041.6 of the final rule governing the underwriting and provision of covered short-term loans, these two definitions and the related commentary are being omitted from the final rule. 234 definitions to comment 2(b)-1 of the final rule in order to provide examples of the rule of construction, and streamlined certain other proposed commentary as described above. Section 1041.3 Scope of coverage; exclusions; exemptions The primary purpose of proposed part 1041 was to identify and adopt rules to prevent unfair and abusive practices as defined in section 1031 of the Dodd-Frank Act in connection with certain consumer credit transactions. Based upon its research, outreach, and analysis of available data, the Bureau proposed to identify such practices with respect to two categories of loans to which it proposed to apply this rule: (1) consumer loans with a duration of 45 days or less; and (2) consumer loans with a duration of more than 45 days that have a total cost of credit above a certain threshold and that are either repayable directly from the consumer’s income stream, as set forth in proposed § 1041.3(c), or are secured by the consumer’s motor vehicle, as set forth in proposed § 1041.3(d). In the proposal, the Bureau tentatively concluded that it is an unfair and abusive practice for a lender to make a covered short-term loan without determining that the consumer has the ability to repay the loan. The Bureau likewise tentatively concluded that it is an unfair and abusive practice for a lender to make a covered longer-term loan without determining the consumer’s ability to repay the loan. Accordingly, the Bureau proposed to apply the protections of proposed part 1041 to both categories of loans. In particular, proposed §§ 1041.5 and 1041.9 would have required that, before making a covered loan, a lender must determine that the consumer has the ability to repay the loan. Proposed §§ 1041.6 and 1041.10 would have imposed certain limitations on repeat borrowing, depending on the type of covered loan. Proposed §§ 1041.7, 1041.11, and 1041.12 would have 235 provided for alternative requirements that would allow lenders to make covered loans, in certain limited situations, without first determining that the consumer has the ability to repay the loan. Proposed § 1041.14 would have imposed consumer protections related to repeated lenderinitiated attempts to withdraw payments from consumers’ accounts in connection with covered loans. Proposed § 1041.15 would have required lenders to provide notices to consumers before attempting to withdraw payments on covered loans from consumers’ accounts. Proposed §§ 1041.16 and 1041.17 would have required lenders to check and report borrowing history and loan information to certain information systems with respect to most covered loans. Proposed § 1041.18 would have required lenders to keep certain records on the covered loans that they make. And proposed § 1041.19 would have prohibited actions taken to evade the requirements of proposed part 1041. The Bureau did not propose to extend coverage to several other types of loans and specifically proposed excluding, to the extent they would otherwise be covered under proposed § 1041.3, certain purchase money security interest loans, certain loans secured by real estate, credit cards, student loans, non-recourse pawn loans, and overdraft services and lines of credit. The Bureau likewise proposed not to cover loans that have a term of longer than 45 days if they are not secured by a leveraged payment mechanism or vehicle security or if they have a total cost of credit below a rate of 36 percent per annum. By finalizing application of the underwriting requirements with respect to certain categories of loans as described above, and excluding certain other types of loans from the reach of the rule, the Bureau does not mean to signal any definitive conclusion that it could not be an unfair or abusive practice to make any other types of loans, such as loans that are not covered by 236 part 1041, without reasonably assessing a consumer’s ability to repay. Moreover, this rule does not supersede or limit any protections imposed by other laws, such as the Military Lending Act and implementing regulations. The coverage limits in the rule simply reflect the fact that these are the types of loans the Bureau has studied in depth to date and has chosen to address within the scope of the proposal. Indeed, the Bureau issued, concurrently with the proposal, a Request for Information (RFI), which solicited information and evidence to help assess whether there are other categories of loans for which lenders do not determine the consumer’s ability to repay that may pose risks to consumers. The Bureau also sought comment in response to the RFI as to whether other lender practices associated with covered loans may warrant further action by the Bureau. The Bureau thus is reinforcing the point that all covered persons within the meaning of the Dodd-Frank Act have a legal duty not to engage in unfair, deceptive, or abusive acts or practices. The Bureau is explicitly authorized to consider, on a case-by-case basis, through its supervisory or enforcement activities, whether practices akin to those addressed here are unfair, deceptive, or abusive in connection with loans not covered by the rule. The Bureau also is emphasizing that it may decide to engage in future rulemaking with respect to other types of loans or other types of practices associated with covered loans at a later date. 3(a) General In proposed § 1041.3(a), the Bureau provided that proposed part 1041 would apply to a lender that makes covered loans. The Bureau received no specific comments on proposed § 1041.3(a), and is finalizing this provision as proposed except that it has adopted language as 237 discussed above in connection with the definition of lender in § 1041.2(a)(13) to refer to a person who “extends credit by making covered loans.” 3(b) Covered loan In the proposal, the Bureau noted that section 1031(b) of the Dodd-Frank Act empowers it to prescribe rules to identify and prevent unfair, deceptive, or abusive acts or practices associated with consumer financial products or services. Section 1002(5) of the Dodd-Frank Act defines such products or services as those offered or provided for use by consumers primarily for personal, family, or household purposes or, in certain circumstances, those delivered, offered, or provided in connection with another such consumer financial product or service. Proposed § 1041.3(b) would have provided, generally, that a covered loan means closed-end or open-end credit that is extended to a consumer primarily for personal, family, or household purposes that is not excluded by § 1041.3(e). By proposing to apply the rule only to loans that are extended to consumers primarily for personal, family, or household purposes, the Bureau intended it not to apply to loans that are made primarily for a business, commercial, or agricultural purpose. But the proposal explained that a lender would violate proposed part 1041 if it extended a loan ostensibly for a business purpose and failed to comply with the requirements of proposed part 1041 for a loan that is, in fact, primarily for personal, family, or household purposes. In this regard, the Bureau referenced the section-by-section analysis of proposed § 1041.19, which provided further discussion of evasion issues. Proposed comment 3(b)-1 would have clarified that whether a loan is covered is generally based on the loan terms at the time of consummation. Proposed comment 3(b)-2 238 would have clarified that a loan could be a covered loan regardless of whether it is structured as open-end or closed-end credit. Proposed comment 3(b)-3 would have explained that the test for determining the primary purpose of a loan is the same as the test prescribed by Regulation Z § 1026.3(a) and clarified by the related commentary in supplement I to part 1026. The Bureau stated that lenders are already familiar with the Regulation Z test and that it would be appropriate to apply that same test here to maintain consistency in interpretation across credit markets, though the Bureau also requested comment on whether more tailored guidance would be useful here as the related commentary in supplement I to part 1026, on which lenders would be permitted to rely in interpreting proposed § 1041.3(b), did not discuss particular situations that may arise in the markets that would be covered by proposed part 1041. One commenter noted that while business loans are outside the scope of the rule, many small business owners use their personal vehicles to secure title loans for their businesses, and asserted that it will be difficult for lenders to differentiate the purposes of a loan in such instances. Another commenter suggested that provisions should be added to ensure that loans are made for personal use only. More generally, one commenter stated that the breadth of the definition of covered loan would enhance the burden that the proposed rule would impose on credit unions. In response, the Bureau notes that its experience with these markets has made it aware that the distinction between business and household purposes is necessarily fact-specific, yet the basic distinction is embedded as a jurisdictional matter in many consumer financial laws and has long been regarded as a sensible line to draw. Further, the concern about the breadth of this definition as affecting credit unions is addressed substantially by the measures adopted in the 239 final rule to reduce burdens for lenders, along with the exclusions and exemptions that have been adopted, including the conditional exemption for alternative loans. The Bureau is finalizing § 1041.3(b) as proposed. The commentary is finalized as proposed, except proposed comment 3(b)-1, which the Bureau is not finalizing. That comment had proposed that whether a loan is covered is generally determined based on the loan terms at the time of consummation. As noted below, final comment 3(b)(3)-3 makes clear that a loan may become a covered longer-term loan at any such time as both requirements of § 1041.3(b)(3)(i) and (ii) are met, even if they were not met when the loan was initially made. 3(b)(1) Proposed § 1041.3(b)(1) would have brought within the scope of proposed part 1041 those loans in which the consumer is required to repay substantially the entire amount due under the loan within 45 days of either consummation or the advance of loan proceeds. Loans of this type, as they exist in the market today, typically take the form of single-payment loans, including payday loans, vehicle title loans, and deposit advance products. However, coverage under proposed § 1041.3(b)(1) was not limited to single-payment products, but rather included any single-advance loan with a term of 45 days or less and any multi-advance loan where repayment is required within 45 days of a credit draw.433 Under proposed § 1041.2(a)(6), this type of covered loan was defined as a covered short-term loan. Specifically, proposed § 1041.3(b)(1) prescribed different tests for determining whether a loan is a covered short-term loan based on whether or not the loan is closed-end credit that does 433 While application of the 45-day duration limit for covered short-term loans varies based on whether the loan is a single- or multiple-advance loan, the Bureau often used the phrase “within 45 days of consummation” throughout the proposal and in the final rule as a shorthand way of referring to coverage criteria of both types of loans. 240 not provide for multiple advances to consumers. For this type of credit, a loan would be a covered short-term loan if the consumer is required to repay substantially the entire amount of the loan within 45 days of consummation. For all other types of loans, a loan would be a covered short-term loan if the consumer is required to repay substantially the entire amount of an advance within 45 days of the advance. As proposed comment 3(b)(1)-1 explained, a loan does not provide for multiple advances to a consumer if the loan provides for full disbursement of the loan proceeds only through disbursement on a single specific date. The Bureau stated that a different test to determine whether a loan is a covered short-term loan is appropriate for loans that provide for multiple advances to consumers, because open-end credit and closed-end credit providing for multiple advances may be consummated long before the consumer incurs debt that must be repaid. If, for example, the consumer waited more than 45 days after consummation to draw on an open-end line, but the loan agreement required the consumer to repay the full amount of the draw within 45 days of the draw, the loan would not be practically different than a closed-end loan repayable within 45 days of consummation. The Bureau preliminarily found that it is appropriate to treat the loans the same for the purposes of proposed § 1041.3(b)(1). As the Bureau described in part II of the proposal, the terms of short-term loans are often tied to the date the consumer receives his or her paycheck or benefits payment. While pay periods typically vary from one week to one month, and expense cycles are typically one month, the Bureau proposed 45 days as the upper bound for covered short-term loans in order to accommodate loans that are made shortly before a consumer’s monthly income is received and that extend beyond the immediate income payment to the next income payment. These 241 circumstances could result in loans that are somewhat longer than a month in duration, but the Bureau believed that they nonetheless pose similar risks of harm to consumers as loans with durations of a month or less. The Bureau also considered proposing to define covered short-term loans as loans that are substantially repayable within either 30 days of consummation or advance, 60 days of consummation or advance, or 90 days of consummation or advance. The Bureau, nonetheless, did not propose the 30-day period because, as described above, some loans for some consumers who are paid on a monthly basis can be slightly longer than 30 days, yet still would essentially constitute a one-pay-cycle, one-expense-cycle loan. The Bureau stated that it did not propose either the 60-day or 90-day period because loans with those terms encompass multiple income and expense cycles, and thus may present somewhat different risks to consumers, though such loans would have been covered longer-term loans if they met the criteria set forth in proposed § 1041.3(b)(2). As discussed in the proposal, the Bureau proposed to treat longer-term loans, as defined in proposed § 1041.3(b)(2), as covered loans only if the total cost of credit exceeds a rate of 36 percent per annum and if the lender or service provider obtains a leveraged payment mechanism or vehicle security as defined in proposed § 1041.3(c) and (d). The Bureau did not propose similar limitations with respect to the definition of covered short-term loans because the evidence available to the Bureau seemed to suggest that the structure and short-term nature of these loans give rise to consumer harm even in the absence of costs above the 36 percent threshold or particular means of repayment. 242 Proposed comment 3(b)(1)-2 noted that both open-end credit and closed-end credit may provide for multiple advances to consumers. The comment explained that open-end credit is self-replenishing even though the plan itself has a fixed expiration date, as long as during the plan’s existence the consumer may use the line, repay, and reuse the credit. Likewise, closedend credit may consist of a series of advances. For example, under a closed-end commitment, the lender might agree to lend a fixed total amount in a series of advances as needed by the consumer, and once the consumer has borrowed the maximum, no more is advanced under that particular agreement, even if there has been repayment of a portion of the debt. Proposed comment 3(b)(1)-3 explained that a determination of whether a loan is substantially repayable within 45 days requires assessment of the specific facts and circumstances of the loan. Proposed comment 3(b)(1)-4 provided guidance on determining whether loans that have alternative, ambiguous, or unusual payment schedules would fall within the definition. The comment explained that the key principle in determining whether a loan would be a covered short-term loan or a covered longer-term loan is whether, under applicable law, the consumer would be considered to be in breach of the terms of the loan agreement if the consumer failed to repay substantially the entire amount of the loan within 45 days of consummation. As noted above, § 1041.3(b)(1) provides the substance of the definition of covered shortterm loan as referenced in § 1041.2(a)(10) of the final rule. The limited comments on this provision are presented and addressed in the section-by-section analysis of that definition. For the reasons stated there, the Bureau is finalizing § 1041.3(b)(1) as proposed, with only nonsubstantive language changes. One modification has been made in the commentary, however, to 243 address comments received about deposit advance products. New comment 3(b)(1)-4 in the final rule states that a loan or advance is substantially repayable within 45 days of consummation or advance if the lender has the right to be repaid through a sweep or withdrawal of any qualifying electronic deposit made into the consumer’s account within 45 days of consummation or advance. A loan or advance described in this paragraph is substantially repayable within 45 days of consummation or advance even if no qualifying electronic deposit is actually made into or withdrawn by the lender from the consumer’s account. This comment was added to address more explicitly a deposit advance product in which the lender can claim all the income coming in to the account, as it comes in, for the purpose of repaying the loan, regardless of whether income in fact comes in during the first 45 days after a particular advance. Proposed comment 3(b)(1)-4 thus has been renumbered as comment 3(b)(1)-5 of the final rule. 3(b)(2) Proposed § 1041.3(b)(2) would have brought within the scope of proposed part 1041 several types of loans for which, in contrast to loans covered under proposed § 1041.3(b)(1), the consumer is not required to repay substantially the entire amount of the loan or advance within 45 days of consummation or advance. Specifically, proposed § 1041.3(b)(2) extended coverage to longer-term loans with a total cost of credit exceeding a rate of 36 percent per annum if the lender or service provider also obtains a leveraged payment mechanism as defined in proposed § 1041.3(c) or vehicle security as defined in proposed § 1041.3(d) in connection with the loan before, at the same time, or within 72 hours after the consumer receives the entire amount of funds that the consumer is entitled to receive. Under proposed § 1041.2(a)(8), this type of covered loan would be defined as a covered longer-term loan. 244 As discussed above in connection with § 1041.2(a)(7), the Bureau defined a sub-category of covered longer-term loans that would be subject to certain tailored provisions in proposed §§ 1041.6, 1041.9, and 1041.10 because they involved balloon-payment structures that the Bureau believed posed particular risks to consumers. The Bureau proposed to cover such longerterm balloon-payment loans only if they exceeded the general rate threshold and involved leveraged payment mechanisms or vehicle security, but specifically sought comment on whether such products should be subject to the rule more generally in light of the particular concerns about balloon payment structures. In light of the Bureau’s decision to differentiate which parts of the rule apply to longerterm balloon-payment loans and more generally to longer-term loans, the Bureau has decided to make the two categories mutually exclusive and to describe them separately in § 1041.3(b)(2) and § (b)(3) of the final rule, respectively. Accordingly, the Bureau is finalizing § 1041.3(b)(2) to define longer-term balloon-payment loans, incorporating the language of proposed § 1041.2(a)(7) as further revised in various respects. First, for purposes of greater clarity in ordering § 1041.3(b) of the final rule, the Bureau is separating out its treatment of covered longer-term balloon-payment loans (in § 1041.3(b)(2)) from its treatment of all other covered longer-term loans (in § 1041.3(b)(3)). As described in greater detail below in Market Concerns—Underwriting and in the section-by-section analysis of § 1041.4, the Bureau has decided to restructure these provisions in this way because it has decided in the final rule to subject covered longer-term balloon-payment loans both to the underwriting criteria and the payment requirements of the final rule, but to apply only the payment requirements to other types of covered longer-term loans. 245 This organization reflects in part the comments received from industry and trade groups who contended that the Bureau’s concerns about re-borrowing for covered longer-term loans were most applicable to loans with balloon-payment structures. They therefore argued that any ability-to-repay restrictions and underwriting criteria should be limited to longer-term balloonpayment loans. These comments reinforced the Bureau’s preliminary view that concerns about the re-borrowing of covered longer-term balloon-payment loans were most similar to the concerns it had about the re-borrowing of covered short-term loans. As described more fully below in the section on Market Concerns—Underwriting, the Bureau has observed longer-term loans involving balloon payments where the lender does not reasonably assess the borrower’s ability to repay before making the loan, and has observed in these circumstances the same types of consumer harms that it has observed when lenders fail to make a reasonable assessment of the borrower’s ability to repay before making covered short-term loans. Nonetheless, the Bureau also maintains its concerns about lender practices in the market for other covered longer-term loans, and emphasizes that it retains supervision and enforcement authority to oversee such lenders for unfair, deceptive, or abusive acts or practices. As discussed further below, for a number of reasons the Bureau has decided not to address the underwriting of all covered longer-term loans at this time. Nonetheless, as discussed above in the section-by-section analysis of § 1041.2(a)(7) of the final rule, the Bureau is concerned that covered longer-term balloon-payment loans have a loan structure that poses many of the same risks and harms to consumers as with covered short-term loans, and could be adapted in some manner as a loan product intended to circumvent the underwriting criteria for covered short-term loans. Therefore, in § 1041.5 of the final rule, the specific underwriting criteria that 246 apply to covered short-term loans are, with certain modifications, made applicable to covered longer-term balloon-payment loans also (without regard to interest rate or the taking of a leveraged payment mechanism). And along with other covered longer-term loans, these loans remain covered by the sections of the final rule on payment practices as well. Given this resolution of the considerations raised by the comments and based on the Bureau’s further consideration and analysis of the market, the Bureau is finalizing § 1041.3(b)(2) in parallel with § 1041.3(b)(1), since both types of loans—covered short-term loans and covered longer-term balloon loans—are subject to the same underwriting criteria and payment requirements as prescribed in the final rule. As noted above in the discussion of § 1041.2(a)(7), in conjunction with making the definition of covered longer-term balloon-payment loan into a separate category in its own right rather than a subcategory of the general definition of covered longer-term loan, the Bureau has decided to subject such loans to an expansion in scope as compared to the proposal, since longerterm balloon-payment loans are now being covered by both the underwriting and payment provisions of the final rule without regard to whether the loans exceed a particular threshold for the cost of credit or involve the taking of a leveraged payment mechanism or vehicle security. The Bureau had specifically sought comment as to whether to cover longer-term balloonpayment loans regardless of these two conditions, and has concluded that it is appropriate to do so in light of concerns about the risks and harms that balloon-payment structures pose to consumers and of potential industry evolution to circumvent the rule, as set out more extensively below in Market Concerns—Underwriting. 247 The Bureau has also revised the definition of covered longer-term balloon-payment loan to address different types of loan structures in more detail. As discussed above in connection with § 1041.2(a)(7), the proposal would generally have defined the term to include loans that require repayment in a single payment or that require at least one payment that is more than twice as large as any other payment(s) under the loan. The Bureau based the twice-as-large threshold on the definition of balloon payment under Regulation Z, but with some modification in details. However, the Bureau did not expressly address whether covered longer-term balloonpayment loans could be both closed-end and open-end credit. After further consideration of the policy concerns that prompted the Bureau to apply the underwriting requirements in subpart B to covered longer-term balloon-payment loans, the Bureau has concluded that it is appropriate to define that term to include both closed-end and open-end loans that involve the kinds of large irregular payments that were described in the proposed definition. In light of the fact that such loans could be structured a number of ways, the Bureau finds it helpful for purposes of implementation and compliance to build out the definition to more expressly address different types of structures. The Bureau has done this by structuring § 1041.3(b)(2) to be similar to the covered-short-term definition in § 1041.3(b)(1), but with longer time frames and descriptions of additional potential payment structures. Specifically, the revised definition for covered longer-term balloon-payment loans separately addresses closed-end loans that do not provide for multiple advances from other loans (both closed-end and open-end) that do involve multiple advances. With regard to the former set of loans, § 1041.3(b)(2)(i) defines a covered longer-term balloon-payment loan to include those where the consumer is required to repay the entire balance of the loan more than 45 days after 248 consummation in a single payment or to repay such loan through at least one payment that is more than twice as large as any other payment(s). With regard to multiple-advance loans, the revised definition focuses on either of two types of payment structures. Under the first structure, the consumer is required to repay substantially the entire amount of an advance more than 45 days after the advance is made or is required to make at least one payment on the advance that is more than twice as large as any other payment(s). Under the second structure, the consumer is paying the required minimum payments but may not fully amortize the outstanding balance by a specified date or time, and the amount of the final payment to repay the outstanding balance at such time could be more than twice the amount of other minimum payments under the plan. The contours of this definition are thus very similar to those for covered short-term loans, which pose the same kinds of risks and harms for consumers, and its focus on payments that are more than twice as large as other payments is generally consistent with how balloon-payment loans are defined and treated under Regulation Z. The Bureau believes retaining that payment size threshold will promote consistency and reduce the risk of confusion among consumers, industry, and regulators. Along with finalizing § 1041.3(b)(2) as just stated, the Bureau has also built out the related commentary to incorporate the original commentary to proposed § 1041.2(a)(7) and concepts that were already used in the definition of covered short-term loan, as well as to elaborate further on language that has been added to the final rule. As now adopted, comment 3(b)(2)-1 specifies that a closed-end loan is considered to be a covered longer-term balloonpayment loan if the consumer must repay the entire amount of the loan in a single payment which is due more than 45 days after the loan was consummated, or to repay substantially the 249 entire amount of any advance in a single payment more than 45 days after the funds on the loan were advanced, or is required to pay at least one payment that is more than twice as large as any other payment(s). Comment 3(b)(2)-2 states that for purposes of § 1041.3(b)(2)(i) and (ii), all required payments of principal and any charges (or charges only, depending on the loan features) due under the loan are used to determine whether a particular payment is more than twice as large as another payment, regardless of whether the payments have changed during the loan term due to rate adjustments or other payment changes permitted or required under the loan. Comment 3(b)(2)-3 discusses charges for actual unanticipated late payments, for exceeding a credit limit, or for delinquency, default, or a similar occurrence that may be added to a payment, and notes that they are excluded from the determination of whether the loan is repayable in a single payment or a particular payment is more than twice as large as another payment. Likewise, sums that are accelerated and due upon default are excluded from the determination of whether the loan is repayable in a single payment or a particular payment is more than twice as large as another payment. These three comments are based on prior comments to proposed § 1041.2(a)(7), with certain revisions made for consistency and form. Comment 3(b)(2)-4 is new and provides that open-end loans are considered to be covered longer-term balloon-payment loans under § 1041.3(b)(2)(ii) if: either the loan has a billing cycle with more than 45 days and the full balance is due in each billing period, or the credit plan is structured such that paying the required minimum payment may not fully amortize the outstanding balance by a specified date or time, and the amount of the final payment to repay the outstanding balance at such time could be more than twice the amount of other minimum payments under the plan. An example is provided to show how this works for an open-end loan, 250 in light of particular credit limits, monthly billing cycles, minimum payments due, fees or interest, and payments made, to determine whether the credit plan is a covered loan and why. 3(b)(3) As noted above, proposed § 1041.3(b)(2) encompassed both covered longer-term balloon-payment loans and certain other covered longer-term loans. Because the Bureau is finalizing a separate definition of covered longer-term balloon-payment loans in § 1041.3(b)(2), new § 1041.3(b)(3) of the final rule addresses covered loans that are neither covered short-term loans nor covered longer-term balloon-payment loans, but rather are covered longer-term loans that are only subject to provisions of the rule relating to payment practices. Specifically, proposed § 1041.3(b)(2) would have extended coverage to longer-term loans with a total cost of credit exceeding a rate of 36 percent per annum if the lender or service provider also obtains a leveraged payment mechanism as defined in proposed § 1041.3(c) or vehicle security as defined in proposed § 1041.3(d) in connection with the loan before, at the same time, or within 72 hours after the consumer receives the entire amount of funds that the consumer is entitled to receive. Under proposed § 1041.2(a)(8), this type of covered loan would have been defined as a covered longer-term loan. The Bureau received extensive comments on covered longer-term loans, but key changes in the final rule mitigate most of the points made in those comments. As discussed above in connection with § 1041.2(a)(8), many commenters offered views on the prongs of the definition of covered longer-term loan as triggers for whether such loans should be subject not only to the payment requirements of part 1041 but also its underwriting requirements. As just discussed above and discussed more fully in part I and in Market Concerns—Underwriting, the Bureau has 251 decided not to apply these underwriting requirements to longer-term loans unless they involve balloon payments as defined in §§ 1041.2(a)(7) and 1041.3(b)(2). However, the Bureau believes that such longer-term loans may still pose substantial risk to consumers with regard to certain lender payment practices, and therefore is finalizing subpart C of the rule to apply to covered longer-term loans. It thus remains relevant to describe the parameters of such loans in § 1041.3(b)(3) of the final rule, which continues to provide the substantive content for the parallel definition of covered longer-term loans in § 1041.2(a)(8) of the final rule. In light of this decision about the policy interventions, the Bureau has also decided to narrow the definition of covered longer-term loans relative to the proposal both by relaxing the rate threshold and narrowing the focus to only loans involving the taking of a leveraged payment mechanism. Thus, § 1041.3(b)(3) of the final rule defines covered longer-term loans as loans that do not meet the definition of covered short-term loans under § 1041.3(b)(1) or of covered longer-term balloon-payment loans under § 1401.3(b)(2); for all remaining covered loans, two further limitations that were contained in the proposed rule apply, so that a loan only becomes a covered longer-term loan if both of the following conditions are also satisfied: the cost of credit for the loan exceeds a rate of 36 percent per annum, as measured in specified ways; and the lender or service provider obtains a leveraged payment mechanism as defined in § 1041.3(c) of the final rule. As described above in connection with the definition of cost of credit in § 1041.2(a)(6), the Bureau has decided to relax the rate threshold in the final rule by basing the threshold on the annual percentage rate as defined in Regulation Z rather than the total cost of credit concept used in the Military Lending Act. The final rule retains the numeric threshold of 36 percent, however, 252 since, as the proposal explained more fully, that annual rate is grounded in many established precedents of Federal and State law. With regard to the taking of leveraged payment mechanisms or vehicle security as part of the definition of covered longer-term loan, as discussed in more detail below in connection with § 1041.3(c), the Bureau has narrowed the definition to focus solely on loans that involve types of leveraged payment mechanisms that enable a lender to pull funds directly from a consumer’s account. Accordingly, a loan that involves vehicle security may be a covered longer-term loan if it involves a leveraged payment mechanism under § 1041.3(c), but not because it involves vehicle security in its own right. The final rule also modifies and clarifies certain details of timing about when status as a covered longer-term loan is determined, in light of the fact that such loans are only subject to the payment requirements under the final rule. With regard to the rate threshold, it is measured at the time of consummation for closed-end credit. For open-end credit, it is measured at consummation and, if the cost of credit at consummation is not more than 36 percent per annum, again at the end of each billing cycle for open-end credit. Once open-end credit meets the threshold, it is treated as doing so for the duration of the plan. The rule also provides a rule for calculating the cost of credit in any billing cycle in which a lender imposes a charge included in the cost of credit where the principal balance is $0. The definition of leveraged payment mechanisms is also truncated, as mechanisms based on access to employer payments or payroll deduction repayments are no longer germane to a policy intervention that is limited solely to the payment practices in § 1041.8 of the final rule. Also, vehicle security is no longer relevant to determining coverage of longer-term loans. The Bureau has also omitted language providing a 253 72-hour window for determining coverage as a longer-term loan from the final rule, as that was driven largely by the need for certainty on underwriting. In short, the two major modifications to this provision as it had been set forth in the proposal are further clarification of how the 36 percent rate is measured for open-end credit and the removal of any references to vehicle security and other employment-based sources of repayment. The commentary to proposed § 1041.3(b)(2) has been extensively revised in light of the other restructuring that has occurred in § 1041.3(b) of the final rule. To summarize briefly, comments 3(b)(3)-1 to 3(b)(3)-3 and 3(b)(3)(ii)-1 to 3(b)(3)(ii)-2 largely recapitulate the provisions of § 1041.3(b)(3) of the final rule in greater detail, as well as clarifying their practical application through a series of examples. Two key points of clarification, however, concern timing. First, comment 3(b)(3)-3 makes clear that a loan may become a covered longer-term loan at any such time as both requirements of § 1041.3(b)(3)(i) and (ii) are met, even if they were not met when the loan was initially made. Second, comment 3(b)(3)(ii)-1 states that the condition in § 1041.3(b)(3)(ii) is satisfied if a lender or service provider obtains a leveraged payment mechanism before, at the same time as, or after the consumer receives the entire amount of funds that the consumer is entitled to receive under the loan, regardless of the means by which the lender or service provider obtains a leveraged payment mechanism. For the reasons stated in view of the comments, the Bureau is finalizing § 1041.3(b)(3) and the commentary as described above. 3(c) Leveraged payment mechanism Proposed § 1041.3(c) would have set forth three ways that a lender or a service provider could obtain a leveraged payment mechanism that, if other conditions were met under proposed 254 § 1041.3(b)(2), would bring a longer-term loan within the proposed coverage of proposed part 1041. Specifically, the proposal would have treated a lender as having obtained a leveraged payment mechanism if the lender or service provider had the right to initiate a transfer of money from the consumer’s account to repay the loan, the contractual right to obtain payment from the consumer’s employer or other payor of expected income, or required the consumer to repay the loan through payroll deduction or deduction from another source of income. In all three cases, the consumer would be required, under the terms of an agreement with the lender or service provider, to cede autonomy over the consumer’s account or income stream in a way that the Bureau believed, as stated in the proposal, would change incentives to determine the consumer’s ability to repay the loan and can exacerbate the harms the consumer experiences if the consumer does not have the ability to repay the loan and still meet the consumer’s basic living expenses and major financial obligations. As explained in the section-by-section analysis of proposed §§ 1041.8 and 1041.9, the Bureau preliminarily found that it is an unfair and abusive practice for a lender to make such a loan without determining that the consumer has the ability to repay. Proposed § 1041.3(c)(1) generally would have provided that a lender or a service provider obtains a leveraged payment mechanism if it has the right to initiate a transfer of money, through any means, from a consumer’s account (as defined in proposed § 1041.2(a)(1)) to satisfy an obligation on a loan. For example, this would occur with a post-dated check or preauthorization for recurring electronic fund transfers. However, the proposed regulation did not define leveraged payment mechanism to include situations in which the lender or service provider initiates a one-time electronic fund transfer immediately after the consumer authorizes such transfer. 255 In the proposal, the functionality of this determination was that it served as one of three preconditions to the underwriting of such covered longer-term loans, along with the provisions of proposed § 1041.3(c)(2) and (c)(3). In light of other changes to the proposed rule, however, the final rule is no longer covering the underwriting of covered longer-term loans (other than balloon-payment loans), but simply determining whether they are subject to the intervention for payment practices in § 1041.8 of the final rule. As described above, as a result of the decision to apply only the rule’s payment requirements to covered-longer term loans, the Bureau is not finalizing the provisions of proposed § 1041.3(c)(2) and (c)(3), which covered payment directly from the employer and repayment through payroll deduction, respectively, as they are no longer germane to the purpose of this policy intervention. With the elimination of those two provisions, § 1041.3(c)(1) is being reorganized more simply as just part of § 1041.3(c) of the final rule to focus on forms of leveraged payment mechanism that involve direct access to consumers’ transaction accounts. Proposed § 1041.3(c)(1) generally would have provided that a lender or a service provider obtains a leveraged payment mechanism if it has the right to initiate a transfer of money, through any means, from a consumer’s account (as defined in proposed § 1041.2(a)(1)) to satisfy an obligation on a loan. For example, this would occur with a post-dated check or preauthorization for recurring electronic fund transfers. However, the proposed regulation did not define leveraged payment mechanism to include situations in which the lender or service provider initiates a one-time electronic fund transfer immediately after the consumer authorizes such transfer. 256 As proposed comment 3(c)(1)-1 explained, the key principle that makes a payment mechanism leveraged is whether the lender has the ability to “pull” funds from a consumer’s account without any intervening action or further assent by the consumer. In those cases, the lender’s ability to pull payments from the consumer’s account gives the lender the ability to time and initiate is to coincide with expected income flows into the consumer’s account. This means that the lender may be able to continue to obtain payment (as long as the consumer receives income and maintains the account) even if the consumer does not have the ability to repay the loan while meeting his or her major financial obligations and basic living expenses. In contrast, the Bureau stated in the section-by-section analysis of proposed § 1041.3(c)(1) that a payment mechanism in which the consumer “pushes” funds from his or her account to the lender does not provide the lender leverage over the account in a way that changes the lender’s incentives to determine the consumer’s ability to repay the loan or exacerbates the harms the consumer experiences if the consumer does not have the ability to repay the loan. Proposed comment 3(c)(1)-2 provided examples of the types of authorizations for lenderinitiated transfers that constitute leveraged payment mechanisms. These include checks written by the consumer, authorizations for electronic fund transfers (other than immediate one-time transfers as discussed further below), authorizations to create or present remotely created checks, and authorizations for certain transfers by account-holding institutions (including a right of setoff). Proposed comment 3(c)(1)-4 explained that a lender does not obtain a leveraged payment mechanism if a consumer authorizes a third party to transfer money from the consumer’s account to a lender as long as the transfer is not made pursuant to an incentive or instruction from, or duty to, a lender or service provider. Proposed comment 3(c)(1)-3 contained similar language. 257 As noted above, proposed § 1041.3(c)(1) provided that a lender or service provider does not obtain a leveraged payment mechanism by initiating a one-time electronic fund transfer immediately after the consumer authorizes the transfer. This provision is similar to what the Bureau proposed in § 1041.15(b), which exempts lenders from providing the payment notice when initiating a single immediate payment transfer at the consumer’s request, as that term is defined in proposed § 1041.14(a)(2), and is also similar to what the Bureau proposed in § 1041.14(d), which permits lenders to initiate a single immediate payment transfer at the consumer’s request even after the prohibition in proposed § 1041.14(b) on initiating further payment transfers has been triggered. Accordingly, proposed comment 3(c)(1)-3 clarified that if the loan agreement between the parties does not otherwise provide for the lender or service provider to initiate a transfer without further consumer action, the consumer may authorize a one-time transfer without causing the loan to be a covered loan. Proposed comment 3(c)(1)-3 further clarified that the term “immediately” means that the lender initiates the transfer after the authorization with as little delay as possible, which in most circumstances will be within a few minutes. Proposed comment 3(c)(1)-4 took the opposite perspective, noting that a lender or service provider does not initiate a transfer of money from a consumer’s account if the consumer authorizes a third party, such as a bank’s automatic bill pay service, to initiate a transfer of money from the consumer’s account to a lender or service provider as long as the third party does not transfer the money pursuant to an incentive or instruction from, or duty to, a lender or service provider. In the proposal, the Bureau noted that it anticipated that scenarios involving authorizations for immediate one-time transfers would only arise in certain discrete situations. 258 For closed-end loans, a lender would be permitted to obtain a leveraged payment mechanism more than 72 hours after the consumer has received the entirety of the loan proceeds without the loan becoming a covered loan. Thus, in the closed-end context, this exception would only be relevant if the consumer was required to make a payment within 72 hours of receiving the loan proceeds—a situation which is unlikely to occur. However, the Bureau acknowledged that the situation may be more likely to occur with open-end credit. According to the proposal, longerterm open-end loans could be covered loans if the lender obtained a leveraged payment mechanism within 72 hours of the consumer receiving the full amount of the funds which the consumer is entitled to receive under the loan. Thus, if a consumer only partially drew down the credit plan, but the consumer was required to make a payment, a one-time electronic fund transfer could trigger coverage without the one-time immediate transfer exception. The Bureau received a few comments on § 1041.3(c)(1) of the proposed rule and the related commentary. One commenter contended that the definition of leveraged payment mechanism is overly broad as between different types of push and pull transactions. Another commenter claimed that the Bureau was improperly attributing motive to the practices of different types of lenders that were using the same leveraged payment mechanisms, that its treatment of leveraged payment mechanisms would have more than a minimal effect on lenders that were already engaged in substantial underwriting, and that the proposed rule and commentary were misaligned with respect to transactions that push or pull money from the consumer’s account. In response to these comments, the Bureau concludes that, in general, its definition is reasonably calibrated to address the core practice at issue here, which is a lender or service 259 provider establishing a right to initiate payment directly from the consumer without any intervening action or further assent from the consumer, subject to certain narrow limitations. The definition of leveraged payment mechanism thus is not overbroad for the purposes served by the rule. As for the final set of comments, the Bureau did not undertake any inquiry or determine any of these issues based on speculation about the motivations of particular lenders; rather, it presumed that lenders that secure leveraged payment mechanisms do so for a mix of reasons. The Bureau also acknowledges at least some tension between the proposed rule and the related commentary in their treatment of push and pull transactions from a consumer’s account. On further consideration, however, the Bureau has concluded that with the focus now solely on payment practices, push transactions are no longer germane to the analysis and thus has revised proposed comments 3(c)(1)-1 and 3(c)(1)-4 accordingly. In light of these comments received and the responses, the Bureau is finalizing proposed § 1041.3(c)(1) as part of § 1041.3(c), and is revising the definition of leveraged payment mechanism to align more closely with the rule’s payment provisions. Specifically, the Bureau is revising the proposed language that would have excluded a one-time immediate transfer from the definition. Under the definition as finalized, the exception applies if the lender initiates a single immediate payment transfer at the consumer’s request, as defined in § 1041.8(a)(2). As discussed in the section-by-section analysis of §§ 1041.8 and 1041.9, transfers meeting the definition of a single immediate payment transfer at the consumer’s request are excluded from the cap on failed payment attempts and the payment notice requirements. The Bureau has concluded that using the same definition for purposes of excluding certain transfers from the definition of leveraged payment mechanism is important for the consistency of the rule. 260 One practical result of this revision is that, whereas the proposed exclusion from the definition of leveraged payment mechanism would have applied only to a one-time electronic fund transfer, the exclusion as finalized permits the lender to initiate an electronic fund transfer or process a signature check without triggering coverage under § 1041.3(b)(3), provided that the lender initiates the transfer or processes the signature check in accordance with the timing and other conditions in § 1041.8(a)(2). The Bureau notes, however, that the definition of single immediate payment transfer at the consumer’s request applies only to the first time that a lender initiates the electronic fund transfer or processes the signature check pursuant to the exception. It does not apply to the re-presentment or re-submission of a transfer or signature check that is returned for nonsufficient funds. If a transfer or signature check is returned, the lender could still work with the consumer to obtain payment in cash or to set up another transfer meeting the definition of single immediate payment transfer at the consumer’s request. The Bureau is finalizing the remainder of the commentary to this provision, which is reordered as comments 3(c)-1 to 3(c)-4 of the final rule, with revisions to the language consistent with the revisions made to the definition of leverage payment mechanism in § 1041.3(c). 3(d) Exclusions for certain credit transactions As discussed above, the Bureau decided to narrow how part 1041 applies to covered longer-term loans to focus only on payment practices. Accordingly, the detailed discussion of vehicle security that appeared in proposed § 1041.3(d) in connection with the definition of covered longer-term loan under proposed § 1041.3(b)(2) is no longer germane to the final rule. As noted in the section-by-section analysis of § 1041.2(a)(19) of the final rule, the Bureau has now moved certain language from proposed § 1041.3(d) describing vehicle security to § 261 1041.2(a)(19) of the final rule, since vehicle security is relevant to application to § 1041.6 of the final rule. Thus the remainder of § 1041.3 is being renumbered, and all references to the provisions of proposed § 1041.3(e) have now been finalized as § 1041.3(d), with further revisions and additions as described below. Proposed § 1041.3(e) would have excluded specific types of credit from part 1041, specifically purchase money security interest loans extended solely for the purchase of a good, real estate secured loans, certain credit cards, student loans, non-recourse pawn loans in which the consumer does not possess the pledged collateral, and overdraft services and overdraft lines of credit. The Bureau found as a preliminary matter that notwithstanding the potential term, cost of credit, repayment structure, or security of these loans, they arise in distinct markets that may pose a somewhat different set of concerns for consumers. At the same time, the Bureau was concerned about the risk that these exclusions could create avenues for evasion of the proposed rule. In the Accompanying RFI, the Bureau also solicited information and additional evidence to support further assessment of whether other categories of loans may pose risks to consumers where lenders do not determine the consumer’s ability to repay. The Bureau also emphasized that it may determine in a particular supervisory or enforcement matter or in a later rulemaking, in light of evidence available at the time, that the failure to assess ability to repay when making a loan excluded from coverage here may nonetheless be an unfair or abusive act or practice. The Bureau did not receive any comments on the brief opening language in § 1041.3(e) of the proposed rule, and is finalizing the language which notes that the exclusions listed in § 1041.3(d) of the final rule apply to certain transactions, with slight modifications for clarity. The Bureau did, however, receive some general comments about the topic of exclusions 262 from the scope of coverage of the proposed rule. First, various consumer groups argued that there should be no exclusions or exemptions from coverage under the rule, which would weaken its effectiveness. A “fintech” company urged the Bureau to develop a “sandbox” type of model to allow innovation and to encourage the development of alternative loan models. Another such company offered a more complicated and prescriptive regulatory scheme establishing a safe harbor, lifting income verification requirements for loans with low loss rates and loans with amortizing payment plans, and full relief from cooling-off periods if borrowers repay their loans on time with their own money. One commenter during the SBREFA process argued for a broad exemption from the rule for payday lenders in States that permit such loans pursuant to existing regulatory frameworks governing payday lending. Another sought an exemption for Tribal lenders, asserting that the Bureau lacked statutory authority to treat them as covered by the rule. Many finance companies, and others commenting on their behalf, offered reasons why the Bureau should omit traditional installment loans from coverage under the rule; they also presented different formulations of how this result could be achieved. The Bureau does not agree that the exclusions listed in the proposal should be eliminated, for all the reasons set out in the discussion of those specific exclusions below (and notes that a further exclusion and two conditional exemptions have been added to or revised from the proposed rule). As for the notion of a “sandbox” approach to financial innovation, the Bureau has developed its own approach to these issues, having created and operated its Project Catalyst for several years now as a means of carrying out the Bureau’s statutory objective to ensure that “markets for consumer financial products and services operate transparently and efficiently to 263 facilitate access and innovation.”434 The suggestion that a distinct and highly prescriptive regulatory approach should be adopted in preference to the framework actually set out in the proposal is not supported by any data or analysis of this market. The arguments for an exemption of payday lender in those States where they are permitted to make such loans are directly contrary to all of the data and analysis contained in the extended discussions above in part II and below in Market Concerns—Underwriting. All of the risks and harms that the Bureau has identified from covered loans occur, by definition, in those States that authorize such lending, rather than in the 15 States and the District of Columbia that have effectively banned such lending under their State laws. The arguments raised on behalf of Tribal lenders have also been raised in Tribal consultations that the Bureau has held with federally recognized Indian tribes, as discussed in part III, and rest on what the Bureau believes is a misreading of the statutes and of governing Federal law and precedents governing the scope of Tribal immunity.435 As for the points raised by finance companies and others about traditional installment loans, they are largely being addressed by various modifications to the proposed rule, including by not imposing underwriting requirements for covered longer-term loans (other than covered longer-term balloon-payment loans), by adopting the exclusions and conditional exemptions, and, as some commenters suggested, by adopting the definition of cost of credit under TILA in place of the definition of total cost of credit in the proposed rule. 434 12 U.S.C. 5511(b)(5). More information about Project Catalyst is available on the Bureau’s website at https://www.consumerfinance.gov/about-us/project-catalyst/ (last visited Sept. 24, 2017). 435 See, e.g., CFPB v. Great Plains Lending, 846 F.3d 1049 (9th Cir. 2017), reh’g denied (Apr. 5, 2017) (court of appeals affirmed district court ruling that Tribal Lending Entities must comply with civil investigative demands issued by the CFPB); see also Otoe-Missouria Tribe of Indians v. New York State Dep’t of Fin. Servs., 769 F.3d 105, 107 (2d Cir. 2014); Donovan v. Coeur d’Alene Tribal Farms, 751 F.2d 1113, 1115 (9th Cir. 1985). 264 3(d)(1) Certain purchase money security interest loans Proposed § 1041.3(e)(1) would have excluded from coverage under proposed part 1041 loans extended for the sole and express purpose of financing a consumer’s initial purchase of a good when the good being purchased secures the loan. Accordingly, loans made solely to finance the purchase of, for example, motor vehicles, televisions, household appliances, or furniture would not be subject to the consumer protections imposed by proposed part 1041 to the extent the loans are secured by the good being purchased. Proposed comment 3(e)(1)-1 explained the test for determining whether a loan is made solely for the purpose of financing a consumer’s initial purchase of a good. If the item financed is not a good or if the amount financed is greater than the cost of acquiring the good, the loan is not solely for the purpose of financing the initial purchase of the good. Proposed comment 3(e)(1)-1 further explained that refinances of credit extended for the purchase of a good do not fall within this exclusion and may be subject to the requirements of proposed part 1041. Purchase money loans are typically treated differently than non-purchase money loans under the law. The FTC’s Credit Practices Rule generally prohibits consumer credit in which a lender takes a nonpossessory security interest in household goods but makes an exception for purchase money security interests.436 The Federal Bankruptcy Code, the UCC, and some other State laws also apply different standards to purchase money security interests. This differential treatment facilitates the financing of the initial purchase of relatively expensive goods, which many consumers would not be able to afford without a purchase money loan. In the proposal, the Bureau stated that it had not yet determined whether purchase money loans pose similar risks 436 16 CFR 444.2(a)(4). 265 to consumers as the loans covered by proposed part 1041. Accordingly, the Bureau proposed not to cover such loans at this time. A number of commenters expressed concern about the proposal’s use of a sole purpose test for determining when a loan made to finance the consumer’s initial purchase of a good gives rise to a purchase money security interest. Other alternatives were suggested, including a primary purpose test or perhaps the definition used in the UCC adopted in many States. Some commenters expressed concerns about motor vehicle purchases, in particular, noting that where the amount financed includes not simply the vehicle itself, but also the costs of ancillary products such as an extended service contract or a warranty, or other related costs such as taxes, tags, and title, it may be unclear whether the loan would lose its status as a purchase money security interest loan and become a covered loan instead. Others contended that covering the refinancing of credit that was extended for the purchase of a good could seem inconsistent with the terms of the exclusion itself, and could also bring back within the proposed rule’s scope of coverage many motor vehicle loans where the total cost of credit would exceed a rate of 36 percent per annum. These commenters again were particularly concerned about motor vehicle loans, which they noted often exceed a 100 percent lien-to-value ratio because additional products, such as add-on products like extended warranties, are often financed along with the price of the vehicle. In response to these comments, the Bureau streamlined and added language to proposed comment 3(e)(1)-1 to specify that a loan qualifies for this exclusion even if the amount financed under the loan includes Federal, State, or local taxes or amounts required to be paid under applicable State and Federal licensing and registration requirements. The Bureau recognized that these mandatory and largely unavoidable items should not cause a loan to lose its excluded 266 status. Yet the same considerations do not apply to ancillary products that are being sold along with a vehicle or other household good, but are not themselves the good in which the lender takes a security interest as a condition of the credit. As to the concern about refinances of credit extended for the purchase of a good, and especially the concern that this provision could bring back within the proposed rule’s scope of coverage many motor vehicle loans where the total cost of credit would exceed a rate of 36 percent per annum, the Bureau concluded that other changes made elsewhere in the final rule largely mitigate these concerns. In particular, the Bureau notes that the definition of total cost of credit in § 1041.2(a)(18) of the proposed rule has now been replaced with the definition of cost of credit in § 1041.2(a)(6) of the final rule, which aligns this term with Regulation Z. The Bureau also notes that these concerns about refinancing are most applicable to covered longer-term loans, which are no longer subject to underwriting criteria in the final rule (with the exception of covered longer-term balloon-payment loans). And though they are subject to the payment provisions, other changes in the coverage and the scope of the exceptions for certain payment transfers mitigate the effects for credit unions, in particular, that were the source of many of the comments on this issue. For these reasons, the Bureau is finalizing the regulation text as proposed, and the revised commentary as explained above as § 1041.3(d)(1) in the final rule. 3(d)(2) Real estate secured credit Proposed § 1041.3(e)(2) would have excluded from coverage under proposed part 1041 loans that are secured by real property, or by personal property used as a dwelling, and in which the lender records or perfects the security interest. The Bureau stated that even without this exclusion, very few real estate secured loans would meet the coverage criteria set forth in 267 proposed § 1041.3(b). Nonetheless, the Bureau preliminarily found that a categorical exclusion would be appropriate. For the most part, these loans are already subject to Federal consumer protection laws, including, for most closed-end loans, ability-to-repay requirements under Regulation Z § 1026.43. The proposed requirement that the security interest in the real estate be recorded or perfected also strongly discourages attempts to use this exclusion for sham or evasive purposes. Recording or perfecting a security interest in real estate is not a cursory exercise for a lender—recording fees are often charged and documentation is required. As proposed comment 3(e)(2)-1 explained, if the lender does not record or otherwise perfect the security interest in the property during the term of the loan, the loan does not fall under this exclusion and may be subject to the requirements of proposed part 1041. The Bureau did not receive any comments on this portion of the proposed rule, and is finalizing this exclusion and the commentary as proposed, with formatting changes only. 3(d)(3) Credit cards Proposed § 1041.3(e)(3) would have excluded from coverage under proposed part 1041 credit card accounts meeting the definition of credit card account under an open-end (not homesecured) consumer credit plan in Regulation Z § 1026.2(a)(15)(ii), rather than products meeting the more general definition of credit card accounts under Regulation Z § 1026.2(a)(15). By focusing on the narrower category, the exclusion would apply only to credit card accounts that are subject to the Credit CARD Act of 2009,437 which provides various heightened safeguards for consumers. These protections include a limitation that card issuers cannot open a credit card account or increase a credit line on a card account unless the card issuer first considers the 437 Public Law 111–24, 123 Stat. 1734 (2009). 268 consumer’s ability to repay the required payments under the terms of the account, as well as other protections such as limitations on fees during the first year after account opening, late fee restrictions, and a requirement that card issuers give consumers a reasonable amount of time to pay their bill.438 The Bureau preliminarily found that potential consumer harms related to credit card accounts are more appropriately addressed by the CARD Act, its implementing regulations, and other applicable law. At the same time, if the Bureau were to craft a broad exclusion for all credit cards as generally defined under Regulation Z, the Bureau would be concerned that a lender seeking to evade the requirements of the rule might seek to structure a product in a way that is designed to take advantage of this exclusion. The Bureau therefore proposed a narrower definition, focusing only on those credit card accounts that are subject to the full range of protections under the CARD Act and its implementing regulations. Among other requirements, the regulations imposing the CARD Act prescribe a different ability-to-repay standard that lenders must follow, and the Bureau found as a preliminary matter that the combined consumer protections governing credit card accounts subject to the CARD Act are sufficient for that type of credit. One commenter stated that all credit cards should be excluded from coverage under the rule, not just those subject to the CARD Act. Another industry commenter found it noteworthy that credit cards are not covered under the rule even though they can result in a cycle of debt. Consumer groups argued that this exclusion should be narrowed to lower-cost mainstream credit 438 15 U.S.C. 1665e; see also 12 CFR 1026.51(a); Supplement I to 12 CFR part 1026. 269 cards in harmony with the provisions of the Military Lending Act and implementing regulations. Other narrowing categories were also suggested in that comment. For all the reasons stated in the proposal, the Bureau does not find it sensible to expand coverage in this exclusion beyond those credit cards that are subject to the various heightened safeguards and protections for consumers in the CARD Act. At the same time, the reasons for drawing the boundaries of this exclusion around that particular universe of credit cards also militate against narrowing the scope of the exclusion further. Accordingly, the Bureau is finalizing this exclusion as proposed, with formatting changes only. The Bureau notes that “hybrid prepaid-credit card” products, which are treated as open-end (not home-secured) consumer credit plans under the final prepaid accounts rule, will be excluded from the scope of this final rule under § 1041.3(d)(3).439 3(d)(4) Student loans Proposed § 1041.3(e)(4) would have excluded from coverage under proposed part 1041 loans made, insured, or guaranteed pursuant to a Federal student loan program, and private education loans. The Bureau stated that even without this exclusion, very few student loans would meet the coverage criteria set forth in proposed § 1041.3(b). Nonetheless, the Bureau preliminarily determined that a categorical exclusion is appropriate. Federal student loans are provided to students or parents meeting eligibility criteria established by Federal law and regulations, such that the protections afforded by this proposed rule would be unnecessary. Private student loans are sometimes made to students based on their future potential ability to repay (as distinguished from their current ability), but they are typically co-signed by a party 439 81 FR 83934 (Nov. 22, 2016). 270 with financial capacity. These loans raise discrete issues that may warrant further attention in the future, but the Bureau found as a preliminary matter that they were not appropriately considered along with the types of loans at issue in this rulemaking. The Bureau stated in the proposal that it would continue to monitor the student loan servicing market for trends and developments; for unfair, deceptive, or abusive practices; and to evaluate possible policy responses, including potential rulemaking. Consumer groups contended that student loans should not be excluded from coverage under the rule. They noted that the effect of deleting this exclusion would likely be limited to private education loans, since the total cost of credit for Federal student loans in the proposed rule would likely not exceed a rate of 36 percent per annum. The Bureau continues to judge that student loans are specialized in nature, are subject to certain other regulatory constraints more specifically contoured to the loan product, and are generally not appropriately considered among the types of loans at issue here. The Bureau did not receive any other comments on this portion of the proposed rule, and is finalizing this exclusion as proposed, with formatting changes only. 3(d)(5) Non-recourse pawn loans Proposed § 1041.3(e)(5) generally would have excluded from coverage, under proposed part 1041, loans secured by pawned property in which the lender has sole physical possession and use of the pawned property for the entire term of loan, and for which the lender’s sole recourse if the consumer does not redeem the pawned property is the retention and disposal of the property. Proposed comment 3(e)(5)-1 explained that if any consumer, including a co-signor or guarantor, is personally liable for the difference between the outstanding loan balance and the 271 value of the pawned property, then the loan does not fall under this exclusion and may be subject to the requirements of proposed part 1041. The Bureau preliminarily found that bona fide, non-recourse pawn loans generally pose somewhat different risks to consumers than loans covered under proposed part 1041. As described in part II, non-recourse pawn loans involve the consumer physically relinquishing control of the item that secures the loan during the term of the loan. The Bureau stated that consumers may be more likely to understand and appreciate the risks associated with physically turning over an item to the lender when they are required to do so at consummation. Moreover, in most situations, the loss of a non-recourse pawned item over which the lender has sole physical possession during the term of the loan is less likely to affect the rest of the consumer’s finances than is either a leveraged payment mechanism or vehicle security. For instance, a pawned item of this nature may be valuable to the consumer, but the consumer most likely does not rely on the pawned item for transportation to work or to pay basic living expenses or major financial obligations. Otherwise, the consumer likely would not have pawned the item under those terms. Finally, because the loans are non-recourse, in the event that a consumer is unable to repay the loan, the lender must accept the pawned item as fully satisfying the debt, without further collection activity on any remaining debt obligations. In all of these ways, the Bureau stated in the proposal that pawn transactions appear to differ significantly from the secured loans that would be covered under proposed part 1041. One commenter claimed that the same reasons for excluding non-recourse pawn loans applies to vehicle title loans, and that vehicle title loans may even be preferred by consumers as the consumer retains the use of the vehicle and they can be less costly. Another similarly argued 272 that the Bureau ignored the principle of a level playing field among different financial products by excluding high-cost alternatives like pawn loans, which can be even more costly at times than payday loans. Consumer groups suggested that the exclusion should be narrowed only to pawn loans where the loan does not exceed the fair market value of the good. Another commenter representing pawnbrokers argued that the exclusion for pawn loans is justified because pawn transactions function as marketed, they are less likely than other loan products to affect the rest of the consumer’s finances, consumers do not experience very high default rates or aggressive collection efforts, certain other harms identified in the proposal do not occur in the pawn market, State and local government regulation is working well, consumers are given clear disclosures on their pawn ticket, and loan terms are longer than the typical 14-day payday loan. The Bureau does not find that these comments justify any modifications to this provision, and therefore finalizes the exclusion and the commentary as proposed, with formatting changes only. The first two comments do not provide any tangible support for eliminating the rationale for the exclusion of non-recourse pawn loans, and issues involving vehicle title loans are addressed elsewhere, as in Market Concerns—Underwriting, which describes the special risks and harms to consumers of repossession of their vehicle, which would potentially cause them to lose their basic transportation to work and to manage their everyday affairs. The suggestion that certain pawn loans should be covered loans depending on the relationship between the amount of the loan and the fair market value of the good would introduce needless complexity into the rule without discernible benefits. The Bureau notes that non-recourse pawn loans had previously been referenced in the definition of non-covered bridge loan in proposed § 1041.2(a)(13), which 273 has now been omitted from the final rule. To the extent that provision would have restricted the making of such loans in connection with the underwriting criteria for covered longer-term loans, those provisions are not being included in the final rule. To the extent that provision would have restricted the making of such loans in connection with the requirements in the rule for making covered short-term or longer-term balloon-payment loans, the Bureau concludes that various other changes made in §§ 1041.5 and 1041.6 address the subject of those restrictions in ways that obviate the need for defining the term non-covered bridge loan. However, note that any type of loan, including pawn loans, if used to bridge between multiple covered short-term loans or covered longer-term balloon-payment loans, are factors which could indicate that a lender’s ability-to-repay determinations are unreasonable. See comment 5(b)-2. 3(d)(6) Overdraft services and lines of credit Proposed § 1041.3(e)(6) would have excluded from coverage under proposed part 1041 overdraft services on deposit accounts as defined in 12 CFR 1005.17(a), as well as payments of overdrafts pursuant to a line of credit subject to Regulation Z, 12 CFR part 1026. Proposed comment 3(e)(6)-1 noted that institutions could rely on the commentary to 12 CFR 1005.17(a) in determining whether credit is an overdraft service or an overdraft line of credit that is excluded from the requirements of part 1041. Overdraft services generally operate on a consumer’s deposit account as a negative balance, where the consumer’s bank processes and pays certain payment transactions for which the consumer lacks sufficient funds in the account and imposes a fee for the service as an alternative to either refusing to authorize the payment (in the case of most debit and ATM transactions and ACH payments initiated from the consumer’s account) or rejecting the payment and charging a non-sufficient funds fee (in the case of other ACH 274 payments as well as paper checks). Overdraft services have been treated separately from the provisions of Regulation Z in certain circumstances, and are subject to specific rules under EFTA and the Truth in Savings Act (TISA) and their respective implementing regulations.440 In contrast, overdraft lines of credit are separate open-end lines of credit under Regulation Z that have been linked to a consumer’s deposit account to provide automatic credit draws to cover the processing of payments for which the funds in the deposit account are insufficient. As discussed above in part II, the Bureau is engaged in research and other activity in anticipation of a separate rulemaking on overdraft products and practices.441 Given that overdraft services and overdraft lines of credit involve complex overlays with rules about payment processing, deposit accounts, set-off rights, and other forms of depository account access, the Bureau preliminarily found that any discussion of whether additional regulatory protections are warranted for those two products should be reserved for that rulemaking. Accordingly, the Bureau proposed excluding both types of overdraft products from the scope of this rule, using definitional language from Regulation E to distinguish both overdraft services and overdraft lines of credit from other types of depository credit products. One industry commenter argued that the Bureau ignored the principle of a level playing field among different financial products by excluding high-cost alternatives like overdraft, which can be even more costly at times than payday loans. Consumer groups argued that the Bureau should eliminate this exclusion or limit it in various ways. The Bureau maintains the analysis presented in the proposed rule to conclude that overdraft services and lines of credit are unique 440 441 74 FR 59033 (Nov. 17, 2009) (EFTA); 70 FR 29582 (May 24, 2005) (TISA). CFPB Study of Overdraft Programs White Paper; Checking Account Overdraft. 275 products with a distinct regulatory history and treatment, which should be excluded from this rule and addressed on their own as a matter of supervision, enforcement, and regulation. The Bureau also did not find persuasive the suggestion that overdraft services and lines of credit should be covered in some partial manner, which would introduce needless complexity into the rule without discernible benefits. Having received no other comments on this portion of the proposed rule, the Bureau is finalizing this exclusion and the commentary as proposed, with formatting changes only. 3(d)(7) Wage advance programs Based on prior discussions with various stakeholders, the Bureau solicited and received comments in the proposal in connection with the definition of lender under proposed § 1041.2(a)(11) about some newly formed companies that are seeking to develop programs that provide innovative access to consumers’ wages in ways that do not seem to pose the kinds of risks and harms presented by covered loans. Certain of these companies, but by no means all of them, are part of the “fintech” wave. Some are developing new products as an outgrowth of businesses focusing mainly on payroll processing, for example, whereas others are not associated with consumers’ employers but rather are focused primarily on devising new means of advising consumers about how to improve their approach to cash management. The Bureau has consistently expressed interest in encouraging more experimentation in this space. In particular, a number of these innovative financial products are seeking to assist consumers in finding ways to draw on the accrued cash value of wages they have earned but not yet been paid. Some of these products are doing so without imposing any fees or finance charges, other than a charge for participating in the program that is designed to cover processing 276 costs. Others are developing different models that may involve fees or advances on wages not yet earned. The Bureau notes that some efforts to give consumers access to accrued wages may not be credit at all. For instance, when an employer allows an employee to draw accrued wages ahead of a scheduled payday and then later reduces the employee’s paycheck by the amount drawn, there is a quite plausible argument that the transaction does not involve “credit” because the employee may not be incurring a debt at all. This is especially likely where the employer does not reserve any recourse upon the payment made to the employee other than the corresponding reduction in the employee’s paycheck. Other initiatives are structured in more complicated ways that are more likely to constitute “credit” under the definition set forth in § 1041.2(a)(11) and Regulation Z. For example, if an employer cannot simply reduce the amount of an employee’s paycheck because payroll processing has already begun, there may be a need for a mechanism for the consumer to repay the funds after they are deposited in the consumer’s account. The Bureau has decided in new § 1041.3(d)(7) to exclude such wage advance programs—to the extent they constitute credit—from coverage under the rule if they meet certain additional conditions. The Bureau notes that the payment of accrued wages on a periodic basis, such as bi-weekly or monthly, appears to be largely driven by efficiency concerns with payroll processing and employers’ cash management. In addition, the Bureau believes that the kinds of risks and harms that the Bureau has identified with making covered loans, which are often unaffordable as a result of the identified unfair and abusive practice, may not be present where these types of innovative financial products are subject to appropriate safeguards. Accordingly, 277 where advances of wages constitute credit, the Bureau is adopting § 1041.3(d)(7) to exclude them from part 1041 if the advances are made by an employer, as defined in the Fair Labor Standards Act, 29 U.S.C. 203(d), or by the employer’s business partner, to the employer’s employees, provided that the following conditions apply:  The employee is not required to pay any charges or fees in connection with such an advance from the employer or the employer’s business partner, other than a charge for participating in the program; and  The entity advancing the funds warrants that it has no legal or contractual claim or remedy against the employee based on the employee’s failure to repay in the event the amount advanced is not repaid in full; will not engage in any debt collection activities if the advance is not deducted directly from wages or otherwise repaid on the scheduled date; will not place the amount advanced as a debt with or sell the debt to a third party; and will not report the debt to a consumer reporting agency concerning the amount advanced. The Bureau has considered the comments as well as its own analysis of this evolving marketplace and has concluded that new and innovative financial products that meet these conditions will tend not to produce the kinds of risks and harms that the Bureau’s final rule is seeking to address with respect to covered loans. At the same time, nothing prevents the Bureau from reconsidering these assumptions in a future rulemaking if there is evidence that such products are harming consumers. The Bureau has also adopted new commentary. Comment 3(d)(7)-1 notes that wage 278 advance programs must be offered by the employee’s employer or the employer’s business partner, and examples are provided of such business partners, which could include companies that are involved in providing payroll processing, accounting services, or benefits programs to the employer. Comment 3(d)(7)(i)-1 specifies that the advance must be made only against accrued wages and must not exceed the amount of the employee’s accrued wages, and provides further definition around the meaning of accrued wages. Comment 3(d)(7)(ii)(B)-1 clarifies that though the entity advancing the funds is required to warrant that it has no legal or contractual claim or remedy against the consumer based on the consumer’s failure to repay in the event the amount advanced is not repaid in full, this provision does not prevent the entity from obtaining a one-time authorization to seek repayment from the consumer’s transaction account. For these reasons, the Bureau is adopting the exclusion for wage advance programs as described in § 1041.3(d)(7) of the final rule and the related commentary. 3(d)(8) No-cost advances As discussed above in connection with § 1041.3(d)(7), the Bureau noted in the proposal, in connection with its discussion of the definition of lender in proposed § 1041.2(a)(11), that some newly formed companies are providing products or services that allow consumers to draw on wages they have earned but not yet been paid. Some of these companies are providing advances of funds and are doing so without charging any fees or finance charges, for instance by relying on voluntary tips. The proposal noted that others were seeking repayment and compensation through electronic transfers from the consumer’s account. The Bureau sought comment on whether to exclude such entities and similar products from coverage under the rule. The Bureau received limited comments on this issue, perhaps reflecting that it represents 279 a fairly new business model in the marketplace, with some championing the potential benefits for consumers and others maintaining that no exclusions—or at least no additional exclusions— should be created to the rule as it was proposed. Some comments described in more detail how the evolution of these products was unfolding, how they operate, and how they may affect the marketplace and consumers. The Bureau has also had discussions with stakeholders in connection with its other functions, such as market monitoring, supervision, and general outreach, that have informed its views and understanding of these new products and methods of providing access to funds for more consumers. As discussed above in connection with § 1041.3(d)(7), the Bureau is aware that some of these products provide access to the consumer’s own funds in the form of earned wages already accrued but not yet paid out because of administrative and payroll processes historically developed by employers, whereas other products rely on estimates of wages likely to be accrued, or accrued on average, and may make advances against expected wages that are not already earned and accrued. The Bureau has carefully considered the comments it has received on these issues, as well as other information about the market that it has gleaned from the course of its regular activities. The Bureau has addressed certain wage advance programs offered by employers or their business partners in § 1041.3(d)(7), as discussed above. In addition, after further weighing the potential benefits to consumers of this relatively new approach, the Bureau has decided to create a specific exclusion in § 1041.3(d)(8) of the final rule to apply to no-cost advances, regardless of whether they are offered by an employer or its business partner. The exclusion contains similar conditions to § 1041.3(d)(7), except that it applies to advances of funds where the consumer is not required to pay any charge or fee (even a fee for participating in the 280 program), and it is not limited to the accrued cash value of the employee’s wages. Like § 1041.3(d)(7), the exclusion is further limited to situations in which the entity advancing the funds warrants to the consumer as part of the contract between the parties (i) that it has no legal or contractual claim or remedy against the consumer based on the consumer’s failure to repay in the event the amount advanced is not repaid in full; and (ii) that with respect to the amount advanced to the consumer, the entity advancing the funds will not engage in any debt collection activities, place the debt with or sell the debt to a third party, or report the debt to a consumer reporting agency if the advance is not repaid on the scheduled date. The exclusion in § 1041.3(d)(8) is thus designed to apply to programs relying solely on a “tips” model or otherwise providing emergency assistance at no cost to consumers. The Bureau estimates, based on its experience with the marketplace for different types of small-dollar loans, that products meeting the conditions of § 1041.3(d)(8) are likely to benefit consumers and unlikely to lead to the risks and harms described below in Market Concerns—Underwriting. Unlike the proposal, the Bureau has decided not to confine such no-fee advances solely to the employer-employee context, as the very specific features of their product structure makes an exclusion from the rule for them likely to be beneficial for consumers across the spectrum. At the same time, nothing prevents the Bureau from reconsidering these assumptions in a future rulemaking if there is evidence that such products are harming consumers. New comment 3(d)(8)-1 further provides that though an entity advancing the funds is required to warrant that it has no legal or contractual claim or remedy against the consumer based on the consumer’s failure to repay in the event the amount advanced is not repaid in full, 281 this provision does not prevent the entity from obtaining a one-time authorization to seek repayment from the consumer’s transaction account. For these reasons, the Bureau is adopting the exclusion for no-cost advances as described in § 1041.3(d)(8) of the final rule and the related commentary. 3(e) Conditional exemption for alternative loans In § 1041.11 of the proposed rule, the Bureau set forth a conditional exemption for loans with a term of between 46 days and 180 days, if they satisfied a set of conditions that generally followed those established by the NCUA under the Payday Alternative Loan (PAL) Program as described above in part II. The proposal did not, however, contain a comparable exemption for PAL loans with durations between 30 and 45 days, with 30 days being the minimum duration permitted for a PAL loan. Loans that met the conditions of the proposed conditional exemption would have been exempted from the proposed underwriting criteria applicable to covered longerterm loans, but still would have been subject to the requirements on payment practices and the notice requirements. The Bureau received many general comments on the proposed exemption for PAL loans offered by credit unions and for comparable loan products if offered by other lenders. Some commenters argued that credit unions, as a class of entity, should be entirely exempted from all coverage under the rule. Others asked for more tailored exemptions for certain credit unions, such as for those with assets totaling less than $10 billion. Still others requested that credit unions be relieved of specific obligations under the rule, such as from compliance and record retention provisions (because their prudential regulators already address those matters); or from payment regulations for internal collections that do not incur fees; or from underwriting 282 requirements for Community Development Financial Institutions (CDFIs) that provide beneficial credit and financial services to underserved markets and populations. By contrast, other commenters did not think the Bureau could or should create any special provisions for credit unions in particular. But some consumer and legal aid groups were supportive of the PAL program, which they viewed as beneficial to consumers and not easily subject to manipulation. Some asserted that the PAL program was too constrained to support any broad provision of such loans, which were unlikely to yield a reasonable rate of return and thus not likely to generate a substantial volume of loans or to be sustainable for other lenders that are not depository institutions. Others argued that the proposed rule contained provisions that would go beyond the terms of the PAL program and increase complexity, and these additional provisions should be scaled back to mirror the PAL program more closely. Some commenters contended that the PAL program itself imposed a usury limit, which would be improper if adopted by the Bureau. As discussed earlier, the Bureau has decided not to finalize the specific underwriting criteria with respect to covered longer-term loans (other than covered longer-term balloonpayment loans) at this time. However, the Bureau has decided, for the reasons explained below, to create a conditional exemption to the rule that applies to any alternative loan, which is a term that is defined more specifically below. In brief, an alternative loan is a covered loan that meets certain conditions and requirements that are generally consistent with the provisions of the PAL program as authorized and administered by the NCUA, including any such loan made by a Federal credit union that is in compliance with that program. The conditions and requirements of the exemption are modified in certain respects relative to the proposal to reflect that the 283 conditional exemption now also encompasses loans of less than 45 days in duration to create a more comprehensive lending framework, unlike the coverage initially described in the proposed rule. In creating this exception, the Bureau agrees with the commenters that concluded, after observing the PAL program over time, that program is generally beneficial to consumers and not easily subject to manipulation in ways that would create risks and harms to consumers. At the same time, the Bureau recognizes that one of the objectives set forth in the DoddFrank Act is for Federal consumer financial law to be enforced consistently without regard to the status of a person as a depository institution.442 Consistent with that objective, the Bureau has set forth the elements of alternative loans in general form, so that lenders other than Federal credit unions—including both banks and other types of financial institutions—can offer comparable loans in accordance with essentially the same conditions and requirements. By doing so, the Bureau is making it possible for more lenders to offer this product, which will offer the opportunity to test the prediction made by some commenters that these loans would not scale if offered by lenders that are not depository institutions—a point on which the Bureau is not yet convinced either way. The conditional exemption for alternative loans contained in § 1041.3(e) of the final rule is adopted pursuant to the Bureau’s exemption authority in section 1022(b)(3) of the Dodd-Frank Act to “conditionally or unconditionally exempt any class of covered persons, service providers, or consumer financial products or services, from any . . . rule issued under this title.”443 In this respect, Congress gave the Bureau broad latitude, simply stating that it should do so “as [it] 442 See 12 U.S.C. 5511(b)(4) (“Federal consumer financial law is enforced consistently, without regard to the status of a person as a depository institution, in order to promote fair competition.”). 443 12 U.S.C. 5512 (b)(3)(A). 284 deems necessary or appropriate to carry out the purposes and objectives of this title.”444 The statutory language thus indicates that the Bureau should evaluate the case for creating such an exemption in light of its general purposes and objectives as Congress articulated them in section 1021 of the Dodd-Frank Act. In addition, when the Bureau exercises its exemption authority under section 1022(b)(3) of the Dodd-Frank Act, it is further required to take into consideration, as appropriate, three additional statutory factors: (i) the total assets of the class of covered persons; (ii) the volume of transactions involving consumer financial products or services in which the class of covered persons engages; and (iii) existing provisions of law which are applicable to the consumer financial product or service and the extent to which such provisions provide consumers with adequate protections.445 Here, the Bureau perceives tangible benefit for consumers and for lenders by preserving the framework of the PAL program, which as discussed in part II has had some success in generating approximately $134.7 million in originations in 2016—up 9.7 percent from the 2015 levels—with relatively low costs of credit and relatively low levels of charge-offs for this particular market. In particular, the Bureau agrees with those commenters that noted the distinct elements of the PAL program, including the specified product features, are not configured to give rise to the kinds of risks and harms that are more evident with covered short-term loans or covered longer-term balloon-payment loans. In short, the PAL product thus far seems to be beneficial for consumers, and a conditional exemption to make such loans more broadly available to the public appears consistent with the Bureau’s purpose “of ensuring that all 444 445 12 U.S.C. 5512(b)(3)(A). 12 U.S.C. 5512(b)(3)(B)(i) – (iii). 285 consumers have access to markets for consumer financial products and services.”446 Likewise, it seems consistent also with the Bureau’s objective of ensuring that “markets for consumer financial products and services operate transparently and efficiently to facilitate access and innovation,” and the competition that alternative loans could provide to other types of covered loans may be helpful in protecting consumers “from unfair . . . or abusive acts and practices.”447 Turning to the statutory factors set out in section 1022(b)(3), the assets of the expected class of lenders is likely to remain relatively small in light of the thousands of smaller credit unions, as also is the volume of transactions, which many commenters did not seem to expect would scale into much larger loan programs, though the Bureau is not yet convinced on this point either way. In addition, the PAL program itself is regulated and overseen by NCUA with respect to the credit unions who offer it, which means that “existing provisions of law . . . are applicable to [it]” and it is reasonable at this time to judge that “such provisions provide consumers with adequate protection” in using this loan product, as Congress indicated was germane to determining the justifications for an exemption.448 Moreover, under the general terms of § 1041.3(e), which allows all lenders to make alternative loans regardless of whether they are credit unions, the Bureau and other regulators, including State regulators, stand wellpositioned to monitor the development of this loan product over time, and to make adjustments if the current experience of these loans as generally beneficial for consumers were perceived to be changing in ways that created greater consumer risks and harms. 446 12 U.S.C. 5511(a). 12 U.S.C. 5511(b)(5) and (b)(2). 448 12 U.S.C. 5512(b)(3)(B)(iii). 447 286 The Bureau decided to create this conditional exemption in order to recognize that the NCUA is currently operating and supervising this established loan program for credit unions and to avoid duplicative overlap of requirements that could foster confusion and create undue burdens for certain lenders, in light of the Bureau’s conclusion that loans made on terms that are generally consistent with the PAL program do not pose the same kinds of risks and harms for consumers as the types of covered loans addressed by this rule.449 It also judges this approach to be superior to the broader scope of exemptions urged by various commenters, such as a complete exemption from the rule for all loans of all types made by credit unions (rather than just PAL loans), or even a conditional exemption from certain portions of the rule for all loans of all types made by credit unions. As for the comment that these loans impose a usury cap, the Bureau has explained elsewhere that an actual usury cap would flatly prohibit certain loans from being made based directly on the interest rate being charged, whereas the exemption provided here would merely allow such loans to avoid triggering certain conditions of making such loans—most notably, the requirement that the lender reasonably assess the borrower’s ability to repay the loan according to its terms but also the provisions concerning payment practices. For all of these reasons, the Bureau is finalizing this provision and the related commentary with several modifications. First, in response to comments suggesting that various conditions for alternative loans as stated in the proposed rule would render this loan product too burdensome and complex, the Bureau has eliminated certain conditions for such loans in the final rule. In particular, among the conditions added in the proposal that now are dropped are: 449 See 12 U.S.C. 5512(b)(3)(B) (in deciding whether to issue an exemption, “the Bureau shall, as appropriate, take into consideration . . . existing provisions of law which are applicable to the consumer financial product or service and the extent to which such provisions provide consumers with adequate protection”). 287 required monthly payments; rules on charging fees; required checking of affiliate records; certain additional requirements, such as prohibitions on prepayment penalties and sweeping of accounts in certain circumstances, as well as required information furnishing. Second, certain changes have been made to take account of the fact that proposed § 1041.11 had applied only to covered longer-term loans, whereas § 1041.3(e) of the final rule applies to covered loans more generally. The language of each prong of § 1041.3(e)(1)-(4) of the final rule is set out below, and immediately thereafter any changes made from the proposed language to the text of the final rule are specified and explained. Again, as a prefatory matter, an alternative loan is a covered loan that meets all four of these sets of conditions and requirements. 3(e)(1) Loan term conditions (1) Loan term conditions. An alternative loan must satisfy the following conditions: (i) The loan is not structured as open-end credit, as defined in § 1041.2(a)(16); (ii) The loan has a term of not less than one month and not more than six months; (iii) The principal of the loan is not less than $200 and not more than $1,000; (iv) The loan is repayable in two or more payments, all of which payments are substantially equal in amount and fall due in substantially equal intervals, and the loan amortizes completely during the term of the loan; and (v) The loan carries a cost of credit (excluding any application fees) of not more than the interest rate permissible for Federal credit unions to charge under regulations issued by the National Credit Union Administration at 12 CFR 701.21(c)(7)(iii), and any application fees charged to the consumer reflect the actual costs associated with processing the application and do 288 not exceed the application fees permissible for Federal credit unions to charge under regulations issued by the National Credit Union Administration at 12 CFR 701.21(c)(7)(iii). This language of the final rule had originated in § 1041.11(a) of the proposed rule. The name of the exemption has been revised from a conditional exemption for certain covered longer-term loans up to six months in duration to a conditional exemption for alternative loans. The term of the loan is modified from “not more than six months” to “not less than one month and no more than six months,” again to reflect the change made in this exemption to encompass the broader set of all covered loans, rather than just covered longer-term loans. The other conditions, including the $200 floor and the $1,000 cap, are maintained because they are consistent with the requirements of the PAL program. The prior condition that the loan be repayable in two or more payments “due no less frequently than monthly” is now changed to omit the quoted language because the term of these loans may now be shorter than was the case in the proposal. The amortization provision is broken out and simplified to provide more flexibility around the payment schedule and allocation, which again reflects the fact that many of these loans may now be covered short-term loans. Finally, the prior language around total cost of credit is now replaced with cost of credit, which is consistent with TILA and Regulation Z and is responsive to suggestions made by several commenters; the permissible interest rate on such products is that set by the NCUA for the PAL program; any application fees charged to the consumer must reflect the actual associated costs and comply with the provisions of any NCUA regulations; and the lender does not impose any charges other than the rate and application fees permitted by the NCUA for the PAL program. 3(e)(2) Borrowing history condition 289 “Prior to making an alternative loan under § 1041.3(e), the lender must determine from its records that the loan would not result in the consumer being indebted on more than three outstanding loans made under this section from the lender within a period of 180 days. The lender must also make no more than one alternative loan under § 1041.3(e) at a time to a consumer.” Aside from conforming language changes, the only substantive revision here is to excise references to affiliates of the lenders, consistent with the NCUA’s practice in administering the PAL program. 3(e)(3) Income documentation condition “In making an alternative loan under § 1041.3(e), the lender must maintain and comply with policies and procedures for documenting proof of recurring income.” This prong contains minor conforming language changes only. 3(e)(4) Safe harbor “Loans made by Federal credit unions in compliance with the conditions set forth by the National Credit Union Administration at 12 CFR 701.21(c)(7)(iii) for a Payday Alternative Loan are deemed to be in compliance with the requirements and conditions of paragraphs (e)(1), (e)(2), and (e)(3) of this section.” This prong contains entirely new language, replacing what had been “additional requirements” in § 1041.11(e) of the proposed rule. Those additional requirements tailored by the NCUA for credit unions and included in the original proposal would be cumbersome in various respects for all lenders to adopt, including provisions on additional information furnishing, restrictions on sweeps and set-offs as means of a depository institution collecting on 290 the loan, and prepayment penalties. The safe harbor provided for Federal credit unions in compliance with NCUA’s requirements for the PAL program, however, reflects the fact that to qualify for the safe harbor, a credit union would be obligated to comply with all of the additional requirements of the PAL program. Having considered the comments received, the Bureau concludes that it is appropriate to finalize § 1041.3(e) for all the reasons discussed above. The Bureau also is finalizing proposed comment 3(d)(8)-1 as comment 3(e)-1 of the final rule, which notes that this provision does not confer on the lenders of such loans any exemption from the requirements of other applicable laws, including State laws. This comment also clarifies that all lenders, including Federal credit unions and persons that are not Federal credit unions, are permitted to make loans under the specific terms in § 1041.3(e), provided that such loans are permissible under other applicable laws, including State laws. The remainder of the commentary is being carried forward from the proposed rule with revisions, all made to align them with the modified language in § 1041.3(e) of the final rule. The proposed comments previously designated as 11(a)-1 to (11)(e)(1)(ii)-2 are now renumbered as comments 3(e)(1)-1 to 3(e)(3)-1 in the final rule. 3(f) Conditional exemption for accommodation loans In the proposal, in connection with the discussion of the proposed definition of lender in § 1041.2(a)(11), the Bureau noted that some stakeholders had suggested narrowing the definition of lender to avoid covering lenders that are primarily focused on other types of lending or other types of financial services, but on occasion make covered loans as a means of accommodating their existing customers. The stakeholders posited that such loans would be likely to operate differently from loans made as a primary line of business, for instance because the lenders who 291 make them have information about consumers’ financial situations from their primary lines of business and because their incentives in making the loans is to preserve their customer relationships, and thus may not pose the same risks and harms as other types of covered loans. The Bureau solicited comments on this suggestion. The Bureau had also proposed a more detailed provision, in proposed § 1041.12, in order to provide a conditional exemption for certain covered longer-term loans that would be made through accommodation lending programs and would be underwritten to achieve an annual portfolio default rate of not more than five percent. The proposal would have allowed a lender to make such loans without meeting the specific underwriting criteria contained in the proposed rule, though proposed § 1041.12 laid out its own detailed provisions applicable to the making of such loans. Notably, the Bureau found that the feedback it received on this provision overlapped considerably with the comments submitted in response to the question the Bureau had asked with respect to the definition of lender about providing an exception based on de minimis lending. Many commenters expressed their views favoring a de minimis exemption. Several of them urged that the Bureau should set parameters for the exemption based both on loan volume and the percentage of revenue derived from such loans. More specific suggestions ranged from caps of 100 to several thousand loans per year; one commenter suggested 2,000 loans per year that yield no more than five percent of revenue; others urged a cap of 2,500 loans per year that yield no more than 10 percent of revenue. The Bureau also received a number of comments on proposed § 1041.12 and proposed comments 12(a)-1 to (12)(f)(1)(ii)-2. Banking organizations argued that the Bureau should 292 exempt types of institutions rather than types of loans, and that because community banks are responsible providers of small loans, they should be conditionally exempted from coverage. Many commenters were also critical of the provisions of proposed § 1041.12, which they viewed as so cumbersome as to discourage many institutions from engaging in this type of lending. These comments focused particularly on the back-end requirements and calculations included in the proposal. Some commenters noted the guidance already in place from other banking regulators that had suppressed such lending at the banks, and predicted that the proposal would exacerbate those difficulties. State bank regulators, in particular, advocated in favor of a de minimis threshold to preserve such lending by smaller community banks as beneficial to consumers, especially in rural areas and as a way to provide alternatives if the effect of the rule would be to cause consolidation in the small-dollar lending market. Consumer groups generally opposed exemptions to the rule but acknowledged that a properly structured de minimis provision would be unlikely to create much if any harm to consumers. As stated earlier, the Bureau has decided not to finalize the ability-to-repay requirements with respect to covered longer-term loans (other than covered longer-term balloon-payment loans) at this time. However, as a result of reviewing and analyzing the public input on the issue of accommodation lending more generally, the Bureau has determined to create a conditional exemption that is applicable to accommodation loans that have been traditionally made primarily by community banks and credit unions. At the same time, in line with the Dodd-Frank Act’s goal of enforcing Federal consumer financial law without regard to a financial company’s status 293 as a depository institution, 450 the Bureau has set forth the elements of accommodation loans in general form such that any lender whose covered loan originations fall below the thresholds set in final § 1041.3(f) can qualify for the conditional exemption. In part, the Bureau is reaching this conclusion based on its review of the comments received, which indicated that lenders would find the approach taken in proposed § 1041.12 to be cumbersome or even unworkable for lenders. Whether or not this was objectively demonstrable for most lenders, it was clear that the proposed approach would have been taken as a discouraging factor for those deciding whether or not to make such loans. Moreover, the Bureau concluded that loans made as an occasional accommodation to existing customers were not likely to pose the same risks and harms as other types of covered loans, because such loans would be likely to operate differently and carry different incentives for the lender as compared to loans made as a primary line of business. As discussed in the preceding section on alternative loans, when the Bureau exercises its exemption authority under section 1022(b)(3) of the Dodd-Frank Act to create an exemption for “any class of covered persons, service providers, or consumer financial products or services, from any . . . rule issued under this title,” it has broad latitude that Congress conferred upon it to do so.451 Again, Congress simply said that the Bureau should exercise this authority “as [it] deems necessary or appropriate to carry out the purposes and objectives of this title,”452 and the Bureau’s general purposes and objectives are stated in section 1021 of the Dodd-Frank Act. In addition, when the Bureau exercises its exemption authority under section 1022(b)(3) of the 450 See 12 U.S.C. 5511(b)(4) (“Federal consumer financial law is enforced consistently, without regard to the status of a person as a depository institution, in order to promote fair competition.”). 451 12 U.S.C. 5512 (b)(3)(A). 452 12 U.S.C. 5512(b)(3)(A). 294 Dodd-Frank Act, it is further required, as appropriate, to take into consideration three statutory factors: “(i) the total assets of the class of covered persons; (ii) the volume of transactions involving consumer financial products or services in which the class of covered persons engages; and (iii) existing provisions of law which are applicable to the consumer financial product or service and the extent to which such provisions provide consumers with adequate protections.”453 Here, too, it appears that Congress intended the Bureau to do so in view of its purposes and objectives as set forth in the Dodd-Frank Act. Here, the Bureau perceives tangible benefit for consumers and for lenders to be able to maintain access to individualized loans of the kind permitted by this provision and in line with the traditions and experience of community banks over many years, which have generally underwritten these loans as an accommodation on an individualized basis in light of their existing customer relationships. In this manner, the conditional exemption would help ensure “that all consumers have access to markets for consumer financial products and services,”454 which is a principal purpose of the Dodd-Frank Act, and would not be restricted in their existing access to such traditional loan products. At the same time, this conditional exemption would enable the Bureau “to reduce unwarranted regulatory burdens”455 on these longstanding loan products made to existing bank customers on an individualized basis in light of their existing customer relationships, without posing any of the kinds of risks and harms to consumers that exist with the types of covered loans addressed by this rule. 453 12 U.S.C. 5512(b)(3)(B)(i) – (iii). 12 U.S.C. 5511(a). 455 12 U.S.C. 5511(b)(3). 454 295 And though the provisions of § 1041.3(f) are written in general terms to be applicable to lenders that are not themselves depository institutions, it does not appear likely that these provisions would be open to wide-scale abuse, precisely because the loan and revenue restrictions are set at a de minimis level that would tend to limit the scope of any predatory behavior. Assessing the matter against the three additional statutory factors as well, then, the assets of these lenders availing themselves of this provision would likely be limited; the volume of transactions would be small, by definition and design; and Federal consumer financial law, as implemented through the Bureau’s continuing supervisory and enforcement authorities and by other means as provided in the statute, would maintain consumer protections in the broader market despite this slight restriction on coverage under the rule. Therefore, as stated in § 1041.3(f), this provision will conditionally exempt any accommodation loan from coverage under the final rule. That category is defined to apply to a covered loan made by any lender where the lender and its affiliates collectively have made 2,500 or fewer covered loans in the current calendar year and also made 2,500 or fewer covered loans in the preceding calendar year; and during the most recent completed tax year in which the lender was in operation, if applicable, the lender and any affiliates that were in operation and used the same tax year derived no more than 10 percent of their receipts from covered short-term and longer-term balloon-payment loans, or if the lender was not in operation in a prior tax year, the lender reasonably anticipates that the lender and any of its affiliates that use the same tax year will, during the current tax year, derive no more than 10 percent of their receipts from covered short-term loans and covered longer-term balloon-payment loans. Comment 3(f)-1 of the final rule provides an example of the application of this provision to a sample lender. 296 Although, in general, all covered loans and the receipts from those loans would count toward the thresholds in § 1041.3(f) for the number of loans per year and for receipts, § 1041.3(f) allows lenders not to count toward either threshold covered longer-term loans for which the conditional exclusion for transfers in § 1041.8(a)(1)(ii) applies to all transfers for payments made under the loan. As explained in the section-by-section discussion of § 1041.8(a)(1)(ii), when the lender is the account-holder, that provision excludes certain transfers from the definition of payment transfer if, pursuant to the terms of the loan agreement or account agreement, the lender (1) does not charge the consumer any fee, other than a late fee under the loan agreement, in the event that the lender initiates a transfer of funds from the consumer’s account in connection with the covered loan for an amount that the account lacks sufficient funds to cover; and (2) does not close the consumer’s account in response to a negative balance that results from a transfer of funds initiated in connection with the covered loan. These conditions provide substantial protection against the harms targeted by the provisions in §§ 1041.8 and 1041.9. As a result, loans for which all payment transfers are excluded under § 1041.8(a)(1)(ii) from the definition of payment transfer are not subject to either the prohibition in § 1041.8(b) on initiating more than two consecutive failed payment transfers or the requirement in § 1041.9(b) to provide payment notices prior to initiating certain payment withdrawals. Since those loans carry with them substantial protection against the harms targeted in subpart C and would not be subject to those provisions, the Bureau believes that it is simpler not to count them for purposes of § 1041.3(f) either. The Bureau had sought comment about the appropriate parameters of this conditional exemption, which is designed to be a de minimis provision to allow only a certain amount of 297 lending of this kind to accommodate customers as a distinct sidelight to the institution’s main lines of business. Once again, the purpose of this provision is to accommodate existing customers through what traditionally have been loans that were underwritten on an individualized basis for existing customers. It was not proposed, and is not being adopted, to stimulate the development of a model for loans that are offered in high volumes. As for the parameters that the Bureau decided on, they closely reflect the submissions received in the comment process, with both the overall loan limit (2,500 per year) and the revenue limit (no more than 10 percent of receipts) intended to keep loans made pursuant to this exemption to a very limited part of the lender’s overall business. Each of the two provisions operates together to achieve that joint objective, which would not necessarily be achieved by either component operating in isolation. The Bureau decided to create this conditional exemption in order to respond to the persuasive points made by the commenters about the benefits that would flow from preserving this modest amount of latitude to be able to contour specialized loans as an accommodation to individual customers. That is especially so in view of the unlikelihood that this practice would pose the same kinds of risks and harms that the Bureau recognized with covered short-term loans and covered longer-term balloon-payment loans as described below in Market Concerns— Underwriting. The adoption of this conditional exemption also evinces the Bureau’s recognition of the input it has heard from many stakeholders over the years, particularly from depository institutions, who have regularly supplied the Bureau with details about their perspective that smaller depository lenders such as community banks and credit unions have a long history and tradition of making loans to accommodate their existing customers for various personal reasons, 298 such as minor expenses related to some type of family event. These loans are typically underwritten, customized, made for small amounts and at reasonable cost, and generate low levels of defaults. Although this type of accommodation lending is often quite specialized and individualized, it could be construed to overlap in certain ways with the covered loans encompassed by the rule. The conditional exemption that is now finalized in § 1041.3(f) provides an effective method of addressing legitimate concerns about the potentially detrimental consequences of that overlap for consumers. 3(g) Receipts The Bureau has added a new definition of the term receipts, which § 1041.3(g) of the final rule defines to mean total income (or, in the case of a sole proprietorship, gross income) plus cost of goods sold as these terms are defined and reported on Internal Revenue Service (IRS) tax return forms (such as Form 1120 for corporations; Form 1120S and Schedule K for S corporations; Form 1120, Form 1065, or Form 1040 for LLCs; Form 1065 and Schedule K for partnerships; and Form 1040, Schedule C for sole proprietorships). Receipts do not include net capital gains or losses; taxes collected for and remitted to a taxing authority if included in gross or total income, such as sales or other taxes collected from customers but excluding taxes levied on the entity or its employees; or amounts collected for another (but fees earned in connection with such collections are receipts). Items such as subcontractor costs, reimbursements for purchases a contractor makes at a customer’s request, and employee-based costs such as payroll taxes are included in receipts. This definition of receipts is modeled on the definitions of the 299 same term in the Bureau’s larger participant rulemakings for the consumer reporting456 and debt collection markets,457 which in turn were based in part on the Small Business Administration’s definition of receipts at 13 CFR 121.104. The Bureau is adding this definition to clarify how the term is used in § 1041.3(f) in the course of describing accommodation loans, and to reduce the risk of confusion among consumers, industry, and regulators. 3(h) Tax year The Bureau has added a new definition of the term tax year, which § 1041.3(h) of the final rule defines to have the same meaning attributed to this term by the IRS as set forth in IRS Publication 538, which provides that a tax year is an annual accounting period for keeping records and reporting income and expenses. The Bureau is adding this definition to clarify how the term is used in § 1041.3(f) in the course of describing accommodation loans, and to reduce the risk of confusion among consumers, industry, and regulators. Subpart B—Underwriting Overview of the Bureau’s Approach in the Proposal and in the Final Rule The Bureau proposed to identify an unfair and abusive practice with respect to the making of covered short-term loans pursuant to its authority to “prescribe rules . . . identifying as unlawful unfair, deceptive, or abusive acts or practices.”458 The proposal explained the Bureau’s preliminary view that it is both an unfair and abusive practice for a lender to make such a loan without reasonably determining that the consumer will have the ability to repay the loan. To 456 77 FR 42874 (July 20, 2012). 77 FR 65775 (Oct. 31. 2012). 458 Public Law 111–203, section 1031(b), 124 Stat. 1376 (2010). 457 300 avoid committing this unfair and abusive practice, the Bureau stated that a lender would have to make a reasonable assessment that the consumer has the ability to repay the loan. The proposal would have established a set of requirements to prevent the unlawful practice by requiring lenders to follow certain specified underwriting practices in assessing whether the consumer has the ability to repay the loan, as well as imposing certain limitations on rapid re-borrowing. The Bureau proposed the ability-to-repay requirements under its authority to prescribe rules for “the purpose of preventing unfair and abusive acts or practices.”459 The proposal would have further relied on section 1022(b)(3) of the Dodd-Frank Act460 to exempt certain covered short-term loans from the ability-to-repay requirements if the loans satisfied a set of conditions designed to avoid the harms that can result from unaffordable loans, including the harms that can flow from extended sequences of multiple loans in rapid succession. Accordingly, lenders seeking to make covered short-term loans would have the choice, on a case-by-case basis, either to comply with the ability-to-repay requirements according to the specified underwriting criteria or to make loans that meet the conditions set forth in the proposed exemption—conditions that are specifically designed as an alternative means to protect consumers against the harms that can result from unaffordable loans. As detailed further below, the Bureau has carefully considered its own research, analysis performed by others, and the public comments received with respect to this rulemaking and is now finalizing its finding that failing to reasonably determine whether consumers have the ability to repay covered short-term loans according to their terms is an unfair and abusive 459 460 12 U.S.C. 5531(b). 12 U.S.C. 5512(b)(3). 301 practice. These sources establish that unaffordable covered short-term loans generate severe harms for a substantial population of consumers. The Bureau has made the judgment that the harms and risks of such loans can be addressed most effectively by requiring lenders to underwrite such loans in accordance with specific criteria and thus not to make such a loan without reasonably determining that the consumer has the ability to repay the loan according to its terms. The Bureau has also retained the conditional exemption, while noting that the conditions on such loans, which are specifically designed as an alternative means to protect consumers against the harms that can result from unaffordable loans, will likely prompt lenders to consider more carefully their criteria for making such loans as well, given that defaults and delinquencies can no longer be offset by the revenues from repeated re-borrowing. The Bureau has modified various details of the proposed rule with respect to the underwriting criteria for the ability-to-repay requirement and the conditional exemption to strike a better balance among compliance burdens and other concerns, but has maintained the basic framework that was initially set forth in the proposed rule. The Bureau also proposed to identify the same unfair and abusive practice with respect to the failure to assess consumers’ ability to repay certain longer-term loans, including both installment and balloon-payment structures, as long as the loans exceeded certain price thresholds and involved the taking of either a leveraged payment mechanism or vehicle security. The Bureau proposed to subject these covered longer-term loans to underwriting requirements similar to those for covered short-term loans, as well as proposing two exemptions for loans that satisfied different sets of conditions designed to avoid the risks and harms that can result from unaffordable loans. 302 As detailed further below, the Bureau has carefully considered its own research, analysis performed by others, and the public comments received with respect to the proposed treatment of covered longer-term loans, and has decided to take a bifurcated approach at this time to concerns about unfair or abusive underwriting of longer-term loans. With regard to balloon payment structures, the Bureau finds that failing to reasonably assess whether consumers have the ability to repay covered longer-term balloon-payment loans according to specific underwriting criteria is an unfair and abusive practice. Because they require large lump-sum or irregular payments, these loans impose financial hardships and payment shocks on consumers that are similar to those posed by short-term loans over just one or two income cycles. Indeed, the Bureau’s analysis of longer-term balloon-payment loans in the market for vehicle title loans found that borrowers experienced high default rates—notably higher than for similar loans with amortizing installment payments. The Bureau also has concluded that the outcomes between a singlepayment loan with a term of 46 or more days is unlikely to be much different for consumers than an identical loan with a term of 45 days, and is concerned that failing to cover longer-term balloon-payment loans would induce lenders to slightly extend the terms of their existing shortterm lump-sum loans in an effort to evade coverage under the final rule, as occurred in this market in response to regulations adopted under the Military Lending Act. For these reasons, the Bureau is finalizing its finding that failing to reasonably assess whether consumers have the ability to repay covered longer-term balloon-payment loans is an unfair and abusive practice. The Bureau has made the judgment that these risks and harms can be addressed most effectively—as with covered short-term loans—by requiring lenders to underwrite such loans in accordance with specified criteria and thus not to make such a loan 303 without reasonably determining that the consumer has the ability to repay the loan according to its terms. After having sought comment on the issue of whether longer-term balloon-payment loans should be covered regardless of price or the taking of a leveraged payment mechanism or vehicle security, the Bureau has decided, in light of the risks to consumers, to apply the rule to all such loans, aside from certain exclusions and exemptions described above in § 1041.3 of the final rule. The Bureau has decided, however, not to move forward with its primary finding that it is an unfair and abusive practice to make certain higher-cost longer-term installment loans without making a reasonable determination that the consumer will have the ability to repay the loan, and, accordingly, its prescription of underwriting requirements designed to prevent that practice. The Bureau has decided to defer this aspect of the proposal for further consideration in a later rulemaking. After consideration of the research and the public comments, the Bureau has concluded that further analysis and outreach are warranted with respect to such loans, as well as other types of credit products on which the Bureau sought comment as part of the Request for Information. While such loans differ in certain ways from the loans covered in this final rule, the Bureau remains concerned that failing to underwrite such products may nonetheless pose substantial risk for consumers. The Bureau will continue to gather evidence about the risks and harms of such products for consideration as a general matter in a later rulemaking, and will continue in the meantime to scrutinize such lending for potential unfair, deceptive, or abusive acts or practices pursuant to its supervisory and enforcement authority. And, as detailed in subpart C below, the Bureau has concluded that it is appropriate to apply certain limitations and disclosure requirements concerning payment practices (and related 304 recordkeeping requirements) to longer-term installment loans with a cost of credit above 36 percent that involve the taking of a leveraged payment mechanism. The predicate for the identification of an unfair and abusive practice in the Bureau’s proposal—and thus for the preventive ability-to-repay requirements—was a set of preliminary findings about the consumers who use storefront and online payday loans, single-payment vehicle title loans, and other covered short-term loans, and the impact on those consumers of the practice of making such loans without assessing the consumers’ ability to repay. The preliminary findings as set forth in the proposal, the comments that the Bureau received on them, and the Bureau’s responses to those comments as the foundation of its final rule are all discussed below in the following section referred to as Market Concerns—Underwriting. Further in the discussion below, the Bureau also addresses the same issues with respect to covered longer-term balloon-payment loans. Market Concerns—Underwriting Short-Term Loans In the proposal, the Bureau stated its concern that lending practices in the markets for storefront and online payday lending, single-payment vehicle title loans, and other covered shortterm loans are causing harm to many consumers who use these products. Those harms include default, delinquency, and re-borrowing, as well as various collateral harms from making unaffordable payments. This section reviews the available evidence with respect to the consumers who use covered short-term loans, their reasons for doing so, and the outcomes they experience. It also reviews the lender practices that contribute to these outcomes. The discussion begins with the main points presented in this section of the proposal, stated in 305 summary form, and provides a high-level overview of the general responses offered by the commenters. More specific issues and comments are then treated in more detail in the succeeding subsections. In the proposal, the Bureau’s preliminary views were stated in summary form as follows: • Lower-income, lower-savings consumers. Consumers who use these products tend to come from lower- or moderate-income households. They generally do not have any savings to fall back on, and they have very limited access to other sources of credit; indeed, typically they have sought unsuccessfully to obtain other, lower cost, credit before turning to a short-term loan. The commenters generally validated these factual points, though many disputed the inferences and conclusions to be drawn from these points, whereas others agreed with them. Individual commenters generally validated the factual descriptions of these characteristics of borrowers as well. • Consumers in financial difficulty. Some consumers turn to these products because they have experienced a sudden drop in income (“income shock”) or a large unexpected expense (“expense shock”). Other borrowers are in circumstances in which their expenses consistently outstrip their income. A sizable percentage of users report that they would have taken a loan on almost any terms offered. Again, the commenters generally validated these points as a factual matter, but disputed the inferences and conclusions to be drawn therefrom. • Loans do not function as marketed. Lenders market single-payment products as shortterm loans designed to provide a bridge to the consumer’s next payday or other income receipt. In practice, however, the amounts due on these loans consume such a large portion of the consumer’s paycheck or other periodic income source as to be unaffordable for most consumers 306 seeking to recover from an income or expense shock, and even more so for consumers with a chronic income shortfall. Lenders actively encourage consumers either simply to pay the finance charges due and roll over the loan instead of repaying the loan in full (or effectively roll over the loan by engaging in back-to-back transactions or returning to re-borrow in no more than a few days after repaying the loan). Indeed, lenders are dependent upon such re-borrowing for a substantial portion of their revenue and would lose money if each borrower repaid the loan when it was due without re-borrowing. The commenters tended to recharacterize these points rather than disputing them as a factual matter, though many industry commenters disagreed that these loans should be considered “unaffordable” for “most” consumers if many consumers manage to repay them after borrowing once or twice. Others contended that these loans should not be considered “unaffordable” if they are repaid eventually, even after re-borrowing multiple times in extended loan sequences. The commenters on all sides generally did not dispute the nature of the underlying business model as resting on repeat re-borrowing that lenders actively encourage, though they sharply disputed whether this model benefited or harmed consumers. • Very high re-borrowing rates. Most borrowers find it necessary to re-borrow when their loan comes due or shortly after repaying their loan, as other expenses come due. This reborrowing occurs both with payday loans and with single-payment vehicle title loans. The Bureau found that 56 percent of payday loans are borrowed on the same day and 85 percent of these loans are re-borrowed within a month. Fifty percent of all new storefront payday loans are followed by at least three more loans and 33 percent are followed by six more loans. While single-payment vehicle title loans are often for somewhat longer durations than payday loans, typically with terms of one month, re-borrowing tends to occur sooner and longer sequences of 307 loans are more common. The Bureau found that 83 percent of single-payment vehicle title loans are re-borrowed on the same day and 85 percent of them are re-borrowed within a month. Over half (56 percent) of all new single-payment vehicle title loans are followed by at least three more loans, and more than a third (36 percent) are followed by six or more loans. Of the payday loans made to borrowers paid weekly, bi-weekly, or semi-monthly, over 20 percent are in loan sequences of 20 loans or more and over 40 percent of loans made to borrowers paid monthly are in loan sequences of comparable durations (i.e., 10 or more monthly loans). The commenters did not challenge the thrust of these points as demonstrating a high incidence of re-borrowing, which is a point that was reinforced by consumer groups and was illustrated by many individual commenters as well. • Consumers do not expect lengthy loan sequences. Many consumers who take out a payday loan do not expect to re-borrow to the extent that they do. This is especially true of those consumers who end up in extended cycles of indebtedness. Research shows that many consumers who take out loans are able to accurately predict how long it will take them to get out of debt, especially if they repay immediately or re-borrow only once, but a substantial population of consumers is not able to do so, and for those consumers who end up in extended loan sequences, there is little correlation between predictions and behavior. A study on this topic found that as many as 43 percent of borrowers may have underestimated the length of time to repayment by two weeks or more.461 The study found that consumers who have borrowed heavily in the recent past are even more likely to underestimate how long it will take to repay the 461 See Ronald Mann, “Assessing the Optimism of Payday Loan Borrowers,” 21 Sup. Ct. Econ. Rev. 105 (2013). 308 loan.462 Consumers’ difficulty in this regard may be exacerbated by the fact that such loans involve a basic mismatch between how they are marketed as short-term credit and appear designed to function as long sequences of re-borrowing, which regularly occurs for a number of consumers. This disparity can create difficulties for consumers in being able to estimate accurately how long they will remain in debt and how much they will ultimately pay for the initial extension of credit. Research into consumer decision-making also helps explain why consumers may re-borrow more than they expect. For example, people under stress, including consumers in financial crisis, tend to become very focused on their immediate problems and think less about the future. Consumers also tend to underestimate their future expenses, and may be overly optimistic about their ability to recover from the shock they have experienced or to bring their expenses in line with their incomes. These points were sharply disputed by the commenters, and will be discussed further below. • Very high default rates and collateral harms. Some consumers do succeed in repaying short-term loans without re-borrowing, and others eventually repay the loan after re-borrowing multiple times. But research shows that approximately 20 percent of payday loan sequences and 33 percent of single-payment vehicle title loan sequences end up with the consumer defaulting. Consumers who default can become subject to often aggressive and psychologically harmful debt collection efforts. While delinquent, they may also seek to avoid default in ways that lead to a loss of control over budgeting for their other needs and expenses. In addition, 20 percent of single-payment vehicle title loan sequences end with borrowers losing their cars or trucks to repossession. Even borrowers who have not yet defaulted may incur penalty fees, late fees, or 462 See id. 309 overdraft fees along the way and may find themselves struggling to pay other bills or meet their basic living expenses. Commenters generally did not dispute that consumers may feel the effects of these negative collateral consequences of such loans and of delinquency and default, though industry commenters tended to downplay them and some argued that any such harms were outweighed by the economic benefits of such loans. Individual commenters validated this account of the negative collateral consequences of such loans as reflecting their own experiences. Many others countered that they had successful experiences with these loans and that they were benefited more than they were harmed by these experiences. • Harms occur despite existing regulation. The research indicates that in the States that have authorized payday and other short-term loans, these harms persist despite existing regulatory frameworks. Indeed, payday loans do not legally exist in many States, so by definition the harms identified by the Bureau’s research flow from such loans in those States where they are offered pursuant to existing regulatory frameworks. Even in those States where such loans are offered pursuant to somewhat different conditions, these distinctions do not appear to eliminate the harms that flow from the structure of such loans. In particular, the Bureau is concerned that existing caps on the amount that a consumer can borrow, rollover limitations, and short cooling-off periods still appear to leave many consumers vulnerable to the specific harms discussed above relating to default, delinquency, re-borrowing, and other collateral harms from attempting to avoid the other injuries by making unaffordable payments. Industry commenters took issue with these concerns and disputed this characterization of the effects of such loans. 310 In the proposal, the Bureau also reviewed the available evidence underlying each of these preliminary views. The Bureau sought and received comments on its review of the evidence, and those comments are reviewed and addressed in the discussion below. Based on the reasons set forth in each of the segments in this part, which respond to the comments and present further analysis that the Bureau has engaged in to consider these matters further, the Bureau now adopts as its findings underlying the final rule its views as stated in this initial summary overview, with certain modifications as set forth below. a. Borrower Characteristics and Circumstances of Borrowing As the Bureau laid out in the proposal, borrowers who take out payday, single-payment vehicle title, and other covered short-term loans are typically low-to-moderate income consumers who are looking for quick access to cash, who have little to no savings, who often have poor credit histories, and who have limited access to other forms of credit. Comments received from industry participants, trade associations, and individual users of these loans noted that this description of the borrower population does not describe all of the people who use these loans. That is so, of course, but the Bureau’s discussion in the proposal was not intended as an exhaustive account of the entire universe of borrowers. Instead, it merely represented many of the recurring borrower characteristics that the Bureau found based on its experience with such loans over the past several years and based on data from a number of studies as discussed further below. In the proposal, the Bureau had found preliminarily that the desire borrowers have for immediate cash may be the result of an emergency expense or an unanticipated drop in income. The comments received from industry participants, trade associations, and individual users of 311 these loans strongly reinforce the basis for this finding. Many comments describe the function that these loans perform as coping with income and expense shocks—that is, with unexpected, temporary expenses or shortfalls in income. These comments cited surveys and studies to bolster this point, including one survey that noted 86 percent of borrowers strongly or somewhat agreed that their use of a payday loan was to cope with an unexpected expense. Many other comments, including comments from individual users of these loans, offered anecdotal accounts of the personal reasons many borrowers have for taking out these loans, including a wide variety of circumstances that can create such income or expense shocks. Comments received from consumer groups were also in agreement on these points and further underscored a shared understanding that this impetus drives much of the demand for such loans. The comments received from industry participants, trade associations, and individual users of these loans made a different point as well. One trade association, for example, noted that many consumers use such loans for “income smoothing” or to create a better match between income and expenses in the face of income and expense volatility—that is, where the consumer’s income or expenses fluctuate over the course of the year, such that credit is needed during times of lower income or higher expenses to tide the consumer over until times of higher income or lower expenses. Many reasons were given by commenters, including a high volume of individual commenters, for such income and expense volatility, and the following examples are merely illustrative of the broader and more widespread phenomenon: people who work on commission; people scheduled to receive one-time or intermittent income supplements, such as holiday bonuses; people who work irregular hours, including many contractor or part-time workers; people who have seasonal opportunities to earn extra income by working additional 312 hours; or circumstances that may arise that create the need or the opportunity to satisfy in full some other outstanding debt that is pressing. Comments from consumer groups echoed these accounts of how these economic situations drive a certain amount of the demand for such loans. The nature and weight of these comments thus lend further support to the preliminary findings that the Bureau had made on these issues. In the proposal, the Bureau also noted that many borrowers who take out payday or single-payment vehicle title loans are consumers whose living expenses routinely exceed their income. This category of borrowers may consistently experience negative residual income, or to use a common phrase, find that they routinely have “too much month at the end of the money” and take out such loans in an effort to bolster their income—an effort that often proves to be unsuccessful when they are later unable to repay the loan according to its terms. Various commenters agreed with this account of some borrowers, and some of the individual commenters likewise described their own experiences in this vein. In addition, some commenters noted that certain borrowers may use these kinds of loans to manage accumulated debt, preferring to use the proceeds of the loan to pay down other debt for which nonpayment or default would be more costly alternatives. This was not frequently cited as a reason why many borrowers decide to take out such loans, but it may explain occasional instances. 1. Borrower Characteristics In the proposal, the Bureau noted that a number of studies have focused on the characteristics of payday borrowers. For instance, the FDIC and the U.S. Census Bureau have undertaken several special supplements to the Current Population Survey (CPS Supplement); the 313 proposal cited the most recent available data from 2013, which found that 46 percent of payday borrowers (including storefront and online borrowers) have a family income of under $30,000.463 The latest edition of the Survey has more recent data from 2015, which finds that the updated figure is 49 percent.464 A study covering a mix of storefront and online payday borrowers similarly found that 49 percent had income of $25,000 or less.465 Other analyses of administrative data that include the income borrowers reported to lenders show similar results.466 A 2012 survey administered by the Center for Financial Services Innovation (CFSI) to learn more about users of small-dollar credit products including payday loans, pawn loans, direct deposit advances, installment loans, and auto title loans found that 43 percent of small-dollar credit consumers had a household income between $0 and $25,000, compared to 26 percent of 463 Fed. Deposit Ins. Corp., “2013 FDIC National Survey of Unbanked and Underbanked Households: Appendices,” at appendix. D–12a (Oct. 2014), available at https://www.fdic.gov/householdsurvey/2013/2013appendix.pdf. 464 Fed. Deposit Ins. Corp., “2015 FDIC National Survey of Unbanked and Underbanked Households,” (Oct. 20, 2016), available at https://www.fdic.gov/householdsurvey/2015/2015report.pdf (Calculations made using custom data tool.). 465 Pew Charitable Trusts, “Payday Lending in America: Who Borrows, Where They Borrow, and Why,” at 35 exhibit 14 (Report 1, 2012), available at http://www.pewtrusts.org/~/media/legacy/uploadedfiles/pcs_assets/2012/pewpaydaylendingreportpdf.pdf. 466 CFPB Payday Loans and Deposit Advance Products White Paper, at 18 (reporting that based on confidential supervisory data of a number of storefront payday lenders, borrowers had a reported median annual income of $22,476 at the time of application (not necessarily household income)). Similarly, data from several State regulatory agencies indicate that average incomes range from about $31,000 (Delaware) to slightly over $36,000 (Washington). See Letter from Robert A. Glen, Del. State Bank Comm’r to Hon. Bryan Townsend, Chairman, S. Banking and Bus. Comm. and Hon. Bryon H. Short, Chairman, H. Econ. Dev./Banking/Ins./Commerce Comm. (enclosing Veritec Solutions, “State of Delaware Short-term Consumer Loan Program - Report on Delaware Short-term Consumer Loan Activity For the Year Ending December 31, 2014,” at 6 (Mar. 12, 2015), available at http://banking.delaware.gov/pdfs/annual/Short_Term_Consumer_Loan_Database_2014_Operations_Report.pdf; Wash. Dep’t of Fin. Insts., “2014 Payday Lending Report,” at 6 (2014), available at http://www.dfi.wa.gov/sites/default/files/reports/2014-payday-lending-report.pdf; nonPrime 101 found the median income for online payday borrowers to be $30,000. nonPrime101, “Report 1: Profiling Internet Small-Dollar Lending,” at 7 (2014), available at https://www.nonprime101.com/wp-content/uploads/2013/10/Clarity-ServicesProfiling-Internet-Small-Dollar-Lending.pdf. 314 non-small-dollar credit consumers.467 The mean annual household income for those making use of such products was $32,000, compared to $40,000 for those not using such products. Other studies and survey evidence presented by commenters were broadly consistent with the data and analysis contained in the studies that the Bureau had cited on this point. Additionally, the Bureau found in its analysis of confidential supervisory data that 18 percent of storefront borrowers relied on Social Security or some other form of government benefits or public assistance.468 The FDIC study further found that payday borrowers are disproportionately Hispanic or African-American (with borrowing rates two to three times higher respectively than for non-Hispanic whites) and that unmarried female-headed families are more than twice as likely as married couples to be payday borrowers.469 The CFSI study discussed above upheld this general assessment with regard to race, with African-American and Hispanic borrowers over-represented among such borrowers. 470 The commenters did not take issue with these points, and various submissions across the broad spectrum of stakeholders, including both industry participants and consumer groups, consistently reinforced the point that these loans disproportionately go to minority borrowers. 467 Rob Levy & Joshua Sledge, “A Complex Portrait: An Examination of Small-Dollar Credit Consumers,” (Ctr. for Fin. Servs. Innovation, 2012), available at https://www.fdic.gov/news/conferences/consumersymposium/2012/A%20Complex%20Portrait.pdf. 468 CFPB Payday Loans and Deposit Advance Products White Paper, at 18. 469 Fed. Deposit Ins. Corp., “2015 FDIC National Survey of Unbanked and Underbanked Households,” (Oct. 20, 2016), available at https://www.fdic.gov/householdsurvey/2015/2015report.pdf (Calculations made using custom data tool.). 470 Rob Levy & Joshua Sledge, “A Complex Portrait: An Examination of Small-Dollar Credit Consumers,” (Ctr. for Fin. Servs. Innovation, 2012), available at https://www.fdic.gov/news/conferences/consumersymposium/2012/A%20Complex%20Portrait.pdf. 315 The demographic profiles of single-payment vehicle title borrowers appear to be roughly comparable to the demographics of payday borrowers.471 Calculations from the CPS Supplement indicate that 44 percent of title borrowers have annual family incomes under $30,000.472 Another survey likewise found that 54 percent of title borrowers reported incomes below $30,000, compared with 60 percent for payday borrowers.473 Commenters presented some data to suggest that various borrowers are more educated and that many are middle-aged, but these results did not alter the great weight of the overall survey data on this point. And as with payday borrowers, data from the CPS Supplement show vehicle title borrowers to be disproportionately African-American or Hispanic, and more likely to live in unmarried female-headed families.474 Similarly, a survey of borrowers in three States conducted by academic researchers found that title borrowers were disproportionately female and minority. Over 58 percent of title borrowers were female. African-Americans were over-represented among borrowers compared to their share of their States’ population at large. Hispanic borrowers were over-represented in two of the three States; however, these borrowers were under-represented in Texas, the State with the highest proportion of Hispanic residents in the 471 None of the sources of information on the characteristics of vehicle title borrowers that the Bureau is aware of distinguishes between borrowers taking out single-payment and installment vehicle title loans. The statistics provided here are for borrowers taking out either type of vehicle title loan. 472 Fed. Deposit Ins. Corp., “2015 FDIC National Survey of Unbanked and Underbanked Households,” (Oct. 20, 2016), available at https://www.fdic.gov/householdsurvey/2015/2015report.pdf (Calculations made using custom data tool.). 473 Pew Charitable Trusts, “Auto Title Loans: Market Practices and Borrowers Experiences,” at 28 (2015), available at http://www.pewtrusts.org/~/media/assets/2015/03/autotitleloansreport.pdf; Pew Charitable Trusts, “Payday Lending in America: Who Borrows, Where They Borrow, and Why,” at 35 (Report 1, 2012), available at http://www.pewtrusts.org/~/media/legacy/uploadedfiles/pcs_assets/2012/pewpaydaylendingreportpdf.pdf. 474 Fed. Deposit Ins. Corp., “2015 FDIC National Survey of Unbanked and Underbanked Households,” (Oct. 20, 2016), available at https://www.fdic.gov/householdsurvey/2015/2015report.pdf (Calculations made using custom data tool.). 316 study.475 Commenters generally did not take issue with these points, and various submissions from both industry participants and consumer groups support the view that they are an accurate reflection of the borrower population. One commenter contended that the data did not show vehicle title borrowers to be disproportionately minority consumers, though this view did not seem to take into account the composition of the total population in the States that were surveyed. As noted in the proposal, studies of payday borrowers’ credit histories show both poor credit histories and recent credit-seeking activity. One academic paper that matched administrative data from one storefront payday lender to credit bureau data found that the median credit score for a payday applicant was in the bottom 15 percent of credit scores overall. 476 The median applicant had one open credit card, but 80 percent of applicants had either no credit card or no credit available on a card. The average borrower had 5.2 credit inquiries on her credit report over the preceding 12 months before her initial application for a payday loan (three times the number for the general population), but obtained only 1.4 accounts on average. This suggests that borrowers made repeated but generally unsuccessful efforts to obtain additional other forms of credit prior to initiating a payday loan. While typical payday borrowers may have one or more credit cards, they are unlikely to have unused credit; in fact, they are often delinquent on one or more cards, and have often experienced multiple overdrafts and/or NSFs on their 475 Kathryn Fritzdixon et al., “Dude, Where’s My Car Title?: The Law Behavior and Economics of Title Lending Markets,” 2014 U. IL L. Rev. 1013, at 1029–1030 (2014). 476 Neil Bhutta et al., “Consumer Borrowing after Payday Loan Bans,” 59 J. of L. and Econ. 225, at 231–233 (2016). Note that the credit score used in this analysis was the Equifax Risk Score which ranges from 280–850. Frederic Huynh, “FICO Score Distribution,” FICO Blog (Apr. 15, 2013), http://www.fico.com/en/blogs/risk-compliance/ficoscore-distribution-remains-mixed/. 317 checking accounts.477 A recent report analyzing credit scores of borrowers from five large storefront payday lenders provides corroborative support, finding that the average borrower had a VantageScore 3.0478 score of 532 and that over 85 percent of borrowers had a score below 600, indicating high credit risk.479 By way of comparison, the national average VantageScore is 669 and only 30 percent of consumers have a VantageScore below 600.480 The proposal also cited reports using data from a specialty consumer reporting agency, which indicate that online borrowers have comparable credit scores to storefront borrowers (a mean VantageScore 3.0 score of 525 versus 532 for storefront).481 Another study based on the data from the same specialty consumer reporting agency and an accompanying survey of online small-dollar credit borrowers reported that 79 percent of those surveyed had been denied traditional credit in the past year due to having a low or no credit score, 62 percent had already sought assistance from family and friends, and 24 percent reported having negotiated with a creditor to whom they owed money.482 Moreover, heavy use of online payday loans seems to be correlated with more strenuous credit-seeking: compared to light (bottom quartile) users of 477 Neil Bhutta et al., “Consumer Borrowing after Payday Loan Bans,” 59 J. of L. and Econ. 225, at 231–233 (2016). A VantageScore 3.0 score is a credit score created by an eponymous joint venture of the three major credit reporting companies; scores lie on the range 300-850. 479 nonPrime101, “Report 8: Can Storefront Payday Borrowers Become Installment Loan Borrowers?,” at 7 (2015), available at https://www.nonprime101.com/blog/can-storefront-payday-borrowers-become-installment-loanborrowers/. 480 Experian, “What is Your State of Credit,” (2015), available at http://www.experian.com/live-credit-smart/stateof-credit-2015.html 481 nonPrime101, “Report 8: Can Storefront Payday Borrowers Become Installment Loan Borrowers?,” at 5 (2015), available at https://www.nonprime101.com/blog/can-storefront-payday-borrowers-become-installment-loanborrowers/. Twenty percent of online borrowers are unable to be scored; for storefront borrowers the percentage of unscorable consumers is negligible. However, this may partly reflect the limited quality of the data online lenders obtain and/or report about their customers and resulting inability to obtain a credit report match. 482 Stephen Nunez et al., “Online Payday and Installment Loans: Who Uses Them and Why?, at 44, 51, 60 (MDRC, 2016), available at http://www.mdrc.org/sites/default/files/online_payday_2016_FR.pdf. 478 318 online loans, heavy (top quartile) users were more likely to have been denied credit in the past year (87 percent of heavy users compared to 68 percent of light users).483 In the proposal, the Bureau also noted that other surveys of payday borrowers added to the picture of consumers in financial distress. For example, in a survey of payday borrowers published in 2009, fewer than half reported having any savings or reserve funds.484 Almost a third of borrowers (31.8 percent) reported monthly debt-to-income payments of 30 percent or higher, and more than a third (36.4 percent) of borrowers reported that they regularly spend all the income they receive. Similarly, a 2010 survey found that over 80 percent of payday borrowers reported making at least one late payment on a bill in the preceding three months, and approximately one quarter reported frequently paying bills late. Approximately half reported bouncing at least one check in the previous three months, and 30 percent reported doing so more than once.485 Furthermore, a 2012 survey found that 58 percent of payday borrowers report that they struggled to pay their bills on time. More than a third (37 percent) said they would have taken out a loan on almost any terms offered. This figure rises to 46 percent when the respondent rated his or her financial situation as particularly poor.486 483 Stephen Nunez et al., “Online Payday and Installment Loans: Who Uses Them and Why?, at 38 tbl. 6 (MDRC, 2016), available at http://www.mdrc.org/sites/default/files/online_payday_2016_FR.pdf. 484 Gregory Elliehausen, “An Analysis of Consumers’ Use of Payday Loans,” at 29 (Geo. Wash. Sch. of Bus., Monograph No. 41, 2009), available at https://www.researchgate.net/publication/237554300_AN_ANALYSIS_OF_CONSUMERS%27_USE_OF_PAYDAY _LOANS. 485 Jonathan Zinman, “Restricting Consumer Credit Access: Household Survey Evidence on Effects Around the Oregon Rate Cap,” at 20 tbl. 1 (Dartmouth College, 2008), available at http://www.dartmouth.edu/~jzinman/Papers/Zinman_RestrictingAccess_oct08.pdf. 486 See Pew Charitable Trusts, “Payday Lending in America: How Borrowers Choose and Repay Payday Loans,” at 20 (Report 2, 2013), http://www.pewtrusts.org/en/research-and-analysis/reports/2013/02/19/how-borrowerschoose-and-repay-payday-loans. 319 A large number of comments received from industry participants, trade associations, consumer groups, academics, and individual users of these loans extensively reinforced this picture of the financial situation for many storefront and online borrowers. Industry participants and trade associations presented their understanding of the characteristics of the borrower population as being marked by poor credit histories, an acute need for credit, aggressive efforts to seek credit, and general unavailability of other means of credit for many of these borrowers. In many of the comments, these characteristics were described in particular detail and emphasized as making the case to show the need for the availability of such loans. Many individual users of these loans also related their own personal stories and situations, which were typically marked by these same features of their financial histories that demonstrated their need for credit products. Despite these points of general agreement, many industry participants, trade associations, individual users of such loans, and some academics submitted comments that vigorously disagreed with what they regarded as assumptions the Bureau had made in the proposal about payday and vehicle title borrowers. In their view, the Bureau was wrongly portraying these consumers as financially unsophisticated and incapable of acting in their own best interests. On the contrary, many of these commenters stated, such borrowers are often very knowledgeable about the costs and terms of such loans. Their decision to take out a payday or vehicle title loan was represented, in many instances, as being based on a rational judgment that access to this form of credit is far more valuable than reducing the risks and costs associated with their indebtedness. The Bureau recognizes that the characteristics of individual users of payday and single320 payment vehicle title loans are differentiated in many and various ways. Much of the debate here represents different characterizations and opinions about potential conclusions drawn from the facts, rather than direct disagreements about the facts themselves. These issues are important and they are considered further in the discussions of unfairness and abusiveness under final § 1041.4. 2. Circumstances of Borrowing The proposal discussed several surveys that have asked borrowers why they took out their loans or for what purpose they used the loan proceeds, and noted that these are challenging questions to study. Any survey that asks about past behavior or events runs some risk of recall errors.487 In addition, the fact that money is fungible makes this question more complicated. For example, a consumer who has an unexpected expense may not feel the effect fully until weeks later, depending on the timing of the unexpected expense relative to other expenses and to the receipt of income. In that circumstance, a borrower may say either that she took out the loan because of the unexpected expense, or that she took out the loan to cover regular expenses. Perhaps because of this difficulty, results across surveys are somewhat inconsistent, with one finding high levels of unexpected expenses, while others find that payday loans are used primarily to pay for regular expenses. In the first survey discussed in the proposal, a 2007 survey of payday borrowers, the most common reason cited for taking out a loan was “an unexpected expense that could not be postponed,” with 71 percent of respondents strongly agreeing with this reason and 16 percent 487 See generally David Grimes and Kenneth F. Schulz, “Bias and Causal Associations in Observational Research,” 359 Lancet 9302, at 248–252. (2002); see E. Hassan, “Recall Bias Can Be a Threat to Retrospective and Prospective Research Designs,” 3 Internet J. of Epidemiology 2 (2005) (for a more specific discussion of recall bias). 321 somewhat agreeing.488 A 2012 survey of payday loan borrowers, by contrast, found that 69 percent of respondents took their first payday loan to cover a recurring expense, such as utilities, rent, or credit card bills, and only 16 percent took their first loan for an unexpected expense.489 The 2012 CFSI survey of alternative small-dollar credit products, discussed earlier in this section asked separate questions about what borrowers used the loan proceeds for and what precipitated the loan.490 Responses were reported for “very short term” and “short term” credit; “very short term” referred to payday, pawn, and deposit advance products. Respondents could report up to three reasons for what precipitated the loan; the most common reason given for very-short-term borrowing (approximately 37 percent of respondents) was “I had a bill or payment due before my paycheck arrived,” which the authors of the report on the survey results interpreted as a mismatch in the timing of income and expenses. Unexpected expenses were cited by 30 percent of very-short-term borrowers, and approximately 27 percent reported unexpected drops in income. Approximately 34 percent reported that their general living expenses were consistently more than their income. Respondents could also report up to three uses for the funds; the most common answers related to paying for routine expenses, with about 40 percent reporting the funds were used to “pay utility bills,” about 40 percent reporting the funds were used to pay “general living expenses,” and about 20 percent saying the funds were used to pay rent. Of all 488 Gregory Elliehausen, “An Analysis of Consumers’ Use of Payday Loans,” at 35 (Geo. Wash. Sch. of Bus., Monograph No. 41, 2009), available at https://www.researchgate.net/publication/237554300_AN_ANALYSIS_OF_CONSUMERS%27_USE_OF_PAYDAY _LOANS. 489 Pew Charitable Trusts, “Payday Lending in America: Who Borrows, Where They Borrow, and Why,” at 5 (Report 1, 2012), available at http://www.pewtrusts.org/~/media/legacy/uploadedfiles/pcs_assets/2012/pewpaydaylendingreportpdf.pdf. 490 Rob Levy & Joshua Sledge, “A Complex Portrait: An Examination of Small-Dollar Credit Consumers,” (Ctr. for Fin. Servs. Innovation, 2012), available at https://www.fdic.gov/news/conferences/consumersymposium/2012/A%20Complex%20Portrait.pdf. 322 the reasons for borrowing, consistent shortfalls in income relative to expenses was the response most highly correlated with consumers who reported repeated usage or rollovers. A survey of 768 online payday users conducted in 2015 and drawn from a large administrative database of payday borrowers looked at similar questions, and compared the answers of heavy and light users of online loans.491 Based on consumers’ self-reported borrowing history, they were segmented into heavy users (users with borrowing frequency in the top quartile of the dataset) and light users (bottom quartile). Heavy users were much more likely to report that they “[i]n past three months, often or always ran out of money before the end of the month” (60 percent versus 34 percent). In addition, heavy users were nearly twice as likely as light users to state their primary reason for seeking their most recent payday loan as being to pay for “regular expenses such as utilities, car payment, credit card bill, or prescriptions” (49 percent versus 28 percent). Heavy users were less than half as likely as light users to state their reason as being to pay for an “unexpected expense or emergency” (21 percent versus 43 percent). Notably, 18 percent of heavy users stated that their primary reason for seeking a payday loan online was that they “had a storefront loan, needed another [loan]” as compared to just over one percent of light users. One industry commenter asserted that a significant share of vehicle title loan borrowers were small business owners who use these loans for business, rather than personal uses. The commenter pointed to one study that cited anonymous “industry sources” who claimed that 25- 491 Stephen Nunez et al., “Online Payday and Installment Loans: Who Uses Them and Why?, (MDRC, 2016), available at http://www.mdrc.org/sites/default/files/online_payday_2016_FR.pdf (A demand-side analysis from linked administrative, survey, and qualitative interview data.). 323 30 percent of title borrowers were small businesses492 and another study that cited an unpublished lender survey which found that about 20 percent of borrowers were selfemployed.493 Evidence was not provided by the commenter to document the share of vehicle title loan borrowers who are either self-employed or small business owners; however, the Bureau notes that it is important to distinguish between borrowers who may be small business owners but may not necessarily use a title loan for a business purpose. For example, one survey of title loan borrowers found that while 16 percent of title loan borrowers were self-employed, only 6 percent of title loan borrowers state that they took the loan for a business expense.494 The study’s authors concluded that “…it seems like business credit is not a significant portion of the loans.”495 Another survey found that 20 percent of title loan borrowers are self-employed, and an additional 3 percent were both self-employed and worked for an employer. In that survey, 3 percent of title loan borrowers reported the loan was for a business expense and 2 percent reported the loan was for a mix of personal and business use.496 Some commenters agreed with the Bureau that the results across surveys are somewhat inconsistent, perhaps because of methodological issues. Industry commenters predictably chose to place more emphasis on the results that accorded with their arguments that these loans help consumers cope with financial shocks or allow smoothing of income. By contrast, consumer 492 Todd. J. Zywicki, “Consumer Use and Government Regulation of Title Pledge Lending, 22 Loyola Cons. Law. Rev. 4 (2010), available at http://lawecommons.luc.edu/cgi/viewcontent.cgi?article=1053&context=lclr. 493 Jim Hawkins, “Credit on Wheels: The Law and Business of Auto-Title Lending,” 69 Wash. & Lee L. Rev. 535, 545 (2012). 494 Kathryn Fritzdixon et al., “Dude, Where’s My Car Title?: The Law Behavior and Economics of Title Lending Markets,” 2014 U. IL L. Rev. 1013, 1033 (2014). 495 Kathryn Fritzdixon et al., “Dude, Where’s My Car Title?: The Law Behavior and Economics of Title Lending Markets,” 2014 U. IL L. Rev. 1013, 1036 (2014). 496 See Pew Charitable Trusts, “Auto Title Loans,” at 29 (March 2015), available at http://www.pewtrusts.org/~/media/assets/2015/03/autotitleloansreport.pdf. 324 groups predictably took the opposite perspective. They contended that these loans do present special risks and harms for consumers that outweigh the benefits of access to such loans without being subject to any underwriting, especially for those consumers who experience chronic shortfalls of income. Both groups of commenters chose to downplay the results that tended to undermine their arguments. On the whole, these comments do not call into question the Bureau’s treatment of the factual issues here, but go more to the potential characterization of those facts or the inferences to be drawn from them. Those issues are discussed further in the section-by-section analysis for § 1041.4 below. A number of comments from industry participants and trade associations faulted the Bureau for not undertaking to conduct its own surveys of borrowers to gauge the circumstances that lead them to use payday, title, or other covered short-term loans. Although the Bureau had reviewed and analyzed at least four different surveys of such borrowers conducted over the past decade, as discussed above, these commenters stated that the Bureau would have furthered its understanding by speaking with and hearing directly from such borrowers. Nonetheless, many of these commenters offered further non-survey information of this kind by referencing the consumer narratives in thousands of individual consumer complaints about payday, title, and other covered loans that have been filed with the Bureau (which also include a substantial number of debt collection complaints stemming from such loans). They also pointed to individual responses that have been filed about such loans on the Bureau’s online “Tell Your Story” function, where some number of individual borrowers have explained how they use such loans, often describing the benefits and challenges they have experienced as a result. 325 In addition, a large volume of comments—totaling well over a million comments about the proposal, both pro and con—were filed with the Bureau by individual users of payday and vehicle title loans. Many of these commenters described their own personal experiences with these loans, and others offered their perspectives. The Bureau has reviewed these comments and has carefully considered the stories they told. These comments include a large number of positive accounts of how people successfully used such loans to address shortfalls or cope with emergencies and concerns about the possibility of access to such loans being removed. The comments included fewer but still a very sizable number of other accounts, much more negative in tone, of how consumers who took out such loans became trapped in long cycles of repeated reborrowing that led to financial distress, marked by problems such as budgetary distortions, high collateral costs, the loss of depository accounts and other services, ultimate default on the loans, and the loss of other assets such as people’s homes and their vehicles. Some of these comments came from the individual consumers themselves, while many came from friends, family members, clergy, legal aid attorneys, neighbors, or others who were concerned about the impact the loans had on consumers whom they knew, and in some cases whom they had helped to mitigate the negative experience through financial assistance, counseling, or legal assistance. The enormous volume of such individual comments itself helps to provide considerably more information about borrowers that helps to supplement the prior survey data discussed in the proposal. It appears that various parties on both sides of these issues went to great lengths to solicit such a large number of comment submissions by and about individual users of such loans. The substantial volume and variation of individual comments have further added to the Bureau’s understanding of the wide variety of circumstances in which such borrowing occurs. 326 They underscore the Bureau’s recognition that not only the personal characteristics, but also the particularized circumstances, of individual users of payday and single-payment vehicle title loans can be quite differentiated from one another across the market. Nonetheless, the focus of this rule is on how the identified lender practice of making such loans without reasonably assessing the borrower’s ability to repay the loan according to its terms affects this broad and diverse universe of consumers. b. Lender Practices As described in the proposal, the business model of lenders who make payday and singlepayment vehicle title loans is predicated on the lenders’ ability to secure extensive re-borrowing. As recounted in the Background section, the typical storefront payday loan has a principal amount of $350, and the consumer pays a typical fee of 15 percent of the principal amount. For a consumer who takes out such a loan and repays it when it is due without re-borrowing, this means the typical loan would produce roughly $50 in revenue to the lender. Lenders would thus require a large number of “one-and-done” consumers to cover their overhead and acquisition costs and generate profits. However, because lenders are able to induce a large percentage of borrowers to repeatedly re-borrow, lenders have built a model in which the typical storefront lender, as discussed in part II above, has two or three employees serving around 500 customers per year. Online lenders do not have the same overhead costs, but they have been willing to pay substantial acquisition costs to lead generators and to incur substantial fraud losses, all of which can only be sufficiently offset by their ability to secure more than a single fee—and often many repeated fees—from their borrowers. 327 In the proposal, the Bureau used the term “re-borrow” to refer to situations in which consumers either roll over a loan (which means they pay a fee to defer payment of the principal for an additional period of time), or take out a new loan within a short period time following a previous loan. Re-borrowing can occur concurrently with repayment in back-to-back transactions or can occur shortly thereafter. In the proposal, the Bureau stated its reasons for concluding that re-borrowing often indicates that the previous loan was beyond the consumer’s ability to repay while meeting the consumer’s other major financial obligations and basic living expenses. As discussed in more detail in the section-by-section analysis of § 1041.6, the Bureau proposed and now concludes that it is appropriate to consider loans to be re-borrowings when the second loan is taken out within 30 days of the consumer being indebted on a previous loan. While the Bureau’s 2014 Data Point used a 14-day period and the Small Business Review Panel Outline used a 60-day period, the Bureau used a 30-day period in its proposal to align the time frame with consumer expense cycles, which are typically a month in length. This duration was designed to account for the fact that where repaying a loan causes a shortfall, the effect is most likely to be experienced within a 30-day period in which monthly expenses for matters such as housing and other debts come due. The Bureau recognizes that some re-borrowing that occurs after a 30-day period may be attributable to the spillover effects of an unaffordable loan and that some re-borrowing that occurs within the 30-day period may be attributable to a new need that arises unrelated to the impact of repaying the short-term loan. Thus, while other periods could plausibly be used to determine when a follow-on loan constitutes re-borrowing, the Bureau believes that the 30-day period provides the most appropriate period for these purposes. In fact, the evidence presented below suggests that for any of these three potential time frames, though 328 the percentage varies somewhat, the number of loans that occur as part of extended loan sequences of 10 loans or more is around half of all payday loans. Accordingly, this section, Market Concerns—Underwriting, uses a 30-day period to determine whether a loan is part of a loan sequence. The proposal noted that the majority of lending revenue earned by storefront payday lenders and lenders that make single-payment vehicle title loans comes from borrowers who reborrow multiple times and become enmeshed in long loan sequences. Based on the Bureau’s data analysis, approximately half of all payday loans are in sequences that contain 10 loans or more, depending on the time frame that is used to define the sequence.497 Looking just at loans made to borrowers who are paid weekly, bi-weekly, or semi-monthly, more than 20 percent of loans are in sequences that are 20 loans or longer. Similarly, the Bureau found that about half of all single-payment vehicle title loans are in sequences of 10 loans or more, and over two-thirds of them are in sequences of at least seven loans.498 The commenters did not take serious issue with this data analysis, and the Bureau finds these particular facts to be of great significance in assessing the justifications for regulatory measures that would address the consequent harms experienced by consumers. Commenters on all sides of the proposal did not seriously take issue with the account presented in the proposal of the basic business model in the marketplace for payday and singlepayment vehicle title loans. They did have widely divergent views about whether they would 497 This is true regardless of whether sequence is defined using either a 14-day, 30-day, or 60-day period to determine whether loans are within the same loan sequence. Using the 14-day period, just under half of these loans (47 percent) are in sequences that contain 10 loans or more. Using a longer period, more than half of these loans (30 days, 53 percent; 60 days, 59 percent) are in sequences that contain 10 loans or more. 498 CFPB Single-Payment Vehicle Title Lending, at 14. 329 characterize these facts as beneficial or pernicious, or what consequences they perceive as resulting from this business model. One credit union trade association stated its view that such lending takes advantage of consumers and exacerbates bad financial situations and thus it favored curbs on payday lending. Consumer groups and numerous individual borrowers echoed this view. Industry participants, other trade associations, and many other individual borrowers took the position, explicitly or implicitly, that the benefits experienced by successful users of these loans outweighed the costs incurred by those who engaged in repeat re-borrowing with consequent negative outcomes and collateral consequences. As discussed below, the Bureau has considered the comments submitted on the proposal and continues to believe that both the short term and the single-payment structure of these loans contributes to the long loan sequences that borrowers take out. Various lender practices exacerbate the problem by marketing to borrowers who are particularly likely to wind up in long sequences of loans, by failing to screen out borrowers who are likely to wind up in long-term debt or to establish guardrails to avoid long-term indebtedness, and by actively encouraging borrowers to continue to re-borrow when their single-payment loans come due. 1. Loan Structure The proposal described how the single-payment structure and short duration of these loans makes them difficult to repay. Within the space of a single income or expense cycle, a consumer with little to no savings cushion and who has borrowed to meet an unexpected expense or income shortfall, or who chronically runs short of funds, is unlikely to have the available cash needed to repay the full amount borrowed plus the finance charge on the loan when it is due and to cover other ongoing expenses. This is true for loans of a very short duration regardless of how 330 the loan may be categorized. Loans of this type, as they exist in the market today, typically take the form of single-payment loans, including payday loans and vehicle title loans, though other types of credit products are possible.499 Because the focus of the Bureau’s research has been on payday and vehicle title loans, the discussion in Market Concerns—Underwriting centers on those types of products. The size of single-payment loan repayment amounts (measured as loan principal plus finance charges owed) relative to the borrower’s next paycheck gives some sense of how difficult repayment may be. The Bureau’s storefront payday loan data shows that the average borrower being paid on a bi-weekly basis would need to devote 37 percent of her bi-weekly paycheck to repaying the loan. Single-payment vehicle title borrowers face an even greater challenge. In the data analyzed by the Bureau, the median borrower’s payment on a 30-day loan is equal to 49 percent of monthly income,500 and the Bureau finds it especially significant as indicating the severe challenges and potential for negative outcomes associated with these loans. The commenters did not offer any data that disagreed with this analysis of how the loan structure works in practice. Industry commenters did assert, however, that the structure of these loans is not intended or designed as a means of exploiting consumers, but rather has evolved as 499 In the past, a number of depository institutions have also offered deposit advance products. A small number of institutions still offer similar products. Like payday loans, deposit advances are typically structured as short-term loans. However, deposit advances do not have a pre-determined repayment date. Instead, deposit advance agreements typically stipulate that repayment will automatically be taken out of the borrower’s next qualifying electronic deposit. Deposit advances are typically requested through online banking or over the phone, although at some institutions they may be requested at a branch. As described in more detail in the CFPB Payday Loans and Deposit Advance Products White Paper, the Bureau’s research demonstrated similar borrowing patterns in both deposit advance products and payday loans. See CFPB Payday Loans and Deposit Advance Products White Paper, at 32-42. 500 The data used for this calculation is described in CFPB Data Point: Payday Lending, at 10-15 and in CFPB Report on Supplemental Findings. 331 needed to comply with the directives of State law and State regulation of this lending market. As a historical matter, this appears to be incorrect; indeed, another commenter is the founder of the company who helped to initiate the payday lending industry, W. Allan Jones. The comment notes that the “traditional ‘payday loan’ product” was first developed by his company in 1993 in Tennessee and then became the basis for legislation and regulation that has spread to a majority of States, with various modifications and refinements. As noted above in part II.A, however, another large payday lender—QC Financial—began making payday loans in Kansas in 1992 under an existing provision of that state’s existing consumer lending structure and that same year at least one State regulator formally held that deferred presentment activities constituted consumer lending subject to the State’s consumer credit laws.501 Other accounts of the history of payday lending generally tend to reinforce these historical accounts that modern payday lending began emerging in the early 1990s as a variant of check-cashing stores whereby the check casher would cash and hold consumers’ personal checks for a fee for several days—until payday— before cashing them.502 The laws of States, particularly those that had adopted the Uniform Consumer Credit Code (UCCC) including Kansas and Colorado, permitted lenders to retain a minimum finance charge on loans ranging in the 1990’s from about $15 to $25 per loan regardless of State rate caps, and payday lenders used those provisions to make payday loans. In other States, and later in UCCC States, more specific statutes were enacted to authorize and 501 QC Holdings, Inc., Registration Statement (Form S–1), at 1 (May 7, 2004); see, e.g., Laura Udis, Adm’r Colo. Dep’t of Law, Unif. Consumer Credit Code, “Check Cashing Entities Which Provide Funds In Return For A PostDated Check Or Similar Deferred Payment Arrangement And Which Impose A Check Cashing Charge Or Fee May Be Consumer Lenders Subject To The Colorado Uniform Consumer Credit Code,” Administrative Interpretation No. 3.104–9201 (June 23, 1992) (on file). 502 Pew Charitable Trusts, “A Short History of Payday Lending Law,” (July 18, 2012), available at http://www.pewtrusts.org/en/research-and-analysis/analysis/2012/07/a-short-history-of-payday-lending-law. 332 regulate what had become payday lending. No doubt the structure of such loan products over time is affected by and tends to conform to State laws and regulations, but the point here is that the key features of the loan structure, which tend to make these loans difficult to repay for a significant population of borrowers, are core to this financial product and are fairly consistent across time and geography. Regardless of the historical background, however, one implication of the suggestion put forward by these commenters appears to be that the intended consequence of this loan product is to produce cycles of re-borrowing or extended loan sequences for many consumers that exceed the permissible short-term loan periods adopted under State law. The explanation seems to be that the actual borrowing needs of consumers extend beyond the permissible loan periods permitted by State law. If that is so, then the inherent nature of this mismatched product imposes large forecasting risks on the consumer, which may often lead to unexpected harms. And even if the claim instead is that the loan structure manages to co-exist with the formal constraints imposed by State law, this justification does little to minimize the risks and harms to the substantial population of consumers who find themselves trapped in extended loan sequences. 2. Marketing The proposal also noted that the general positioning of short-term products in marketing and advertising materials as a solution to an immediate liquidity challenge attracts consumers facing these problems, encouraging them to focus on short-term relief rather than the likelihood that they are taking on a new longer-term debt. Lenders position the purpose of the loan as being for use “until next payday” or to “tide over” the consumer until she receives her next 333 paycheck.503 These types of product characterizations can encourage consumers to think of these loans as easy to repay, a fast solution to a temporary cash shortfall, and a short-term obligation, all of which lessen the risk in the consumer’s mind that the loan will become a long-term debt cycle. Indeed, one study reporting consumer focus group feedback noted that some participants reported that the marketing made it seem like payday loans were “a way to get a cash infusion without creating an additional bill.”504 As discussed in the proposal, in addition to presenting loans as short-term solutions, rather than potentially long-term obligations, lender advertising often focuses on how quickly and easily consumers can obtain a loan. An academic paper reviewing the advertisements of Texas storefront and online payday and vehicle title lenders found that the speed of getting a loan is the most frequently advertised feature in both online (100 percent) and storefront (50 percent) 503 See, e.g., Speedy Cash, “Payday Loan”, https://www.speedycash.com/payday-loans (last visited Sept. 24, 2017) (“A Speedy Cash payday loan may be a solution to help keep you afloat until your next pay day.”); Check Into Cash, “Our Loan Process,” https://checkintocash.com/payday-loans/ (last visited Sept. 24, 2017) (“A payday loan is a small dollar short-term advance used as an option to help a person with small, often unexpected expenses.”); Cash America, “Cash Advance/Short-term Loans,” http://www.cashamerica.com/LoanOptions/CashAdvances.aspx (last visited May 18, 2016) (noting that “a short-term loan, payday advance or a deferred deposit transaction—can help tide you over until your next payday” and that “A single payday advance is typically for two to four weeks. However, borrowers often use these loans over a period of months, which can be expensive. Payday advances are not recommended as long-term financial solutions.”); Cmty. Fin. Servcs. Ass’n of Am., “Is A Payday Advance Appropriate For You?,” http://cfsaa.com/what-is-a-payday-advance/is-a-payday-advance-appropriate-for-you.aspx (last visited May 18, 2016) (The national trade association representing storefront payday lenders analogizes a payday loan to “a cost-efficient ‘financial taxi’ to get from one payday to another when a consumer is faced with a small, short-term cash need.” The website elaborates that, “Just as a taxi is a convenient and valuable service for short distance transportation, a payday advance is a convenient and reasonably-priced service that should be used to meet small-dollar, short-term needs. A taxi service, however, is not economical for long-distance travel, and a payday advance is inappropriate when used as a long-term credit solution for ongoing budget management.”). 504 Pew Charitable Trusts, “Payday Lending in America: How Borrowers Choose and Repay Payday Loans,” at 22 (Report 2, 2013), available at http://www.pewtrusts.org/en/research-and-analysis/reports/2013/02/19/howborrowers-choose-and-repay-payday-loans (“To some focus group respondents, a payday loan, as marketed, did not seem as if it would add to their recurring debt, because it was a short-term loan to provide quick cash rather than an additional obligation. They were already in debt and struggling with regular expenses, and a payday loan seemed like a way to get a cash infusion without creating an additional bill.”). 334 payday and title loans.505 Advertising that is focused on immediacy and speed capitalizes on the sense of urgency borrowers feel when facing a cash shortfall. Indeed, the names of many payday and vehicle title lenders include the words (in different spellings) “speedy,” “cash,” “easy,” and “quick,” thus emphasizing their rapid and simple loan funding. All of the commenters generally agreed as a factual matter that the marketing and offering of such loans is typically marked by ease, speed, and convenience, which are touted as positive attributes of such loans that make them desirable credit products from the standpoint of potential borrowers. Yet industry participants and trade associations broadly disputed what they viewed as the Bureau’s perspective on the potential implications of this marketing analysis, as suggesting that many borrowers lack knowledge or awareness about the nature, costs, and overall effects of these loans. Consumer advocates, on the other hand, contended that the manner in which these loans are being marketed affects the likelihood that borrowers will tend to view them as short-term obligations that will not have long-term effects on their overall financial position, which often leads consumers to experience the negative outcomes associated with unexpectedly ending up in extended loan sequences. 3. Failure to Assess Ability to Repay As discussed in the proposal, the typical loan process for storefront payday, online payday, and single-payment vehicle title lenders generally involves gathering some basic information about borrowers before making a loan. Lenders normally do collect income 505 Jim Hawkins, “Using Advertisements to Diagnose Behavioral Market Failure in Payday Lending Markets,” 51 Wake Forest L. Rev. 57, at 71 (2016). The next most advertised features in online content are simple application process and no credit check/bad credit OK (both at 97 percent). For storefront lenders, the ability to get a high loan amount was the second most highly advertised content. 335 information, although the information they collect may just be self-reported or “stated” income. Payday lenders collect information to ensure the borrower has a checking account, and title lenders need information about the vehicle that will provide the security for the loan. Some lenders access consumer reports prepared by specialty consumer reporting agencies and engage in sophisticated screening of applicants, and at least some lenders turn down the majority of applicants to whom they have not previously made loans. One of the primary purposes of this screening, however, is to avoid fraud and other “first payment defaults,” not to make any kind of determination that borrowers will be able to repay the loan without re-borrowing. These lenders generally do not obtain any information about the borrower’s existing obligations or living expenses, which means that they cannot and do not prevent those with expenses chronically exceeding income, or those who have suffered from an income or expense shock from which they need substantially more time to recover than the term of the loan, from taking on additional obligations in the form of payday or similar loans. Thus, lenders’ failure to assess the borrower’s ability to repay the loan permits those consumers who are least able to repay the loans, and consequently are most likely to re-borrow, to obtain them. Lending to borrowers who cannot repay their loans would generally not be profitable in a traditional lending market, but as described elsewhere in this section, the factors that funnel consumers into cycles of repeat re-borrowing turn the traditional model on its head by creating incentives for lenders to actually want to make loans to borrowers who cannot afford to repay them when due if instead the consequence is that these borrowers are likely to find themselves re-borrowing repeatedly. Although industry stakeholders have argued that lenders making shortterm loans already take steps to assess “ability to repay” and will always do so out of economic 336 self-interest, the Bureau believes that this refers narrowly to whether the consumer will default up front on the loan, rather than whether the consumer has the capacity to repay the loan without having to re-borrow and while meeting other financial obligations and basic living expenses. The fact that lenders often do not perform additional underwriting when borrowers are rolling over a loan, or are returning to borrow again soon after repaying a prior loan, further shows that lenders do not see re-borrowing as a sign of borrowers’ financial distress or as an outcome to be avoided. Rather, repeated re-borrowing may be perceived as a preferred outcome for the lender or even as an outcome that is a crucial underpinning to the business model in this loan market. For the most part, commenters did not take issue with the tenets of this factual description of the typical underwriting process for such loans, though some lenders contended that they do not intentionally seek out potential customers who are likely to have to re-borrow multiple times. As noted, however, this approach is consistent with the basic business model for such loans as described above. Industry participants and trade associations did dispute one perceived implication of this discussion by asserting that long loan sequences, at least standing alone, cannot simply be assumed to be harmful or to demonstrate a consumer’s inability to repay these loans, as many factors may bear on those outcomes. This point is discussed further below. 4. Encouraging Long Loan Sequences In the proposal, the Bureau recounted its assessment of the market by noting that lenders attract borrowers in financial crisis, encourage them to think of the loans as a short-term solution, and fail to screen out those for whom the loans are likely to become a long-term debt cycle. After that, lenders then actively encourage borrowers to re-borrow and continue to be indebted rather than pay down or pay off their loans. Although storefront payday lenders typically take a 337 post-dated check, which could be presented in a manner timed to coincide with deposit of the borrower’s paycheck or government benefits, lenders usually encourage or even require borrowers to come back to the store to redeem the check and pay in cash.506 When the borrowers return, they are typically presented by lender employees with two salient options: repay the loan in full, or simply pay a fee to roll over the loan (where permitted under State law). If the consumer does not return, some lenders may reach out to the customer but ultimately the lender will proceed to attempt to collect by cashing the check. On a $300 loan at a typical charge of $15 per $100 borrowed, the cost to defer the due date for another 14 days until the next payday is $45, while repaying in full would cost $345, which may leave the borrower with insufficient remaining income to cover expenses over the ensuing month and therefore tends to prompt reborrowing. Requiring repayment in person gives staff at the stores the opportunity to frame for borrowers a choice between repaying in full or just paying the finance charge, which may be coupled with encouragement guiding them to choose the less immediately painful option of paying just the finance charge and rolling the loan over for another term. Based on its experience from supervising payday lenders over the past several years, the Bureau has observed 506 The Bureau believes from its experience in conducting examinations of storefront payday lenders and its outreach that cash repayments on payday and vehicle title loans are prevalent, even when borrowers provide post-dated checks or ACH authorizations for repayment. The Bureau has developed evidence from reviewing a number of payday lenders subject to supervisory examination in 2014 that the majority of them call each borrower a few days before payment is due to remind them to come to the store and pay the loan in cash. As an example, one storefront lender requires borrowers to come in to the store to repay. Its website states: “All payday loans must be repaid with either cash or money order. Upon payment, we will return your original check to you.” Others give borrowers “appointment” or “reminder” cards to return to make a payment in cash. In addition, vehicle title loans do not require a bank account as a condition of the loan, and borrowers without a checking account must return to storefront title locations to make payments. 338 that storefront employees are generally incentivized to maximize the store’s loan volume and the data suggest that re-borrowing is a crucial means of achieving this goal.507 As laid out in the proposal, the Bureau’s research shows that payday borrowers rarely reborrow a smaller amount than the initial loan. Doing so would effectively amortize their loans by reducing the principal amount owed over time, thereby reducing their costs and the expected length of their loan sequences. Rather than encouraging borrowers to make amortizing payments that would reduce their financial exposure over time, lenders encourage borrowers to pay the minimum amount and re-borrow the full amount of the earlier loan, thereby contributing to this outcome. In fact, as discussed in the proposal, some online payday loans automatically roll the loan over at the end of its term unless the consumer takes affirmative action in advance of the due date, such as notifying the lender in writing at least three days before the due date. As some industry commenters noted, single-payment vehicle title borrowers who take out multiple loans in a sequence are more likely than payday borrowers who taken out multiple loans in a sequence to reduce the loan amount from the beginning to end of that sequence. After excluding for single loan sequences for which this analysis is not applicable, 37 percent of single-payment vehicle title loan sequences have declining loan amounts compared to just 15 percent of payday loan sequences. This greater likelihood of declining loan amounts for single-payment vehicle title loans compared to payday loans may also be influenced by the larger median size of title loans, which is $694, as compared to the median size of payday loans, which is $350. However, this still indicates that a large majority of single payment vehicle title loan borrowers have constant 507 Most storefront lenders examined by the Bureau employ simple incentives that reward employees and store managers for loan volumes. 339 or increasing loan amounts over the course of a sequence. In addition, the Bureau’s analysis shows that those single payment vehicle title loan borrowers who do reduce their loan amounts during a sequence only do so for a median of about $200, which is less than a third of the median loan amount of about $700.508 This may reflect the effects of certain State laws regulating vehicle title loans that require some reduction in loan size across a loan sequence. Lenders also actively encourage borrowers who they know are struggling to repay their loans to roll over and continue to borrow. In the Bureau’s work over the past several years to monitor the operations and compliance of such lenders, including supervisory examinations and enforcement actions, the Bureau has found evidence that lenders maintain training materials that promote borrowing by struggling borrowers.509 In one enforcement action, the Bureau found that if a borrower did not repay in full or pay to roll over the loan on time, personnel would initiate collections. Store personnel or collectors would then offer new loans as a source of relief from the collections activities. This approach, which was understood to create a “cycle of debt,” was depicted graphically as part of the standard “loan process” in the company’s new hire training manual. The Bureau is aware of similar practices in the single-payment vehicle title lending market, where store employees offer borrowers additional cash during courtesy calls and when calling about past-due accounts, and company training materials instruct employees to “turn collections calls into sales calls” and encourage delinquent borrowers to refinance to avoid default and repossession of their vehicles. 508 See CFPB Single-Payment Vehicle Title Lending, at 18. Press Release, Bureau of Consumer Fin. Prot., “CFPB Takes Action Against Ace Cash Express for Pushing Payday Borrowers Into Cycle of Debt,” (July 10, 2014), available at http://www.consumerfinance.gov/newsroom/cfpb-takes-action-against-ace-cash-express-for-pushing-paydayborrowers-into-cycle-of-debt/. 509 340 It also appears that lenders do little to affirmatively promote the use of “off ramps” or other alternative repayment options, even when those are required by law to be made available to borrowers. Such alternative repayment plans could help at least some borrowers avoid lengthy cycles of re-borrowing. Lenders that belong to one of the two national trade associations for storefront payday lenders have agreed to offer an extended payment plan to borrowers, but only if the borrower makes a request at least one day prior to the date on which the loan is due.510 (The second national trade association reports that its members provide an extended payment plan option, but details on that option are not available.) In addition, about 18 States require payday lenders to offer repayment plans to borrowers who encounter difficulty in repaying payday loans. The usage rate of these repayment plans varies widely, but in all cases it is relatively low.511 One explanation for the low take-up rate on these repayment plans may be that 510 Cmty. Fin. Srvcs. Ass’n of Am., “CFSA Member Best Practices,” http://cfsaa.com/cfsa-member-bestpractices.aspx (last visited May 18, 2016); Cmty. Fin. Srvcs. Ass’n of Am., “What Is an Extended Payment Plan?,” http://cfsaa.com/cfsa-member-best-practices/what-is-an-extended-payment-plan.aspx (last visited May 18, 2016); Fin. Srvc. Ctrs. of Am., Inc., “FiSCA Best Practices,” http://www.fisca.org/Content/NavigationMenu/AboutFISCA/CodesofConduct/default.htm (last visited May 18, 2016). 511 Washington permits borrowers to request a no-cost installment repayment schedule prior to default. In 2014, 14 percent of payday loans were converted to installment loans. Wash. Dep’t of Fin. Insts., “2014 Payday Lending Report,” at 7 (2014), available at http://www.dfi.wa.gov/sites/default/files/reports/2014-payday-lending-report.pdf Illinois allows payday loan borrowers to request a repayment plan with 26 days after default. Between 2006 and 2013, the total number of repayment plans requested was less than 1 percent of the total number of loans made in the same period. Ill. Dep’t. of Fin. & Prof. Reg., “Illinois Trends Report All Consumer Loan Products Through December 2015,” at 19 (Apr. 14, 2016), available at http://www.idfpr.com/DFI/CCD/pdfs/IL_Trends_Report%202015-%20FINAL.pdf?ActID=1204&ChapterID=20). In Colorado, in 2009, 21 percent of eligible loans were converted to repayment plans before statutory changes repealed the repayment plan. State of Colorado, Dep’t of Law, Office of the Att’y Gen., “2009 Deferred Deposit Lenders Annual Report,” at 2 (2009) (hereinafter Colorado 2009 Deferred Deposit Lenders Annual Report), available at http://www.coloradoattorneygeneral.gov/sites/default/files/contentuploads/cp/ConsumerCreditUnit/UCCC/AnnualR eportComposites/2009_ddl_composite.pdf. In Utah, six percent of borrowers entered into an extended payment plan. G. Edward Leary, Comm’r of Fin. Insts. for the State of Utah to Hon. Gary R. Herbert, Governor, and the Legislature, (Report of the Commissioner of Financial Institutions for the Period July 1, 2013 to June 30, 2014), at 135, (Oct. 2, 2014) available at http://dfi.utah.gov/wp-content/uploads/sites/29/2015/06/Annual1.pdf. Florida law also requires lenders to extend the loan term on the outstanding loan by sixty days at no additional cost for 341 certain lenders disparage the plans or fail to promote their availability.512 By discouraging the use of repayment plans, lenders make it more likely that such consumers will instead re-borrow. The Bureau’s supervisory examinations uncovered evidence that one or more payday lenders train their employees not to mention repayment plans until after the employees have offered renewals, and then only to mention repayment plans if borrowers specifically ask about them. In general, most of the commenters did not take issue with this factual account of the mechanics or incentives that lead to a high incidence of rolling over such loans, and much of what they said tended to confirm it. In particular, industry commenters acknowledged that incentive programs for their employees based on net revenue are widespread in the industry. Such programs are not illegal, of course, but given the structure of these loans as described above, this suggests that employees are being incentivized to encourage re-borrowing and extended loan sequences by having borrowers roll their loans over repeatedly. Industry participants, trade associations, and some individual users of such loans did argue, however, about the implications of this analysis. One of their claims is that many consumers have an actual borrowing need that extends beyond the loan period permitted under State law, and thus repeated re-borrowing may be a means of synchronizing the consumer’s borrowers who indicate that they are unable to repay the loan when due and agree to attend credit counseling. Although 84 percent of loans were made to borrowers with seven or more loans in 2014, fewer than 0.5 percent of all loans were granted a cost-free term extension. See Brandon Coleman & Delvin Davis, “Perfect Storm: Payday Lenders Harm Florida Consumer Despite State Law,” at 4 (Ctr. for Responsible Lending, 2016), available at http://www.responsiblelending.org/sites/default/files/nodes/files/researchpublication/crl_perfect_storm_florida_mar2016_0.pdf 512 Colorado’s 2009 annual report of payday loan activity noted lenders’ self-reporting of practices to restrict borrowers from obtaining the number of loans needed to be eligible for a repayment plan or imposing cooling-off periods on borrowers who elect to take a repayment plan. Colorado 2009 Deferred Deposit Lenders Annual Report. This evidence was from Colorado under the state’s 2007 statute which required lenders to offer borrowers a no-cost repayment plan after the third balloon loan. The law was changed in 2010 to prohibit balloon loans, as discussed in part II. 342 borrowing needs to the specific contours of the loan product. In particular, they contended that re-borrowing may be beneficial to consumers as part of longer-term strategies around income smoothing or debt management, a point that is discussed further below. 5. Payment mechanisms and vehicle title The proposal noted that where lenders can collect payments through post-dated checks or ACH authorizations, or obtain security interests in borrowers’ vehicles, these mechanisms also can be used to encourage borrowers to re-borrow, as a way to avoid what otherwise could be negative consequences if the lender were to cash the check or repossess the vehicle. For example, consumers may feel significantly increased pressure to return to a storefront to roll over a payday or vehicle title loan that includes such features. They may do so rather than risk incurring new fees in connection with an attempt to deposit the consumer’s post-dated check, such as an overdraft or NSF fee from the bank and a returned-item fee from the lender if the check were to bounce or risk suffering the repossession of their vehicle. The pressure can be especially acute when the lender obtains security in the borrower’s vehicle. The proposal also noted that in cases where consumers do ultimately default on their loans, and these mechanisms are at last effectuated, they often magnify the total harm that consumers suffer from losing their access to essential transportation. Consumers often will have additional account and lender fees assessed against them, and some will end up having their bank accounts closed. When this occurs, they will have to bear the many attendant costs of becoming stranded outside the banking system, which include greater inconvenience, higher costs, reduced safety of their funds, and the loss of the other advantages of a standard banking relationship. These harms are very real for many consumers. For example, as discussed in more detail 343 below in Market Concerns—Payments, the Bureau’s research has found that 36 percent of borrowers who took out online payday or payday installment loans and had at least one failed payment during an eighteen-month period had their checking accounts closed by the bank by the end of that period, a rate that is four times greater than the closure rate for accounts that only had NSFs from non-payday transactions.513 For accounts with failed online payday loan transactions, account closures typically occur within 90 days of the last observed online payday loan transaction; in fact, 74 percent of account closures in these situations occur within 90 days of the first NSF return triggered by an online payday or payday installment lender.514 This suggests that the online loan played a role in the closure of the account, or that payment attempts failed because the account was already headed towards closure, or both.515 In general, the commenters did not challenge the Bureau’s factual account of how these payment mechanisms can lead to these collateral consequences that harm consumers. Industry commenters did disagree, however, with the premise that these harms were caused by the use of covered short-term loans. Some disagreed about the overall magnitude of these harms, stating that there is no evidence that covered short-term loans actually cause account closures or NSF fees, as stated in the proposal, and arguing that the Bureau overstated the extent to which consumers who default are subjected to NSF fees or fees resulting from bounced checks. But 513 CFPB Online Payday Loan Payments, at 12. CFPB Online Payday Loan Payments, at 23. 515 See also Complaint at 14, Baptiste v. J.P. Morgan Chase Bank, No. No. 12–04889 (E.D.N.Y. Oct. 1, 2012) (alleging plaintiff’s bank account was closed with a negative balance of $641.95, which consisted entirely of bank’s fees triggered by the payday lenders’ payment attempts); id. at 20-21 (alleging plaintiff’s bank account was closed with a negative balance of $1,784.50, which consisted entirely of banks fees triggered by the payday lender’s payment attempts and payments provided to the lenders through overdraft, and that plaintiff was subsequently turned down from opening a new checking account at another bank because of a negative ChexSystems report stemming from the account closure). 514 344 they did not present any convincing data to refute what the Bureau had observed from its own research and experience, and the assertion that online loans may have performed more poorly than storefront loans in these respects was not persuasive. Although the Bureau did not purport to find that the evidence in its data was determinative as to causation, the relationship between the consumer experience on such loans and the borrower outcomes was strongly reinforced by the data and logical as to the connection between them. c. Patterns of Lending and Extended Loan Sequences The Bureau’s proposal described how borrower characteristics, the circumstances of borrowing, the structure of the short-term loans, and the practices of the lenders together lead to dramatic negative outcomes for many payday and single-payment vehicle title borrowers. There is strong evidence that a meaningful share of borrowers who take out payday and single-payment vehicle title loans end up with very long sequences of loans, and the loans made to borrowers with these negative outcomes make up a majority of all the loans made by these lenders.516 Long loan sequences lead to very high total costs of borrowing. Each single-payment loan carries the same cost as the initial loan that the borrower took out. For a storefront borrower 516 In addition to the array of empirical evidence demonstrating this finding, industry stakeholders themselves have expressly or implicitly acknowledged the dependency of most storefront payday lenders’ business models on repeat borrowing. A June 20, 2013 letter to the Bureau from an attorney for a national trade association representing storefront payday lenders asserted that, “[i]n any large, mature payday loan portfolio, loans to repeat borrowers generally constitute between 70 and 90 percent of the portfolio, and for some lenders, even more,” and that “[t]he borrowers most likely to roll over a payday loan are, first, those who have already done so, and second, those who have had un-rolled-over loans in the immediately preceding loan period.” Letter from Hilary B. Miller to Bureau of Consumer Fin. Prot. (June 20, 2013), available at http://files.consumerfinance.gov/f/201308_cfpb_cfsa-informationquality-act-petition-to-CFPB.pdf. The letter asserted challenges under the Information Quality Act to the Bureau’s published White Paper (2013); see also Letter from Ron Borzekowski & B. Corey Stone, Jr., Bureau of Consumer Fin. Prot., to Hilary B. Miller (Aug. 19, 2013), available at https://encrypted.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=3&ved=0ahUKEwjEzu_EuMDWAh UGYiYKHY00ASEQFggvMAI&url=http%3A%2F%2Ffiles.consumerfinance.gov%2Ff%2F201308_cfpb_cfsaresponse.pdf&usg=AFQjCNF8PpFfXq_pt-lFOJtot1tRX_Or6A (Bureau’s response to the challenge). 345 who takes out the average-sized payday loan of $350 with a typical fee of $15 per $100, each reborrowing by rolling over the loan means paying additional fees of $52.50. After just three reborrowings, the borrower will have paid more than $150 simply to defer payment of the original principal amount by an additional period ranging from six weeks to three months. As noted in the proposal, the cost of re-borrowing for title borrowers is even more dramatic, given the higher price and larger size of those loans. The Bureau’s data indicates that the median loan size for single-payment vehicle title loans is $694. One study found that the most common rate charged on the typical 30-day title loan is $25 per $100 borrowed, which is a common State limit and equates to an APR of 300 percent.517 A typical instance of re-borrowing thus means that the consumer pays a fee of around $175. After just three re-borrowings, a consumer will typically have paid about $525 simply to defer payment of the original principal amount by three months. The proposal cited evidence for the prevalence of long sequences of payday and title loans, which comes from the Bureau’s own work, from analysis by independent researchers and analysts commissioned by industry, and from statements by industry stakeholders. The Bureau has published several analyses of storefront payday loan borrowing.518 Two of these have focused on the length of loan sequences that borrowers take out. In these publications, the Bureau defined a loan sequence as a series of loans where each loan was taken out either on the day the prior loan was repaid or within some number of days from when the loan was repaid. 517 Pew Charitable Trusts, “Auto Title Loans: Market Practices and Borrower Experiences,” at 11, 34 n.15 (2015), available at http://www.pewtrusts.org/~/media/assets/2015/03/autotitleloansreport.pdf. 518 See generally CFPB Data Point: Payday Lending; CFPB Payday Loans and Deposit Advance Products White Paper. 346 The Bureau’s 2014 Data Point used a 14-day window to define a sequence of loans. Those data have been further refined in the CFPB Report on Supplemental Findings and shows that when a borrower who is not currently in a loan sequence takes out a payday loan, borrowers wind up taking out at least four loans in a row before repaying 43 percent of the time, take out at least seven loans in a row before repaying 27 percent of the time, and take out at least 10 loans in a row before repaying 19 percent of the time.519 In the CFPB Report on Supplemental Findings, the Bureau re-analyzed the data using 30-day and 60-day definitions of sequences. The results are similar, although using longer windows leads to longer sequences of more loans. Using the 30-day definition of a sequence, 50 percent of new loan sequences contain at least four loans, 33 percent of sequences contain at least seven loans, and 24 percent of sequences contain at least 10 loans.520 Borrowers who take out a fourth loan in a sequence have a 66 percent likelihood of taking out at least three more loans, for a total sequence length of seven loans. And such borrowers have a 48 percent likelihood of taking out at least six more loans, for a total sequence length of 10 loans.521 These findings are mirrored in other analyses. During the SBREFA process, one participant submitted an analysis prepared by Charles River Associates (CRA) of loan data from 519 See CFPB Report on Supplemental Findings. CFPB Report on Supplemental Findings. In proposed § 1041.6 the Bureau is proposing some limitations on loans made within a sequence, and in proposed § 1041.2(a)(12), the Bureau is proposing to define a sequence to include loans made within 30 days of one another. The Bureau believes that this is a more appropriate definition of sequence than using either a shorter or longer time horizon for the reasons set forth in the section-by-section analyses of proposed §§ 1041.2(a)(12) and 1041.6. For these same reasons, the Bureau believes that the findings contained in the CFPB Report on Supplemental Findings and cited in text provide the most accurate quantification of the degree of harm resulting from cycles of indebtedness. 521 These figures are calculated simply by taking the share of sequences that are at least seven (or ten) loans long and diving by the share of sequences that are at least four loans long. 520 347 several small storefront payday lenders.522 Using a 60-day sequence as its definition, CRA found patterns of borrowing very similar to those that the Bureau had found. Compared to the Bureau’s results using a 60-day sequence definition, in the CRA analysis there were more loans where the borrower defaulted on the first loan or repaid without re-borrowing (roughly 44 percent versus 25 percent), and fewer loans that had 11 or more loans in the sequence, but otherwise the patterns were nearly identical.523 Similarly, in an analysis funded by an industry research organization, researchers found a mean sequence length, using a 30-day sequence definition, of nearly seven loans.524 This is slightly higher than the mean 30-day sequence length in the Bureau’s analysis (5.9 loans). Analysis of a multi-lender, multi-year dataset by a research group affiliated with a specialty consumer reporting agency found that over a period of approximately four years the average borrower had at least one sequence of nine loans; that 25 percent of borrowers had at least one loan sequence of 11 loans; and that 10 percent of borrowers had at least one loan sequence of 22 loans.525 Looking at these same borrowers for a period of 11 months—one month longer than the duration analyzed by the Bureau—the researchers found that on average the longest sequence these borrowers experienced over the 11 months was 5.3 loans, that 25 percent of borrowers had 522 Arthur Baines et al., “Economic Impact on Small Lenders of the Payday Lending Rules Under Consideration by the CFPB,” Charles River Associates, (2015), available at http://www.crai.com/publication/economic-impact-smalllenders-payday-lending-rules-under-consideration-cfpb. The CRA analysis states that it used the same methodology as the Bureau. 523 See generally CFPB Report on Supplemental Findings. 524 Marc Anthony Fusaro & Patricia J. Cirillo, “Do Payday Loans Trap Consumers in a Cycle of Debt?,” at 23 (2011), available at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=1960776. 525 nonPrime 101, “Report 7B: Searching for Harm in Storefront Payday Lending, A Critical Analysis of the CFPB’s ‘Debt Trap’ Data,” at 60 tbl. C-1 (2016), available at https://www.nonprime101.com/wpcontent/uploads/2016/02/Report-7-B-Searching-for-Harm-in-Storefront-Payday-Lending-nonPrime101.pdf. Sequences are defined based on the borrower pay period, with a loan taken out before a pay period has elapsed since the last loan was repaid being considered part of the same loan sequence. 348 a sequence of at least seven loans, and that 10 percent of borrowers had a sequence of at least 12 loans.526 This research group also identified a core of users with extremely persistent borrowing, and found that 30 percent of borrowers who took out a loan in the first month of the four-year period also took out a loan in the last month.527 The median time in debt for this group of extremely persistent borrowers was over 1,000 days, which is more than half of the four-year period. The median borrower in this group of extremely persistent borrowers had at least one loan sequence of 23 loans long or longer (which was nearly two years for borrowers who were paid monthly). Perhaps most notable, almost one out of ten members of this research group (nine percent) borrowed continuously for the entire four-year period.528 In the proposal, the Bureau also presented its analysis of single-payment vehicle title loans according to the same basic methodology.529 Using a 30-day definition of loan sequences, the Bureau found that short-term single-payment vehicle title loans had loan sequences that were 526 nonPrime 101, “Report 7B: Searching for Harm in Storefront Payday Lending, A Critical Analysis of the CFPB’s ‘Debt Trap’ Data,” at 60 tbl. C-1 (2016), available at https://www.nonprime101.com/wpcontent/uploads/2016/02/Report-7-B-Searching-for-Harm-in-Storefront-Payday-Lending-nonPrime101.pdf. The researchers were able to link borrowers across the five lenders in their dataset and include within a sequence loans taking out from different lenders. Following borrowers across multiple lenders did not materially increase the average length of the longest sequence but did increase the length of sequences for the top decile by one to two loans. Compare id. at tbl. C–2 with tbl. C–1. The author of the report focus on loan sequences where a borrower pays more in fees than the principal amount of the loan as sequences that cause consumer harm. The Bureau does not believe that this is the correct metric for determining whether a borrower has suffered harm. 527 nonprime 101, “Report 7C: A Balanced View of Storefront Payday Lending,” (2016), available at https://www.nonprime101.com/wp-content/uploads/2016/03/Report-7-C-A-Balanced-View-of-Storefront-PaydayBorrowing-Patterns-3https://www.nonprime101.com/wp-content/uploads/2016/03/Report-7-C-A-Balanced-View-ofStorefront-Payday-Borrowing-Patterns-3.28.pdf.28.pdf. 528 nonprime 101, “Report 7C: A Balanced View of Storefront Payday Lending,” at tbl. 2 (2016), available at https://www.nonprime101.com/wp-content/uploads/2016/03/Report-7-C-A-Balanced-View-of-Storefront-PaydayBorrowing-Patterns-3https://www.nonprime101.com/wp-content/uploads/2016/03/Report-7-C-A-Balanced-View-ofStorefront-Payday-Borrowing-Patterns-3.28.pdf.28.pdf. A study of borrowers in Florida claims that after the first year, over 20 percent of borrowers never use payday loans again and 50 percent of borrowers no longer use payday loans after two years. Floridians for Financial Choice, “The Florida Model: Baseless and Biased Attacks are Dangerously Wrong on Florida Payday Lending,” at 5 (2016), available at http://financialchoicefl.com/wpcontent/uploads/2016/05/FloridaModelReport.pdf. 529 See generally CFPB Single-Payment Vehicle Title Report. 349 similar to payday loans. More than half (56 percent) of these sequences contained at least four loans; 36 percent contained seven or more loans; and 23 percent had 10 or more loans. The Bureau’s analysis found that title borrowers were less likely than those using payday loans to repay a loan without re-borrowing or defaulting. Only 12 percent of single-payment vehicle title loan sequences consisted of a single loan that was repaid without subsequent re-borrowing, compared to 22 percent of payday loan sequences.530 Other sources on title lending are more limited than for payday lending, but are generally consistent. For instance, the Tennessee Department of Financial Institutions publishes a biennial report on 30-day single-payment vehicle title loans. The most recent report shows very similar results to those the Bureau found in its research, with 66 percent of borrowers taking out four or more loans in row, 40 percent taking out more than seven loans in a row, and 24 percent taking out more than 10 loans in a row.531 Some commenters noted data showing that vehicle title borrowers use re-borrowing to self-amortize their principal balance to a greater extent than payday borrowers do, which they suggested is evidence that title re-borrowing is not injurious. As noted previously, while it is true that more title borrowers in multi-loan sequences have declining loan balances than do payday borrowers in multi-loan sequences, this is likely the result of title loans starting out at much larger amounts. More salient is the fact that 63 percent of multi-loan sequences of title loans are for principal amounts that either remain unchanged or actually increase during the 530 CFPB Single-Payment Vehicle Title Lending, at 11; CFPB Report on Supplemental Findings, at 121. Letter from Greg Gonzales, Comm’r,Tennessee Dep’t of Fin. Insts., to Hon. Bill Haslam, Governor and Hon. Members of the 109th General Assembly, at 8 (Apr. 12, 2016) (Report on the Title Pledge Industry), available at http://www.tennessee.gov/assets/entities/tdfi/attachments/Title_Pledge_Report_2016_Final_Draft_Apr_6_2016.pdf. 531 350 sequence, and that even those title loan sequences that do have a decline in loan amount over time only have a median decline of about $200 from beginning to end of the sequence, which is less than one-third of the average total amount of these loans. And the default rate remains high even for amortizing multi-loan sequences of title loans, at 22 percent, which is slightly higher than the default rate for payday loans (20 percent), even though the latter amortize less often. All of this suggests that even if title borrowers can somewhat reduce the larger principal amount of their loans over time, it remains difficult to succeed in digging themselves out of the debts they have incurred with these loans. In addition to direct measures of the length of loan sequences, the cumulative number of loans that borrowers take out provides ample indirect evidence that they are often getting stuck in a long-term debt cycle. The Bureau has measured total borrowing by payday borrowers in two ways. In one study, the Bureau took a snapshot of borrowers in lenders’ portfolios at a point in time (measured as borrowing in a particular month) and tracked them for an additional 11 months (for a total of 12 months) to assess overall loan use. This study found that the median borrowing level was 10 loans over the course of a year, and more than half of the borrowers had loans outstanding for more than half of the year.532 In another study, the Bureau measured the total number of loans taken out by borrowers beginning new loan sequences. It found that these borrowers had lower total borrowing than borrowers who may have been mid-sequence at the beginning of the period, but the median number of loans for the new borrowers was six loans over a slightly shorter (11-month) period.533 Research by others finds similar results, with 532 533 CFPB Payday Loans and Deposit Advance Products White Paper, at 23. CFPB Data Point: Payday Lending, at 10-15. 351 average or median borrowing, using various data sources and various samples, of six to 13 loans per year.534 One commenter provided further data on the length of time consumers use payday loans, which gave more particulars about multi-year indebtedness in States with payday lending, such as South Carolina and Florida. The Florida data showed that over 40 percent of all consumers who took out one or more payday loans in 2012 continued to use the product three years later, and about a third of all consumers who took one or more payday loans in 2012 continued to use the product five years later. The South Carolina data provided similar information, but reported findings for consumers by borrowing intensity. It tended to show that those with the greatest intensity of borrowing were the least likely to end the borrowing relationship over a three-year period. Separately, a report on payday lending market trends by a specialty consumer reporting agency finds that over half of all loans are made to existing customers rather than consumers who have not used payday loans before.535 This report concludes that “even though new customers are critical, existing customers are the most productive.”536 The proposal also noted that, given differences in the regulatory context and the overall nature of the market, less information is available about online lending than storefront lending. 534 Paige Marta Skiba and Jeremy Tobacman, “Payday Loans, Uncertainty, and Discounting: Explaining Patterns of Borrowing, Repayment, and Default,” (Vand. L. and Econ., Research Paper No. 08–33, 2008). (finding an average of 5.5 loans per year for payday borrowers). A study of Oklahoma payday borrowing found an average of eight loans per year. Uriah King and Leslie Parrish, “Payday Loans, Inc.: Short on Credit, Long on Debt,” at 1 (Ctr. for Responsible Lending, 2011), available at http://www.responsiblelending.org/payday-lending/researchanalysis/payday-loan-inc.pdf; Michael A. Stegman, Payday Lending, 21 J. of Econ. Perspectives 169, at 176 (2007) (finding a median of 8–12 loans per year). 535 See generally Clarity Services, Inc., “2017 Subprime Lending Trends: Insights into Consumers & the Industry,” (2017), available at https://www.clarityservices.com/wp-content/uploads/2017/03/Subprime-Lending-Report-2017Clarity-Services-3.28.17.pdf. 536 Clarity Services, Inc., “2017 Subprime Lending Trends: Insights into Consumers & the Industry,” at 8 (2017), available at https://www.clarityservices.com/wp-content/uploads/2017/03/Subprime-Lending-Report-2017-ClarityServices-3.28.17.pdf. 352 Borrowers who take out payday loans online are likely to change lenders more frequently than storefront borrowers, so that absent comprehensive data that allows borrowing patterns to be tracked across all lenders, measuring the duration of loan sequences becomes much more challenging. The limited information that is available suggests that online borrowers take out fewer loans than storefront borrowers, but that borrowing is highly likely to be under-counted. A report commissioned by an online lender trade association, using data from three online lenders making single-payment payday loans, reported an average loan length of 20 days and an average of 73 days in debt per year.537 The report averages the medians of the three lenders’ data, which makes interpretation of these values difficult; still, these findings indicate that borrowers take out three to four loans per year at these lenders. Additional analysis is available based on the records of a specialty consumer reporting agency. The records show similar loans per borrower, 2.9, but over a multi-year period.538These loans, however, are not primarily single-payment payday loans. A small number are installment loans, while most are “hybrid” loans with a typical duration of roughly four pay cycles. In addition, this statistic likely understates usage because online lenders may not report all of the loans they make, and some may only report the first loan they make to a borrower. Borrowers may also be more likely to change lenders online and, as many lenders do not report to the 537 G. Michael Flores, “The State of Online Short-Term Lending, Second Annual Statistical Analysis Report,” Bretton-Woods, Inc., at 5 (Feb. 28, 2014), available at http://onlinelendersalliance.org/wpcontent/uploads/2015/07/2015-Bretton-Woods-Online-Lending-Study-FINAL.pdf (commissioned by the Online Lenders Alliance). 538 nonPrime 101, Report 7–A, “How Persistent in the Borrower-Lender Relationship in Payday Lending?”, at 6 tbl. 1 (2015) available at https://www.nonprime101.com/how-persistent-is-the-borrower-lender-relationship-in-paydaylending-2/. 353 specialty consumer reporting agency that provided the data for the analysis, when borrowers change lenders their subsequent loans often may not be in the data analyzed. Although many industry commenters disputed the significance of these findings, they offered little evidence that was inconsistent with the data presented by the Bureau. One commenter disputed the accuracy of the Bureau’s statement that 69 percent of payday loan sequences which end in default are multi-loan sequences and offered its own analysis based on its own customer data, which presented somewhat lower numbers but was largely consistent with the data presented by the Bureau. Still other commenters cited a petition that purported to show data errors relating to the Bureau’s White Paper on payday loans and deposit advance products that was used to draw conclusions about the prevalence of re-borrowing, which they argued was based on an unrepresentative sample weighted heavily toward repeat users. The Bureau has addressed this criticism previously, and explained that the methodology used in the White Paper, which took a snapshot of borrowers at the beginning of a twelve-month observation period and followed those borrowers over the ensuing eleven months, is an appropriate method of assessing borrowing intensity even though it is true that any such snapshot will be disproportionately composed of repeat borrowers because they comprise the bulk of payday lenders’ business. At the same time, the Bureau has conducted an alternative analysis which tracks the borrowing experience of fresh borrowers and it is that analysis on which the Bureau is principally relying here for covered short-term loans. Another study was cited to suggest that cost does not drive the cycle of debt because it found that borrowers who were given no-fee loans had re-borrowing rates that were comparable 354 to those who were given loans with normal fees.539 The upshot of this study, however, tended to show that the single-payment loan structure was instead a sufficient driver of the debt cycle, even without regard to the size of the fees that were charged. In fact, this study actually tends to refute the claim made elsewhere by industry commenters that the Bureau is trying to evade the statutory prohibition on imposing a usury cap by addressing price, since price alone does not seem to drive the cycle of debt that is a primary source of the harms resulting from these loans— rather, it is the single-payment loan structure that does so. Many industry participants and trade associations contended that, standing alone, multiple loan sequences cannot be presumed to be harmful to consumers. In particular, one trade association stated that where an income or expense shock cannot be resolved at once, reborrowing in extended loan sequences can be an effective longer-term strategy of income smoothing or debt management until the consumer’s financial situation improves. Thus reborrowing cannot be presumed to be necessarily irrational or harmful, depending on the circumstances. This commenter also cited studies that examined the credit scores of payday borrowers and reported finding better outcomes for longer-term borrowers than for those who are limited to shorter loan durations, and also that reported finding better outcomes for consumers in States with less restrictive payday lending laws than for those in States with more restrictive laws. These issues are important and they are discussed further in § 1041.4 below. A coalition of consumer groups was in agreement as a factual matter that many consumers of payday and single-payment vehicle title loans end up in extended loan sequences, 539 Marc A. Fusaro and Patricia J. Cirillo, “Do Payday Loans Trap Consumers in a Cycle of Debt?” (Ark. Tech U. & Cypress Research Group, 2011). 355 and many individual commenters described their own personal experiences and perspectives on this point. They observed that borrowers in these situations do in fact suffer many if not all of the harmful collateral consequences described in the proposal, which merely compound their existing financial difficulties and leave them worse off than they were before they took out such loans. Once again, however, putting aside the starkly different conclusions that commenters were drawing from the data, the basic accuracy of the data presented in the proposal on the patterns of lending and extended loan sequences was generally acknowledged. The arguments for and against the validity of their respective conclusions are considered further in the sectionby-section analysis for § 1041.4 below. d. Consumer Expectations and Understanding of Loan Sequences As discussed in the proposal, extended sequences of loans raise tangible concerns about the market for short-term loans. These concerns are exacerbated by the empirical evidence on consumer understanding of such loans. The available evidence indicates that many of the borrowers who take out long sequences of payday loans and single-payment vehicle title loans do not anticipate at the outset that they will end up experiencing those long sequences. Measuring consumers’ expectations about re-borrowing is inherently challenging. When answering survey questions about loan repayment, there is the risk that borrowers may conflate repaying an individual loan with completing an extended sequence of borrowing. Asking borrowers retrospective questions about their expectations at the time they started borrowing is likely to suffer from recall problems, as people have difficulty remembering what they expected at some time in the past. The recall problem is likely to be compounded by respondents tending to want to avoid admitting that they have made a mistake. Asking about expectations for future 356 borrowing may also be imperfect, as some consumers may not be thinking explicitly about how many times they will roll a loan over when taking out their first loan. Merely asking the question may cause people to think about it and focus on it more than they otherwise would have. Two studies discussed in the proposal have asked payday and vehicle title borrowers at the time they took out their loans about their expectations about re-borrowing, either the behavior of the average borrower or their own borrowing, and compared their responses with actual repayment behavior of the overall borrower population.540 One 2009 survey of payday borrowers found that over 40 percent of borrowers thought that the average borrower would have a loan outstanding for only two weeks, and another 25 percent said four weeks. Translating weeks into loans, the four-week response likely reflects borrowers who believe the average number of loans that a borrower will take out before repaying is either one loan or two loans, depending on how many respondents were paid bi-weekly as opposed to monthly. The report did not provide data on actual re-borrowing, but based on analysis performed by the Bureau and others, these results suggest that respondents were, on average, somewhat optimistic about reborrowing behavior.541 However, it is difficult to be certain that some survey respondents did not conflate the time during which the loans are outstanding with the contract term of individual loans. This may be so because the researchers asked borrowers, “What’s your best guess of how 540 Kathryn Fritzdixon et al., “Dude, Where’s My Car Title?: The Law Behavior and Economics of Title Lending Markets,” 2014 U. IL L. Rev. 1013 (2014); Marianne Bertrand and Adair Morse, “Information Disclosure, Cognitive Biases and Payday Borrowing,” 66 J. of Fin. 1865 (2011). 541 Marianne Bertrand and Adair Morse, “Information Disclosure, Cognitive Biases and Payday Borrowing,” 66 J. of Fin. 1865 (2011). Based on the Bureau’s analysis, approximately 50–55 percent of loan sequences, measured using a 14-day sequence definition, end after one or two loans, including sequences that end in default. See also CFPB Data Point: Payday Lending, at 11; CFPB Report on Supplemental Findings, at chapter 5. Using a relatively short re-borrowing period seems more likely to match how respondents interpret the survey question, but that is speculative. Translating loans to weeks is complicated by the fact that loan terms vary depending on borrowers’ pay frequency; four weeks is two loans for a borrower paid bi-weekly, but only one loan for a borrower paid monthly. 357 long it takes the average person to pay back in full a $300 payday loan?” Some borrowers may have interpreted this question to refer to the specific loan being taken out, rather than subsequent rollovers. People’s beliefs about their own re-borrowing behavior could also vary from their beliefs about average borrowing behavior by others. This study also did not specifically distinguish other borrowers from the subset of borrowers who end up in extended loan sequences. Another study discussed in the proposal was a study of single-payment vehicle title borrowers, where researchers surveyed borrowers about their expectations about how long it would take to repay the loan.542 The report did not have data on borrowing, but compared the responses with the distribution of repayment times reported by the Tennessee Department of Financial Institutions. The report found that the entire population of borrowers was slightly optimistic, on average, in their predictions.543 The two studies just described compared borrowers’ predictions of average borrowing with overall average borrowing levels, which is only informative about how accurate borrowers’ predictions are about the average. By contrast, a 2014 study by Professor Ronald Mann544, which was discussed in the proposal, did attempt to survey borrowers at the point at which they were borrowing. This survey asked them about their expectations for repaying their loans and compared their responses with their subsequent actual borrowing behavior, using loan records to 542 Kathryn Fritzdixon et al., “Dude, Where’s My Car Title?: The Law Behavior and Economics of Title Lending Markets,” 2014 U. IL L. Rev. 1013, at 1029–1030 (2014). 543 As noted above, the Bureau found that the re-borrowing patterns in data analyzed by the Bureau are very similar to those reported by the Tennessee Department of Financial Institutions. 544 Ronald Mann, “Assessing the Optimism of Payday Loan Borrowers,” 21 Supreme Court Econ. Rev. 105 (2013). 358 measure how accurate their predictions were. The results described in the report, combined with subsequent analysis that Professor Mann shared with Bureau staff, show the following:545 First, and most significant, many fewer borrowers expected to experience long sequences of loans than actually did experience long sequences. Focusing on the borrowers who ended up borrowing for more than 150 days, it is notable that none predicted they would be in debt for even 100 days.546 And of those who ended up borrowing for more than 100 days, only a very small fraction predicted that outcome.547 Indeed, the vast majority of those who borrowed for more than 100 days actually expected to borrow for less than 50 days.548 Borrowers who experienced long sequences of loans do not appear to have expected those long sequences when they made their initial borrowing decision; in fact they had not predicted that their sequences would be longer than the average predicted by borrowers overall. And while some borrowers did expect long sequences, those borrowers were more likely to err in their predictions; as Mann 545 The Bureau notes that Professor Mann draws different interpretations from his analysis than does the Bureau in certain instances, as explained below, and industry stakeholders, including SERs, have cited Mann’s study as support for their criticism of the Small Business Review Panel Outline. Much of this criticism is based on Professor Mann’s finding that “about 60 percent of borrowers accurately predict how long it will take them finally to repay their payday loans.” Ronald Mann, “Assessing the Optimism of Payday Loan Borrowers,” 21 Supreme Court Econ. Rev. 105, at 105 (2013). The Bureau notes, however, that this was largely driven by the fact that many borrowers predicted that they would not remain in debt for longer than one or two loans, and in fact this prediction was accurate for many such borrowers. But it did not address the much larger forecasting problems experienced by other borrowers, particularly those who ended up in extended loan sequences. 546 See Attachment to E-mail from Ronald Mann, Professor, Columbia Law School, to Jialan Wang & Jesse Leary, Bureau of Consumer Fin. Prot. (Sept. 24, 2013, 1:32 EDT), at 17. Correspondence between Bureau staff and Professor Mann was included as related material in the public docket supporting the proposed rule as published in the Federal Register on July 22, 2016. 547 See Attachment to E-mail from Ronald Mann, Professor, Columbia Law School, to Jialan Wang & Jesse Leary, Bureau of Consumer Fin. Prot. (Sept. 24, 2013, 1:32 EDT), at 17. 548 See Attachment to E-mail from Ronald Mann, Professor, Columbia Law School, to Jialan Wang & Jesse Leary, Bureau of Consumer Fin. Prot. (Sept. 24, 2013, 1:32 EDT), at 17. The same point can be made from another angle as well. Only 10 percent of borrowers expected to be in debt for more than 70 days (five two-week loans), and only 5 percent expected to be in debt for more than 110 days (roughly eight two- week loans), yet the actual numbers were substantially higher. See Ronald Mann, “Assessing the Optimism of Payday Loan Borrowers,” 21 Supreme Court Econ. Rev. 105, at 122 (2013) Indeed, approximately 12 percent of borrowers still remained in debt after 200 days (14 two-week loans). See comment letter submitted by Prof. Ronald Mann, at 2. 359 noted, “both the likelihood of unexpectedly late payment and the proportionate size of the error increase substantially with the length of the borrower’s prediction.”549 Second, Mann’s analysis suggests that past borrowing experience is not indicative of increased understanding of product use. In fact, those who had borrowed the most in the past did not do a better job of predicting their future use; they were actually more likely to underestimate how long it would take them to repay fully. As Mann noted in his paper, “heavy users of the product tend to be those that understand least what is likely to happen to them.”550 Finally, Mann’s research also indicated that about as many consumers underestimated how long they would need to re-borrow as those who overestimated it, which suggested they have difficulty predicting the extent to which they will need to re-borrow. In particular, the Bureau’s analysis of the data underlying Mann’s paper determined that there was not a correlation between borrowers’ predicted length of re-borrowing and their actual length of reborrowing.551 Professor Mann, in an email to the Bureau, confirmed that his data showed no significant relationship between the predicted number of days and the days to clearance.552 This point was reinforced in his survey results by the fact that fully 20 percent of the borrowers who 549 Ronald Mann, “Assessing the Optimism of Payday Loan Borrowers,” 21 Supreme Court Econ. Rev. 105, at 127 (2013) 550 See Ronald Mann, “Assessing the Optimism of Payday Loan Borrowers,” 21 Supreme Court Econ. Rev. 105, at 127 (2013). 551 Attachment to E-mail from Ronald Mann, Professor, Columbia Law School, to Jialan Wang & Jesse Leary, Bureau of Consumer Fin. Prot. (Sept. 24, 2013, 1:32 EDT), at 17. 552 E-mail from Ronald Mann, Professor, Columbia Law School, to Jialan Wang & Jesse Leary, Bureau of Consumer Fin. Prot. (Sept. 24, 2013, 1:32 EDT). 360 responded were not even able to offer any prediction at all about their expected duration of indebtedness.553 Professor Mann submitted a comment about his paper, which took issue with the Bureau’s analysis of its findings. He contended his research shows instead that most payday borrowers expected some repeated sequences of loans, most of them accurately predicted the length of the sequence that they would borrow, and they did not systematically err on the optimistic side. The Bureau acknowledges these findings, and does not believe they are inconsistent with the interpretation provided here. Mann also noted that the Bureau placed its main emphasis not on the entire universe of borrowers, but on the group of borrowers who continued borrowing over the period for which he had access to the loan data, where his research showed that many of those borrowers did not anticipate that they would end up in such extended loan sequences. He further acknowledged that “the absolute size of the errors is largest for those with the longest sequences.”554 He went on to state that this finding suggests “that the borrowers who have borrowed the most are those who are in the most dire financial distress, and consequently least able to predict their future liquidity.”555 He also noted that the errors of estimation these borrowers tend to make are unsystematic and do not consist either of regular underestimation or regular overestimation of their subsequent duration of borrowing.556 The discussion of these survey findings thus seems to reflect more of a difference in emphasis than a disagreement over the facts. Professor Mann’s interpretation appears most 553 Ronald Mann, “Assessing the Optimism of Payday Loan Borrowers,” 21 Supreme Court Econ. Rev. 105, at 121 (2013). 554 Prof. Ronald Mann comment letter, at 3. 555 Prof. Ronald Mann comment letter, at 3. 556 Prof. Ronald Mann comment letter, at 3. 361 applicable to those borrowers who remain in debt for a relatively short period, who constitute a majority of all borrowers, and who do not appear to systematically fail to appreciate what will happen to them when they re-borrow. The Bureau does not disagree with this point. Instead, it emphasizes the subset of borrowers who are its principal concern, which consists of those longer-term borrowers who find themselves in extended loan sequences and thereby experience the various harms that are associated with a longer cycle of indebtedness. For those borrowers, the picture is quite different, and their ability to estimate accurately what will happen to them when they take out a payday loan is more limited, as Mann noted in his paper and in the comment he submitted.557 For example, of the borrowers who remained in debt at least 140 days (10 biweekly loans), it appears that all (100 percent) underestimated their times in debt, with the average borrower in this group spending 119 more days in debt than anticipated (equivalent to 8.5 unanticipated rollovers). Of those borrowers who spent 90 or more days in debt (i.e., those most directly affected by the rule’s limits on re-borrowing under the § 1041.6), it appears that more than 95 percent underestimated their time in debt, spending an average of 92 more days in debt than anticipated (equivalent to 6.5 unanticipated rollovers). Additionally, a line of “best fit” provided by Professor Mann describing the relationship between a borrower’s expected time in debt and the actual time in debt experienced by that borrower shows effectively zero slope (indicating no correlation between a borrower’s expectations and outcomes). In other words, while many individuals appear to have anticipated short durations of use with reasonable accuracy (highlighted by Mann’s interpretation), virtually none properly anticipated long 557 Ronald Mann, “Assessing the Optimism of Payday Loan Borrowers,” 21 Supreme Court Econ. Rev. 105, at 127 (2013); Prof. Ronald Mann comment letter, at 2. 362 durations (which is the market failure described here).558 For further discussion on the Mann data, see the Section 1022(b)(2) Analysis in part VII below. Professor Mann’s comment also referred to two other surveys of payday borrowers that the Bureau discussed in its proposal. A trade association commissioned the two surveys, which suggest that consumers are able to predict their borrowing patterns.559 Both studies, as the Bureau had noted and as Professor Mann acknowledged, are less reliable in their design than the original Mann study because they focus only on borrowers who had successfully repaid a recent loan, which clearly would have biased the results of those surveys, because that approach would tend to under-sample borrowers who are in extended loan sequences. In addition, by entirely omitting borrowers whose loan sequences ended in default, these studies would have skewed the sample in other respects as well. At a minimum, the majority of borrowers who are light users of payday loans are likely to experience such loans very differently from the significant subset of borrowers (who are a minority of all borrowers, though the loans made to them constitute an overall majority of these loans) who find that they end up in extended loan sequences and suffer the various negative consequences of that predicament. 558 It should be noted that Professor Mann did not provide his data to the Bureau, either prior to the proposal, or in his comment in response to the proposal. In place of these data, the Bureau is relying on the charts and graphs he provided in his correspondence with and presentation to the Bureau. Amongst other things, these graphs depict the distribution of borrowers’ expectations and outcomes, but as they are scatterplots, counting the number of observations in areas of heavy mass (e.g., expecting no rollovers) is difficult. As such the analysis provided here may be somewhat imprecise. 559 Tarrance Group et al., “Borrower and Voter Views of Payday Loans,” Cmty. Fin. Servs. Ass’n of America (2016), available at http://www.tarrance.com/docs/CFSA-BorrowerandVoterSurvey-AnalysisF03.03.16.pdf; Harris Interactive, “Payday Loans and the Borrower Experience,” Cmty. Fin. Servs. Ass’n of America (2013), available at http://cfsaa.com/Portals/0/Harris_Interactive/CFSA_HarrisPoll_SurveyResults.pdf. The trade association and SERs have cited this survey in support of their critiques of the Bureau’s Small Business Review Panel Outline. 363 These surveys, which were very similar to each other, were conducted in 2013 and 2016 of storefront payday borrowers who had recently repaid a loan and had not taken another loan within a specified period of time. Of these borrowers, 94 to 96 percent reported that when they took out the loan they understood well or very well “how long it would take to completely repay the loan” and a similar percentage reported that they, in fact, were able to repay their loan in the amount of time they expected. These surveys suffer from the challenge of asking people to describe their expectations about borrowing at some time in the past, which may lead to recall problems, as described earlier. In light of the sampling bias discussed above and the challenge inherent in the survey design, the Bureau concludes that these studies do not undermine the evidence above indicating that especially those consumers who engage in long-term reborrowing through extended loan sequences are generally not able to predict accurately the number of times that they will need to re-borrow. As discussed in the proposal, several factors may contribute to consumers’ lack of understanding of the risk of re-borrowing that will result from loans that prove unaffordable. As explained above in the section on lender practices, there is a mismatch between how these products are marketed and described by industry and how they actually operate in practice. Although lenders present the loans as a temporary bridge option, only a minority of payday loans are repaid without any re-borrowing. These loans often produce lengthy cycles of rollovers or new loans taken out shortly after the prior loans are repaid. Not surprisingly, many borrowers (especially those who end up in extended loan sequences) are not able to tell when they take out the first loan how long their cycles will last and how much they will ultimately pay for the initial disbursement of cash. Even borrowers who believe they will be unable to repay the loan 364 immediately—and therefore expect some amount of re-borrowing—are generally unable to predict accurately how many times they will re-borrow and at what cost, unless they manage to repay the loan fairly quickly. And, as noted above, borrowers who end up re-borrowing many times are especially susceptible to inaccurate predictions. Moreover, as noted in the proposal, research suggests that financial distress can be one of the factors in borrowers’ decision-making. As discussed above, payday and single-payment vehicle title loan borrowers are often in financial distress at the time they take out the loans. Their long-term financial condition is typically very poor. For example, as described above, studies find that both storefront and online payday borrowers have little to no savings and very low credit scores, which is a sign of overall distressed financial condition. They may have credit cards but likely do not have unused credit, are often delinquent on one or more cards, and have often experienced multiple overdrafts and/or NSFs on their checking accounts.560 They typically have tried and failed to obtain other forms of credit before turning to a payday lender, or they otherwise may perceive that such other options would not be available to them and there is no time to comparison shop when facing an imminent liquidity crisis. Research has shown that when people are under pressure they tend to focus on the immediate problem they are confronting and discount other considerations, including the longerterm implications of their actions. Researchers sometimes refer to this phenomenon as “tunneling,” evoking the tunnel-vision decision-making that people may tend to engage in as 560 See Neil Bhutta et al., “Payday Loan Choices and Consequences,” at 15–16 (Apr. 2, 2014), available at http://www.calcfa.com/docs/PaydayLoanChoicesandConsequences.pdf; Neil Bhutta et al., “Payday Loan Choices and Consequences,” 47 J. of Money, Credit and Banking 223 (2015); CFPB Online Payday Loan Payments, at 3–4; Brian Baugh, “What Happens When Payday Borrowers Are Cut Off From Payday Lending? A Natural Experiment,) Payday Lending? A Natural Experiment,) (Ph.D. dissertation, Ohio State University, 2015), available at http://fisher.osu.edu/supplements/10/16174/Baugh.pdf. 365 they confront such situations. Consumers experiencing a financial crisis, as they often are when they are deciding whether or not to take out these kinds of loans, can be prime examples of this behavior.561 Even when consumers are not facing a crisis, research shows that they tend to underestimate their near-term expenditures562 and, when estimating how much financial “slack” they will have in the future, tend to discount even the expenditures they do expect to incur.563 Finally, regardless of their financial situation, research suggests that consumers may generally have unrealistic expectations about their future earnings, their future expenses, and their ability to save money to repay future obligations. Much research has documented that consumers in many contexts demonstrate optimism bias about future events and their own future performance. Without attempting to specify how frequently these considerations may affect individual borrower behavior, it is enough here to note that they are supported in the academic literature and are consistent with the observed behavior of those who use covered short-term loans.564 As discussed in the proposal, each of these behavioral biases is exacerbated when facing a financial crisis, and taken together they can contribute to affecting the decision-making of 561 See generally Sendhil Mullainathan & Eldar Shafir, “Scarcity: The New Science of Having Less and How It Defines Our Lives,” (Picador, 2014). 562 Johanna Peetz & Roger Buehler, “When Distance Pays Off: The Role of Construal Level in Spending,” Predictions, 48 J. of Experimental Soc. Psychol. 395 (2012); Johanna Peetz & Roger Buehler, “Is the A Budget Fallacy? The Role of Savings Goals in the Prediction of Personal Spending,” 34 Personality and Social Psychol. Bull. 1579 (2009); Gulden Ulkuman et al., “Will I Spend More in 12 Months or a Year? The Effects of Ease of Estimation and Confidence on Budget Estimates,” 35 J. of Consumer Research 245, at 249 (2008). 563 Jonathan Z. Berman et al., “Expense Neglect in Forecasting Personal Finances,” at 5–6 (2014) (forthcoming publication in J. of Marketing Res.), available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2542805. 564 The foundational works on optimism bias come from the behavioral economics literature on forecasting. See, e.g., Daniel Kahneman & Amos Tversky, “Intuitive Prediction: Biases and Corrective Procedures,” 12 TIMS Studies in Mgmt. Science 313 (1979); Roger Buehler et al., “Exploring the “Planning Fallacy”: Why People Underestimate their Task Completion Times,” 67 J. Personality & Soc. Psychol. 366 (1994); Roger Buehler et al., “Inside the Planning Fallacy: The Causes and Consequences of Optimistic Time Prediction, in Heuristics and Biases: The Psychology of Intuitive Judgment,” at 250–70 (Thomas Gilovich, Dale Griffin, & Daniel Kahneman eds., 2002). Nonetheless, it is worth noting that many of the same behaviors and outcomes can be derived from other economic models based on the premise that consumers in similar situations behave rationally in light of their circumstances. 366 consumers who are considering taking out a payday loan, a single-payment vehicle title loan, or some other covered short-term loan. The effect of these behavioral biases may cause consumers to fail to make an accurate assessment of the likely duration of indebtedness, and, consequently, the total costs they will pay as a result of taking out the loan. Tunneling also may cause consumers not to focus sufficiently on the future implications of taking out a loan. To the extent consumers do comprehend what will happen when the loan comes due—or when future loans come due in extended loan sequences—underestimation of future expenditures and optimism bias can cause them to misunderstand the likelihood of repeated re-borrowing. These effects could be attributable to their belief that they are more likely to be able to repay the loan without defaulting or re-borrowing than they actually are. And consumers who recognize at origination that they will have difficulty paying back the loan and that they may need to roll the loan over or re-borrow once or twice may still underestimate the likelihood that they will wind up rolling over or re-borrowing multiple times and the increasingly high costs of doing so. Regardless of the underlying explanation, the empirical evidence indicates that many borrowers who find themselves ending up in extended loan sequences did not expect that outcome—with their predictive abilities diminishing as the loan sequences become more extended. In this regard, it is notable that one survey found that payday and vehicle title borrowers were more likely to underestimate the cost and amount of time in debt than borrowers 367 of other products examined in the survey, including pawn loans, deposit advance products, and installment loans.565 The commenters on this discussion in the proposal expressed sharply divergent views. Some industry commenters stated their belief that consumers make rational decisions and many of them do expect to re-borrow when they take out covered short-term loans. Others noted that this argument fails to come to grips with the key problem that the Bureau has focused on in its analysis—known to economists as a “right tail” problem—which rests on the fact that a subset constituting a substantial population of payday borrowers are the ones who do not seem to expect but yet experience the most extreme negative outcomes with these loans. Other industry participants and trade associations criticized the Bureau for not conducting its own surveys of payday and title borrowers, and contended that such surveys would have shown that borrowers are generally well informed about their decisions to obtain such loans. And a large number of comments from individual users of these loans were in accord with these views, presenting their own experiences with such loans as positive and as having benefited their financial situations. Other industry commenters pointed out what they regarded as a low volume of consumer complaints about this product, which they viewed as inconsistent with the notion that many borrowers are surprised by experiencing unexpected negative outcomes with these loans. Yet it is equally plausible that those borrowers who find themselves in extended loan sequences may be embarrassed and therefore may be less likely to submit complaints about their situation. This is 565 Rob Levy & Joshua Sledge, “A Complex Portrait: An Examination of Small-Dollar Credit Consumers,” (Ctr. for Fin. Servs. Innovation, 2012), available at https://www.fdic.gov/news/conferences/consumersymposium/2012/A%20Complex%20Portrait.pdf. 368 consistent with survey results that show many confirmed borrowers nonetheless deny having taken out a payday loan.566 Borrowers may also blame themselves for having gotten themselves caught up in a cycle of debt authorized by State law, which may also explain why they would be unlikely to file a complaint with a government agency or a government official. In addition, the Bureau has noted previously that a relatively high proportion of debt collection complaints it receives are about payday loans—a much higher proportion, for example, than for mortgages or auto loans or student loans.567 From its consumer complaint data, the Bureau observed that from November 2013 through December 2016 more than 31,000 debt collection complaints cited payday loans as the underlying debt. More than 11 percent of the complaints that the Bureau has handled about debt collection stem directly from payday loans.568 And in any event, it is not at all clear that the Bureau receives a low number of consumer complaints about payday loans once they are normalized in comparison to other credit products. For example, in 2016, the Bureau received approximately 4,400 complaints in which consumers reported “payday loan” as the complaint product and about 26,600 complaints about credit cards.569 Yet there are only about 12 million payday loan borrowers annually, and approximately 156 million consumers have one or more credit cards.570 Therefore, by way of 566 Gregory Elliehausen and Edward C. Lawrence, “Payday Advance Credit in America: An Analysis of Customer Demand,” (Geo. U., McDonough Sch. of Bus., Monograph No. 35, 2001). 567 Bureau of Consumer Fin. Prot., “Monthly Complaint Report, Vol. 18,” (Dec. 2016), available at https://www.consumerfinance.gov/data-research/research-reports/monthly-complaint-report-vol-18/. 568 Bureau of Consumer Fin. Prot., “Monthly Complaint Report, Vol. 18,” at 12 (Dec. 2016), available at https://www.consumerfinance.gov/data-research/research-reports/monthly-complaint-report-vol-18/. 569 Bureau of Consumer Fin. Prot., “Consumer Response Annual Report, January 1–December 31, 2016,” at 27, 33 (Mar. 2017), available at https://www.consumerfinance.gov/documents/3368/201703_cfpb_Consumer-ResponseAnnual-Report-2016.PDF. 570 Bureau staff estimate based on finding that 63 percent of American adults hold an open credit card and Census population estimates. Bureau of Consumer Fin. Prot., “The Consumer Credit Card Market Report,” at 36 (Dec. 2015), available at http://files.consumerfinance.gov/f/201512_cfpb_report-the-consumer-credit-card-market.pdf; 369 comparison, for every 10,000 payday loan borrowers, the Bureau received about 3.7 complaints, while for every 10,000 credit cardholders, the Bureau received about 1.7 complaints. In addition, some faith leaders and faith groups of many denominations from around the country collected and submitted comments, which underscored the point that many borrowers may direct their personal complaints or dissatisfactions with their experiences elsewhere than to government officials. Indeed, some of the faith leaders who commented on the proposal mentioned their intentions or efforts to develop their own safer loan products in response to the crises related to them by such borrowers. Various commenters, including some academics such as Professor Mann whose views are discussed above, also cited research that they viewed as showing that such borrowers understand the nature of the product, including the fact that they may remain indebted beyond the initial term of the loan, with many able to predict accurately (within two weeks) how long it will take to repay their loan or loans. They cited various studies to make the point that consumers are in a better position to understand and act in their own interests than are policymakers who are more removed from the conditions of their daily lives. Some of these commenters were particularly critical of what they viewed as the erroneous assumptions and, even more broadly, the misguided U.S. Census Bureau, “Annual Estimates of Resident Population for Selected Age Groups by Sex for the United States, States, Counties, and Puerto Rico Commonwealth and Municipios: April 1, 2010 to July 1, 2016,” (June 2017), available at https://factfinder.census.gov/bkmk/table/1.0/en/PEP/2016/PEPAGESEX. Other estimates of the number of credit card holders have been higher, meaning that 1.7 complaints per 10,000 credit card holders would be a high estimate. The U.S. Census Bureau estimated there were 160 million credit card holders in 2012, and researchers at the Federal Reserve Bank of Boston estimated that 72.1 percent of U.S. consumers held at least one credit card in 2014. U.S. Census Bureau, “Statistical Abstract of the United States: 2012,” at 740 tbl.1188 (Aug. 2011), available at https://www.census.gov/library/publications/2011/compendia/statab/131ed.html; Claire Greene et al., “The 2014 Survey of Consumer Payment Choice: Summary Results,” at 18 (Fed. Reserve Bank of Boston, No. 16–3, 2016), available at https://www.bostonfed.org/-/media/Documents/researchdatareport/pdf/rdr1603.pdf. And as noted above in the text, additional complaints related to both payday loans and credit cards are submitted as debt collection complaints with “payday loan” or “credit card” listed as the type of debt. 370 general approach taken by behavioral economists. They argued that any such approach to policymaking is not well grounded and runs counter to their preferred view that consumer behavior instead is marked by rational expectations and clear insight into decision-making about financial choices. By contrast, many consumer groups and some researchers took a very different view. They tended to agree with the points presented in the proposal about how behavioral characteristics can undermine decision-making for borrowers of these loans, especially for those in financial distress. In their view, these factors can and often do lead to misjudgments by many consumers of the likelihood that they may find themselves caught up in extended loan sequences and experiencing many of the harmful collateral consequences that were described in the proposal. They suggested that both the research and the personal experiences of many borrowers suggest that this picture of a substantial number of consumers is generally accurate, especially for those consumers who find that they have ended up in extended loan sequences. As the Bureau had noted in the proposal, the patterns of behavior and outcomes in this market are broadly consistent with a number of cognitive biases that are described and documented in the academic literature on behavioral economics. Yet it is important to note that the Bureau’s intervention is motivated by the observed pattern of outcomes in the market, and not by any settled viewpoint on the varying theories about the underlying rationality of the decisions that may lead to them. That is, the Bureau does not and need not take a position here on the types of behavioral motivations that may drive the observed outcomes, for it is the outcomes themselves that are problematic, regardless of how economists may attempt to explain them. In fact, both the rational agent models generally favored by industry comments and the 371 more behavioral models favored by consumer groups and some researchers could very well lead to these same observed outcomes. The Bureau has weighed these conflicting comments and concludes that the discussion of these issues in the proposal remains generally accurate and is supported by considerable research and data on how payday and title loans operate in actual practice and how these loans are experienced by consumers. The data do seem to indicate that a significant group of consumers do not accurately predict the duration of their borrowing. This is particularly true, notably, for the subset of consumers who do in fact end up in extended loan sequences. These findings, and not any definitive judgment about the validity of behavioral economics or other theories of consumer behavior, provide the foundation on which this rule is based. Finally, though certain commenters have expressed concern that the Bureau had not heard sufficiently from individual users of these loans, the Bureau has now received and reviewed a high volume of individual comments that were submitted as part of this rulemaking process. e. Delinquency and Default The proposal also addressed the specific topics of delinquency and default on payday and single-payment vehicle title loans. In addition to the various harms caused by unanticipated loan sequences, the Bureau was concerned that many borrowers suffer other harms from unaffordable loans in the form of the collateral costs that come from being delinquent or defaulting on the loans. Many borrowers, when faced with unaffordable payments, will be late in making loan payments, and may ultimately cease making payments altogether and default on their loans.571 571 This discussion uses the term “default” to refer to borrowers who do not repay their loans. Precise definitions will vary across analyses, depending on specific circumstances and data availability. 372 They may take out multiple loans before defaulting, either because they are simply delaying the inevitable or because their financial situation deteriorates over time to the point where they become delinquent and eventually default rather than continuing to pay additional re-borrowing fees. For example, the evidence from the CFPB Report on Supplemental Findings shows that approximately two-thirds of payday loan sequences ending in default are multi-loan sequences in which the borrower has rolled over or re-borrowed at least once before defaulting. And nearly half of the consumers who experienced either a default or a 30-day delinquency already had monthly fees exceeding $60 before their first default or 30-day delinquency occurred. While the Bureau noted in the proposal that it is not aware of any data directly measuring the number of late payments across the industry, studies of what happens when payments are so late that the lenders deposit the consumers’ original post-dated checks suggest that late payment rates are relatively high. For example, one study of payday borrowers in Texas found that in 10 percent of all loans, the post-dated checks were deposited and bounced. 572 Looking at the borrower level, the study found that half of all borrowers had a check that was deposited and bounced over the course of the year following their first payday loan.573 An analysis of data collected in North Dakota showed a lower, but still high, rate of lenders depositing checks that later bounced or trying to collect loan payment via an ACH payment request that failed. It showed that 39 percent of new borrowers experienced a failed loan payment of this type within a year after their first payday loan, and 44 percent did so within the first two years after their first 572 Paige Marta Skiba and Jeremy Tobacman, “Payday Loans, Uncertainty, and Discounting: Explaining Patterns of Borrowing, Repayment, and Default,” at 33 tbl. 2 (Vand. L. and Econ., Research Paper No. 08–33, 2008). The study did not separately report the percentage of loans on which the checks that were deposited were paid. 573 These results are limited to borrowers paid on a bi-weekly schedule. 373 payday loan.574 In a public filing, one large storefront payday lender reported a lower rate (6.5 percent) of depositing checks, of which nearly two-thirds were returned for insufficient funds.575 In the Bureau’s analysis of ACH payments initiated by online payday and payday installment lenders, half of online borrowers had at least one overdraft or NSF transaction related to their loans over 18 months. These borrowers’ depository accounts incurred an average total of $185 in fees.576 As the Bureau noted in the proposal, bounced checks and failed ACH payments can be quite costly for borrowers. The median bank fee for an NSF transaction is $34.00, which is equivalent to the cost of a rollover on a $300 storefront loan.577 If the lender makes repeated attempts to collect using these methods, this leads to repeated fees being incurred by the borrower. The Bureau’s research indicates that when one attempt fails, online payday lenders make a second attempt to collect 75 percent of the time but are unsuccessful in 70 percent of those cases. The failure rate increases with each subsequent attempt.578 In addition to incurring NSF fees from a bank, in many cases when a check bounces the consumer can be charged a returned check fee by the lender. This means the borrower would be incurring duplicative and additional fees for the same failed transaction. In this connection, it 574 Susanna Montezemolo & Sarah Wolff, “Payday Mayday: Visible and Invisible Payday Defaults,” at 4 (Ctr. for Responsible Lending, 2015), available at http://www.responsiblelending.org/sites/default/files/nodes/files/researchpublication/finalpaydaymayday_defaults.pdf. 575 “For the years ended December 31, 2011 and 2010, we deposited customer checks or presented an Automated Clearing House (“ACH”) authorization for approximately 6.7 percent and 6.5 percent, respectively, of all the customer checks and ACHs we received and we were unable to collect approximately 63 percent and 64 percent, respectively, of these deposited customer checks or presented ACHs.” Advance America 2011 10–K. Borrowerlevel rates of deposited checks were not reported. 576 CFPB Online Payday Loan Payments, at 10–11. 577 CFPB Study of Overdraft Programs, at 52. 578 CFPB Online Payday Loan Payments, at 3–4; see generally Market Concerns—Payments. 374 should be noted that lender-imposed late fees are subject to certain restrictions in some but not all States.579 The proposal also noted that default can also be quite costly for borrowers. These costs vary with the type of loan and the channel through which the borrower took out the loan. As discussed above, default may come after a lender has already made repeated and expensive attempts to collect from the borrower’s deposit account, such that a borrower may ultimately find it necessary to close the account. In other instances, the borrower’s bank or credit union may close the account if the balance is driven negative and the borrower is unable for an extended period of time to return the balance to positive. And borrowers of single-payment vehicle title loans stand to suffer even greater harms from default, as it may lead to the repossession of their vehicle. In addition to the direct costs of the loss of an asset, the deprivation of their vehicle can seriously disrupt people’s lives and put at risk their ability to remain employed or to manage their ordinary affairs as a practical matter. Yet another consequence of these setbacks could be personal bankruptcy in some cases. Default rates on individual payday loans appear at first glance to be fairly low. This figure is three percent in the data the Bureau has analyzed, and the commenters are in accord about this figure.580 But because so many borrowers respond to the unaffordability of these loans by re-borrowing in sequences of loans rather than by defaulting immediately, a more meaningful 579 Most States limit returned item fees on payday loans to a single fee of $15–$40; $25 is the most common returned-item fee limit. Most States do not permit lenders to charge a late fee on a payday loan, although Delaware permits a late fee of five percent and several States’ laws are silent on the question of late fees. 580 Default here is defined as a loan not being repaid as of the end of the period covered by the data or 30 days after the maturity date of the loan, whichever was later. The default rate was slightly higher [four percent] for new loans that are not part of an existing loan sequence, which could reflect an intention by some borrowers to take out a loan and not repay, or the mechanical fact that borrowers with a high probability of defaulting for some other reason are less likely to have a long sequence of loans. 375 measure of default is the share of loan sequences that end in default. The Bureau’s data show that, using a 30-day definition of a loan sequence, fully 20 percent of loan sequences end in default. A recent report based on a multi-lender dataset showed similar results, with a three percent loan-level default rate and a 16 percent sequence-level default rate.581 Other researchers have found similarly high levels of default. One study of Texas borrowers found that 4.7 percent of loans were charged off, while 30 percent of borrowers had a loan charged off in their first year of borrowing. 582 Default rates on single-payment vehicle title loans are higher than those on storefront payday loans; in addition, initial single-payment vehicle title loans are more likely than storefront payday loans to result in a default. In the data analyzed by the Bureau, the default rate on all title loans is six percent, and the sequence-level default rate is 33 percent.583 Over half of all defaults occur in single-payment vehicle title loan sequences that consist of three or fewer loans. Nine percent of single-payment vehicle title loan sequences consist of single loans that end in default, compared to six percent of payday loan sequences.584 The Bureau’s research suggests that title lenders repossess a vehicle slightly more than half the time when a borrower defaults on a loan. In the data the Bureau has analyzed, three percent of all single-payment vehicle title loans lead to repossession, which represents approximately 50 581 nonprime101, “Report 3: Measure of Reduced Form Relationship between the Payment-Income Ratio and the Default Probability,” at 6 (2015), available at https://www.nonprime101.com/wp-content/uploads/2015/02/ClarityServices-Measure-of-Reduced-Form-Relationship-Final-21715rev.pdf. This analysis defines sequences based on the pay frequency of the borrower, so some loans that would be considered part of the same sequence using a 30-day definition are not considered part of the same sequence in this analysis. 582 Paige Marta Skiba and Jeremy Tobacman, “Payday Loans, Uncertainty, and Discounting: Explaining Patterns of Borrowing, Repayment, and Default,” at 33 tbl. 2 (Vand. L. and Econ., Research Paper No. 08–33, 2008). Again, these results are limited to borrowers paid bi-weekly. 583 CFPB Single-Payment Vehicle Title Lending, at 23. 584 CFPB Single-Payment Vehicle Title Lending, at 11; CFPB Report on Supplemental Findings, at 120. 376 percent of loans on which the borrower defaulted. At the sequence level, 20 percent of sequences end up with the borrower’s vehicle being repossessed. In other words, one in five borrowers is unable to escape their debt on these loans without losing their car or truck. Some industry and trade association commenters posited that the Bureau had overstated the default and repossession rates on vehicle title loans. Companies argued that the Bureau had erroneously stated a higher repossession rate than their own data showed, with one commenter estimating its own short-term title loan sequence repossession rate at 8.4 percent. Others contended that the Bureau’s repossession rates were much higher than those reported through other sources, such as regulator reports in States like Idaho and Texas. In arguing that the Bureau had overstated the default and repossession rates, one trade group also cited a study which had concluded that the rates were lower. The study relied on a handful of State regulator reports in addition to “industry sources.” Yet the difference seems to trace to the fact that default and repossession rates are typically reported at the loan level rather than the sequence level. The Bureau’s loan-level data is actually fairly similar to the figures cited by these commenters. But the Bureau believes that sequence level is a more appropriate indicator, since it captures experience at the level of the borrower. Put differently, sequence level more appropriately indicates outcomes for particular consumers, rather than for particular lenders; from this standpoint, a loan that is rolled over three times before defaulting should not be miscounted as three “successfully” repaid loans and one default. As noted previously, over 80 percent of single-payment vehicle title loans were re-borrowed on the same day as a previous loan was repaid. Regardless, to the extent any one company has lower repossession rates than the average, that fact does not put in question the averages that the Bureau used, because inevitably there will 377 be companies that are both above and below the average. The Bureau also notes that the study discussed above cited by a trade group, which relies on undefined “industry sources” and a handful of State regulator reports to criticize the Bureau’s data on default and repossession rates, relied on far less robust loan level data than the Bureau used to arrive at the figures it cited in the Bureau’s supplemental research report and in the proposal. One commenter noted that because the vehicles put up for collateral on these loans are usually old and heavily used, lenders often do not repossess the vehicle because it is not worth the trouble. This commenter also argued that the impact of repossession is not significant, based on a study indicating that less than 15 percent of consumers whose vehicles are repossessed would not find alternative means of transportation, which again is at odds with the information presented in other studies that have been cited.585 Another commenter asserted that the stress created by the threat of vehicle repossession is no worse than other stresses felt by consumers in financial difficulties, though it is difficult to know how much to credit this claim. The proposal further noted that borrowers of all types of covered loans are also likely to be subject to collection efforts, which can take aggressive forms. From its consumer complaint data, the Bureau observed that from November 2013 through December 2016 more than 31,000 debt collection complaints cited payday loans as the underlying debt. More than 11 percent of the complaints that the Bureau has handled about debt collection stem directly from payday loans.586 These collections efforts can include harmful and harassing conduct, such as repeated 585 Kathryn Fritzdixon et al., “Dude, Where’s My Car Title?: The Law Behavior and Economics of Title Lending Markets,” 2014 U. IL L. Rev. 1013, at 1038 (2014). 586 Bureau of Consumer Fin. Prot., “Monthly Complaint Report, Vol. 18,” at 12 (Dec. 2016), available at https://www.consumerfinance.gov/data-research/research-reports/monthly-complaint-report-vol-18/. 378 phone calls from collectors to the borrower’s home or place of work, the harassment of family and friends, and in-person visits to consumers’ homes and worksites. Some of this conduct, depending on the facts and circumstances, may be illegal. Aggressive calling to the borrower’s workplace can put at risk the borrower’s employment and jeopardize future earnings. Many of these practices can cause psychological distress and anxiety for borrowers who are already under the strain of financial pressure. In fact, the Bureau’s enforcement and supervisory examination processes have uncovered evidence of numerous illegal collection practices by payday lenders, including practices of the kinds just described. These have included: illegal third-party calls, illegal home visits for collection purposes, false threats to add new fees, false threats of legal action or referral to a nonexistent in-house “collections department,” and deceptive messages regarding non-existent “special promotions” to induce borrowers to return calls.587 In addition, lenders and trade associations contended that the Bureau had overstated the extent of harm, noting that they do not typically report nonpayment of these kinds of loans to consumer reporting agencies, which can interfere with the consumer’s access to credit, and that this lack of reporting would obviate any harm that the borrower would suffer on that front. Nonetheless, debt collectors can and do report unpaid debts to the consumer reporting companies even when the original creditors do not, and the aggressive collection tactics that the Bureau has identified with respect to unpaid payday loans through its investigations and numerous 587 See Bureau of Consumer Fin. Prot., “Supervisory Highlights,” at 17–19 (Spring 2014), available at http://files.consumerfinance.gov/f/201405_cfpb_supervisory-highlights-spring-2014.pdf. 379 enforcement actions suggest that this may be a common collateral consequence of default on these loans as well.588 The potential consequences of the loss of a vehicle depend on the transportation needs of the borrower’s household and the available transportation alternatives. According to two surveys of title loan borrowers, 15 percent of all borrowers report that they would have no way to get to work or school if they lost their vehicle to repossession.589 Using an 8 percent repossession rate, one industry commenter asserted that only about one percent of title loan borrowers would thus lose critical transportation, by multiplying 15 percent times 8 percent. However, the survey author specifically warns against doing this, noting that “a borrower whose car is repossessed probably has lower wealth and income than a borrower whose car is not repossessed, and is therefore probably more likely to lack another way of getting to work.”590 More than one-third (35 percent) of borrowers pledge the title to the only working vehicle in the household.591 Even those with a second vehicle or the ability to get rides from friends or take public transportation would presumably experience significant inconvenience or even hardship from the loss of a vehicle. This hardship goes beyond simply getting to work or school, and would as a practical 588 See, e.g., In the Matter of Money Tree, Inc., No. 2016–CFPB–0028; In the Matter of EZCORP, Inc., No. 2015– CFPB–0031; CFPB v. NDG Financial Corp., No. 15–05211 (S.D.N.Y. 2015); In the Matter of ACE Cash Express, Inc., No. 2014–CFPB–0008; In the Matter of Westlake Servs., LLC, No. 2015–CFPB–0026. The Bureau has also taken actions against debt collectors, some of which collect in part on small-dollar loans. See, e.g., CFPB v. MacKinnon, et al., No. 16–00880 (W.D.N.Y. 2016). 589 Kathryn Fritzdixon et al., “Dude, Where’s My Car Title?: The Law Behavior and Economics of Title Lending Markets,” 2014 U. IL L. Rev. 1013, 1029–1030 (2014); Pew Charitable Trusts, “Auto Title Loans: Market Practices and Borrowers’ Experiences,” at 14 (2015), available at http://www.pewtrusts.org/~/media/Assets/2015/03/AutoTitleLoansReport.pdf?la=en; 590 Kathryn Fritzdixon et al., “Dude, Where’s My Car Title?,” 2014 U. IL L. Rev. at 1038 n.137. 591 Pew Charitable Trusts, “Auto Title Loans: Market Practices and Borrowers’ Experiences,” (2015), available at http://www.pewtrusts.org/~/media/Assets/2015/03/AutoTitleLoansReport.pdf?la=en. 380 matter also adversely affect the borrower’s ability to conduct their ordinary household affairs, such as obtaining food or medicine or other necessary services. In the proposal, the Bureau noted that it analyzed online payday and payday installment lenders’ attempts to withdraw payments from borrowers’ deposit accounts, and found that six percent of payment attempts that were not preceded by a failed payment attempt themselves fail, incurring NSF fees.592 Another six percent avoid failure, despite a lack of sufficient available funds in the borrower’s account, but only because the borrower’s depository institution makes the payment as an overdraft, in which case the borrower was likely to be charged a fee that is generally similar in magnitude to an NSF fee. The Bureau could not determine default rates from these data. As noted in the proposal, when borrowers obtain a payday or title loan, they may fail to appreciate the extent of the risk that they will default and the costs associated with default. Although consumers may well understand the concept and possibility of default, in general, they are unlikely, when they are deciding whether to take out a loan, to be fully aware of the extent of the risk and severity of the harms that would occur if they were to default or what it would take to avoid default. They may be overly focused on their immediate needs relative to the longerterm picture. The lender’s marketing materials may have succeeded in convincing the consumer of the value of a loan to bridge financial shortfalls until their next paycheck. Some of the remedies a lender might invoke to address situations of nonpayment, such as repeatedly 592 The bank’s analysis includes both online and storefront lenders. Storefront lenders normally collect payment in cash and only deposit checks or submit ACH requests for payment when a borrower has failed to pay in person. These check presentments and ACH payment requests, where the borrower has already failed to make the agreedupon payment, have a higher rate of insufficient funds. 381 attempting to collect from a borrower’s checking account or using remotely created checks, may be unknown or quite unfamiliar to many borrowers. Realizing that these measures are even a possibility would depend on the borrower investigating what would happen in the case of an event they typically do not expect to occur, such as a default. Industry commenters contended that consumers tend to be highly knowledgeable about the nature, costs, and overall effects of payday and single-payment vehicle title loans. Yet they generally did not address the points raised here about the level of awareness and familiarity that these consumers would tend to have about the risks and costs of these other, more collateral consequences of delinquency and default. Consumer groups, by contrast, supported the view that these collateral consequences are part of the true overall cost of payday and title loans and that they are largely unforeseen by most consumers. f. Collateral Harms from Making Unaffordable Payments The proposal further elucidated other harms associated with payday and title loans, in addition to the harms associated with delinquency and default, by describing how borrowers who take out these loans may experience other financial hardships as a result of making payments on unaffordable loans. These harms may occur whether or not the borrower also experiences delinquency or default somewhere along the way, which means they could in many cases be experienced in addition to the harms otherwise experienced from these situations. These further harms can arise where the borrower feels compelled to prioritize payment on the loan and does not wish to re-borrow. This course of action may result in defaulting on other obligations or forgoing basic living expenses. If a lender has taken a security interest in the borrower’s vehicle, for example, and the borrower does not wish to re-borrow, then the borrower 382 is likely to feel compelled to prioritize payments on the title loan over other bills or crucial expenditures, because of the substantial leverage that the threat of repossession gives to the lender. The repayment mechanisms for other short-term loans can also cause borrowers to lose control over their own finances. If a lender has the ability to withdraw payment directly from a borrower’s checking account, the borrower may lose control over the order in which she would prefer her payments to be made and thus may be unable to choose to make essential expenditures before repaying the covered loan. This is especially likely to happen when the lender is able to time the withdrawal to align with the borrower’s payday or with the specific day when the borrower is scheduled to receive periodic income. Moreover, even if a title borrower does not have her vehicle repossessed, the threat of repossession in itself may cause tangible harm to borrowers. It may cause them to forgo other essential expenditures in order to make a payment they cannot afford in order to avoid repossession.593 And there may be psychological harm in addition to the stress associated with the possible loss of a vehicle. Lenders recognize that consumers often have a “pride of ownership” in their vehicle and, as discussed above, one or more lenders are willing to exceed their maximum loan amount guidelines by considering the vehicle’s sentimental or use value to the consumer when they are assessing the amount of funds they will lend. 593 As the D.C. Circuit observed of consumers loans secured by interests in household goods, “[c]onsumers threatened with the loss of their most basic possessions become desperate and peculiarly vulnerable to any suggested ‘ways out.’ As a result, ‘creditors are in a prime position to urge debtors to take steps which may worsen their financial circumstances.’ The consumer may default on other debts or agree to enter refinancing agreements which may reduce or defer monthly payments on a short-term basis but at the cost of increasing the consumer's total longterm debt obligation.” AFSA, 767 F.2d at 974 (1985) (internal citation omitted). 383 The Bureau noted in the proposal that it is not able to directly observe the harms that borrowers suffer from making unaffordable payments. But it stands to reason that when loans are made without regard to the consumer’s ability to repay and the lender secures the ability to debit a consumer’s account or repossess a vehicle, many borrowers are suffering harms from making unaffordable payments at certain times, and perhaps frequently. The commenters had vigorous reactions to this discussion in the proposal. On the effects that vehicle title borrowers feel based on their concern about losing their transportation, industry commenters argued that the Bureau had overstated its points. They emphasized that these loans are typically non-recourse loans in many States, which puts some specific limits on the harm experienced by borrowers. In the proposal, the Bureau had observed that this result would still expose the borrower to consider threat of harm if they end up losing their primary (and in many instances their sole) means of transportation to work and to manage their everyday affairs. Moreover, the Bureau notes these comments omit the issue of what harms exist in States where vehicle title loans are recourse. The Bureau notes the receipt of a comment letter from two consumer advocacy groups that discussed in detail the laws and lender practices in Arizona, where a robust vehicle title loan market exists. They wrote that in Arizona lenders are permitted to sue for deficiency balances after repossession; lenders can collect a “reasonable amount” for the cost of collection and court and attorneys’ fees related to repossession; and that as of 2015, nine of out of 10 largest title lenders still required borrowers to provide bank account access to get loans secured by vehicles.594 Furthermore, these commenters countered that borrowers often 594 The Bureau notes that an industry trade group argued that lenders generally do not pursue deficiencies even when it is legal to do so. However, in substantiating this assertion the trade group essentially cites itself as evidence for 384 can find other means of transportation, citing what they present as a supportive survey. Their interpretation of the data is not convincing, however, as even the authors of the survey cautioned against making simplistic calculations about factors and probabilities that are intertwined in the analysis, and which thus may considerably understate the incidence of hardship. One industry commenter pointed to a survey which showed that though a majority of title loan borrowers would prioritize their title loan payment over that of a credit card, very few of these borrowers would prioritize a title loan payment over rent, utilities, groceries, or other expenses. However, the author of this survey clearly states that because of an extremely small sample size, his findings are anecdotal and are not representative of borrowers either in the local area surveyed or nationally.595 The industry commenters further noted that as many as half of the title borrowers who default do so on their first payment, and they construed this occurrence as a strategic default which demonstrates that these borrowers did not confront any particular hardship by facing unaffordable payments that could cause them to lose their vehicle. Yet the notion that a borrower would make the conscious decision to employ this approach as a means of “selling” their vehicle, where they likely will receive a sharply reduced price for it and expose themselves to the other related risks discussed here, seems strained and implausible. That is especially the case insofar as doing so would needlessly incur the risks and costs of various potential penalty the proposition (i.e., the trade group cites language from a study that itself cites language from the same trade group’s website regarding best practices around repossession). 595 Jim Hawkins, “Credit on Wheels: The Law and Business of Auto-Title Lending,” 69 Wash. & Lee L. Rev. 535, 541 (2012). 385 fees, late fees, towing fees, and the like that could occur (depending on the provisions of State law) when lenders carry out a repossession of the vehicle. Industry and trade association commenters also suggested that the proposal is improperly paternalistic by attempting to substitute the judgment of the Bureau for the judgments made by individual consumers about how best to address the risks of collateral harms from making unaffordable payments. Difficult choices that consumers have to make about how to meet their obligations may be temporarily eased by the ability to access these loans and utilize the proceeds, at least for those consumers who do not end up experiencing the kinds of negative collateral consequences described above from delinquencies and defaults, and perhaps for some other borrowers as well. It also can substitute a new creditor with more limited recourse for an existing creditor with greater leverage, such as a landlord or a utility company. Although the addition of a payday or title loan obligation to the already-constrained mix of obligations can lead to the kind of budgeting distortions described by the proposal, it might instead lead to more immediate financial latitude to navigate those choices and avoid the impending harms of delinquency or default on other pre-existing obligations. This narrative was echoed by comments from a large number of individual users of such loans, who described the benefits they experienced by having access to the loan proceeds for immediate use while finding various ways to avert the negative collateral consequences described in the proposal. Consumer groups, on the other hand, strongly urged the view that payday and title loans often lead to harms similar to those described in the proposal for a significant set of borrowers. This position was buttressed by submissions from and about a sizeable number of individual borrowers as well, which included narratives describing extreme financial dislocations flowing 386 directly from harms cause by unaffordable payments. Although the proceeds of such loans do offer a temporary infusion of flexibility into the borrower’s financial situation, that brief breathing spell is generally followed almost immediately thereafter by having to confront similar financial conditions as before but now with the looming or actual threat of these harmful collateral consequences being felt as well. Again, in contrast to the viewpoint that repeated reborrowing may be consciously intended as a means of addressing financial shortfalls over a longer period of time, the consumer groups contended that extended loan sequences often reflect the inherent pressures of the initial financial need, now exacerbated by having to confront unaffordable payments on the new loan. And many individual users of such loans described their own negative experiences in ways that were consistent with the difficult situations and outcomes that can result from having to deal with unaffordable payments. Once again, the factual observations presented in the proposal on the kinds of collateral harms that can arise for payday and title borrowers who struggle to pursue potential alternatives to making unaffordable payments, as opposed to defaulting on these loans, were not seriously contested. The disagreement among the commenters was instead over the inferences to be drawn from these facts in context of other facts and potential benefits that they presented as bearing on their views of overall consumer welfare, and thus the broader conclusions to be drawn for purposes of deciding whether or not to support the proposed rule. Those contextual matters are important and will be discussed further in § 1041.4 below. g. Harms Remain Under Existing Regulatory Approaches As stated in the proposal, based on the Bureau’s analysis and outreach, the harms that it has observed from payday loans, single-payment vehicle title loans, and other covered short-term 387 loans persist in these markets despite existing regulatory frameworks. This formulation, of course, is something of a tautology, since if the harms the Bureau perceives to exist do in fact exist, they clearly do so despite the impact of existing regulatory frameworks that fail to prevent or mitigate them. Nonetheless, in the proposal the Bureau stated that existing regulatory frameworks in those States that have authorized payday and/or title lending still leave many consumers vulnerable to the specific harms discussed above relating to default, delinquency, reborrowing, and the collateral harms that result from attempting to avoid these other injuries by making unaffordable payments. Several different factors have complicated State efforts to effectively apply their regulatory frameworks to payday and title loans. For example, lenders may adjust their product offerings or their licensing status to avoid State law restrictions, such as by shifting from payday loans to vehicle title or installment loans or open-end credit or by obtaining licenses under State mortgage lending laws.596 As noted earlier, the State regulatory frameworks grew up around the pre-existing models of single-payment payday loans, but have evolved in certain respects over the past two decades. States also have faced challenges in applying their laws to certain online lenders, including lenders claiming Tribal affiliation or offshore lenders.597 As discussed above in part II, States have adopted a variety of different approaches for regulating short-term loans. For example, 15 States and the District of Columbia have interest 596 As discussed in part II, payday lenders in Ohio began making loans under the State’s Mortgage Loan Act and Credit Service Organization Act following the 2008 adoption of the Short-Term Lender Act, which limited interest and fees to 28 percent APR among other requirements, and a public referendum the same year voting down the reinstatement of the State’s Check-Cashing Lender Law, under which payday lenders had been making loans at higher rates. 597 A recent report summarizes these legal actions and advisory notices. See Diane Standaert & Brandon Coleman, Ending the Cycle of Evasion: Effective State and Federal Payday Lending Enforcement (2015), http://www.responsiblelending.org/payday-lending/research-analysis/crl_payday_enforcement_brief_nov2015.pdf. 388 rate caps or other restrictions that, in effect, prohibit payday lending and thereby limit access to this form of credit. Although consumers in these States may still be exposed to potential harms from short-term lending, such as online loans made by lenders that claim immunity from these State laws or from loans obtained in neighboring States, these provisions provide strong protections for consumers by substantially reducing their exposure to the harms they can incur from these loans. Again, as discussed above, these harms flow from the term and the singlepayment structure of these loans, which along with certain lender practices expose a substantial population of consumers to the risks and harms they experience, such as ending up in extended loan sequences. As explained in greater detail in part II above and in the section-by-section analysis for § 1041.5, the 35 States that permit payday loans in some form have taken a variety of different approaches to regulating such loans. Some States have restrictions on rollovers or other reborrowing. Among other things, these restrictions may include caps on the total number of permissible loans in a given period, or cooling-off periods between loans. Some States prohibit a lender from making a payday loan to a borrower who already has an outstanding payday loan. Some States have adopted provisions with minimum income requirements. For example, some States provide that a payday loan cannot exceed a percentage (most commonly 25 percent) of a consumer’s gross monthly income. Some State payday or title lending statutes require that the lender consider a consumer’s ability to repay the loan before making a loan, though none of them specifies what steps lenders must take to determine whether the consumer has the ability to repay a loan. Some States require that consumers have the opportunity to repay a short-term loan through an extended payment plan over the course of a longer period of time. And some 389 jurisdictions require lenders to provide specific disclosures in order to alert borrowers of potential risks. While the proposal noted that these provisions may have been designed to target some of the same or similar potential harms identified above, these provisions do not appear to have had a significant impact on reducing the incidences of re-borrowing and other harms that confront consumers of these loans. In particular, as discussed above, the Bureau’s primary concern about payday and title loans is that many consumers end up re-borrowing over and over again, turning what was ostensibly a short-term loan into a long-term cycle of debt with many negative collateral consequences. The Bureau’s analysis of borrowing patterns in different States that permit payday loans indicates that most States have very similar rates of re-borrowing, with about 80 percent of loans followed by another loan within 30 days, regardless of the terms of the specific restrictions that are in place.598 In particular, laws that prevent direct rollovers of payday loans, as well as laws that impose very short cooling-off periods between loans, such as Florida’s prohibition on same-day re-borrowing, have had very little impact on re-borrowing rates measured over periods longer than one day. The 30-day re-borrowing rate in all States that prohibit rollovers is 80 percent, and in Florida the rate is 89 percent. Some States, however, do stand out as having substantially lower re-borrowing rates than other States. These include Washington, which limits borrowers to no more than eight payday loans in a rolling 12-month period and has a 30-day re-borrowing rate of 63 percent, and Virginia, which imposes a minimum loan length of two pay periods and imposes a 45-day cooling-off period once a borrower has had five loans in a rolling six-month 598 CFPB Report on Supplemental Findings, at Chapter 4. 390 period, and has a 30-day re-borrowing rate of 61 percent (though title loans have claimed much greater market share in the wake of these restrictions on payday loans). Likewise, the Bureau explained in the proposal the basis for its view that disclosures would be insufficient to adequately reduce the harm that consumers suffer when lenders do not reasonably determine consumers’ ability to repay the loan according to its terms, which rested on two primary reasons. First, the Bureau noted that it is difficult for disclosures to address the underlying incentives in this market for lenders to encourage borrowers to re-borrow and take out extended loan sequences. As the Bureau discussed in the proposal, the prevailing business model in the short-term loan market involves lenders deriving a very high percentage of their revenues from extended loan sequences. The Bureau noted that while enhanced disclosures would provide more information to consumers, the Bureau believed that the single-payment structure of these loans, along with their high cost, would cause them to remain unaffordable for most consumers. The Bureau believed that, as a result, lenders would have no greater incentive to underwrite them more rigorously, and lenders would remain dependent on long-term loan sequences for revenues. Second, the Bureau noted in the proposal that empirical evidence suggests that disclosures may have only modest impacts on consumer borrowing patterns for short-term loans generally and negligible impacts on whether consumers re-borrow. The Bureau stated that evidence from a field trial of several disclosures designed specifically to warn of the risks and costs of re-borrowing showed that these disclosures had a marginal effect on the total volume of 391 payday borrowing.599 The Bureau observed that its analysis of similar disclosures implemented by the State of Texas showed a reduction in loan volume of 13 percent after the disclosure requirement went into effect, relative to the loan volume changes for the study period in comparison States, but further showed that the probability of re-borrowing on a payday loan declined by only approximately two percent once the disclosure was put in place.600 The Bureau noted that the analysis thus tended to confirm the fairly limited magnitude of the effects from the field trial. For these reasons, the Bureau stated in the proposal that evidence indicates the core harms to consumers in this credit market remain even after a disclosure regime is put in place. The Bureau also repeated its observation that consumers have a very high probability of winding up in a very long loan sequence once they have taken out only a few loans in a row.601 The Bureau noted that the contrast of the very high likelihood that a consumer will wind up in a longterm debt cycle after taking out only a few loans, with the nearly negligible impact of a disclosure on consumer re-borrowing patterns, provides further evidence of the insufficiency of disclosures to address what the Bureau perceives to be one of the core harms to consumers here. The issues around the sufficiency of disclosures, and whether it is likely that further disclosures would adequately address the harms that the Bureau has identified with payday and singlepayment vehicle-title loans, are discussed further in the section-by-section analysis for § 1041.5. 599 Marianne Bertrand & Adair Morse, “Information Disclosure, Cognitive Biases and Payday Borrowing and Payday Borrowing,” 66 J. Fin. 1865 (2011), available at http://onlinelibrary.wiley.com/doi/10.1111/j.15406261.2011.01698.x/full. 600 See CFPB Report on Supplemental Findings, at 73. 601 As discussed above in Market Concerns—Underwriting, a borrower who takes out a fourth loan in a sequence has a 66 percent likelihood of taking out at least three more loans, for a total sequence length of seven loans, and a 57 percent likelihood of taking out at least six more loans, for a total sequence length of 10 loans. 392 The proposal also discussed the SBREFA process, and noted that many participants urged the Bureau to reconsider the proposals under consideration and to consider deferring to existing regulation of these credit markets by the States or to adopt Federal regulations that are modeled on the laws or regulations of certain States. In the Small Business Review Panel Report, the Panel recommended that the Bureau continue to consider whether regulations in place at the State level are sufficient to address concerns about unaffordable loan payments. The Panel also recommended that the Bureau consider whether existing State laws and regulations could provide a model for elements of the Federal regulation. The SBA Office of Advocacy raised similar issues and suggested that the Bureau should defer to State payday lending laws. The Bureau has examined State laws closely in connection with its work on the final rule, as discussed in part II above, and the Bureau has taken guidance from what it has learned from its consideration of those differing frameworks. The Bureau has also consulted with various State regulators and State Attorneys General on these issues over the course of its original research on these topics, its formulation of the SBREFA framework, its conduct of the SBREFA process, its formulation of the proposal, and its work since to finalize the rule. The Bureau has also considered the comments that it has received from all parties, including State regulators and State Attorneys General and the SBA Office of Advocacy, which conflict with one another in a great many respects on the topics and arguments that have already been addressed in this discussion. All of this consideration of the State legal and regulatory frameworks has been applicable to the Bureau’s consideration of how it should approach its formulation of underwriting processes, restrictions on rollovers, and the use of cooling-off periods. 393 For those States with strong usury caps, of course, it bears repeating that the Bureau is not authorized to mirror those provisions because it is expressly barred by statute from imposing any usury cap on these loans. The Bureau has recognized this explicit restriction and carefully followed it in promulgating this rule, which does not prohibit any loan from being made based on the interest rate charged on the loan. Some of the industry commenters and trade associations have disputed this point in connection with certain provisions of the proposal, but have not explained how any loans are being prohibited on that basis. Industry participants and trade associations commented extensively on the fact that payday and single-payment vehicle title loans are subject to significant regulation already in the remaining States, even without any new regulation being proposed by the Bureau. They pointed to specific State frameworks as examples of how these products are regulated adequately and as providing access to credit without posing undue problems for borrowers. One trade association, for example, specifically cited Florida’s regulatory framework as allowing consumers in that State to use such products productively and successfully, while generating few complaints. Florida Congressional representatives made the same point. Other commenters, including some of the State Attorneys General, pointed to regulatory models in other States and drew similar conclusions. The Bureau has carefully assessed these State frameworks in considering how to respond to the comments received on the proposal and whether and how to modify the proposal in formulating the provisions of the final rule. For example, despite Colorado’s 2010 payday lending reforms that set a six-month minimum loan term for payday loans and reduced the annual percentage rates, concerns remain about sustained use and ability to repay the loans. A recent report based on State regulator data 394 noted that in 2015, the average borrower “took out 3.3 loans from the same lender over the course of the year, with a growing percentage of consumers (14.7 percent) being in debt every day for 12 consecutive months. Also one in four payday loans show signs of distress by delinquency or default.”602 In 2010, the State of Washington amended its payday lending law to limit borrowers to no more than eight loans in a rolling 12-month period, add an extended repayment plan that borrowers could take any time before default, and add a database that all lenders must use to report loans and check before new loans are made.603 The State regulator has issued yearly reports; with the most recent report being from calendar year 2015. There is no specific abilityto-repay requirement other than the loan amount cannot exceed 30 percent of the borrower’s gross monthly income or a maximum of $700 with no review of expenses.604 The 2015 report contains three highlights in particular. First, borrowing patterns continue to reflect a small number of borrowers responsible for most of the State’s payday loans. For payday loans originated in calendar year 2015, about one-quarter (25.38 percent) of borrowers took out about half (49.59 percent) of the total loans.605 Second, about a quarter of borrowers—26.62 percent— reached the eight-loan cap during 2015.606 Note that the cap is based on a rolling 12-month 602 Delvin Davis, Center for Responsible Lending, “Mile High Money: Payday Stores Target Colorado Communities of Color,” at 1 (Aug. 2017), available at http://www.responsiblelending.org/sites/default/files/nodes/files/researchpublication/crl-mile-high-money-aug2017.pdf. 603 All references are to the current Washington State Department of Financial Institutions report except where otherwise noted. Wash. State Dep’t. of Fin. Insts., “2015 Payday Lending Report,” at 4 (2015), available at http://www.dfi.wa.gov/sites/default/files/reports/2015-payday-lending-report.pdf. 604 Wash. State Dep’t. of Fin. Insts., “2015 Payday Lending Report,” at 4 (2015), available at http://www.dfi.wa.gov/sites/default/files/reports/2015-payday-lending-report.pdf. 605 Wash. State Dep’t. of Fin. Insts., “2015 Payday Lending Report,” at 8 (2015), available at http://www.dfi.wa.gov/sites/default/files/reports/2015-payday-lending-report.pdf; (Borrower Loan Frequency table). 606 Wash. State Dep’t. of Fin. Insts., “2015 Payday Lending Report,” at 7 (2015), available at http://www.dfi.wa.gov/sites/default/files/reports/2015-payday-lending-report.pdf. 395 period rather than a calendar year and some of these loans may have been originated in 2014. Also, note that some borrowers may be seeking loans online through unlicensed lenders that are not included in the State’s database. Third, 12.35 percent of loans were converted to an extended repayment plan (known as an installment loan plan) at some point in 2015. Borrowers may convert a payday loan to an installment loan plan at any time prior to default at no charge, with 90 to 180 days to repay based on the loan amount.607 Missouri’s regulatory framework offers an illustrative example that bears on the Bureau’s decision to require specific underwriting criteria under § 1041.5, a set of requirements that many commenters have criticized as unduly prescriptive and unnecessarily burdensome. By contrast, Missouri law requires small-dollar lenders to consider the borrower’s financial ability to reasonably repay under the terms of the loan contract, but does not specify how lenders may go about satisfying this requirement.608 The unsatisfactory result of this law, which fails to specify how lenders must satisfy the ability-to-repay requirement and thus allows lenders to exercise latitude in this regard, was starkly illustrated in a recent Missouri case that addressed the practical results of this framework. In a debt collection case, an appeals court judge concluded that the law, “which was designed for unsecured loans of five hundred dollars or less, has through the allowance of practically unlimited interest rates charged on the loans allowed the companies that provide these loans to use the court system to collect amounts from debtors far beyond anything that could be deemed consistent with the statute’s original purpose,” thus 607 Wash. State Dep’t. of Fin. Insts., “2015 Payday Lending Report,” at 4, 7 (2015), available at http://www.dfi.wa.gov/sites/default/files/reports/2015-payday-lending-report.pdf. 608 Mo. Rev. Stat. sec. 408.500(7). 396 providing “a clear example of predatory lending.”609 The judge then presented examples from the factual record in the case as follows: “Class member, D.W., took out a $100 loan from CSI. A judgment was entered against him for $705.18; the garnishment is still pending. So far, $3174.81 has been collected, and a balance of $4105.77 remains. Class member, S.S., took out an $80 loan from CSI. A judgment was entered against her for $2137.68; the garnishment is still pending. So far, $5346.41 has been collected, and a balance of $19,643.48 remains. Class member, C.R., took out a $155 loan from CSI. A judgment was entered against her for $1686.93; the garnishment is still pending. So far, $9566.15 has been collected, and a balance of $2162.07 remains.”610 The judge went on to provide four other similar examples, all of which were apparently deemed by the lender to satisfy its own conception of an ability-to-repay standard, even though the judge found that “the amount the lenders are collecting or are attempting to collect on these types of loans shocks the conscience” and were “beyond the ability of many debtors to ever pay off.”611 In addition, many industry participants and trade associations pointed out that payday and title lending are already regulated at the Federal level to some degree. They noted, for example, that the following laws already apply to such loans: the Truth in Lending Act, the Electronic Transfer Act, the Equal Credit Opportunity Act, the Fair Debt Collection Practices Act, and the 609 Hollins v. Capital Solutions Investments, Inc., 477 S.W.3d 19, 27 (Mo. Ct. App. 2015) (Dowd, J., concurring). Id. 611 Id. at 27-28. 610 397 Gramm-Leach-Bliley Act, among others. Many of these statutes have implementing regulations as well, thus adding to the pre-existing coverage of these loans under Federal law. And as recounted in part III, the Bureau has, in fact, engaged in extensive supervisory and enforcement activity with respect to payday loans and payday lenders under various provisions of the Federal consumer laws. These commenters often recognized that the Dodd-Frank Act confers separate and additional authority on the Bureau to promulgate rules to address unfair, deceptive, or abusive acts or practices, but contended that this authority should be used sparingly in light of the many statutes and regulations that already apply to such loans. In contrast, the consumer groups and other commenters drew a very different conclusion from their review of the State regulatory frameworks. They noted that more than 90 million people live in States without payday loans—where the State usury caps are viewed as effectively prohibiting such loans from being made as a practical matter—and observed that many of these consumers manage to deal with their cash shortfalls without resort to such loans. The same commenters contended that these consumers are not harmed by the absence of payday loans and instead are able to serve their financial needs through other credit products that are less risky. In their view, the alternatives available to potential borrowers in need of short-term credit are more diverse and more extensive than industry commenters have suggested. This market, as they describe it, is much broader than payday and single-payment vehicle title loans; it also comprises products such as credit cards, subprime credit cards, certain bank and credit union products, nonrecourse pawn loans, employer funds, charitable funds, and payment plans that are often made available by utilities and others. They also suggested that other non-credit strategies, such as debt counseling and credit counseling, can be productive alternatives to payday and title loans. 398 There was a wide gap in perspectives between these consumer groups and the industry commenters, who generally contended that these borrowers have a very limited range of alternative sources of credit available to them, other than payday and title loans, and are adversely affected when they lack access to these types of covered short-term loans. This disagreement is important and is considered further in the section-by-section analysis for § 1041.4 below in the discussions of unfairness and abusiveness. In sum, the Bureau has considered all of the comments received about the effects of the existing legal and regulatory frameworks, including the State frameworks, on the issues addressed in the proposal. Based on the Bureau’s analysis of the factual data as noted above, the regulatory frameworks in most States that allow and regulate payday, title, and other covered short-term loans do not appear to have had a significant impact on reducing the amounts of default, delinquency, re-borrowing, and the other collateral harms from making unaffordable payments that confront consumers of these loans. Nor have other existing regulatory frameworks had a significant impact in mitigating those harms to consumers. For these and the other reasons discussed above, the Bureau concludes that federal intervention in these markets is warranted at this time. Longer-Term Balloon-Payment Loans As stated in the proposal, some longer-term payday installment loans and vehicle title loans are structured either to be repaid in a single lump-sum payment or to require a large balloon payment, often as a final payment of all principal due following a series of smaller interest-only payments. Unsurprisingly, many consumers find making such a payment as challenging as making the single payment under a traditional, two-week payday loan, and such 399 loans frequently result in default or re-borrowing. The Bureau concludes that consumers are likely to be adversely affected by the practice of making these loans without reasonably assessing the borrower’s ability to repay the loan while paying for basic living expenses and other major financial obligations. And while there does not appear to be a large market of longer-term balloon-payment loans today, the Bureau is concerned that the market for such loans might grow if it only regulated the underwriting of covered shortterm loans. Based on the evolution in small-dollar loan markets after the Military Lending Act was enacted and the initial regulations implementing the MLA were adopted, the Bureau is concerned that lenders would gravitate toward making non-underwritten balloon-payment loans that slightly exceed the time limits in the definition for covered short-term loans, resulting in similar risks and harms to consumers from default, delinquency, re-borrowing, and the collateral consequences of forgoing basic living expenses or major financial obligations to avoid default. The Bureau received comments specifically on covered longer-term loans involving balloon payments. Several industry commenters stated that the Bureau’s concerns about reborrowing for covered longer-term loans should have focused primarily on loans with balloon payments, and argued that any restrictions should thus be limited to balloon-payment loans. The Bureau agrees with these commenters that the re-borrowing concerns with these loans are similar to the Bureau’s concerns regarding covered short-term loans, and highlight similar problems from making covered longer-term balloon-payment loans without reasonably assessing the borrower’s ability to repay. The thrust of these industry comments thus has tended to reinforce the judgment the Bureau has now made to address the underwriting of covered longer-term 400 balloon-payment loans in this rule.612 As discussed more fully in the section-by-section analysis of §§ 1041.2(a)(7) and 1041.3(b)(2) of the final rule, the Bureau had proposed to define a covered longer-term balloonpayment loan to mean a covered longer-term loan that, in essence, is repayable either in a single lump-sum payment or requires at least one payment that is more than twice as large as any other payment.613 After consideration of comments received concerning whether to maintain the proposal’s approach to limiting coverage of such balloon-payment structures to those products that exceed a rate threshold and involved the taking of a leveraged payment mechanism or vehicle security, the Bureau has decided to adopt a more expansive definition that includes all such payment structures regardless of price or other factors, unless they are specifically excluded or exempted under § 1041.3 of the final rule. Because relatively few covered longer-term balloon-payment loans appear in the market today, the Bureau is supplementing its analysis in this section with relevant information it has on related types of covered longer-term loans—such as hybrid payday loans, payday installment 612 The Bureau acknowledges that its determination to address the underwriting of all covered longer-term balloonpayment loans in the final rule does represent an expansion of coverage over the proposal in certain respects, which are that it would cover all such loans regardless of their cost, and regardless of whether the lender obtained a leveraged payment mechanism or vehicle security. Given that the prevalence of these kinds of loans with a balloonpayment structure is limited, however, the Bureau finds from its experience and analysis of these loan markets that the incidence of low-cost longer-term balloon-payment loans (or high-cost longer-term balloon-payment loans that do not have a leveraged payment mechanism or vehicle security) is relatively insignificant. 613 To be precise, the term “covered longer-term balloon-payment loan” is defined in § 1041.2(a)(7) of the final rule to mean a loan described in § 1041.2(b)(2) of the final rule, which is a covered loan that is not a covered short-term loan and: “(i) for closed-end credit, the consumer is required to repay the entire balance of the loan in a single payment more than 45 days after consummation, is required to repay substantially the entire amount of any advance in a single payment more than 45 days after the advance, or is required to pay at least one payment that is more than twice as large as any other payment(s); or (ii) for open-end credit, the consumer is required to repay substantially the entire amount of any advance at the end of a payment billing cycle that exceeds 45 days, or the credit plan is structured such that paying the required minimum payments may not fully amortize the outstanding balance by a specified date or time, and the amount of the final payment to repay the outstanding balance at such time could be more than twice the amount of other minimum payments under the plan.” Id. 401 loans, and vehicle title installment loans. Although these types of loans would not necessarily involve balloon payments per se, the Bureau finds no reason to expect that matters such as borrower characteristics and circumstances of borrowing are likely to differ substantially as between borrowers of longer-term title loans generally, for example, and borrowers of such loans with a balloon-payment structure. The Bureau concludes as follows: • Lower-income, lower-savings consumers in financial difficulty. While there is less research available about the consumers who use these products as compared to covered shortterm loan products, available information suggests that consumers who use hybrid payday, payday installment, and vehicle title installment loans also tend to come frequently from loweror moderate-income households, have little savings or available credit, and have been turned away from other credit products. Their reasons for borrowing and use of loan proceeds are also generally consistent with those of short-term borrowers. • Ability-to-collect business models. Lenders of most covered longer-term loans have built their business model on their ability to collect, rather than the consumers’ ability to repay the loans. Specifically, these lenders generally screen for fraud risk but do not consider consumers’ expenses to determine whether a loan is tailored to what the consumers can actually afford. They tend to rely heavily on pricing structures and on leverage over the consumer’s bank account or vehicle title to protect their own interests, even when the loans prove unaffordable for consumers. Lenders may continue receiving payments even when the consumer is left unable to meet her basic living expenses or major financial obligations. Again, though this tends to be the case for borrowers of covered longer-term loans, it is even more likely to be true of such borrowers if their loans have a balloon-payment structure. 402 • Very high default rates. Defaults are a concern with covered longer-term loans generally, and especially so if those loans reflect a balloon-payment structure. In data from one lender that the Bureau analyzed, about 60 percent of balloon-payment installment loans result in default or refinancing. In general, borrowers experienced very high levels of delinquency and default—in some cases the default rate was over 50 percent at the loan sequence level. Prior to reaching the point of default, borrowers can be exposed to a variety of harms whose likelihood and magnitude are substantially increased because of leveraged payment mechanisms or vehicle security relative to similar loans without these features. • Re-borrowing. The combination of leveraged payment mechanism or vehicle security with an unaffordable balloon payment can compel consumers to re-borrow. They will often have to engage in costly re-borrowing when they are unable to repay the entire loan all at once and extraction of the unaffordable loan payment would leave them unable to cover basic living expenses or major financial obligations. • Consumers do not understand the risks. The Bureau concludes that borrowers do not fully understand or anticipate the consequences that are likely to occur when they take out covered longer-term balloon-payment loans, including both the high likelihood of default and the degree of collateral damage that can occur in connection with unaffordable loans. a. Borrower Characteristics and Circumstances of Borrowing Stand-alone data specifically about payday installment and vehicle title installment borrowers is less robust than for borrowers of covered short-term loans, as discussed above. Yet a number of sources provide combined data for both categories. Both the unique and combined sources suggest that borrowers in these markets generally have low-to-moderate incomes and 403 poor credit histories. Their reasons for borrowing and use of loan proceeds are also generally consistent with those of covered short-term borrowers. 1. Borrower Characteristics As described above, typical payday borrowers have low average incomes ($25,000 to $30,000), poor credit histories, and have often repeatedly sought credit in the months leading up to taking out a payday loan.614 Given the overlap in the set of firms offering these loans, the similar pricing of the products, and certain similarities in the structure of the products (e.g., the high cost and the synchronization of payment due dates with borrower paydays or next deposits of income), the Bureau finds that the characteristics and circumstances of payday installment borrowers are likely to be very similar to those of short-term payday borrowers. To the extent data is available limited to payday installment borrowers, the data confirms this view. For example, from a study of over one million high-cost loans made by four payday installment lenders, both storefront and online, median borrower gross annual income was reported to be $35,057.615 Similarly, administrative data from Colorado and Illinois indicate that 60 percent of the payday installment borrowers in those States have income of $30,000 or below. And a study of online payday installment borrowers, using data from a specialty consumer reporting agency, found a median income of $30,000 and an average VantageScore of 614 Fed. Deposit Ins. Corp., “2013 FDIC National Survey of Unbanked and Underbanked Households” at 15–17 (Oct. 2014), available at https://www.fdic.gov/https://www.fdic.gov/householdsurvey/2013/householdsurvey/2013/. See also Gregory Elliehausen, “An Analysis of Consumers’ Use of Payday Loans,” at 27 (Geo. Wash. Sch. of Bus., Monograph No. 41, 2009), available at https://www.researchgate.net/publication/237554300_AN_ANALYSIS_OF_CONSUMERS%27_USE_OF_PAYDAY _LOANS (61percent of borrowers have household income under $40,000); Jonathan Zinman, “Restricting Consumer Credit Access: Household Survey Evidence on Effects Around the Oregon Rate Cap,” (Dartmouth College, 2008), available at http://www.dartmouth.edu/~jzinman/Papers/Zinman_RestrictingAccess_oct08.pdf. 615 Howard Beales & Anand Goel, “Small Dollar Installment Loans: An Empirical Analysis,” at 12 tbl. 1 (Geo. Wash. Sch. of Bus., 2015), available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2581667. 404 523; each of these was essentially identical as between the levels for storefront payday borrowers and for online payday borrowers.616 The information about vehicle title borrowers that the Bureau has reviewed does not distinguish between single-payment and installment vehicle title borrowers. For the same reasons that the Bureau concludes that the demographic data with respect to short-term payday borrowers can be extrapolated to payday installment borrowers, the Bureau also finds that the demographic data is likely to be similar as between short-term vehicle title borrowers and vehicle title installment borrowers. As discussed above, vehicle-title borrowers across all categories tend to be low-income or moderate-income, with 56 percent having reported incomes below $30,000, and are disproportionately racial and ethnic minorities or members of femaleheaded households.617 2. Circumstances of Borrowing Again, less data is available that focuses specifically on the circumstances of borrowing for users of payday installment and vehicle title installment loans than is available for short-term loans, and the data must be approached with some caution, since studies that seek to examine why consumers took out liquidity loans or for what purpose face a number of challenges. For example, any survey that asks about past behavior or events runs the risk of recall errors, and the fact that money is fungible makes this question even more complicated. For example, a 616 nonPrime101, “Report 8: Can Storefront Payday Borrowers Become Installment Loan Borrowers?,” at 5, 7 (2015), available at https://www.nonprime101.com/blog/can-storefront-payday-borrowers-become-installmentloan-borrowers/. 617 Fed. Deposit Ins. Corp., “2013 FDIC National Survey of Unbanked and Underbanked Households: Appendices,” at appendix. D–12a (Oct. 2014), available at https://www.fdic.gov/householdsurvey/2013/2013appendix.pdf.; Kathryn Fritzdixon et al., “Dude, Where’s My Car Title?: The Law Behavior and Economics of Title Lending Markets,” 2014 U. IL L. Rev. 1013, at 1029–1030 (2014). 405 consumer who has an unexpected expense may not feel the full effect until weeks later, depending on the timing of the unexpected expense relative to other expenses and the receipt of income. In that circumstance, a borrower may say that she took out the loan because of an emergency or may say instead that the loan was taken out to cover regular expenses. A 2012 survey of over 1,100 users of alternative small-dollar credit products asked borrowers separately about what precipitated the loan and what they used the loan proceeds for.618 Responses were reported for “very short term” and “short term” credit, with “short term” referring to non-bank installment loans and vehicle title loans.619 The most common reason borrowers gave for taking out “short term” credit (approximately 36 percent of respondents) was “I had a bill for an unexpected expense (e.g., medical emergency, car broke down).” About 23 percent of respondents said “I had a payment due before my paycheck arrived,” which the authors of the report on the survey results interpret as a mismatch in the timing of income and expenses, and a similar number said their general living expenses were consistently more than their income. The use of funds most commonly identified was to pay for routine expenses, with nearly 30 percent reporting “pay utility bills” and about 20 percent reporting “general living expenses,” but about 25 percent said the use of the money was “car-related,” either purchase or repair. In contrast, participants who took out “very short term” products such as payday and deposit advance products were somewhat more likely to cite “I had a bill or payment due before 618 Rob Levy & Joshua Sledge, “A Complex Portrait: An Examination of Small-Dollar Credit Consumers,” (Ctr. for Fin. Servs. Innovation, 2012), available at https://www.fdic.gov/news/conferences/consumersymposium/2012/A%20Complex%20Portrait.pdf. 619 “Very short term” referred to payday, pawn, and deposit advance products offered by depository institutions. Rob Levy & Joshua Sledge, “A Complex Portrait: An Examination of Small-Dollar Credit Consumers,” at 4 (Ctr. for Fin. Servs. Innovation, 2012), available at https://www.fdic.gov/news/conferences/consumersymposium/2012/A%20Complex%20Portrait.pdf. 406 my paycheck arrived,” or that their general living expenses were consistently more than their incomes as compared to respondents who took out “short term” products, though unexpected expenses were also cited by about 30 percent of the “very short term” respondents. More than 40 percent of “very short term” respondents also reported using the funds to pay for routine expenses, including both paying utility bills and general living expenses. b. Lender Practices 1. Loan Structure As stated in the proposal, some longer-term payday installment loans and vehicle title loans are structured either to be repaid in a single lump-sum payment or to require a large balloon payment, often as a final payment of all principal due following a series of smaller interest-only payments. Unsurprisingly, many consumers find making such a payment as challenging as making the single payment under a traditional, two-week payday loan, and such loans frequently result in default or re-borrowing. 2. Failure to Assess Ability to Repay Many lenders making longer-term balloon-payment loans—like lenders making other types of longer-term loans—have constructed a business model that allow them to offer loans profitably despite very high loan-level and sequence-level default rates. Rather than assessing whether borrowers will have the ability to repay the loans, these lenders engage in limited upfront screening to detect potential fraud and other “first payment defaults,” and otherwise rely heavily on loan features and practices that result in consumers continuing to make payments beyond the point at which they are affordable. These lenders do not seek to prevent those with expenses chronically exceeding income from taking on additional obligations in the form of 407 payday installment or similar loans. Lending to borrowers who cannot repay their loans would generally not be profitable in a traditional lending market, but the key features of these loans— leveraged payment mechanisms, vehicle security, and high cost—turn the traditional model on its head. These product features significantly reduce lenders’ interest in ensuring that payments under a covered longer-term balloon-payment loan are within the consumer’s ability to repay. Some of these consumers may repay the entire loan at the expense of suffering adverse consequences in their inability to keep up with basic living expenses or major financial obligations. Others end up defaulting on their loans at a point later than would otherwise be the case, thus allowing lenders to extract additional revenue on the way ultimately to the same adverse result. Product features that make this possible include the ability to withdraw payments directly from a borrower’s deposit account or the leverage derived from the ability to repossess the borrower’s means of transportation to work and for other everyday activities. The effect is especially strong when the lender times the loan payments to coincide with deposits of the consumer’s periodic income into the account. In these cases, lenders can succeed in extracting payments from the consumer’s account even if they are not affordable to the consumer. The lender’s risk of default is reduced, and the point at which default ultimately occurs is delayed. As a result, the lender’s incentive to invest time or effort into determining whether the consumer will have the ability to make the loan payments is greatly diminished. c. Harms Spurred by Balloon-Payment Loan Structures When these features are combined with a balloon-payment structure, lenders can operate, presumably at a profit, even when borrowers are defaulting on 50 percent of loan sequences. The circumstances of the borrowers and the structure of the loans that require a large balloon 408 payment to be made all at once can lead to dramatic negative outcomes for many borrowers who receive unaffordable loans because the lender does not reasonably assess their ability to repay. The Bureau is particularly concerned about the harms associated with re-borrowing and refinancing; harms associated with default, including vehicle repossession or the loss of a deposit account; and harms that flow from borrowers forgoing basic living expenses or defaulting on other major financial obligations as a result of making unaffordable payments on such loans. In the CFPB Report on Supplemental Findings, the Bureau analyzed several aspects of the re-borrowing and refinancing behavior of borrowers who take out vehicle title installment loans. For a longer-term loan with a balloon payment due at the end, the data analyzed by the Bureau demonstrated a large increase in borrowing around the time of the balloon payment, relative to loans without a balloon-payment feature. Further, for loans with a balloon payment, the re-borrowing was much more likely to occur around the time the balloon payment came due and consumers were less likely to take cash out, suggesting that the unaffordability of the balloon payment is the primary or sole reason for the re-borrowing or refinancing. Specifically, about 60 percent of balloon-payment installment loans resulted in refinancing, re-borrowing, or default. In contrast, nearly 60 percent of comparable fullyamortizing installment loans were repaid without refinancing or re-borrowing. Moreover, the reborrowing often only deepened the consumer’s financial distress. Balloon payments were not only associated with a sharp uptick in re-borrowing, but also with increased incidence of default. Notably, the default rate for balloon-payment vehicle title installment loans that the Bureau analyzed was about three times higher than the default rate for comparable fully-amortizing vehicle title installment loans offered by the same lender. 409 In addition to the harms discussed above, the Bureau is concerned that borrowers who take out these loans may experience other financial hardships as a result of making payments on unaffordable loans. Even if there are sufficient funds in the account, extraction of the payment through leveraged payment mechanisms or vehicle security places control of the timing of the payment with the lender, leading to the risk that the borrower’s remaining funds will not cover their other expenses or obligations. The resulting harms are wide-ranging and, almost by definition, can be quite extreme. Consumers may experience knock-on effects from their failure to meet these other obligations, such as additional fees to resume utility services or late fees on other obligations. This risk is further heightened when lenders time the loan payment due dates to coincide with the consumer’s receipt of income, which is typically the case. Furthermore, even if the consumer’s account lacks sufficient funds available to cover the required loan payment, the lender still may be able to collect the payment from the consumer’s bank by putting the account into an overdraft position. Where that occurs, the consumer will incur overdraft fees and, at many banks, extended overdraft fees. When new funds are deposited into the account, those funds will go to repay the overdraft and not be available to the consumer for other expenses or obligations. Thus, at least certain types of covered longer-term loans—in particular, long-term balloon-payment loans—carry a high degree of risk that if the payment proves unaffordable, the consumer will still be forced to repay the loan and incur further adverse effects, such as penalty fees or legal actions such as vehicle repossession or eviction. The Bureau is not able to directly observe the harms borrowers suffer from making unaffordable payments. The presence of a leveraged payment mechanism or vehicle security, however, each make it highly likely that borrowers who are struggling to pay back the loan will 410 suffer these harms. The very high rates of default on these loans means that many borrowers do struggle to repay these loans, and it is therefore reasonable to infer that many borrowers are also suffering harms from making unaffordable payments.620 d. Consumer Expectations and Understanding The Bureau is concerned about these negative consequences for consumers that flow from covered longer-term balloon-payment loans made without reasonably assessing the borrower’s ability to repay, because there is strong reason to believe that consumers do not understand the likelihood of the risk that such loans will prove unaffordable or the likelihood and extent of the adverse collateral consequences of such unaffordable loans. As an initial matter, the Bureau finds that many consumers fail to understand that lenders making longer-term balloon-payment loans—like lenders making other types of longer-term loans—do not evaluate their ability to repay their loans and instead have built business models that tolerate default rates well in excess of 30 percent, even after many consumers have incurred the further costs of re-borrowing. While the Bureau is unaware of any borrower surveys in these two markets, these two conditions are directly contrary to the practices of lenders in nearly all other credit markets—including other subprime lenders. 620 Wage assignments represent a particularly extreme form of a lender taking the control of a borrower’s funds away from a borrower. When wages are assigned to the lender, the lender does not even need to go through the process of submitting a request for payment to the borrower’s financial institution; the money is simply forwarded to the lender without ever passing through the borrower’s hands. The Bureau is concerned that where loan agreements provide for wage assignments, a lender can continue to obtain payment as long as the consumer receives income, even if the consumer does not have the ability to repay the loan while meeting her major financial obligations and basic living expenses. This concern applies equally to contract provisions that would require the consumer to repay the loan through payroll deductions or deductions from other sources of income, as such provisions would operate in essentially the same way to extract unaffordable payments. These approaches raise concerns that go beyond the scope of this rule, and the Bureau will continue to scrutinize the use of wage assignments in connection with longerterm loans not addressed by the final rule, using its supervision and enforcement authority to identify and address unfair, deceptive, or abusive acts or practices. 411 The Bureau has observed that most borrowers are unlikely to take out a loan they expect to default on, and hence the fact that at least one in three sequences ends in default strongly suggests that borrowers do not understand the degree of risk to which they are exposed with regard to such negative outcomes as default or loss of their vehicle, re-borrowing in connection with unaffordable loans, or having to forgo basic living expenses or major financial obligations. Even if consumers did understand that lenders offering longer-term balloon-payment loans were largely uninterested in their ability to repay, consumers would still be hindered in their ability to anticipate the risks associated with these loans. As discussed above, most borrowers taking out longer-term loans are already in financial distress.621 Many have had a recent unexpected expense, like a car repair or a decline in income, or they may have chronic problems in making ends meet. Even when not facing a crisis, research shows that consumers tend to underestimate their near-term expenditures622 and, when estimating how much financial “slack” they will have in the future, discount even the expenditures they do expect to incur.623 Consumers also tend to underestimate volatility in their own earnings and expenses, especially the risk of unusually low income or high expenses. Such optimism bias tends to have a greater effect when consumers are 621 Rob Levy & Joshua Sledge, “A Complex Portrait: An Examination of Small-Dollar Credit Consumers,” at 12 chart 3 (Ctr. for Fin. Servs. Innovation, 2012), available at https://www.fdic.gov/news/conferences/consumersymposium/2012/A%20Complex%20Portrait.pdf. 622 Gulden Ulkuman et al., “Will I Spend More in 12 Months or a Year? The Effects of Ease of Estimation and Confidence on Budget Estimates,” 35 J. of Consumer Research 245, at 245–246 (2008).; Johanna Peetz & Roger Buehler, “Is the A Budget Fallacy? The Role of Savings Goals in the Prediction of Personal Spending,” 34 Personality and Social Psychol. Bull. 1579 (2009); Johanna Peetz & Roger Buehler, “When Distance Pays Off: The Role of Construal Level in Spending,” Predictions, 48 J. of Experimental Soc. Psychol. 395 (2012). 623 Jonathan Z. Bermann et al., “2015 Expense Neglect in Forecasting Personal Finances,” 53 J. of Marketing Res. 535 (2016). 412 projecting their income and expenses over longer periods.624 Finally, in addition to gaps in consumer expectations about the likelihood that these loans will generally prove unaffordable, the Bureau observes that consumers underestimate the potential damage from default such as secondary fees, loss of vehicle or loss of account, which may tend to cause consumers to underestimate degree of harm that could occur if a loan proved unaffordable. In sum, the Bureau’s analysis of longer-term balloon-payment loans, as supplemented by its analysis of related types of longer-term loans, indicates that many consumers are unable to appreciate the likelihood of the risk and the magnitude of the harm they face from such loans if they are made on unaffordable terms. This is likely to be the case, in particular, with covered longer-term balloon-payment loans made without reasonably assessing the borrower’s ability to repay the loan according to its terms. Section 1041.4 Identification of Unfair and Abusive Practice—Underwriting Preliminary Discussion on Covered Longer-Term Balloon-Payment Loans The bulk of the Bureau’s analysis below is tailored toward covered short-term loans because those loans are the Bureau’s primary source of concern, and the market for which the Bureau has the most evidence. However, the Bureau’s statement of the unfair and abusive practice in § 1041.4 of the final rule also encompasses covered longer-term balloon-payment loans as defined in § 1041.2(a)(7) of the final rule. Accordingly, these loans, like covered shortterm loans, are subject to both the underwriting and payment requirements of the final rule. 624 As noted elsewhere, this discussion is not dependent on a particular endorsement of the tenets of behavioral economics and is likewise consistent with economic models based on rational expectations as applied in the circumstances of the kinds of situations faced by the borrowers of such loans. 413 The Bureau does not believe that currently there is a particularly large market for these loans, which is why most of the Bureau’s evidence is focused on covered short-term loans. But as described above in Market Concerns—Underwriting, where the Bureau has observed covered longer-term loans involving balloon payments for which the lender does not assess borrowers’ ability to repay before making the loan, it has seen the same type of consumer harms and other circumstances that the Bureau has observed when lenders fail to assess ability to repay before making covered short-term loans. Indeed, the Bureau’s analysis of longer-term balloon-payment loans in the market for vehicle title loans found that borrowers experienced high default rates— notably higher than for similar loans with amortizing installment payments.625 If the Bureau were to finalize this rule without including longer-term balloon-payment loans, it also has great concern that the market for longer-term balloon-payment loans, which is currently quite small, could expand dramatically if lenders were to begin to make efforts to circumvent its provisions by making these loans without assessing borrowers’ ability to repay. The result would be that the same type of unfair and abusive practice (just with a slightly different credit product) would persist and impose similar harms on consumers. This scenario is also more than mere speculation. The Military Lending Act was enacted in 2006 and imposed a 36 percent interest-rate cap on certain loans made to servicemembers and their dependents.626 Rules to implement its provisions were adopted,627 and the small-dollar loan 625 Rather than elongate the section-by-section analysis of § 1041.4 by engaging in a separate and distinct analysis of each prong of unfairness and abusiveness for covered longer-term balloon-payment loans, the Bureau would simply note that much of the general analysis is basically the same, except that the substantial risks and harms to consumers of high levels of re-borrowing with unaffordable covered short-term loans would be analogized to the substantial risks and harms to consumers of high levels of defaults with unaffordable covered longer-term balloon-payment loans. 626 Public Law 109–364, 120 Stat. 2266 (2006). 414 industry, in particular, went to some lengths to circumvent the provisions of those rules by making changes in their loan products, such as modifying terms and conditions and extending the duration of such loans.628 The resulting evasion of the rules was successful enough that Congress found it necessary to revisit the law and direct that new rules be adopted to close loopholes that the prior rules had created, which had undermined the purposes of the Act.629 The new regulations were adopted in July 2015 and are now in effect.630 The fact of this recent experience in this very industry underscores the Bureau’s concern that applying the underwriting criteria of this rule to covered longer-term balloon-payment loans is necessary to effectuate its purpose to protect consumers. This point reinforces the Bureau’s view, based on the limited evidence of the small size of the market currently existing for these loans, that the analysis below would apply to covered longer-term balloon-payment loans as well as to covered short-term loans if that market were to expand. Thus, the Bureau has made the judgment to similarly regulate covered longer-term balloon-payment loans. The Bureau did not receive many comments on just the specific portion of the Bureau’s proposal about covered longer-term loans involving balloon payments. However, the Bureau did receive a few. Several industry commenters stated that the Bureau’s concerns about reborrowing for covered longer-term loans should have focused primarily on loans with balloon payments, and argued that any restrictions should thus be limited to balloon-payment loans. These commenters were correct that the Bureau’s concerns regarding re-borrowing, which are 627 72 FR 50580 (Aug. 31, 2007). 79 FR 58602, 58602–06 (Sept. 29, 2014). 629 Public Law 112–239, 126 Stat. 1785 (2013). 630 80 FR 43560 (July 22, 2015). 628 415 similar to the Bureau’s concerns regarding covered short-term loans, were focused primarily on covered longer-term balloon-payment loans. This is one of the reasons why the Bureau is finalizing only this portion of the proposal involving covered longer-term loans, and provides further support for the Bureau’s conclusion that the analysis below relating to covered short-term loans is applicable to covered longer-term balloon-payment loans as well. Having addressed this issue here, the remainder of the discussion in this section of the unfair and abusive practice of making loans without reasonably assessing the borrower’s ability to repay the loan according to its terms will focus exclusively on covered short-term loans. The Bureau’s Approach in the Proposal As the Bureau noted in the proposal, it is standard practice in most consumer lending markets for lenders to assess whether a consumer has the ability to repay a loan before making the loan. In certain markets, Federal law requires this.631 The Bureau did not propose to make a determination whether, as a general rule for all kinds of credit, it is an unfair or abusive practice for any lender to make a loan without making such a determination. Nor did the Bureau propose to resolve that question in this rulemaking. Rather, the focus of the subpart B of the proposed rule was on a more specific set of loans that the Bureau has carefully studied, as discussed in more detail above in part II and in Market Concerns—Underwriting. Based on the evidence presented in the proposal, and pursuant to its authority under section 1031(b) of the Dodd-Frank 631 See, e.g., Dodd-Frank Act section 1411, codified at 15 U.S.C. 1639c(a)(1); CARD Act, 15 U.S.C. 1665e; HPML Rule, 73 FR 44522, 44543 (July 30, 2008). In addition, the OCC has issued numerous guidance documents about the potential for legal liability and reputational risk connected with lending that does not take account of borrowers’ ability to repay. See OCC Advisory Letter 2003–3, Avoiding Predatory and Abusive Lending Practices in Brokered and Purchased Loans (Feb. 21, 2003), available at http://www.occ.gov/static/news-issuances/memos-advisoryletters/2003/advisory-letter-2003-3.pdf; FDIC, Guidance on Supervisory Concerns and Expectations Regarding Deposit Advance Products, 78 FR 70552 (Nov. 26, 2013); OCC, Guidance on Supervisory Concerns and Expectations Regarding Deposit Advance Products, 78 FR 70624 (Nov. 26, 2013). 416 Act, the Bureau proposed to identify it as both an unfair practice and an abusive practice for a lender to make a covered short-term loan without reasonably determining that the consumer will have the ability to repay the loan under its explicit authority to prescribe rules for “the purpose of preventing [unfair and abusive] acts or practices.”632 In this specific context, “ability to repay” was defined in the proposal to mean that the consumer will have the ability to repay the loan without re-borrowing and while meeting the consumer’s major financial obligations and basic living expenses. The Bureau had made preliminary findings and reached preliminary conclusions about the unfairness and the abusiveness of making these loans without such a reasonable determination, based on the specific evidence cited in the proposal, which is discussed further below as well as above in part II and Market Concerns—Underwriting. The Bureau sought comment on the evidence it had presented on these issues and on the preliminary findings and conclusions it had reached in the proposal. It also sought comment on whether making the kinds of loans that meet the conditions set forth in the proposed exemption—conditions that are specifically designed as an alternative means to protect consumers against the harms that can result from unaffordable loans—should not be regarded as an unfair or abusive practice. General Comments The Bureau received a number of general comments about the Bureau’s use of its authority to prohibit unfair, deceptive, or abusive acts or practices (“UDAAP”). The Bureau addresses those more general comments here, but specific comments on the prongs of unfairness or abusiveness are found below. 632 12 U.S.C. 5531(b). 417 Some industry participants suggested that an act or practice can only be deemed unfair, deceptive, or abusive if there is a strong element of wrongdoing or a sense that an unconscionable advantage has been taken, which they asserted did not exist. Many industry participants and trade associations attacked the factual foundation set forth in the proposal as inadequate. And they took particular issue with the framing of the proposal as resting on what they viewed as mere assertions and presuppositions, not clearly grounded in factual findings, as reflected in certain phrasings and characterizations (or even “slogans”). They further viewed this preliminary foundation for the proposal as reflecting bias or prejudgment on the part of the Bureau that improperly colored its approach to these issues. Industry participants and trade associations also highlighted the Bureau’s observation made in the proposal that “the evidence on the effects on consumers of access to storefront payday loans is mixed.” They argued that the Bureau cannot rest any rulemaking that imposes a substantial market intervention, including UDAAP rulemakings, on mixed evidence that is not more clearly definitive of the key points at issue. Accordingly, these commenters again contended that the Bureau was resting its proposed rule on an insufficient factual threshold. Bank and credit union commenters, among others, suggested that the Bureau either lacked—or had failed to provide—data to support the application of the abusiveness standard (or more broadly, the UDAAP standard) in context of the kinds of short-term loans they provide, which would be covered loans under the proposal. Here again, one commenter cited the Bureau’s reliance on “a set of preliminary findings” and what it “believes” to be true as indicative of the Bureau’s lack of supporting data. Another suggested that loans made by community banks that are covered under the proposed rule are not predatory and do not 418 perpetuate a cycle of indebtedness. This commenter noted that community banks have developed a business model that does not rely on rolling over loans and churning fees, that they underwrite all of their own small loans, and that default and vehicle repossession rates associated with these loans are very small. These commenters thus asserted that the Bureau lacks evidence to demonstrate that their practices associated with these loans are unfair, deceptive or abusive. For these and other reasons, community bank and credit union commenters strongly advocated for the Bureau to use its exemption authority to ensure that their lending activities would not be covered under the terms of any final rule, either in whole or in part. Similarly, commenters asserted that the Bureau was acting improperly by resting the proposed rule on its mere “beliefs” and preliminary findings, rather than holding off until the Bureau was in a position to render definitive conclusions on the main points at issue. In particular, they contended that UDAAP rules governing these covered loans could not validly be enacted until after the Bureau makes definitive rulings based on evidence and fact. Some commenters, comprising both industry participants and trade associations, argued that the Dodd-Frank Act does not authorize the Bureau to ban a “product,” but only to “prescribe rules” identifying unlawful UDAAP “acts or practices.” One industry commenter argued that the Bureau had mischaracterized or ignored relevant legal precedent that controls how the Bureau must interpret its UDAAP authority under the Dodd-Frank Act, going so far as to say that Bureau lawyers had a professional responsibility to correct the record, and arguing that the Bureau does not have the authority to invalidate entire contracts or whole products. Other commenters argued that the proposed rule was overbroad insofar as it rested on the sweeping conclusion that all alternative underwriting approaches would be unable to pass muster under the 419 unfair or abusive standards laid out in the statute. Further, they contended that the proposed rule would largely eliminate payday and title loans, which are sources of credit that many consumers have long relied on, all of which would exceed the Bureau’s statutory mandate. One commenter also made the point that the Bureau’s proposal seemed inconsistent with the statutory objective of leveling the playing field for all competitors of consumer financial products by addressing the perceived unfairness of regulating just these covered loans without addressing all of the products that may have similar or equivalent features. Many industry participants and trade associations submitted comments that attacked the broader legal authority of the Bureau to propose any rule governing these types of short-term loans, especially a rule under its UDAAP authority. A few of them argued that the Bureau’s authority is narrowly constrained because the Truth in Lending Act and its implementing regulations provide a pervasive regulatory framework to govern consumer credit transactions. Others argued that when Congress intended to impose ability-to-repay requirements on specific lending markets, it did so explicitly by statute (as it did with mortgages and credit cards), but it did not confer such explicit authority on the Bureau to regulate payday and title loans in this manner. As a consequence, these commenters maintained that the expressio unius canon of statutory construction applies to deny the Bureau any such regulatory authority. Some commenters stated views that conflicted with those set out above. One trade association, in particular, stated that Congress plainly recognized the problems created by unregulated and less regulated lenders, and for that reason conferred on the Bureau new authority to supervise and write rules for the payday lending industry for the first time ever at the Federal level. More generally, consumer groups were strongly supportive of the Bureau’s legal authority 420 to develop and finalize the proposed rule. Rather than viewing other ability-to-repay provisions in Federal consumer law as implied negative restrictions on the Bureau’s authority, these commenters pointed to them and others (such as the Military Lending Act) as embodying a considerable trend of expanding public policy now supporting the principle that consumer lending generally should be premised on the borrower’s ability to repay. They noted that, along with recent Federal law on mortgage and credit card lending, certain States now embody this principle in statute, and many more do so by judicial precedent. They noted that general statements of this principle in Federal and State law tend to define this approach as requiring the lender to establish the borrower’s ability to repay the loan while meeting basic living expenses and without re-borrowing. Approach in the Final Rule and Changes to Language in § 1041.4 The terms “unfair” and “abusive” are defined terms in the Dodd-Frank Act with multiple prongs. Under the Act, the Bureau cannot determine an act or practice to be unlawful unless “the Bureau has a reasonable basis to conclude” that the act or practice “causes or is likely to cause substantial injury to consumers which is not reasonably avoidable by consumers” and “such substantial injury is not outweighed by countervailing benefits to consumers or to competition.”633 The Bureau is expressly authorized to “consider established public policies as evidence” in “determining whether an act or practice is unfair.”634 An “abusive” act or practice is defined, among other things, as one that “takes unreasonable advantage of (A) a lack of understanding on the part of the consumer of the material risks, costs, or conditions of the 633 634 12 U.S.C. 5531(c)(1). 12 U.S.C. 5531(c)(2). 421 product or service; [or of] (B) the inability of the consumer to protect the interests of the consumer in selecting or using a consumer financial product or service.”635 In the proposal, each of the specified prongs of these two terms defined in the statute was discussed separately. Hence the comments that were submitted on these specific legal grounds regarding the Bureau’s approach can be presented and addressed in this format as well, and that discussion is contained in the following sections. But the more general comments on the Bureau’s legal approach to developing ability-to-repay rules under UDAAP to govern covered short-term loans, as those comments were summarized above, can be directly addressed here. To begin with, the commenters’ suggestion that an act or practice can only be deemed unfair, deceptive, or abusive if there is a strong element of wrongdoing or a sense that an unconscionable advantage has been taken is a mischaracterization of the Bureau’s UDAAP authority as prescribed by law. Although public policy is a factor that the Bureau may consider for purposes of identifying unfairness, both the unfairness and abusiveness standards rest upon well-defined elements in the Dodd-Frank Act, and a sense of wrongdoing or unconscionability is not one of them. In fact, the FTC and Congress have explicitly rejected the notion that agencies should be measuring whether an act is “immoral, unethical, oppressive, or unscrupulous” or consistent with public policy to make unfairness findings.636 An abusive practice may require that the person take “unreasonable advantage” of various conditions,637 but that does not require any sense of unconscionability. The commenters do not offer any compelling justification for 635 12 U.S.C. 5531(d)(2)(A) and (B). J. Howard Beales, Former Dir. of Bureau of Consumer Prot., “The FTC’s Use of Unfairness Authority: Its Rise, Fall, and Resurrection,” The Marketing and Public Policy Conference (May 30, 2003). 637 Though taking “unreasonable advantage” is not a prerequisite for an abusiveness finding if a company “materially interferes with the ability of a consumer to understand a term or condition of a consumer financial product or service.” 12 U.S.C. 5531(d)(1). 636 422 their position that the Bureau should, or even is authorized to, supplement the specific statutory prongs that Congress adopted to define the terms “unfair” and “abusive” with these additional and loose concepts that were not incorporated in the statute. Congress was undoubtedly aware of the unconscionability standard when it passed the Consumer Financial Protection Act, and it did not use the language of unconscionability to limit the unfairness or abusiveness standards. Some commenters attacked various preliminary findings and conclusions set forth in the proposal by reacting to language in the proposed rule conveying that, as is true of any proposed notice-and-comment rulemaking, the Bureau always planned to wait to formulate and support its final conclusions only after receiving feedback on its proposal. The Bureau appropriately noted that various factual statements, observations, or conclusions made in the proposal were to be regarded as tentative until they could be and had been evaluated in light of comments and supporting information received through the entire rulemaking process. In fact, the Bureau is required by law to consider and analyze the comments received before deciding whether and how to finalize any regulations. As described in the section-by-section analysis for § 1014.4 and this preamble, now that the Bureau has had the opportunity to consider the high volume of input that it has received from all stakeholders, including extensive individual involvement by members of the public, it is in a position to articulate and justify the types of formal and definitive conclusions necessary to support the final rule. The factual recitation presented above in the discussion of Market Concerns—Underwriting embodies the Bureau’s presentation of and response to commenters’ specific points that were raised about these factual issues. The fact that the Bureau had presented some of its views in the proposal as tentative thus is not improper and was entirely appropriate at that preliminary stage of the rulemaking process. 423 Some commenters took virtually the opposite tack, objecting to statements made in the proposal, or made by the Bureau in the course of wide-ranging discussions on other occasions, as suggesting bias and prejudgment of certain issues underlying the proposed rule. These objections seem to lack foundation or to be based on statements taken out of context, given the considerable efforts the Bureau has undertaken to process, analyze, and digest the heavy volume of comments received and be responsive to them on the merits in formulating the final rule. The Bureau bases its UDAAP findings on the evidence and conclusions as discussed and now adopted in this section and in Market Concerns—Underwriting. Those findings are more explicitly laid out below when describing the comments and analysis that are applicable to the distinct unfairness and abusiveness prongs. As to the statement that the Bureau based its views on “mixed” evidence, in the proposal the Bureau stated that “[i]n reviewing the existing literature, the Bureau believes that the evidence on the impacts of the availability of payday loans on consumer welfare is mixed. A reasonable synthesis appears to be that payday loans benefit consumers in certain circumstances, such as when they are hit by a transitory shock to income or expenses, but that in more general circumstances access to these loans makes consumer worse off. The Bureau reiterates the point made earlier that the proposed rule would not ban payday or other covered short-term loans, and believes that covered short-term loans would still be available in States that allow them to consumers facing a truly short-term need for credit.” In other words, the Bureau did not simply rest its preliminary findings on its determination to take one side of a debate. Instead, the Bureau analyzed the evidence, which naturally differed on methodology and subjects studied, and synthesized it into a preliminary view that payday loans benefit some consumers in certain 424 circumstances, but generally leave many other consumers worse off, while noting that many of the consumers who benefited would still be able to access payday loans under the provisions of the proposed rule. The Bureau finds that the comments received from banks and credit unions and their trade associations were generally well taken. Many bank and credit union loans are likely not covered by the final rule, because the Bureau is not finalizing the proposals on longer-term small-dollar loans at this time. And to the extent that community banks and credit unions make loans that would otherwise be covered on an accommodation basis for their customers, the Bureau’s use of its exemption authority in the final rule assures that these loans also will not be covered (of course, nonbanks making accommodation loans would similarly be exempt). The Bureau agrees that much of the evidence it reviewed related to loans made by nonbanks, and not banks. However, the Bureau did review evidence relating to Deposit Advance Products, made by banks, and concluded that it was consistent with the evidence the Bureau had on nonbank covered loans. Further, there appears to be no logical reason to believe that covered short-term loans, made without assessing borrowers’ ability to repay, would impact consumers differently depending on the lender’s charter. The Bureau thus concludes that based on the evidence it reviewed, it is appropriate to apply this rule to the banks and credit unions that are engaged in making covered loans that do not fall within the exemptions provided in the final rule. Doing so is consistent with the Bureau’s objective of ensuring that “Federal consumer 425 financial law is enforced consistently, without regard to the status of a person as a depository institution, in order to promote fair competition.”638 With respect to the commenter that viewed the Bureau’s proposal as inconsistent with the implicit statutory objective of leveling the playing field for all competitors of consumer financial products because it regulates covered loans without addressing every product that may have similar or equivalent features, the objection is unpersuasive. The Bureau is not required to write rules that cover every product or market all at once, and has the authority to prioritize taking action as it deems appropriate, so long as it has the data and justification for doing so for each instance. For example, the final rule does not cover the underwriting of longer-term loans. This rulemaking also does not cover overdraft services on deposit accounts. Both of those products are distinct from covered short-term loans and may be the subject of separate rulemaking efforts, as well as remaining subject to the Bureau’s oversight through the exercise of its supervisory and enforcement authority. For commenters who argued that the proposed rule was a misuse of the Bureau’s prevention authority, or was too harsh and too prescriptive so as to be disproportionate to the evidence of harm to consumers that the Bureau presented in the proposal, several responses are in order. The initial question is whether the Bureau can show in this final rule that in identifying the practice described in § 1041.4 as unfair and abusive, the Bureau acted within the scope of its express legal authority to adopt rules to identify and prevent unfair and abusive acts or practices—a topic that is covered in detail in the following sections. Comments about whether 638 12 U.S.C. 5511(b)(4). 426 the proposed ability-to-repay requirements are consistent with the Bureau’s prevention authority are addressed in more detail below in the section-by-section analysis of § 1041.5. The Bureau’s determination that the failure of a lender to reasonably determine the consumer’s ability to repay a covered short-term or longer-term balloon-payment loan according to its terms meets the statutory prongs of the Bureau’s “unfair” or “abusive” authority, as discussed further in the following sections, and thus the Bureau is not imposing a ban on any “product” but instead is simply prescribing rules to prevent the acts or practices so identified. The Bureau does not agree with commenters who suggest that the proposed underwriting rules would effectively have banned lenders from making covered loans. The Bureau continues to believe that even under the underwriting rules contained in the proposal, lenders would have been able to continue to make loans to consumers who, in fact, had the ability to repay those loans. In any event, the Bureau has reconsidered certain aspects of the ability-to-repay underwriting provisions presented in the proposal, in response to substantive comments that were received on various details of the proposed underwriting approach, which provisions are being implemented in a somewhat modified form in § 1041.5 below; and the Bureau is finalizing the alternative framework that it has presented for making such loans without all the underwriting criteria specified in § 1041.5, subject to a cap on how much lending could be achieved within this framework. For more details, see the Section 1022(b)(2) Analysis in part VII below and the section-by-section analysis for § 1041.5 of the final rule. 427 More generally, the Bureau’s rule does not invalidate whole products.639 Section 1041.4 identifies an unfair and abusive practice in the market—the making of covered short-term and longer-term loans without reasonably determining borrowers’ ability to repay the loans according to their terms. Other sections of the rule, including §§ 1041.5 and 1041.6, are intended to prevent that existing practice and the associated harms. This approach to UDAAP rulemaking (identification and then prevention) is a consistent and straightforward application of UDAAP precedent, as discussed further in part IV above. As to whether the specified components of the ability-to-repay determinations are disproportionate to the risks posed by such lending, the law does not impose any such proportionality test, as long as the statutory prongs of unfairness and abusiveness are met and the remedy imposed bears a reasonable relationship to addressing the identified practice. Nonetheless, it is again relevant here that, as explained in detail below in the section-by-section analysis of § 1041.5, the final rule has incorporated changes in the specified underwriting criteria to harmonize them more closely with those applicable to credit cards and to render them less demanding than the ability-to-repay test used for making mortgage loans. In particular, the Bureau has reconsidered certain aspects of the ability-to-repay underwriting criteria presented in the proposal in response to substantive comments that were received on various details of its 639 Commenters seem to believe that because section 1036(a)(1)(A) of the Dodd-Frank Act states it is unlawful to “offer or provide to a consumer any financial product or service not in conformity with Federal consumer financial law,” and section 1036(a)(1)(B) separately states that it is unlawful “to engage in any unfair, deceptive, or abusive act or practice,” that Congress intended to limit the Bureau’s UDAAP authority such that it could not be used to ban or invalidate products or services. This reading ignores the definition of Federal consumer financial law, which includes the Dodd-Frank Act itself and “any rule or order prescribed by the Bureau under [the Dodd-Frank Act],” which includes the prohibition against UDAAP as well as UDAAP rules. 12 U.S.C. 5481(14). Thus, the clear meaning of section 1036(a)(1)(A) is to make it unlawful to “offer or provide to a consumer any financial product or service not in conformity” with the prohibition against unfair, deceptive, or abusive acts or practices in section 1036(a)(1)(B). 428 proposed approach, and as a result these criteria are being implemented in a somewhat modified form in § 1041.5 below to take account of and respond to these particular concerns raised by the commenters. In addition, the Bureau’s proposal presented an alternative framework for making such loans, subject to a cap on how much lending could be achieved within this framework. That alternative framework is being adopted in the final rule, subject to certain modifications, as discussed further below in § 1041.6. For these reasons, the Bureau concludes that the approach set forth in the final rule imposes a remedy that bears a reasonable relationship to addressing the unfair and abusive practice identified by the Bureau so that it does not persist in this market. With respect to the commenters who asserted that the TILA or any combination of Federal statutes and regulations impliedly divest the Bureau of the authority to propose any rule governing these types of short-term loans under its UDAAP authority, those provisions do not seem able to bear the weight of the argument. On the contrary, the Dodd-Frank Act plainly gave the Bureau the authority to “prescribe rules” identifying “unfair, deceptive, or abusive acts or practices” that violate Federal law,640 even though Congress was well aware that the TILA, in particular, already was applicable to consumer financial products, such as the covered short-term loans addressed by this rule. Nor has Congress given any indication that it intended to restrict the Bureau from adopting an underwriting approach for this loan market (ability-to-repay underwriting, which is based on the lender making a reasonable determination that the borrower will have the ability to repay the loan) that has found increasing Congressional favor in other markets. The Bureau agrees with the commenters who took the view that Congress has plainly recognized the 640 12 U.S.C. 5531(b). 429 importance of these measures as a means of protecting consumers in two major consumer loan markets (credit cards and mortgages), which tends to support rather than undermine a finding that lending should be premised on the borrower’s ability to repay in the market for these covered loans as well. Commenters arguing otherwise did not provide any case law in support of this argument, and the cases cited by a few commenters involved Congress expressly articulating its intent to limit an agency’s authority in a particular manner, or an agency acting in a manner inconsistent with an express Congressional mandate. Neither applies here. Further the Bureau’s action is not without precedent, as at least one other agency has issued rules to prevent unfair or deceptive practices through an ability-to-repay requirement. Before the Consumer Financial Protection Act was passed into law, the Federal Reserve Board issued a rule under the Home Ownership and Equity Protection Act imposing ability-to-repay requirements for mortgage lenders “to prevent unfairness, deception, and abuse.”641 For these reasons, and as discussed further in the Bureau’s analysis of each of the prongs of the statute addressed below, the Bureau is finalizing its conclusion that it is an unfair and abusive practice for a lender to make a covered short-term or longer-term balloon-payment loans without reasonably determining that the borrowers will have the ability to repay the loans according to their terms. The Bureau made four modifications to proposed § 1041.4. The Bureau has added to the phrase “ability to repay the loan” the phrase “according to its terms,” such that the final statement of the unfair and abusive practice is, in part, the failure to assess that the consumer “will have the ability to repay the loan according to its terms.” The addition was meant to address a common misimpression conveyed by commenters. Many commenters 641 73 FR 44522, 44522-23 (July 30, 2008). 430 claimed that borrowers who cannot pay an originated loan nonetheless do have an ability to repay because they can repay after some amount of re-borrowing. To further reflect the Bureau’s intent, both now and at the stage of the proposal, that lenders should assess the borrower’s ability to repay without re-borrowing, the Bureau has added the phrase “according to its terms.” Second, the Bureau has added covered longer-term balloon-payment loans to the statement of the unfair and abusive practice, as noted above. Third, the Bureau added official commentary, at comment 4-1, clarifying that a lender who complies with § 1041.5 in making a covered short-term loan or a covered longer-term balloon-payment loan has not committed the unfair and abusive practice under § 1041.4. The comment further clarifies that a lender who complies with § 1041.6 in making a covered shortterm loan has not committed the unfair and abusive practice under § 1041.4 and is not subject to § 1041.5. This comment is added to clarify that the combination of §§ 1041.5 and 1041.6 are the Bureau’s intended method for preventing the practice in § 1041.4, that loans made under § 1041.6 are exempt from § 1041.5, and thus, that if a lender complies with § 1041.5 or § 1041.6, a lender would not be in violation of § 1041.4. Fourth, during inter-agency consultations, the Bureau received input from a Federal prudential regulator about the singular nature of the statement of the unfair and abusive practice. The regulator believed that supervisory or enforcement actions of this particular rule should be based on a pattern or practice of activity, rather than an isolated and inadvertent instance, which the regulator believed could deter responsible lenders from making covered loans. In the interest of inter-agency cooperation, the Bureau is adopting the suggestion to pluralize the statement of the unfair and abusive practice. Relatedly, the Bureau does not intend to bring supervisory or 431 enforcement actions against a lender for a single isolated violation of § 1041.5. In the discussion that follows, the Bureau responds to the core arguments raised in comments that were submitted on the Bureau’s proposal. The Bureau has organized the comments received such that all of the core arguments presented by the commenters are addressed in the following analysis of the statutory prongs of whether the identified practice constitutes an “unfair” practice and an “abusive” practice. Unfairness As discussed in the proposal, under section 1031(c)(1) of the Dodd-Frank Act, an act or practice is unfair if it causes or is likely to cause substantial injury to consumers which is not reasonably avoidable by consumers and such injury is not outweighed by countervailing benefits to consumers or to competition. Under section 1031(c)(2), the Bureau may consider established public policies as evidence in making this determination. The proposal preliminarily found that it is an unfair practice for a lender to make a covered short-term loan without reasonably determining that the consumer will have the ability to repay the loan. After issuing the proposal and receiving and reviewing comments, the Bureau is now finalizing that conclusion for covered short-term loans. The Bureau concludes that the practice causes substantial injury in the form of default, delinquency, re-borrowing, and collateral consequences associated with attempts to avoid the other injuries by making unaffordable payments. The data that the Bureau analyzed suggest that, particularly with respect to re-borrowing, the incidence of injury is quite high. The Bureau also concludes that this injury is not reasonably avoidable because a substantial population of borrowers who incur injury—from default, delinquency, re-borrowing, or other collateral consequences from making unaffordable payments—do not anticipate the harm. 432 Lastly, the Bureau concludes that the injury to these borrowers outweighs the countervailing benefits to those and other borrowers benefited by the practice and to competition. The most notable benefit would be greater access to credit for borrowers who lack an ability to repay, but for all the reasons discussed below, the Bureau believes that the harms associated with getting unaffordable credit for a substantial population of consumers outweigh any such benefit. In addition, the Bureau reasonably anticipates that even these borrowers are likely to retain access to some covered short-term loans that comply with the terms of final § 1014.6, subject to the conditions that are imposed in that provision to prevent the risks and harms associated with extended loan sequences. Commenters presented feedback on the Bureau’s preliminary conclusions for each of the three prongs of unfairness. The Bureau addresses the comments on those prongs in turn below. Practice causes or is likely to cause substantial injury The Bureau’s Proposal The proposal noted that the Bureau’s interpretation of the various prongs of the unfairness test is informed by the FTC Act, the FTC Policy Statement on Unfairness, and FTC and other Federal agency rulemakings and related case law.642 Under these authorities, as 642 Over the past several decades, the FTC and Federal banking regulators have promulgated a number of rules addressing acts or practices involving financial products or services that the agencies found to be unfair under the FTC Act (the 1994 amendments to which codified the FTC Policy Statement on Unfairness). For example, in the Credit Practices Rule, the FTC determined that certain features of consumer-credit transactions were unfair, including most wage assignments and security interests in household goods, pyramiding of late charges, and cosigner liability. 49 FR 7740 (March 1, 1984) (codified at 16 CFR 444). The D.C. Circuit upheld the rule as a permissible exercise of unfairness authority. AFSA, 767 F.2d at 957. The Federal Reserve Board adopted a parallel rule applicable to banks in 1985. The Federal Reserve Board’s parallel rule was codified in Regulation AA, 12 CFR part 227, subpart B. Regulation AA has been repealed as of March 21, 2016, following the Dodd-Frank Act’s elimination of the Federal Reserve Board’s rule writing authority under the FTC Act. See 81 FR 8133 (Feb. 18, 2016). In 2009, in the HPML Rule, the Federal Reserve Board found that disregarding a consumer’s repayment ability when extending a higher-priced mortgage loan or HOEPA loan, or failing to verify the consumer’s income, 433 discussed in part IV, “substantial injury” may consist either of a small amount of harm to a large number of individuals or of a larger amount of harm to a smaller number of individuals. In this case, the proposal stated that the practice at issue causes or is likely to cause both—a substantial number of consumers suffer a high degree of harm, and a large number of consumers suffer a lower but still meaningful degree of harm. In the proposal, the Bureau stated its judgment that the practice of making a covered short-term loan without assessing the consumer’s ability to repay the loan according to its terms causes or is likely to cause substantial injury. When a loan is structured to require repayment within a short period of time, the Bureau noted that the payments may outstrip the consumer’s ability to repay since the type of consumers who turn to these products cannot absorb large loan payments on top of their major financial obligations and basic living expenses. If a lender nonetheless makes such loans without determining that the loan payments are within the consumer’s ability to repay, the Bureau stated that it appears the lender’s conduct causes or is likely to cause the injuries described below. The proposal stated that, in the aggregate, the consumers who suffer the greatest injury are those consumers who find it necessary to re-borrow repeatedly and end up in exceedingly long loan sequences. As discussed in the proposal, consumers who become trapped in long loan assets, and obligations used to determine repayment ability, is an unfair practice. See 73 FR 44522 (July 30, 2008). The Federal Reserve Board relied on rulemaking authority pursuant to TILA section 129(l)(2), 15 U.S.C. 1639(l)(2), which incorporated the provisions of HOEPA. The Federal Reserve Board interpreted the HOEPA unfairness standard to be informed by the FTC Act unfairness standard. See 73 FR 44529 (July 30, 2008). That same year, the Federal Reserve Board, the OTS, and the NCUA issued the interagency Subprime Credit Card Practices Rule, in which the agencies concluded that creditors were engaging in certain unfair practices in connection with consumer credit card accounts. See 74 FR 5498 (Jan. 29, 2009). One commenter suggested that the Bureau should not rely on AFSA but instead on Katharine Gibbs School v. FTC, 612 F.2d 658 (2d Cir. 1979), a ruling that AFSA effectively distinguished in a discussion of how the agency should properly go about identifying and specifying unfair acts or practices. The Bureau agrees with the D.C. Circuit’s treatment in AFSA of the ruling in Katharine Gibbs. 434 sequences pay substantial fees for re-borrowing, and they usually do not reduce the principal amount owed when they re-borrow. For example, roughly half of payday loan sequences consist of at least three loans, at which point, in a typical two-week loan, a storefront payday borrower will have paid over a period of eight weeks charges equal to 60 percent or more of the loan amount—and will still owe the full amount originally borrowed. Roughly one-third of consumers re-borrow at least six times, which means that, after three-and-a-half months with a typical two-week loan, the consumer will have paid to the lender a sum equal to 100 percent of the loan amount and made no progress whatsoever in repaying the principal. Almost one-quarter of loan sequences643 consist of at least 10 loans in a row, and 50 percent of all loans are in sequences of 10 loans or more. And looking just at loans made to borrowers who are paid weekly, biweekly, or semi-monthly, approximately 21 percent of loans are in sequences consisting of at least 20 loans. For loans made to borrowers who are paid monthly, 42 percent of loans are in sequences consisting of at least 10 loans. Similarly, for single-payment vehicle title loans, the Bureau found that more than half (56 percent) of loan sequences consist of at least four loans in a row; over a third (36 percent) consist of seven or more loans in a row; and about onefourth (23 percent) had 10 or more loans. The proposal further stated that consumers whose loan sequences are shorter may still suffer meaningful injury from re-borrowing, albeit to a lesser degree than those in longer sequences. Even consumers who re-borrow only once or twice—and, as described in the 643 Note that the one-third of borrowers who re-borrow six times and the one quarter of borrowers who re-borrow 10 times are not separate populations. All of the borrowers who re-borrowed 10 times also re-borrowed six times. 435 proposal, 22 percent of payday and 23 percent of vehicle title loan sequences show this pattern— will still incur significant costs related to re-borrowing or rolling over the loans. The proposal stated that the injuries resulting from default on these loans also appeared to be significant in magnitude. As described in the proposal, 20 percent of payday loan sequences end in default, while 33 percent of single-payment vehicle title sequences end in default. Because covered short-term loans (other than vehicle title loans) are usually accompanied by some specific means of payment collection—typically a postdated check for storefront payday loans and an authorization to submit electronic debits to the consumer’s account for online payday loans—a default means that the lender was unable to secure payment despite using those tools. That means a default is typically preceded by failed attempts to secure payment, which generate bank fees (such as NSF fees) that can put the consumer’s account at risk and lender fees (such as late fees or returned check fees) that add to the consumer’s total indebtedness. Additionally, as discussed in the proposal, where lenders’ attempts to extract money directly from the consumer’s account fail, the lender often will resort to other collection techniques, some of which—such as repeated phone calls, in-person visits to homes and worksites, and lawsuits leading to wage garnishments—can inflict significant financial and psychological damage on consumers.644 The proposal stated that for consumers with a single-payment vehicle title loan, the injury from default can be even greater. In such cases, lenders do not have access to the consumers’ bank account but instead have the ability to repossess the consumer’s vehicle. As discussed in 644 As noted in part IV (Legal Authority), the D.C. Circuit held that psychological harm can form part of the substantial injury along with financial harm. See AFSA, 767 F.2d at 973–74, n.20 (1985). 436 the proposal, almost one in five title loan sequences end with the consumer’s vehicle being repossessed. Consumers whose vehicles are repossessed and who do not have another vehicle may end up either wholly dependent upon public transportation or family or friends to get to work, to shop, or to attend to personal needs. In many personal situations and in many areas of the country, such as rural areas and urban areas without public transportation that is reasonably available, this means they may end up without any effective means of transportation at all. Finally, the proposal stated that the Bureau believes many consumers, regardless of whether they ultimately manage to pay off the loan, suffer collateral consequences as they struggle to make payments that are beyond their ability to repay. For instance, they may be unable to meet their other major financial obligations or may be forced to forgo basic living expenses as a result of prioritizing a loan payment and other loan charges—or having it prioritized for them, in ways they cannot control, by the lender’s exercise of its leveraged payment mechanism. Comments Received The Bureau received many comments from stakeholders on all sides of these issues about whether the identified practice causes or is likely to cause substantial injury to consumers. As an initial matter, the Bureau received a number of comments from industry participants and trade associations on how the Bureau should measure injury before making a determination that a given act or practice is unfair. Several commenters stated that injury should be measured in relation to consumer outcomes in the absence of the act or practice (here payday lending without assessing the borrower’s ability to repay). Commenters argued that the Bureau’s identified injuries should be compared to the alternatives without such loans, including defaulting on other 437 financial obligations, failing to afford basic living expenses, forgoing the purchase of goods and services, and bouncing checks. One commenter argued that the psychological injury from stress caused by the threat of repossession should be offset by the injury of the stress caused by losing electricity, heat, water, or the actual vehicle (assuming the borrower must sell or pawn the vehicle to cover the expense). Another commenter argued that the Bureau failed to identify any “metric” for measuring harm at all, and that without doing so, the Bureau was unable to estimate the scope of harm. Yet another commenter argued that injury should be measured by comparing the cost of covered loans against the cost of alternative loans. A number of industry commenters made the similar argument that covered loans cannot cause substantial injury because they do not hurt, and perhaps improve, overall financial health. They presented various surveys and studies that they viewed as providing support for this point. They also contended that the Bureau had erred by assuming that re-borrowing was necessarily injurious and that sustained and repeated use of these loans was necessarily injurious. Another commenter reported having used the Bureau’s financial well-being survey to compare the scores of its customers with the scores of similarly situated consumers in States that restrict payday lending, and reported finding that its customers had similar or better financial well-being scores. The Bureau also received a number of comments arguing that the Bureau had overstated the scope of harm resulting from and frequency of the re-borrowing, defaults, and repossessions caused by the practice. Similarly, commenters argued that there was no evidence that covered loans cause account closures or NSF fees, as stated in the proposed rule. Those comments are addressed above in Market Concerns—Underwriting. 438 Some commenters suggested that because certain small-dollar loan products usually are underwritten, they have a much lower re-borrowing and default rate. Other industry commenters objected to the premise that repeat borrowing constitutes an injury to consumers at all. They argued that the evidence shows extended borrowing is a net benefit to consumers because borrowers get a temporary reprieve from financial difficulty, or because cash-strapped consumers are able to satisfy necessary expenses. Another commenter pointed to a study finding that borrowers who engage in protracted refinancing have higher credit scores than borrowers who use shorter sequences. Still another commenter claimed that reborrowing for title loans should not be regarded as causing an injury because re-borrowing allows consumers to avoid defaulting on other obligations along with such harms as vehicle repossession. Industry commenters also argued that the Bureau should only count re-borrowing as an injury where consumers did not anticipate that outcome. These commenters cited Professor Mann’s study to suggest that many consumers do anticipate they will need to re-borrow to the degree that they end up actually re-borrowing. Consumer groups, by contrast, disputed that premise both conceptually and factually. In particular, they criticized the Mann study by noting that the harm to consumers that results from paying “exorbitant fees” is incurred most acutely by re-borrowers who pay multiple fees, whether or not they end up defaulting. The Bureau received a number of comments on its conclusion that harm results from default. Some of the industry commenters argued that the Bureau overstated the consequences of default. They contended that many payday loans do not affect credit scores because payday lenders do not furnish information to consumer reporting agencies. Commenters also argued that 439 because some payday lenders may not refer accounts to debt collection, the Bureau overstated the harm of default in that manner as well. Some commenters argued that the adverse effects of debt collection practices should not be considered harm for purposes of this rule because harmful collection practices are addressed separately in the Fair Debt Collections Practices Act. One commenter even argued that borrowers benefit from defaulting on these loans, because it means they were able to get free funds that they never ended up having to repay, supposedly without ever experiencing any other negative consequences. Still another commenter argued that for certain title loans the injury resulting from default can be lower than the injury resulting from default on other types of credit, because many title loans are non-recourse loans, which limits the extent of the injury solely to the impact of vehicle repossession. The Bureau received comments contending that it did not have sufficient evidence to substantiate the collateral consequences associated with payday and title loans that have not been underwritten, in particular the frequency and magnitude of other collateral harms from making unaffordable payments, which the Bureau cited as one of the adverse consequences associated with these loans. Commenters also argued that the Bureau’s claim that consumers are injured because they are not able to absorb loan payments on top of major financial obligations and basic living expenses is circular. They argue that consumers use covered loans because they are unable to pay major financial obligations and basic living expenses, and thus the injury the Bureau identified is pre-existing. In other words, commenters argue that the identified injuries are not caused by the identified practice of making such loans without reasonably assessing the borrower’s ability to repay the loan according to its terms, and are instead, caused by borrowers’ 440 preexisting hardship. Commenters similarly suggested that making ability-to-repay assessments does not correlate to the identified injuries and thus the failure to make such assessments is not the cause of those injuries. The Final Rule After reviewing the comments received, and on further consideration, the Bureau is now concluding that the practice of making covered short-term loans without making a reasonable determination of the consumer’s ability to repay the loan according to its terms causes or is likely to cause substantial injury to consumers. As noted in the proposal, borrowers subject to this practice experience injury when covered short-term loans are made without making a reasonable assessment of their ability to repay and they are unable to cover the loan payment on top of major financial obligations and basic living expenses. These injuries include those associated with default, delinquency, and re-borrowing, as well as the negative collateral consequences of being forced to forgo major financial obligations or basic living expenses to cover the unaffordable loan payment. The frequency and magnitude of these types of harms experienced by consumers was discussed at greater length above in Market Concerns—Underwriting. As stated in that discussion, the Bureau does not find that every borrower is necessarily harmed by this practice, because some portion of borrowers may successfully repay these loans after little or no re-borrowing and without incurring collateral harms from so doing (though it bears noting that many of these successful borrowers presumably would qualify for a loan if the lender first made a reasonable assessment that they have the ability to repay it according to its terms). But the Bureau finds that a substantial population of borrowers is harmed, many severely, when they suffer the kinds of injuries just mentioned, which are discussed at greater length above in Market 441 Concerns—Underwriting, as a result of the identified practice of failing to make a reasonable assessment of the borrower’s ability to repay before making the loan. As noted previously, several commenters asserted that the Bureau should only consider that a practice causes substantial injury after discounting certain benefits that borrowers may get from taking out these loans, or after comparing these loans to all other possible alternatives. That approach is not required by the legal standards regarding unfair practices set forth in the statute, FTC precedent, or case law, and the Bureau has concluded that it is not appropriate here. Adopting the suggested approach would over-complicate the analysis and risk “double-counting” certain countervailing benefits (here first in minimizing the nature of the injury and then again in considering the countervailing benefits for consumers or competition). Following the long history of FTC and other judicial precedent, the Bureau has assessed “substantial injury” and “countervailing benefits” separately, and then weighed the two against each other. In this way, the Bureau will fully comply with the statutory requirements because it will not conclude that the identified practice is unfair until after it has concluded that the practice is “injurious in its net effects” because countervailing benefits for consumers or competition do not outweigh the substantial injury.645 The Bureau conducts that analysis and reaches that conclusion below. Generally, the Bureau measures substantial injury by assessing the aggregate injurious consequences that the specific practice causes or is likely to cause for consumers. So, for the practice at issue in this rule, the magnitude of injury is the aggregate total injurious impact of default, delinquency, re-borrowing, and the collateral consequences caused by making unaffordable payments, all of which are the result of lenders failing to assess borrowers’ ability 645 FTC, Policy Statement on Unfairness, Appended to International Harvester Co., 104 F.T.C. 949, 1070 (1984). 442 to repay before making covered short-term loans. Injury is weighed in the aggregate, rather than simply on a consumer-by-consumer basis; and the practice need not injure every consumer if it affects any substantial number of them or if it imposes severe harm on a smaller number of them. In fact, as acknowledged above, the Bureau recognizes that some consumers do not suffer harm from the practice, and for some consumers who are harmed, the benefits to that one consumer might outweigh the harm. This may be true even of some consumers who could not satisfy the ability-to-repay standard. For example, there may be consumers who encounter a windfall after taking out the loan, but before repaying, such that none of the injuries occurs even though at the time the loan was originated the borrower would not have had an ability to repay. There also could be some consumers whose particular circumstances are such that the benefits of having immediate access to funds outweigh the harms resulting from being unable to repay the loan. The Bureau nonetheless includes the injury associated with those borrowers. Of course, the countervailing benefits to consumers are also measured in the aggregate, and the Bureau includes the benefits even to those consumers who, on net, were injured. As to the specific argument that a practice may only be considered injurious if it is worse than all alternatives, this argument is inconsistent with the statute and not grounded in any precedent. Such a requirement would be akin to the view that as long as an alternative practice can be identified that causes even more injury to consumers, then the practice cannot cause substantial injury. As commenters noted, the Bureau has not calculated a precise total dollar figure for the aggregate injury caused by the practice of making covered loans without making a reasonable determination of the borrower’s ability to repay the loan according to its terms. That calculation 443 would be impractical, and it represents a level of exactitude that has never been required of or attained by the FTC and the prudential regulators in regulating identifiable consumer harms under the terms of their UDAP authorities. However, in assessing the aggregate weight of injury, the Bureau was informed by all of the factual background, data, and evidence canvassed above in Market Concerns—Underwriting. When the impact of default, delinquency, reborrowing, and other negative collateral consequences of making unaffordable payments is aggregated among all borrowers for whom lenders do not assess ability to repay before making a covered short-term loan, the sum of that injury is very substantial. It is worth noting what is not included in the Bureau’s weighing of substantial injury. Several commenters believed that the Bureau was considering all covered short-term loans to be injurious. That is not so. The Bureau has determined, more narrowly, that substantial injury is caused or likely to be caused by making a covered short-term loan without reasonably assessing the consumer’s ability to repay according to its terms. Thus, the Bureau is only counting injury to consumers where the lender did not make a reasonable assessment of the borrower’s ability to repay, which as discussed above leads many consumers to experience the harms from default, delinquency, re-borrowing, and other collateral consequences from attempting to avoid these other injuries by making unaffordable payments.646 The Bureau concludes that, contrary to some commenters’ assertions, re-borrowing should be considered consumer injury when the borrower is forced to do so owing to an inability 646 The Bureau notes that some commenters claimed that certain short-term loans made by community banks and credit unions are underwritten and have much lower re-borrowing and default rates. This is consistent with the logic behind the rule, and provides further evidence that a lender’s failure reasonably to assess ability to repay causes the types of harms that the Bureau has identified. 444 to cover the unaffordable payment, basic living expenses, and major financial obligations. The costs of re-borrowing are not a part of the original loan agreement. When a lender makes a loan without assessing ability to repay, and the borrower ultimately does not have enough funds to cover the unaffordable payment, basic living expenses, and major financial obligations, the consumer is forced to choose between three outcomes (default, re-borrowing, or the default avoidance costs of having to forgo basic living expenses or major financial obligations). Each of these outcomes involves “monetary harm,” which is the most traditional form of injury for unfairness analyses.647 Injury can be acute for borrowers when the lender’s failure to assess ability to repay sets off a chain reaction of multiple rounds of re-borrowing, which incur additional fees and perhaps penalty fees as well. After each new loan, the borrower faces an unrepayable balloon payment that leads the borrower to incur additional fees that were not a part of the original agreement. That the borrower incurs the cost of re-borrowing instead of other injuries as perhaps a least-bad option at that juncture (when compared with default, repossession, or forgoing basic living expenses or major financial obligations), does not make the re-borrowing non-injurious. When the loan comes due, the borrower may be able to incur one type of injury over another, but the borrower does not thereby avoid being injured at all. One commenter provided an illustrative example of a borrower who paid $12,960 to borrow $1,020 in principal because the borrower continued to re-borrow the original principal. Each instance of re-borrowing was the result of a new choice between re-borrowing, default, or forgoing expenses, and each of those decisions 647 FTC Statement on Unfairness, 104 F.T.C. 949 (1984). 445 was forced upon the consumer because the original loan was made without assessing the borrower’s ability to repay the loan according to its terms. Note that the Bureau is not, as some commenters stated, addressing in this rulemaking the sustained use of credit, or long-term indebtedness, standing alone. Such matters could bear scrutiny in particular instances under the Bureau’s supervision or enforcement authority. But for purposes of this rulemaking, continued or repetitious re-borrowing is considered injurious for unfairness purposes here because it imposes new costs on the borrower that were not specified in the original loan agreement, and these costs are caused by the lender’s failure to make a reasonable assessment of the borrower’s ability to repay the original loan according to its terms. The Bureau is unpersuaded by commenters’ claims that protracted refinancing is not harmful because credit scores may actually improve for some borrowers. The study that these commenters cite compares borrowers who roll over covered short-term loans with borrowers who do not. Again, the fact that some borrowers may have positive experiences or some particular form of positive outcomes with these loans is not immaterial, but it fails to address the core point of the data about this market, which shows that for a further substantial population of borrowers, the harms experienced from repeated re-borrowing can be quite severe. Moreover, the possibility that one form of the identified injury may be less injurious than another in one particular respect does not prove that the injury identified is not in fact injurious in other respects. When a lender makes covered loans without assessing ability to repay the loan according to its terms, borrowers may be able to incur one form of injury rather than another from amongst the likely set of injuries—again, default, delinquency, re-borrowing, and the collateral consequences of making unaffordable loan payments—and some may be able to 446 mitigate that injury to an appreciable extent or even to nullify its effects, but many borrowers who have taken out an unaffordable loan will not be able to avoid being gravely injured in this situation.648 Similarly, the argument that re-borrowing on title loans is not injurious because it allows borrowers to avoid default, and thus repossession, is unpersuasive. The potential injuries that consumers face in these situations include default, delinquency, re-borrowing, and the collateral consequences of forgoing other basic living expenses or major financial obligations. In these instances, re-borrowing may be less injurious than another greater injury, but many borrowers will still be injured by the impact of re-borrowing as described at greater length above in Market Concerns—Underwriting, including the collateral consequences of attempting to avoid these other injuries by making unaffordable payments. The Bureau recognizes, as commenters suggest, that some borrowers will be able to anticipate, before they take out the first covered short-term loan, that they may have to reborrow. These industry commenters argue that re-borrowing should not be considered harmful to the extent that borrowers could anticipate it happening. But the most relevant data analyzing borrowers’ ability to anticipate re-borrowing supports the conclusion that a high number of borrowers are not, in fact, able to accurately predict the length of their indebtedness to lenders that offer payday loan products. The 2014 study by Professor Mann that asked borrowers about their expectations for reborrowing then compared those with their actual borrowing experience, yielded insights directly 648 Of course, the Bureau notes that all studies comparing credit score outcomes are subject to the caveat that different creditors use different credit scoring models, which are always changing. 447 relevant for this rule.649 As described in the proposal and the Section 1022(b)(2) Analysis, the study found that borrowers who wound up with very long sequences of loans had very rarely expected those long sequences. See the discussion regarding reasonable avoidability below, and the Section 1022(b)(2) Analysis, for more on the Bureau’s interpretation of the Mann study. Thus, the Bureau continues to believe that the response from these industry commenters glosses over the point that many borrowers are not able to anticipate the nature and the likelihood and the magnitude of the harms that may occur through re-borrowing. To the extent that reborrowing imposes new costs on the borrower that were not part of the costs specified in the original loan agreement—including additional fees and the other collateral consequences of attempting to avoid default by making unaffordable payments while forgoing basic living expenses and major financial obligations—the re-borrowing that occurs can create unexpected harm once the borrower has taken out an initial unaffordable loan. Indeed, many consumers who may anticipate some re-borrowing also seem likely to be unable to anticipate the likelihood and severity of these harms, which is a point the Bureau addresses more fully in the section below on whether injury is reasonably avoidable. Moreover, just as the two prongs of “substantial injury” and “reasonably avoidable” are set out as distinct and independent in the statute, the Bureau concludes that even if some borrowers do accurately predict their length of re-borrowing, this would not change the broader conclusion that the practice causes substantial injury in the aggregate. The Bureau also concludes, as addressed above in Market Concerns—Underwriting, that, contrary to the 649 Robert Mann, Assessing the Optimism of Payday Loan Borrowers, 21 Sup. Ct. Econ. Rev. 105 (2014), and correspondence between prof. Mann and Bureau staff described above in Market Concerns—Underwriting. 448 assertions made by some commenters, it did not significantly overestimate the types of injury caused by default, delinquency, re-borrowing, and the negative collateral consequences of making unaffordable payments when it issued the proposed rule. The Bureau is highly dubious of the claim made by some industry commenters that consumers suffer no harm in the event of a default on a covered loan. The Bureau has seen many examples of payday lenders that engage in strenuous efforts, either on their own behalf or by contracting with debt collectors (or selling the debt to debt buyers), to pursue borrowers for payment in the event of default.650 And the commenters did not present any evidence to show the extent to which lenders of covered short-term loans actually do refrain from seeking to collect on overdue debts. Moreover, nothing prevents such third-party debt collectors or debt buyers from reporting the negative information to consumer reporting agencies, which is a technique some collectors use to facilitate collection.651 In any event, the underlying premise is quite implausible. If there were no real consequences to defaulting on these loans, it is difficult to understand why so many borrowers would engage in repeat re-borrowing, rather than simply defaulting. 650 The Bureau has engaged in many investigations that have led to taking a number of enforcement actions against small-dollar lenders for their illegal debt collection practices that were found to be violations of the statutory prohibition against unfair, deceptive, or abusive acts or practices. See, e.g., In the Matter of Money Tree, Inc., File No. 2016-CFPB-0028; In the Matter of EZCORP, Inc., File No. 2015-CFPB-0031; CFPB v. NDG Financial Corp., Case No. 1:15-cv-05211-CM (S.D.N.Y.); In the Matter of ACE Cash Express, Inc., File No. 2014-CFPB-0008; In the Matter of Westlake Servs., LLC, File No. 2015-CFPB-0026. The Bureau has also taken actions against debt collectors, some of which collect in part on small-dollar loans. See, e.g., CFPB v. MacKinnon, et al., Case No. 1:16cv-00880 (W.D.N.Y.). 651 As for whether harmful debt collection practices can constitute cognizable injury here, it would seem that they can if they flow from the identified practice of making covered short-term loans without reasonably assessing the borrower’s ability to repay the loan according to its terms. Although those practices can be addressed through enforcement or rulemaking under the Fair Debt Collection Practices Act, they are also a natural consequence of the harms that consumers experience from receiving unaffordable loans that they are unable to repay. 449 The Bureau also finds that its assessment of injury should include repossessions resulting from failing to assess ability to repay before making covered vehicle title loans. As noted above, some industry commenters claimed that repossession is not harmful, or not as harmful as the Bureau indicated in its proposal. They rest this argument on two claims. First, they contend that most borrowers can find other means of transportation, citing what they present as a supportive survey, and thus would not be harmed by the loss of their vehicle. Second, they contend that the extent of the direct economic loss that borrowers sustain by having their vehicle repossessed is relatively insignificant. On the first point, the potential consequences of the loss of a vehicle depend on the transportation needs of the borrower’s household and the available transportation alternatives. According to two surveys of title loan borrowers, 15 percent report that they would have no way to get to work or school if they lost their vehicle to repossession.652 For these borrowers, the effects of repossession could thus be catastrophic from an economic standpoint, particular in rural areas or in urban areas where public transportation is not reasonably available. And more than one-third (35 percent) of borrowers pledge the title to the only working vehicle in the household.653 Even those with a second vehicle or who are able to get rides from friends or take public transportation would presumably experience significant inconvenience or even hardship from the loss of a vehicle. This hardship goes beyond simply getting to work or school, and would as a practical matter also adversely affect the borrower’s ability to conduct their ordinary household affairs, such as obtaining food or medicine or other necessary services. The 652 Fritzdixon, et al., at 1029-1030; Pew Charitable Trusts, Auto Title Loans: Market Practices and Borrowers’ Experiences, at 14 (2015), http://www.pewtrusts.org/~/media/assets/2015/03/autotitleloansreport.pdf. 653 Pew 2015. 450 commenters countered that borrowers often can find other means of transportation, citing what they present as a supportive survey. Their interpretation of the data is not convincing, however, as even the authors of the survey cautioned against making simplistic calculations about factors and probabilities that are intertwined in the analysis, and which thus may considerably understate the incidence of hardship, especially for more economically vulnerable populations. As to the second point about the extent of the direct economic loss, the commenters rest this argument either on the low average value of collateralized vehicles or on their claim that some borrowers deliberately choose to liquidate the value of the vehicle by taking out a title loan and then promptly abandoning the vehicle to repossession. While some vehicles used for collateral may not have high value, they still can be crucial as the consumer’s principal means of transportation to and from work or to conduct everyday affairs such as obtaining medical care or buying groceries, medicine, and other essentials. The Bureau describes the harms of repossession in more detail both in Market Concerns—Underwriting and the section-by-section analysis for § 1041.6. The Bureau also finds unpersuasive the assertion made by some commenters that a significant population of consumers would take out a title loan and then intentionally abandon the vehicle instead of just selling it, especially in light of the observations made in Market Concerns—Underwriting that title lenders usually only make loans where the value of the collateral exceeds the principal. Indeed, it appears implausible that consumers would choose to dispose of a vehicle by this means rather than simply selling the vehicle, as the latter approach very likely would usually yield more funds without involving the consumer in any adverse risks or costs of collections activities or repossession fees. It may be that some borrowers take out a 451 title loan and immediately default on it, perhaps even intentionally, and such borrowers may not necessarily experience all of the same harms as other borrowers whose vehicles are repossessed. But no evidence plausibly suggests that this alleged population is at all significant, and thus this fact does not change the Bureau’s overarching conclusion. As for the commenter who argued that the stress associated with repossession is no worse than other forms of financial stress, this argument is speculative and unpersuasive, and at least implicitly acknowledges the fact that potential psychological injury does accompany the threat of repossession. The Bureau also rejects the claim made by some commenters that its arguments about substantial injury are circular because the injuries identified were primarily caused by the original financial hardship that induced the borrower to seek a covered loan, rather than by the covered loan itself. This is a variant on the argument that the real harm to consumers does not flow from the identified practice of failing to underwrite these loans in a reasonable manner but from the fact that many consumers lack the money to meet their obligations. First, to the extent this argument seeks to rely on the benefits provided by access to credit through covered loans in order to cover the borrower’s expenses, or is an exercise in weighing those benefits against the injuries associated with the harm, it is most appropriately treated in the section below on “countervailing benefits.” But more to the point, the Bureau finds that the specific injuries which flow from default, delinquency, re-borrowing, and the collateral consequences of making unaffordable payments, including forgoing major financial obligations or basic living expenses in order to avoid default, are not caused by the borrower’s pre-existing financial hardship for one key reason: these injuries flow from the loan itself and the fact that it was made without reasonably assessing the borrower’s ability to repay the loan according to its terms. These 452 outcomes would not have occurred without the lender engaging in the identified practice of making such loans in such manner. The borrower would have faced other difficulties flowing from her distressed circumstances, but not the harms identified here. In other words, the fact that many consumers are in financial difficulty when they seek out a covered loan—a fact the Bureau has repeatedly recognized—does not mean they are not injured by the identified practice. For certain individual borrowers in particular situations, being able to replace a default on a different obligation with the injury identified in this section might seem to be worthwhile. But the right place to address that potential trade-off is when the analysis turns to assessing whether countervailing benefits outweigh the injury, in the aggregate rather than on an individual borrower basis—matters that are discussed further below. In any event, the pre-existing financial stress of many consumers does not relieve lenders of responsibility for engaging in practices that are unfair or abusive. As the court in FTC v. Neovi stated, the contribution of “independent causal agents . . . do[es] not magically erase the role” of lenders’ in causing the harm.654 When lenders do not assess ability to repay before making loans, they end up making loans to some borrowers who lack the ability to repay. The fact that these borrowers who obtain unaffordable loans will default, become delinquent, reborrow, or experience negative collateral consequences is a natural result of the practice that lenders should expect. In sum, based on the analysis presented here and above in the section on Market Concerns—Underwriting, and upon further consideration after reviewing the high volume of 654 FTC v. Neovi, 604 F.3d 1150, 1155 (9th Cir. 2010). In fact, the argument here is even weaker than that rejected in Neovi, where the claim was that intervening causal factors had rendered the cause identified by the agency insufficiently proximate. Here the alleged causal factor cited by the commenters is not even an intervening factor. 453 comments received from the public, the Bureau concludes that the identified practice causes or is likely to cause substantial injury. Injury not reasonably avoidable The Bureau’s Proposal The second prong of the statutory definition of unfairness is that the “substantial injury” to consumers “is not reasonably avoidable by consumers.” The Bureau proposed to interpret this requirement to mean that unless consumers have reason generally to anticipate the likelihood and severity of the injury and the practical means to avoid it, the injury is not reasonably avoidable. Under the proposed rule, the Bureau stated that in a significant proportion of cases, consumers appear to be unable to reasonably avoid the substantial injuries caused or likely to be caused by the identified practice. Prior to entering into a payday, single-payment vehicle title, or other covered short-term loan, many consumers do not reasonably anticipate the likelihood and severity of the injuries that frequently result from such unaffordable loans, and after entering into the loan, consumers do not have the practical means to avoid the injuries that result from being unable to repay it. As stated in the proposal, many consumers seem unable to reasonably anticipate the likelihood and severity of the consequences of being unable to repay a loan that is unaffordable according to its terms. As discussed in the proposal, the typical consumer is likely generally aware that taking out any loan can lead to adverse consequences if the loan is not repaid, but is not likely to be familiar with all of the harms that can flow from a loan that is made without a reasonable assessment that the borrower will be able to repay it according to its terms. Some additional harms beyond the costs incurred on the loan can include, for example, the risk of 454 accumulating penalty fees on their bank account, the potential loss of their account, or (for title loans), or the risk of aggressive collections. Moreover, even if consumers recognize these harms as possibilities, many are likely not to have sufficient information to understand the frequency with which these adverse effects may occur to borrowers who are affected by the identified practice or the severity of the consequences befalling a typical borrower who obtains an unaffordable loan. An especially compelling example of how consumers may be prone to error in making reasonable evaluations about the injuries to which they are exposed by the identified practice is the substantial number of consumers who re-borrow, many of them repeatedly, prior to eventually defaulting on these loans. But unless consumers are reasonably aware of the likelihood and severity of these injuries, it would not be reasonable for them to make special efforts to avoid such injuries where they are not in position to accurately evaluate the risks. This may be especially the case where the lender qualifies them for a loan without making a reasonable assessment of their ability to repay, as many consumers would be unlikely to expect that lenders would intentionally offer them an unaffordable loan that they would likely be unable to repay. That is not to say that every consumer must understand everything about the potential risks or must be able to anticipate these risks with mathematical precision. Instead, it is only to say that consumers must have a sense of the order of magnitude of the risk, both in terms of its likely frequency and its likely severity. Yet the Bureau also noted in the proposal that in analyzing reasonable avoidability under the FTC Act unfairness standard, the FTC and other 455 agencies have at times focused on factors such as the vulnerability of affected consumers,655 as well as those consumers’ perception of the availability of alternative products.656 Likewise, the Bureau stated that the substantial injury from covered short-term loans may not be reasonably avoidable in part because of the precarious financial situation of many consumers at the time they take out such loans and their belief that searching for potential alternatives will be fruitless and costly. As discussed in the proposal, consumers who take out payday or single-payment vehicle title loans typically have tried and failed to obtain other forms of credit before turning to these covered loans as a last resort. Thus, based on their prior negative experience with attempting to obtain credit, they may reasonably perceive that alternative options would not be available. Consumers facing an imminent liquidity crisis may also reasonably believe that their situation is so dire that they do not have time to shop for alternatives and that doing so could prove costly. 655 See, e.g., FTC Policy Statement on Unfairness, 104 FTC at 1074 (noting that the FTC may consider the “exercise [of] undue influence over highly susceptible classes of purchasers”); Mortgage Assistance Relief Services Rule, 75 FR 75092, 75117 (Dec. 1, 2010) (emphasizing the “financially distressed” condition of consumers “who often are desperate for any solution to their mortgage problems and thus are vulnerable to providers’ purported solutions”); Telemarketing Sales Rule, 75 FR 48458, 48487 (Aug. 10, 2010) (concluding that injury from debt relief programs was not reasonably avoidable in part because “purchasers of debt relief services typically are in serious financial straits and thus are particularly vulnerable” to the “glowing claims” of service providers); Funeral Industry Practices Rule, 47 FR 42260, 42262 (Sept. 24, 1982) (citing characteristics which place the consumer in a disadvantaged bargaining position relative to the funeral director, leaving the consumer vulnerable to unfair and deceptive practices, and causing consumers to have little knowledge of legal requirements and available alternatives). The Funeral Industry Practices Rule and amendments were upheld in the Fourth and Third Circuits. See Harry and Bryant Co. v. FTC, 726 F.2d 993 (4th Cir. 1984); Pennsylvania Funeral Directors Ass’n, Inc. v. FTC, 41 F.3d 81 (3d Cir. 1994). In the Subprime Credit Card Practices Rule—in which three Federal banking regulators identified as unfair certain practices being routinely followed by credit card issuers—the Federal Reserve Board, OTS, and NCUA noted their concern that subprime credit cards “are typically marketed to vulnerable consumers whose credit histories or other characteristics prevent them from obtaining less expensive credit products.” 74 FR 5498, 5539 (Jan. 29, 2009). 656 In the HPML Rule, the Federal Reserve Board discussed how subprime consumers “accept loans knowing they may have difficulty affording the payments because they reasonably believe a more affordable loan will not be available to them,” how “taking more time to shop can be costly, especially for the borrower in a financial pinch,” and how because of these factors “borrowers often make a reasoned decision to accept unfavorable terms.” 73 FR 44522, 44542 (July 30, 2008). 456 The Bureau also stated in the proposal that consumer predictions about their experience with covered short-term loans may be overly optimistic, especially if they are unaware of the risks posed by lenders making these loans without reasonably assessing the borrower’s ability to repay the loan according to its terms. In particular, consumers who experience long sequences of loans often do not expect those long sequences to occur when they make their initial borrowing decision. As detailed above in Market Concerns—Underwriting, empirical evidence suggests that consumers are best able to predict accurately the duration of their borrowing if they repay after little or no re-borrowing, though many underestimate the expected duration while others overestimate it. Notably, borrowers who end up in extended loan sequences are especially likely to err in their predictions of how long their loan sequences will last, usually taking the form of underestimating the expected duration. So consumers are particularly poor at predicting long sequences of loans, a fact that does not appear to differ for those borrowers who have past borrowing experience.657 As discussed in the proposal, the Bureau observes other factors that prevent consumers from reasonably anticipating and avoiding the substantial injury caused by unaffordable shortterm loans. Such loans involve a basic mismatch between how they appear to function as shortterm credit and how they are actually designed and intended by lenders, as part of their business model, to function in long sequences of re-borrowing for a substantial population of consumers. Lenders present these loans as short-term, liquidity-enhancing products that consumers can use 657 As noted in Market Concerns—Underwriting, it appears that some consumers are able to accurately predict that they will need to re-borrow one or two times, and decide to take out the loan regardless of the additional cost of this limited amount of re-borrowing. Accordingly, such costs do not count as substantial injury that is not reasonably avoidable. 457 to bridge an income shortfall until their next paycheck. But in practice, across the universe of borrowers, these loans often do not operate that way. The term of the loan, its balloon-payment structure, and the common use of leveraged payment mechanisms, including vehicle security, all tend to magnify the risks and harms to the borrower. The disparity between how these loans appear to function and how they actually function creates difficulties for consumers in estimating with any accuracy how long they will remain in debt and how much they will ultimately pay for the initial extension of credit. Lenders who make covered short-term loans without reasonably assessing the borrower’s ability to repay the loan according to its terms, to borrowers who often do not reasonably anticipate the likelihood and severity of the risks posed, often further magnify these risks through the way they market the option of repeat borrowing. Payday lenders and title lenders typically present only two options: the re-borrowing option, with its costs limited to another set of fees but no repayment of principal, and the full repayment option of requiring the entire balloon payment to be repaid all at once, with no options offered in between these two. Low-cost repayment or amortization options are typically not presented or are obscured, even where they may be required to be available under State law. Even consumers who are delinquent and have further demonstrated their inability to repay the loan according to its terms are encouraged to reborrow, which leads many consumers to engage in extensive re-borrowing even where they eventually wind up in default. For many re-borrowers, the upshot is that they end up making repeated payments that become increasingly unaffordable in the aggregate over time, even though a substantial number of them still will sustain the harms associated with default. 458 The proposal stated that not only are consumers unable to reasonably anticipate the likelihood and severity of many of these potential harms before entering into a payday or title loan, but after they have entered into a loan, they do not have any practical means to avoid the injuries that will occur if the loan proves to be unaffordable. Consumers who obtain a covered short-term loan that is beyond their ability to repay confront the harms of default, delinquency, re-borrowing, or the collateral consequences of making unaffordable payments that would cause them to miss payments on their major financial obligations and basic living expenses. They can make choices among these competing harms, but once they are facing an unaffordable payment, some form of substantial injury is almost inevitable regardless of what actions they take in that situation. And as discussed in the proposal, lenders engage in a variety of practices that further increase the likelihood and degree of harm, for instance by encouraging additional re-borrowing with its attendant costs even for consumers who are already experiencing substantial difficulties as they are mired in extended loan sequences, and by engaging in payment collection practices that are likely to cause consumers to incur substantial additional fees beyond what they already owe on the terms of the existing loan. Comments Received The Bureau received many comments on whether the substantial injury identified was reasonably avoidable by consumers. A number of commenters opined on the legal standards the Bureau should use when assessing reasonable avoidability. One commenter argued that the proper standard for assessing whether injury is reasonably avoidable is whether the consumer has the ability to anticipate the impending harm and has means to avoid it. In other words, even if consumers do not actually tend to anticipate the likelihood and severity of the impending harm, it 459 could still be viewed as reasonably avoidable as long as knowledge of the impending harm is conceptually attainable. Various parties submitted comments to the Bureau arguing that borrowers can in fact accurately predict the consequences of getting a covered loan. This point is addressed more fully above in the Market Concerns—Underwriting. One commenter claimed that a study showed borrowers who have previously used title loans are more capable of anticipating how long they will be indebted, predicting six or more additional months of indebtedness as compared to consumers who had never used title loans. Some industry commenters also claimed that borrowers must be able to anticipate the consequences of failing to repay a title loan because title loans are simple products, and the use of vehicles as collateral to secure the loan is a defining and obvious feature of these loans. Commenters made similar arguments about payday loans. Various industry commenters claimed that consumers do have the means to avoid the injuries that are caused or likely to be caused by the identified practice. Many of these commenters argued that consumers have the means to avoid the injury simply by forgoing the first covered loan altogether. Commenters argued that such consumers could turn instead to friends and family. They also argued that consumers could instead obtain other forms of credit, such as a traditional non-recourse pawn loan. Others noted that there are further ways to avoid these injuries even after having taken out the first covered loan. Some argued that borrowers could simply budget carefully to ensure timely payment, could take advantage of legal protections that may be available in some States that allow them to lower or extend payments, or could obtain credit counseling or other assistance. Others contended that borrowers could 460 minimize or avoid the harms they experience from these loans by engaging in strategic default, asserting that defaults on such loans do not lead to any further negative consequences for the borrower. Similarly, some commenters claimed that where consumers have consented to leveraged payment mechanisms such as post-dated checks or automatic account withdrawals, they could avoid consequent harms by simply withdrawing their consent at a later point. One commenter asserted that the Bureau falsely assumed that any re-borrowing was a consequence of borrowers having no other credit options. This commenter regarded the data as establishing instead that borrowers do have other options and may have reasons why they would choose to re-borrow even where they can afford to repay the prior loan. In response to the Bureau’s claim that it is reasonable for many consumers in typical circumstances to fail to shop for alternative forms of credit, one commenter argued that whenever alternatives are available, a reasonable consumer would shop for them and obtain them. In other words, even if borrowers do not generally tend to shop for alternatives, any injury could still be reasonably avoidable if consumers could have exercised the ability to shop. Other commenters argued that acts or practices can only be unfair if the lender’s actions alone caused the injury not to be avoidable. In other words, if any of the reasons that consumers could not avoid the harm caused by a lender was not itself also caused by the lender, the act or practice cannot be unfair. Commenters also argued that injury is reasonably avoidable when consumers have a “‘free and informed choice’ not to purchase the product,” citing FTC v. Neovi.658 At least one commenter took the opposite position, arguing that consumers’ financial 658 604 F.3d 1150, 1158 (9th Cir. 2010). 461 situations can give rise to a reasonable conclusion that an injury from the identified practice is not reasonably avoidable. Alternatively, consumer groups observed that whether consumers could have anticipated the injury is irrelevant to whether the injury is reasonably avoidable if consumers lack the means to avoid the injury even if it were to be anticipated. They argued that even if some borrowers can more accurately anticipate the length of their indebtedness, they might nonetheless fail to understand the full range of injuries that can often occur at the end of the sequence, which the Bureau noted in its proposed rule, and which are discussed at greater length above in Market Concerns—Underwriting. Where consumers do not understand that full range of potential harms, such injury is not reasonably avoidable. The Final Rule After reviewing the comments received and taking into account the factual analysis of how such loans work in practice as set forth above in Market Concerns—Underwriting, the Bureau concludes that the substantial injury caused by the identified practice is not reasonably avoidable by consumers. The specific question here is whether the practice at issue causes substantial injury to consumers “which is not reasonably avoidable by consumers.”659 Starting with the established point, already discussed, that there is substantial injury to consumers from making covered shortterm loans without reasonably assessing the borrower’s ability to repay the loan according to its terms. In approaching the “reasonably avoidable” criterion, the Bureau is tasked by Congress to ask whether, if lenders engage in the practice of making these loans available without assessing 659 12 U.S.C. 5531(c)(1)(A). 462 ability to repay, the resulting injuries are reasonably avoidable by consumers acting on their own. As noted above, the Bureau interprets this criterion to mean that unless consumers have reason generally to anticipate the likelihood and severity of the injury, and the practical means to avoid it, the injury is not reasonably avoidable. As also noted earlier, the D.C. Circuit has held that the presence of a market failure or imperfection is highly relevant to the “reasonably avoidable” inquiry, as it may hinder consumers’ free-market decisions and prevent the forces of supply and demand from maximizing benefits and minimizing costs. In addressing this issue, the Bureau does not accept, and the FTC and prudential regulators have never been satisfied with, the notion that injury is avoidable just because a consumer has the right not to enter the market in the first place. No precedent supports the idea that the existence of such a right is by itself an answer to the “reasonably avoidable” issue. Indeed, a consumer generally has a right to decline to initiate the purchase of any product or service, and if the mere existence of that right were the end of the “reasonably avoidable” question, then no act or practice by a seller would ever be subject to regulation on unfairness grounds. The Bureau specifically rejects the arguments advanced by some commenters who contended that acts or practices can only be unfair if the lender’s actions alone caused the injury not to be reasonably avoidable. The practice at issue is the making of covered short-term loans without reasonably assessing the borrower’s ability to repay the loan according to its terms. The making of such loans in this manner—which is an action that is entirely within the lender’s control—is the act that causes injury to consumers, which, as discussed above, is not reasonably avoidable by consumers. The lender need not also be the source that has created all the reasons 463 why that injury is not reasonably avoidable, given the ordinary circumstances of typical consumers, including their general understanding of the likelihood and severity of the risks posed. Nonetheless, as discussed in the proposal and above, as well as in the section on Market Concerns—Underwriting, the Bureau has concluded that the manner in which lenders structure these products—including the term of the loan, its balloon-payment structure, and the common use of leveraged payment mechanisms, and vehicle security—likely contributes significantly to the market failure660 and market imperfections that the Bureau has observed. Commenters opposing the proposed rule who addressed the “reasonably avoidable” criterion generally took the position that the consumers who seek these loans are nonetheless fully capable of reasonably avoiding these injuries in order to protect their own self-interest. Many of these positions were based on their intuitive descriptions or stories about what consumers understand about the risks of loans that they do not have the ability to repay, and how consumer decision-making works. Their intuition is inconsistent with the evidence on which the Bureau has based its findings that the injury is not reasonably avoidable, including survey data showing that past borrowing experience is not indicative of increased understanding of product use. Indeed, those who had borrowed the most in the past did not do a better job of predicting their future use, and as Professor Mann noted, “heavy users of the product tend to be those that understand least what is likely to happen to them.”661 Whereas various commenters cited Professor Mann’s study to show that most consumers are able to make accurate predictions about their extent of re-borrowing, as noted above in 660 661 See Section 1022(b)(2) Analysis in part VII. Mann, Assessing the Optimism, 21 Supreme Court Econ. Rev. at 127. 464 Market Concerns—Underwriting, this was mostly driven by borrowers who anticipate and experience relatively short sequences and manage to repay very quickly. The Bureau appreciates that, as commenters pointed out, Mann’s study, discussed below and in the Section 1022(b)(2) Analysis, suggest that some borrowers are better able to predict their likelihood of re-borrowing. Nonetheless, the Bureau’s primary concern is for those longerterm borrowers who find themselves in extended loan sequences and thereby experience the various harms that are associated with a longer cycle of re-borrowing. For those borrowers, the picture is quite different, and their ability to estimate accurately what will happen to them when they take out a payday loan is quite limited. As Mann noted, very few of those borrowers who experienced the longest sequences anticipated that they would end up in a period of prolonged indebtedness, and in fact “both the likelihood of unexpectedly late payment and the proportionate size of the error increase substantially with the length of the borrower’s prediction.”662 Nor does their accuracy appear to improve with more experience; as he noted in his paper, “heavy users of the product tend to be those that understand least what is likely to happen to them.”663 The further discussion in the comments of Professor Mann’s study, including his own submission, did not alter these results, for as he noted, “the absolute size of the errors is largest for those with the longest sequences,” and “the borrowers who have borrowed the most are those who are in the most dire financial distress, and consequently least able to predict their future liquidity.” And as the Bureau discusses at length in Market Concerns—Underwriting, and in the Section 1022(b)(2) Analysis, multiple different conclusions can be made based on Mann’s 662 663 Id. Id. 465 findings. Certainly, it is possible that many borrowers accurately anticipate their debt durations, as Mann asserts in both his 2013 paper and comment to the proposed rule. However, Mann’s study supports the conclusions that most of those borrowers with long duration sequences did not accurately anticipate this outcome; that a large share of borrowers who anticipated no reborrowing remain in debt for multiple loans, with many being unable to even offer a guess as to the duration of their indebtedness, let alone a precise prediction; and that there appears to be no discernable relationship between borrowers’ individual expectations, and their ultimate outcomes. Indeed, the 2013 Mann study showed that of the borrowers who remained in debt at least 140 days (10 bi-weekly loans), a hundred percent had underestimated their times in debt, with the average borrower in this group spending 119 more days in debt than anticipated (i.e., the equivalent to eight and half unanticipated rollovers.664 Meanwhile, over 95 percent of the borrowers who spent 90 or more days in debt had underestimated their time in debt, spending an average of 92 more days in debt than anticipated (i.e., the equivalent to six and a half unanticipated rollovers). And as described in the proposal, Mann (2014) found that borrowers who wound up with very long sequences of loans had rarely expected those long sequences; that only 40 percent of respondents expected to re-borrow at all even though over 70 percent would actually re-borrow; and, that borrowers did not appear to become better at predicting their own borrowing. Thus, while many individuals appear to have anticipated short durations of use with reasonable accuracy, the Bureau is persuaded that virtually none anticipated long durations with 664 Mann, Ronald. 2013. “Assessing the Optimism of Payday Loan Borrowers.” Sup. Ct. Economic Rev., 21(1): 105-132. 466 anything approaching reasonable accuracy. The harms associated with the long durations outside the scope of the consumers’ anticipation capabilities are precisely the market failure that the final rule seeks to address. The heart of the matter here is consumer perception of risk, and whether borrowers are in position to gauge the likelihood and severity of the risks they incur by taking out covered shortterm loans in the absence of any reasonable assessment of their ability to repay those loans according to their terms. It appears based on the evidence that many consumers do not understand or perceive the probability that certain harms will occur, including the substantial injury that can flow from default, re-borrowing, and the negative collateral consequences of making unaffordable payments as described above in Market Concerns—Underwriting. Other features of these loans—including their term, balloon-payment structure, and the common use of leveraged payment mechanisms or vehicle security—tend to magnify the risks posed when they are obliged to repay the full amount when the loan comes due, on top of all their other existing obligations. Whether consumers can “reasonably avoid” the injuries that flow from the identified practice will depend, in the first instance, on whether they understand the likelihood and the severity of these risks so that they are able to make a reasoned judgment about whether to incur or to forgo such risks. As the Bureau perceives the matter, based on its experience and expertise in addressing consumer financial behavior, the observed evidence described more fully in the Section 1022(b)(2) Analysis and Market Concerns—Underwriting indicates that a large number of consumers do not understand even generally the likelihood and severity of these risks. There are a variety of explanations why consumers will take out covered short-term loans that they actually lack the ability to repay without fully appreciating the nature and magnitude of 467 the risks involved. As the Bureau discussed in connection with the proposed rule, and as described further in Market Concerns—Underwriting and the paragraphs above, the way the product is marketed and presented to them is calculated to obscure the risks. And while many consumers may operate as fully informed rational actors, and thus be able to predict their repayment capacity, those consumers who lack the ability to repay (and thus are most likely to be harmed by the identified practice) tend to be overly optimistic, at least when they are operating under short-term financial stress. The data available from Professor Mann, for example, tends to confirm that a substantial proportion of borrowers—those in extended loan sequences, who are the most vulnerable to harm—have great difficulty in predicting their own repayment capability. And the widespread industry practice of framing covered loans as short-term obligations, even though lenders know that their business model depends on these loans becoming long-term cycles of debt for many consumers, likely exacerbates these misimpressions among borrowers. Some of the particular behavioral obstacles to consumers’ ability to fully understand the magnitude and likelihood of the risks they face, including the difficulties of assessing their likelihood of nonpayment and of appreciating the severity of injury they would face in such an event, are discussed at greater length above in Market Concerns—Underwriting and the Section 1022(b)(2) Analysis. Once again, the economic literature, including studies in the field of behavioral economics but also those modeled on rational expectations, suggests that these considerations are particularly acute for consumers who are under financial stress (such as consumers who lack the ability to repay a covered loan) and under acute time pressure. These considerations, which are well known to economists, may especially degrade the borrower’s ability to reliably evaluate the risks presented in their circumstances. 468 Each of the multiple factors listed in the proposal and above in Market Concerns— Underwriting that may limit consumers’ ability to appreciate the magnitude and severity of risks may operate differently, and to different degrees, on particular consumers. Whether borrowers do not actually have any alternatives, do not perceive any alternatives, do not have time to shop for alternatives, or cannot otherwise anticipate the probability or extent of the harm, it is demonstrably true that a substantial population of consumers to whom industry has traditionally marketed these loans, and who lack the ability to repay, will sign up for a covered loan and, in the aggregate, will suffer substantial injury as a consequence of the identified practice. Stated differently, it is a plausible inference that the substantial injury many reasonable consumers sustain, as actually observed in the marketplace for covered short-term loans, is not in fact avoided by normal consumer decision-making. In its current form, the market does not appear to be self-correcting. Furthermore, once borrowers find themselves obligated on a loan they cannot afford to repay, the resulting injury is generally not reasonably avoidable at any point thereafter. But the Bureau acknowledges that there are limited exceptions to this rule. For example, there may be consumers who encounter a windfall after taking out the loan, but before repaying, such that none of the injuries occurs even though at the time the loan was originated the borrower would not have had an ability to repay. The most common injury is re-borrowing, which operates as a mechanism that is intended (though often unsuccessfully) to manage the potential injuries caused by the identified practice, rather than as an effective escape from injury. Most consumers, after having taken out a covered short-term loan they cannot afford to repay, are confronted with a choice of which injury to incur—default, delinquency, re-borrowing, or collateral consequences 469 of making unaffordable payments, including forgoing essential expenses—or how to minimize the accumulated harm from more than one such injuries. Merely having a choice among an array of injuries does not give borrowers the ability to reasonably avoid any injury. Some industry commenters argued that consumers have other options available to them, so those who re-borrow are choosing to do so. It bears note that this argument is to some extent inconsistent with those made elsewhere by the same and other industry commenters, who argue that borrowers would be left worse off if they did not have access to covered loans because they lack other plausible options. In addition, the Bureau has found that many such alternatives are not widely available to these borrowers, who may not find them to be desirable alternatives in any event. Moreover, here again the Bureau notes that once a consumer has taken out an unaffordable loan, the decision to re-borrow becomes an unsatisfactory choice among the injuries produced by such loans, as just discussed above, rather than an unfettered choice among various alternatives, as might have been the case before the first unaffordable loan was obtained. As for the commenters who suggested consumers can avoid harm by simply defaulting on the loan, this approach would not achieve that objective because the Bureau has identified default as an injury for all the reasons discussed above in Market Concerns—Underwriting. Again, a choice between types of injury is not a mechanism for reasonably avoiding all injury. And the commenters who suggested that such consumers could avoid any further harm by withdrawing their consent to a leveraged payment mechanism they previously granted to the lender are equally wide of the mark. First, for storefront payday loans and other covered shortterm loans that require the borrower to give the lender a post-dated check, it is impractical for the consumer to withdraw consent to that payment mechanism after the loan has been made. 470 Because that mechanism is a condition precedent to making the loan, attempting to withdraw consent later would either be ineffectual or would lead directly to default. As for the leveraged payment mechanism of automated withdrawals from the borrower’s account, such as are commonly granted with on-line covered loans, as discussed in Market Concerns—Payments, consumers experience many practical difficulties in successfully withdrawing their consent afterthe-fact. Even for those borrowers who do manage to avoid that harm, there are other harms attributable to default, as laid out above in Market Concerns—Underwriting. Accordingly, the Bureau concludes that the practice of making covered short-term loans without reasonably assessing the borrower’s ability to repay the loan according to its terms causes substantial injury to consumers, which is not reasonably avoidable by them. Injury not outweighed by countervailing benefits to consumers or to competition The Bureau’s Proposal As noted in part IV and in the proposal, the Bureau’s interpretation of the various prongs of the unfairness test is informed by the FTC Act, the FTC Policy Statement on Unfairness, and FTC and other Federal agency rulemakings and related case law. Under those authorities, it is generally appropriate for purposes of the “countervailing benefits” prong of the unfairness standard to consider both the costs of imposing a remedy and any benefits that consumers enjoy as a result of the practice, but the determination does not require a precise quantitative analysis of the benefits and the costs.665 The Bureau stated in the proposal that it appears that the practice of making payday, 665 FTC Policy Statement on Unfairness; Am. Fin. Svcs. Assoc. v. FTC, 767 F.2d 957, 986 (D.C. Cir. 1985) (“Petitioners would require that the Commission’s predictions or conclusions be based on a rigorous, quantitative economic analysis. There is, however, no basis for imposing such a requirement.”). 471 single-payment vehicle title, and other covered loans without reasonably assessing that the consumer will have the ability to repay the loan according to its terms does not result in benefits to consumers or competition that outweigh the substantial injury that consumers cannot reasonably avoid. As discussed in the proposal and for the reasons stated here, the amount of injury that is caused by the unfair practice, in the aggregate, appears to be quite substantial. Although some consumers may be able to avoid the injury, as noted above, a significant number of consumers who end up in very long loan sequences can incur severe financial injuries that are not reasonably avoidable. Moreover, the proposal stated that some consumers whose short-term loans turn into short- to medium-length loan sequences incur various degrees of injury ranging from modest to severe depending on the particular consumer’s circumstances (such as the specific loan terms, whether and how much the consumer expected to re-borrow, and the extent to which the consumer incurred any collateral harms from making unaffordable payments). In addition, many borrowers who default or become delinquent on the loan also may experience substantial injury that is not reasonably avoidable as a result of the identified practice. Against this very significant amount of harm, the Bureau recognized that it must weigh several potential countervailing benefits to consumers or competition of the practice in assessing whether the practice is unfair. Accordingly, in the proposal the Bureau divided consumers into several groups of different borrowing experiences to analyze whether the practice of extending covered loans without determining that the consumer has the ability to repay the loan yielded countervailing benefits to consumers. The first group consisted of borrowers who repay their loans without re-borrowing. The Bureau referred to these borrowers as “repayers” for purposes of this countervailing benefits 472 analysis. As discussed in the proposal, 22 percent of payday loan sequences and 12 percent of single-payment vehicle title loan sequences end with the consumer repaying the initial loan without re-borrowing. The Bureau stated that many of these consumers may reasonably be determined, before getting a loan, to have the ability to repay their loan, such that the ability-torepay requirement in the proposed rule would not have a significant impact on their eligibility for this type of credit. The Bureau stated that, at most, it would reduce somewhat the speed and convenience of applying for a loan under the current practice, though it was not clear that any such differential would be a material factor for any prospective borrowers. The Bureau stated that, under the status quo, the median borrower lives five miles from the nearest payday store. Consumers generally can obtain payday loans simply by traveling to the store and showing a pay stub and evidence of a checking account; online payday lenders may require even less of a showing in order to extend a loan. For title loans, all that is generally required is that the consumer owns their vehicle outright without any encumbrance. The proposal stated that there could be a significant contraction in the number of payday stores if lenders were required to assess consumers’ ability to repay in the manner required by the proposal, but the Bureau projected that 93 to 95 percent of borrowers would not have to travel more than five additional miles to get a loan. Lenders likely would have to require more information and documentation from the consumer. Indeed, under the proposed rule consumers would have been required in certain circumstances to provide documentation of their income for a longer period of time than their last pay stub. Under the proposal, consumers would also be required to complete a written statement with respect to their expected future income and major financial obligations. 473 Moreover, when a lender makes a loan without determining a consumer’s ability to repay the loan according to its terms, the lender can make the loan upon obtaining a consumer’s pay stub or vehicle title. The Bureau acknowledged in the proposal that lending under the proposed rule may not be so immediate, though automated underwriting systems could achieve similar levels of speed. If lenders assessed consumers’ ability to repay as stated in the proposal, they would secure extrinsic data, such as a consumer report from a nationwide consumer reporting agency, which could slow the process down somewhat. Indeed, under the proposed rule lenders would be required to review the consumer’s borrowing history using the lender’s own records and a report from a registered information system, and lenders would also be required to review a credit report from a nationwide consumer reporting agency. Using this information, along with verified income, under the proposed rule lenders would have to project the consumer’s residual income. As discussed in the analysis contained in the proposal, the proposed rule was designed to enable lenders to obtain electronic income verification, to use a model to estimate rental expenses, and to automate the process of securing additional information and assessing the consumer’s ability to repay. The Bureau anticipated that consumers who are able to demonstrate the ability to repay under the proposed rule would be able to obtain credit to a similar extent as they did in the current market. While the speed and convenience fostered by the current practice may be somewhat reduced for these consumers, the Bureau concluded in the proposal that the proposed requirements would not be overly burdensome in these respects. In particular, the Bureau estimated that the required ability-to-repay determination would take essentially no time 474 for a fully automated electronic system and between 15 and 20 minutes for a fully manual system. While the Bureau stated in the proposal that most repayers would be able to demonstrate their ability to repay under the proposed rule, the Bureau recognized there may be a sub-segment of repayers who could not demonstrate their ability to repay if required to do so by a lender. For them, the current lender practice of making loans without determining their ability to repay could enables them to obtain credit that, by hypothesis, they may actually be able to afford to repay. The Bureau acknowledged that this group of “false negatives” may benefit by being able to obtain covered loans without having to demonstrate their ability to repay in the manner prescribed by the proposed rule. However, the Bureau judged that under the proposed rule lenders would generally be able to identify consumers who are able to repay and that the size of any residual “false negative” population would be small. It assessed this to be especially true to the extent that this class of consumers is disproportionately drawn from the ranks of those whose need to borrow is driven by a temporary mismatch in timing between their income and expenses rather than those who have experienced an income or expense shock or those with a chronic cash shortfall. The Bureau inferred that it is very much in the interest of these borrowers to attempt to demonstrate their ability to repay in order to receive the loan they are seeking, and that lenders will have every incentive to err on the side of finding such ability. Moreover, even if these consumers could not qualify for the loan they would have obtained absent an ability-to-repay requirement, they may still be able to get different credit within their demonstrable ability to repay, such as a smaller loan or a loan with a longer term. For these reasons, the Bureau did not conclude that any “false 475 negative” population resulting from lenders making ability-to-repay assessments would represent a significant amount of countervailing benefit. Finally, the proposal stated that some repayers may not actually be able to afford to repay the loan, but choose to repay it nonetheless, rather than re-borrow or default—which may result in their incurring ancillary costs in connection with another obligation, such as a late fee on a utility bill. Such repayers would not be able to obtain the same loan under the proposed rule that they would have obtained absent an ability-to-repay requirement, but the proposal stated that any benefit they receive under the current practice would appear to be small at most. The second group identified in the proposal consisted of borrowers who eventually default on their loan, either on the first loan or later in a loan sequence after having re-borrowed, perhaps multiple times. The Bureau referred to these borrowers as “defaulters” for purposes of its analysis of countervailing benefits in the proposal. As discussed in the proposal, borrowers of 20 percent of payday and 33 percent of single-payment vehicle title loan sequences fall within this group. For these consumers, the current lender practice of making loans without regard to their ability to repay the loan according to its terms may enable them to obtain what amounts to a temporary “reprieve” from their current situation. They can obtain some ready cash, which may enable them to pay a current bill or current expense. However, the proposal stated that for many consumers, the reprieve can be exceedingly short-lived: 31 percent of payday loan sequences that default are single-loan sequences, and an additional 27 percent of loan sequences that default are two or three loans long (meaning that 58 percent of defaults occur in loan sequences that are one, two, or three loans long). The proposal stated that 29 percent of single-payment vehicle title loan sequences that default are single-loan sequences, and an additional 26 percent of loan sequences 476 that default are two or three loans long (meaning that 55 percent of defaults occur in loan sequences that are one, two, or three loans long). The proposal stated that these consumers thus are merely substituting a payday lender or vehicle title lender for a pre-existing creditor, and in doing so, they end up in a deeper hole by accruing finance charges, late fees, or other charges that are imposed at a high rate. Title loans can have an even more dire consequence for defaulters: 20 percent of them have their vehicle repossessed. The Bureau stated in the proposal that it therefore did not find that defaulters obtain significant benefits from the current lender practice of making loans to them without determining their ability to repay.666 The final and largest group of consumers identified in the proposal consisted of those who neither default nor repay their loans without re-borrowing. Instead, this group of consumers will re-borrow some number of times before eventually repaying the loan. In the proposal, the Bureau referred to consumers with such loan sequences as “re-borrowers” for purposes of its discussion of countervailing benefits. These consumers represent 58 percent of payday loan sequences and 56 percent of title loan sequences. For these consumers, as for the defaulters, the practice of making loans without regard to their ability to repay the loan according to its terms enables them to obtain a temporary reprieve from their current situation. But for this group, the proposal stated that such a reprieve can come at a greater cost and pose a higher likelihood of 666 The Bureau recognizes that some defaulters may not default because they lack the ability to repay, but the Bureau estimates that the percentage of consumers who default despite having the ability to repay the loan is small. Moreover, any benefit such borrowers derive from the loan would not be diminished by the provisions of § 1041.4 precisely because these borrowers do have the ability to repay and thus would qualify for such loans. 477 risk than they would have initially expected, and for many consumers it will come at a substantially greater cost and a much higher likelihood of risk. The proposal stated that some re-borrowers are able to end their borrowing after a relatively small number of additional loans; for example, approximately 22 percent of payday loan sequences and 23 percent of title loan sequences are repaid after the consumer re-borrows once or twice. But even among this group, many consumers do not anticipate before taking out a loan that they will need to re-borrow at all. These consumers cannot reasonably avoid their injuries, and while their injuries may be less severe than the injuries suffered by consumers with extremely long loan sequences, their injuries can nonetheless be substantial, particularly in light of their already precarious finances. Conversely, some of these consumers may expect to reborrow and may accurately predict how many times they will have to re-borrow. For consumers who accurately predict their re-borrowing, the Bureau did not count their re-borrowing costs on the “injury” side of the countervailing benefits scale. The proposal stated that while some re-borrowers end their borrowing after a relatively small number of additional loans, a large majority of re-borrowers end up in significantly longer loan sequences. Of storefront payday loan sequences, for instance, one-third contain seven or more loans, meaning that consumers pay finance charges equal to or greater than 100 percent of the amount borrowed. About a quarter of loan sequences consist of 10 or more loans in succession and even larger aggregate finance charges. For single-payment vehicle title borrowers, the consequences described in the proposal were similarly dramatic: only 23 percent of loan sequences taken out by re-borrowers on title loans are repaid after two or three successive loans, whereas 23 percent of the loan sequences are for 10 or more loans in succession. The 478 Bureau did not find that any significant number of consumers anticipated such lengthy loan sequences, and such empirical research as is available indicates that borrowers who end up in extended loan sequences are the least accurate in predicting the duration of their borrowing. Thus, the Bureau stated its view in the proposal that the substantial injury suffered by the defaulters and those re-borrowers who incurred unanticipated injury—the categories that represent the vast majority of overall borrowers of covered loans—dwarfs any benefits these consumers may receive in terms of a temporary reprieve and also dwarfs the speed and convenience benefits that the repayers may experience. The Bureau acknowledged that any benefits derived by any aforementioned “false negatives” may be reduced under the proposed rule, but it judged that the limited benefits that may be received by this relatively small group are far outweighed by the substantial injuries sustained by the defaulters and re-borrowers, as discussed above. Further, the Bureau stated that under the proposed rule, many borrowers could be led to find more sustainable loan options, such as underwritten credit on terms that are more affordable and better tailored to their budget needs. Turning to the benefits of the practice for competition, the Bureau acknowledged in the proposal that the current practice of lending without regard to consumers’ ability to repay has enabled the payday industry to build a distinctive business model. Under this model, fully half or more of the revenue on these kinds of loans comes from consumers who borrow 10 or more times in succession. This, in turn, has enabled a substantial number of firms to extend such loans from a substantial number of storefront locations. The Bureau estimated that the top 10 storefront payday lenders controlled only about half of the market, and that 3,300 storefront payday lenders were small entities as defined by the SBA. The Bureau also acknowledged that 479 the anticipated effect of limiting lenders to making loans that consumers can actually afford to repay would be to shrink the number of loans per consumer fairly substantially, which may, in turn, result in a more highly concentrated market in some geographic areas. Moreover, the Bureau acknowledged that the practices underlying their current business model enabled lenders to avoid many of the procedural costs that the proposed rule would impose. However, the Bureau did not believe the proposed rule would materially reduce the competitiveness of the payday or title loan markets as a practical matter. As discussed in the proposal, most States in which such lending takes place have established a maximum price for these loans. Although in any given State there are a large number of lenders making these loans, located typically in close proximity to one another, the Bureau preliminarily found from existing research that there is generally no meaningful price competition among these firms. Rather, the Bureau stated that lenders generally charge the maximum possible price allowed in any given State. Lenders that operate in multiple States typically vary their prices from State to State to take full advantage of the parameters that are allowed by local law. Thus, for example, lenders operating in Florida are permitted to charge $10 per $100 loaned, and they do; when those same lenders are lending in South Carolina, they are permitted to charge $15 per $100, and they do that instead. In addition, despite some amount of consolidation that could be expected in the industry, the Bureau preliminarily found that under the proposed rule, based on experience of recent legislative reforms in various States, lenders would likely remain in relatively close proximity to the vast majority of borrowers. In sum, the Bureau stated in the proposal that the benefits of the identified unfair practice for consumers and competition—failing to underwrite covered loans by making a reasonable 480 assessment of the borrower’s ability to repay the loan according to its terms—do not appear to outweigh the substantial injury that is caused or likely to be caused by the practice, and which is not reasonably avoidable by consumers. On the contrary, the Bureau preliminarily determined that the very significant injury caused by the practice outweighs the relatively modest benefits of the practice for consumers or for competition. Comments Received The Bureau received a number of comments on its proposed analysis of whether the substantial injury was outweighed by countervailing benefits to consumers or competition. Several industry participants and trade association commenters contended that this test was simply not met, arguing that the negative effects of the proposed rule would exceed its benefits. They argued that all consumers would be deprived of loans precluded by the rule, not just the “false negatives” or those who may be harmed by them. Some commenters stated their point in a more general way, complaining that the Bureau had failed to present any objective metric or provide hard quantitative evidence to determine the costs and benefits of the identified practice to consumers or to competition in a more rigorous manner. Aside from attacking the general framework of the Bureau’s analysis, commenters also maintained that the Bureau underestimated the costs that the rule would impose on lenders, greatly impeding the industry’s ability to make appropriate covered loans. Some argued that the Bureau should have considered the costs of complying with the rule aggregated with the costs associated with complying with State law requirements. Commenters listed a variety of potential benefits to consumers associated with covered short-term loans, and suggested that the Bureau both understated the benefits and overstated the 481 extent of injury for re-borrowers. The list included that such loans help consumers cope with income shocks, achieve income smoothing, realize an overall improvement in their ability to manage accumulated debt, avoid bounced checks and problems with debt collection firms, reduce delinquency or defaults on other accounts, reduce unemployment, and reduce bankruptcies. Others emphasized that covered short-term loans can allow consumers to avoid riskier and more costly forms of credit, and thus these loans are simply the best and least expensive choice available for cash-strapped consumers with limited credit options. These commenters maintained that such loans allow consumers to avoid the inferior substitutes of even more costly alternatives, such as pawnbrokers, bank overdraft services, credit card cash advances, over-limit credit-card fees, and late-payment fees. As for vehicle title loans, commenters noted that they have the advantage of allowing consumers to tap into an asset to meet current needs and are structured to limit the potential harms to consumers because they are largely non-recourse loans; yet the restrictions posed by mandatory ability-to-repay underwriting would constrict the market for such loans and correspondingly impair the benefits to consumers. Some commenters asserted that studies show that consumer access to payday loans has no negative effect on various measures of consumer financial health. They suggested that credit scores were better for longer-term borrowers as compared to borrowers who engaged in less reborrowing and for borrowers in States with fewer payday loan restrictions as compared to States with greater restrictions, and that some studies conclude that payday lending bans lead to more bounced checks and overdraft fees as well as increased bankruptcy filings. They therefore surmised that covered loans improve the financial well-being of consumers. Several commenters cited as evidence of customer satisfaction the small proportion of complaints submitted to the 482 Bureau about the product, the many positive accounts of covered loan usage in the “Tell Your Story” portion of the Bureau’s website, and substantial product use without substantial levels of complaints to State regulators. Similarly, as stated above in the substantial injury section, a number of commenters believed the identified practice was net beneficial. Many of these commenters argued that borrowers were merely replacing other obligations with a covered short-term loan, and thus the harm of the one was offset by the benefit of being able to pay the other. Some commenters argued that borrowers were not harmed, or were only minimally affected, by defaulting on these loans because those defaults generally do not affect consumers’ credit reports and some lenders do not pursue collection efforts on defaulted loans. The Bureau received a large volume of comments from consumers who attested to the benefits of payday lending from their own personal experiences, though it also received many other comments from individual borrowers and consumer groups complaining about the injuries identified in the proposed rule. One respected academic in the field commented that while economists have generally concluded that payday loans may destroy consumer welfare in some situations and may improve consumer welfare in others, there is disagreement over how many consumers fall in each category. This commenter asserted that the Bureau would only have to resolve this debate about consumer welfare if it were choosing whether to ban payday lending entirely. Many industry commenters, and other commenters including a group of State Attorneys General, argued that by eliminating or limiting access to covered loans, the proposed rule would make consumers worse off because they would be forced to seek more expensive or otherwise more harmful alternatives, and that the Bureau had failed to factor the benefit of being able to 483 avoid these harmful alternatives into its preliminary analysis of unfairness (i.e., countervailing benefits). A number of commenters including a trade group and a university-affiliated research center, among others, argued that consumer demand for credit will continue while the rule will only restrict supply. These comments were made about all of the proposed restrictions on making all three types of covered loans: covered short-term loans, covered longer-term balloonpayment loans, and other covered longer-term loans (i.e., certain high-cost installment loans). And many comments in this vein focused on particular proposed restrictions, with particular emphasis on the 30-day cooling-off periods after a sequence of three loans made under § 1041.5 or § 1041.6, and the limitation on the total number of conditionally exempt covered short-term loans under proposed § 1041.6 to six loans or 90 days of indebtedness in a 12-month period. These commenters asserted that these restrictions would force consumers to substitute alternative forms of credit that are more costly and harmful than covered loans, claiming this to be true of loans ranging from pawn loans, to overdraft, to loans from unlicensed and unregulated online lenders, and even to loans from neighborhood loan sharks. Numerous consumers writing as part of organized letter-writing campaigns raised similar issues, expressing concern about the possibility of not having unlimited access to covered loans and the lack of alternative options. Some commenters referenced or submitted research studies, law review articles, or other analyses of these issues, some of which are described in detail below in the responses to the comments. Some commenters raised one countervailing benefit to the Bureau’s attention that was not included in the proposed rule—that borrowers do not have to undergo a credit check when taking out a covered loan that is originated without underwriting. Others noted that the current 484 practices of many lenders, which do not engage in ability-to-repay underwriting of covered loans, avoids the additional privacy and security risks of maintaining more documentation on borrowers. In addition to the points they made about countervailing benefits for consumers, industry commenters also objected to the Bureau’s analysis of the countervailing benefits to competition. The Bureau received some comments arguing that the Bureau’s statement that there is “generally no meaningful price competition” was inaccurate. Lenders provided assessments of their own market experience that purported to rebut that claim and indicated that covered loans create additional competition for other types of credit. They also argued that the Bureau had not appropriately included in the countervailing benefits the efficiencies of not having to assess the borrower’s ability to repay, which reduce procedural costs to the entity and thus the prices offered to consumers. Commenters further asserted that the Bureau had failed sufficiently to take account of how the identified practice fosters non-price competition among lenders. They also noted that the proposal impedes consumer free choice and that it fails to consider the negative effects it may have on rural consumers. Some commenters emphasized that the proposed rule would lead to market concentration, eliminating thousands of jobs while denying access to a form of credit that millions of consumers currently rely on. Others suggested that lack of clarity over the application of the proposed rule to banks and credit unions could lead them to stop making small-dollar loans to their customers. A coalition of consumer groups commented that the market for short-term small-dollar credit is much broader than the payday and single-payment vehicle title loans covered by this rule. In their analysis, the broader market comprises substitute products they viewed as more 485 advantageous than covered short-term loans, including credit cards, subprime credit cards, certain bank and credit union products, non-recourse pawn loans, employer funds, charitable funds, and payment plans that are often made available by utilities and others. They also suggested that other non-credit strategies, such as debt counseling and credit counseling, should be viewed as preferable alternatives to taking out payday and title loans. They went even further by arguing that payday loans should not even be considered as “credit” to be accessed, as in their view most of these loans generate their own demand through repeated rollovers, rather than meeting the independent credit needs of consumers. The Final Rule After having reviewed and analyzed the comments submitted in response to the proposed rule, the Bureau concludes that though the identified practice of making covered loans without reasonably assessing the borrower’s ability to repay the loan according to its terms presents some countervailing benefits to consumers and competition, those benefits do not outweigh the substantial injury that consumers are unable reasonably to avoid and that stems from the identified practice. Methodology. Again, the Bureau approaches this determination by first weighing substantial injury in the aggregate, then weighing countervailing benefits in the aggregate, and then assessing which of the two predominates. If the benefits predominate, then the practice is not unfair. If the benefits do not predominate, then the practice is unfair. As described above, the substantial injury is incurred through default, delinquency, re-borrowing, and the collateral consequences of 486 making unaffordable payments, including harms from forgoing major financial obligations or basic living expenses in an attempt to avoid these other injuries. It is important to start by recognizing that the Bureau is not assessing the benefits and injury of covered short-term loans. As one academic commenter noted, this would only be necessary if the Bureau were seeking to ban all payday lending in its entirety. Rather, the Bureau is weighing the benefits and injury of the identified practice, which is making such loans without reasonably assessing that borrowers have an ability to repay the loan according to its terms. In other words, the countervailing benefits to consumers consist of the benefits that consumers receive as a result of lenders making these loans without assessing ability to repay (i.e., not having to comply with any of the underwriting criteria of this rule). In weighing the countervailing benefits, the Bureau considers the various costs that a remedy would entail. Costs not incurred to remedy the practice, like costs of complying with independent State law requirements, are not included in the analysis. As the Bureau noted in the proposal, unfairness determinations do not require an exact quantification of costs and benefits. To do so would be impracticable, despite the suggestion made by some commenters that a specific metric or objective quantification was needed to meet the requirements of the statute—a suggestion that was made without any specificity as to methodology and in reliance on no existing precedent. And, in fact, the Bureau has quantified such data as are available about the frequency and extent of re-borrowing, the frequency of default, the frequency of payment failures, the severity of the resulting harms, and various other relevant items, even if some factors (such as the frequency and extent of default avoidance, for example) are not subject to being quantified as a practical matter. 487 At the proposal stage, the Bureau believed that the injury caused by the practice outweighed the benefits to consumers or competition, the latter of which includes the costs associated with complying with the remedy to the extent they would be passed on to consumers (and thus the absence of which is a benefit to consumers). The Bureau has had the chance to process and digest over a million comments that were submitted on the proposed rule and now concludes that this assessment was correct. However, in light of the considerable volume of input received from the public, the Bureau has decided to modify certain parameters of the proposed rule so as to simplify its scope, reduce the potential impact on access to credit, streamline the underwriting process, and add more flexibility within the existing framework. The effect of these adjustments is to reduce the costs associated with complying with the rule and reduce the impact it will have on access to credit, thereby reducing the weight on the countervailing benefits side of the scale. This is so because in assessing the identified practice, the Bureau weighs the injury against the countervailing benefits, and according to the FTC Statement on Unfairness, the costs associated with implementing the remedy (i.e., assessing ability to repay) are included in the benefits that lenders could avoid if they did not have to comply with the underwriting criteria of the final rule. The Bureau’s efforts to ensure that its remedy does not overly restrict access to credit, including adjustments made in § 1041.5 of the final rule that simplify and streamline some of the underwriting criteria that had been contained in the proposal, decrease the costs of the remedy, which in turn reduces the weight that is attributed to the countervailing benefits side of the scale. And the allowance of loans that can be made pursuant to § 1041.6 of the final rule without having to meet those specific underwriting criteria further reduces the weight on this side 488 of the scale. In other words, the Bureau has reacted to commenters who feared the proposed rule was too complex and overly burdensome by reducing complexity and burden. These adjustments affect the balance between consumer injury and countervailing benefits, which results in the injury from the identified practice outweighing the countervailing benefits to consumers by even more than it did at the proposal stage. With these changes, which are described more specifically in the relevant explanation of § 1041.5 of the final rule, the Bureau is reinforced in its conclusion that the substantial injury is not outweighed by countervailing benefits to consumers or to competition. Assessing Benefits to Consumers. To evaluate this assessment in light of the points made by the commenters, it is useful again to divide consumers into several groups of different borrowing experiences, in order to analyze whether and how the practice of making covered short-term loans without reasonably assessing whether the consumer has the ability to repay the loan according to its terms yields countervailing benefits to consumers. Those groups, once again, can be characterized as “repayers,” “defaulters,” and “re-borrowers” for purposes of this analysis. To begin with “repayers,” several commenters stated that the proposed rule would have such a substantial financial impact on lenders that even borrowers who have an ability to repay would not have access to covered loans as a result of the rule. The Bureau acknowledges that some borrowers who might end up repaying their loans because of windfalls or other unexpected developments would be unable to obtain a loan if they cannot meet the ability-to-repay criteria, though it does not anticipate there are large numbers of such consumers. Yet the Bureau stands by its analysis in the proposed rule on how the market will likely consolidate and thus survive as 489 a result of the proposed rule, and thus that lenders will continue to make loans to borrowers who have the ability to repay. Any other conclusion would require the industry to concede that it cannot execute on a successful business model for making these loans unless it can be assured of a relatively large number of borrowers who find themselves caught up in extended loan sequences. The Bureau addresses more specific comments about its analysis of this point in part VII, which considers the benefits, costs, and impacts of the final rule on consumers and covered persons pursuant to section 1022(b)(2)(A) of the Dodd-Frank Act.667 As to whether the rule will drive up prices for borrowers with the ability to repay, the Bureau does not believe it will do so. The Bureau noted in the proposal, and above in Market Concerns—Underwriting, that many covered loans are already offered at the maximum price allowed under State law. Instead of increasing prices, which they typically cannot do, lenders will likely address additional compliance costs and reduced volume by consolidating to some degree, as the Bureau anticipated. The Bureau also has no reason to believe that lenders will be overly conservative and restrictive by lending to an even smaller group of people than the rule would allow. Without evidence to the contrary, the Bureau expects that the industry will act rationally and make those loans that are allowed by the rule. It may be that some lenders will choose to take a conservative approach and decline to lend to borrowers who would be eligible under the rule due to concerns about compliance risk; if so, that would be an unfounded and imprecise reaction to the rule, yet it is a possible outcome in some instances. Even so, the effect on the countervailing benefits determination should be marginal at best. Nonetheless, as set out in the relevant explanation of § 667 12 U.S.C. 5512(b)(2)(A). 490 1041.5 of the final rule, the Bureau has made certain adjustments to streamline and simplify the final rule’s underwriting criteria with the intent of reducing the number of industry participants that would restrict access to credit based on overly conservative assessments of compliance risk. Thus, the Bureau continues to be persuaded that lenders will be able to make covered short-term loans to the population of consumers who have the ability to repay them, and that the “false negative” category of borrowers will be low, especially in light of the adjustments that are made in the final rule to respond to these comments to streamline the underwriting criteria in certain respects. Further, the Bureau notes that the proposed rule, and now the final rule, allows lenders to make some covered loans under the terms set forth in § 1041.6, without all the specific underwriting criteria that would otherwise apply under § 1041.5 because other conditions are imposed that effectively prevent extended loan sequences. Based on the lack of persuasive evidence demonstrating otherwise—and in light of the further changes to the rule that simplify, reduce burden, add flexibility, and ensure broader access to credit—the Bureau concludes that the lending industry should be able to adjust to the rule, and consumers who can afford to repay covered short-term loans according to their terms will generally continue to have access to them. The Bureau thus concludes that restrictions on access to credit for borrowers who have the ability to repay will be minimal. The Bureau also finds that it did not underestimate other benefits to these consumers, such as the speed and convenience associated with lenders not having to underwrite loans by making ability-to-repay determinations. The Bureau continues to maintain the view that the underwriting process for these loans can be largely automated. But as a matter of caution and in response to the comments received, the Bureau decided to make adjustments to further 491 streamline some of the underwriting criteria contained in the proposed rule. For example, as discussed above and in contrast to the proposal, the Bureau has removed some of the complexity around the residual income test, changed the documentation requirements in a variety of ways (including by allowing lenders to rely on consumer statements to authenticate rental expenses), and allowed lenders to take account of income from someone other than the borrower if the borrower has a reasonable expectation of access to that income. Lenders also will be able to assess ability to repay, in the alternative, by using a debt-to-income ratio. And rental expenses can now be based solely on a borrower’s statement without the need to validate such statements through survey or other data. In fact, under the final rule, most borrowers who have the ability to repay typically should be able to get a covered loan without having to present any more documentation of income than a pay stub or a paycheck,668 which commenters indicated is the kind of income documentation that is already required by many lenders. The second group of consumers consists of borrowers who eventually default on their loans, either on the first loan or later in a loan sequence after having re-borrowed, perhaps multiple times. As for these “defaulters” who lack the ability to repay the loan according to its terms, the Bureau did not underestimate the countervailing benefits to them. It is apparent, as a number of commenters attested, that these borrowers typically would not be able to obtain loans under the terms of the final rule. Put another way, the current practice of failing to make a reasonable assessment of whether a borrower can repay a covered loan results in this population 668 This should be true for borrowers unless they wish to rely on matters other than income to demonstrate the ability to repay a covered loan, such as income from another person that is reasonably available for use by the borrower. More specific description of the adjustments made in the final rule to the underwriting requirements contained in the proposed rule can be found in the explanation of § 1041.5 below. 492 of borrowers obtaining loans they do not have the ability to repay, which leads either immediately or eventually to default. As industry commenters noted, losing access to nonunderwritten credit may have consequences for some consumers, including the inability to pay for other needs or obligations or the need to seek out alternative credit options or budgeting strategies. The Bureau considered the impact of the identified practice on access to credit in the proposal, which inherently included the natural consequences of losing access to such nonunderwritten credit. The Bureau continues to regard the current access to credit that would be foreclosed under the ability-to-repay requirement as not an insignificant countervailing benefit. While the vast majority of borrowers who would eventually become defaulters will not be able to obtain covered short-term loans, this forgone benefit must be weighed against the forgone injury. Again, the figures presented in the proposal are instructive in terms of the comparison at issue here. As discussed in the proposal, borrowers of 20 percent of payday and 33 percent of single-payment vehicle title loan sequences fall within this group of “defaulters.” For these consumers, their current access to non-underwritten credit may enable them to obtain a temporary “reprieve” from their current situation by obtaining the cash to pay a current bill or expense. But for many consumers, this reprieve is exceedingly brief: 31 percent of payday loan sequences that default are single-loan sequences, and an additional 27 percent of loan sequences that default are two or three loans long (meaning that 58 percent of defaults occur in loan sequences that are one, two, or three loans long). The proposal also stated that 29 percent of single-payment vehicle title loan sequences that default are single-loan sequences, and an additional 26 percent of loan sequences that default are two or three loans long (meaning that 55 percent of defaults occur in loan sequences that are one, two, or three loans long). Thus these 493 consumers are merely substituting a payday lender or title lender for a pre-existing creditor, and they quickly find themselves in a new and potentially deeper hole by accruing finance charges, late fees, or other charges that are imposed at a high rate as well as the adverse consequences of ultimate default. Title loans can have an even more dire consequence for defaulters: 20 percent of them have their vehicle repossessed, with further adverse consequences, which may be take a severe toll on the consumer’s economic situation if it affects their ability to get to work or carry on a variety of everyday household affairs. The Bureau thus finds that most defaulters do not obtain any significant benefits from the current lender practice of making loans to them without reasonably assessing their ability to repay the loan according to its terms. There is another important point here about the calculus of benefits and injury with respect to “defaulters” that was not discussed in the proposal, yet which underscores the fact that their current access to non-underwritten credit does not benefit them and in fact leads to considerable harm. That is the adverse economic effect of the unsuccessful struggle to repay the unaffordable loan on the remaining population of “defaulters” that were omitted from the above discussion. Note that 58 percent of defaults on payday loans, and 55 percent of defaults on title loans, occur in loan sequences that are one, two, or three loans long. What this leaves aside is that fully 42 percent of default on payday loans, and 45 percent of defaults on title loans, occur after the borrower has already had an extended loan sequence of four or more loans, and then defaults. In many instances, this scenario is strong evidence of consumer mistake, since a consumer who anticipates defaulting should not also incur the high and accumulating costs of reborrowing (which, for a sequence of at least four loans, amounts to more than half of the principal of the original loan, with the total mounting as the sequence extends even further). It is 494 thus quite implausible that these borrowers, who constitute a substantial segment of all “defaulters,” obtain any significant benefits from the current lender practice of making loans to them without reasonably assessing their ability to repay the loan according to its terms. Indeed, quite the contrary is very likely to be the case for the vast majority of these borrowers, and the harm they suffer in these circumstances will generally amount to a very substantial injury. This account provides a strong refutation of the claim by certain commenters that borrowers who default on covered short-term loans do not sustain any substantial injury in light of the corresponding benefits, or that they experience a net benefit because they are able to keep the proceeds of the defaulted loan and perhaps avoid defaulting on some other obligation with more severe consequences. Although that might conceivably be true in some instances, it is implausible in any functioning market that it is likely to be true very often, and that is particularly the case in the context of title loans, where the damaging consequences of vehicle repossession multiply the potential harm even further. So even if there is a small number of such borrowers, it is unlikely to have any material impact on the analysis here. As for the commenters who asserted that default does not affect consumers’ credit reports and sometimes does not lead to debt collection efforts, these are marginal matters when compared to the core harms associated with unaffordable loans that end in default. But in any event, the Bureau’s experience from engaging in supervisory oversight and investigations of these types of lenders have led to numerous enforcement actions demonstrating that many such lenders do seek to collect debts that are due on defaulted loans, which have led to findings of illegal conduct in aggressively 495 seeking to pursue collection of such loans.669 And nothing prevents third party debt collectors or debt buyers from reporting negative information to consumer reporting agencies, which some collectors do to facilitate collection. The third category of consumers is the “re-borrowers” who find themselves in extended loan sequences but eventually manage to find some way to repay the loan, even if only nominally. They are a majority of all borrowers—representing 58 percent of payday loan sequences and 56 percent of title loan sequences. For these consumers, as with the “defaulters,” the identified practice of making loans without reasonably assessing their ability to repay can allow them to obtain a temporary reprieve from the difficulties of their current financial situation. Some commenters suggested that many of them may benefit by literally buying time and to pay off some of their cumulative obligations later rather than sooner and that some financial indicia such as credit scores and bankruptcy filings appear to be more positive for these re-borrowers. It is undoubtedly true that some borrowers who lack the ability to repay may gain an overall benefit from having access to covered short-term loans. Again, these could be borrowers who incur some sort of windfall or positive change in circumstances, or accurately anticipate the extent of their re-borrowing, and may be engaged in either income smoothing or spreading an unexpected cost across a longer time span. In some cases, these borrowers may be substituting a payday lender for some other creditor, such as a landlord or a utility company. It is however, the 669 See, e.g., In the Matter of Money Tree, Inc., File No. 2016-CFPB-0028; In the Matter of EZCORP, Inc., File No. 2015-CFPB-0031; CFPB v. NDG Financial Corp., Case No. 1:15-cv-05211-CM (S.D.N.Y.); In the Matter of ACE Cash Express, Inc., File No. 2014-CFPB-0008; In the Matter of Westlake Servs., LLC, File No. 2015-CFPB-0026. The Bureau has also taken actions against debt collectors, some of which collect in part on small-dollar loans. See, e.g., CFPB v. MacKinnon, et al., Case No. 1:16-cv-00880 (W.D.N.Y.). 496 Bureau’s judgment that the injury to other “re-borrowers” who do not accurately anticipate the length of re-borrowing, and many who find themselves unexpectedly trapped in extended loan sequences, is so substantial as to outweigh the benefits to these other consumers. This point is bolstered by comments received from individual borrowers, consumer groups, and faith groups who related many similar stories about the financial harms sustained by borrowers who found themselves caught up in extended loan sequences—whether or not those sequences ultimately ended in default, as some but not all do. In this regard, it is notable that any such reprieve can pose a higher likelihood of risk and come at a greater cost than many borrowers may have initially expected, and a substantial population of “re-borrowers” can be expected to find that it will come at a much higher likelihood of risk and a substantially greater cost. It is worth restating why this is so. Once again, the dynamic of covered short-term loans is such that once the first loan has been made to a borrower who lacks the ability to repay it, the range of choices open to the borrower is sharply constrained. At the point of taking out the initial loan, the borrower can make a direct choice among competing alternatives as a means of meeting their immediate financial needs, and it is plausible that for some borrowers the decision to take out a covered short-term loan may seem or be superior to other available means of coping with the difficulties of their situation. But after the first loan has been made, the circumstances change significantly. When this first loan comes due, and for any and all subsequent loans, the borrower is no longer at liberty to make an unencumbered choice among competing alternatives. Instead, the borrower now must confront the range of risks and harms that are by now familiar, as they have been set out at length and discussed so often in the proposal and above—default, delinquency, re-borrowing, and the 497 negative collateral consequences of making unaffordable payments, including harms from forgoing major financial obligations or basic living expenses in an attempt to avoid these other injuries. This is the changed situation that borrowers confront as they find themselves facing the constrained choices that lead many of them into extended loan sequences, often unexpectedly, and cause them to bear the high costs of repeatedly rolling over their loans (which, by the time an extended loan sequence reaches seven loans, as one-third of storefront payday loan sequences actually do, means the borrower will have paid charges equal to 100 percent of the original amount borrowed and still owe the full amount of the principal). So while it is certainly likely that some borrowers may choose to take out these loans intentionally to spread a large, unexpected expense across a longer time span, it is equally apparent that many others find themselves in significant trouble if they have taken out such an unaffordable loan as an initial matter, even though they do find a way to manage to pay it back eventually after experiencing the types of harm that accompany the experience of an extended loan sequence. For those borrowers who accurately predict the length of their re-borrowing, the Bureau does not count these costs on the “injury” side of the ledger as against countervailing benefits. In evaluating whether most consumers would or would not be likely to make this choice intentionally and based on accurate predictions, it is relevant here that the evidence suggests that consumers seem to be best able to gauge the expected duration of re-borrowing when the loan sequences are shorter, and such empirical research as is available indicates that borrowers who end up in extended loan sequences are the least accurate in predicting their duration of reborrowing. Again, about one-quarter of storefront payday loan sequences consist of 10 or more 498 loans taken out in succession, and 23 percent of title loan sequences consist of 10 or more loans in succession. The Bureau does not find evidence that any significant number of consumers anticipated such lengthy loan sequences. Another set of considerations that is germane to the circumstances of “re-borrowers” is the effect of lender practices in the market for covered short-term loans. Although these loans are presented and marketed as stand-alone short-term products, lenders are aware (though many consumers likely are not) that only a relatively small number of borrowers repay such loans without any re-borrowing, and their core business model relies on that fact. Moreover, the decision that many lenders have made to offer these loans without reasonably assessing the borrower’s ability to repay the loan according to its terms is the identified practice that causes injury to consumers, which, as discussed above, is not reasonably avoidable by consumers who are often likely to fail to fully understand the likelihood and severity of the risks posed. The Bureau also has concluded that the manner in which lenders structure these products—including the term of the loan, its balloon-payment structure, and the common requirement that the borrower provide a cancelled check or ACH access or provide vehicle security—likely contributes significantly to the result that many borrowers have no good alternatives to ending up in extended loan sequences of repeated re-borrowing that often extend well beyond their initial expectations. It is also worth emphasizing that even these “re-borrowers” who would not have access to most covered short-term loans under § 1041.5 of the final rule, because they lack the ability to repay the loan according to its terms, would have access to loans made subject to the protections found in § 1041.6, with a corresponding reduction in the weight that falls on the countervailing 499 benefits side of the scale. In the end, after aggregating the injury and benefits of these three populations of borrowers, the Bureau believes that the aggregate injury clearly outweighs the aggregate benefits. Substantial groups of consumers suffer acute harm as a result of the various scenarios analyzed above. These outcomes are bolstered by commenters who provided examples of consumers who ended up in extremely long loan sequences and ultimately were required to pay many multiples of the original principal of the loan. Based on the Bureau’s research, 62 percent of these loans were in loan sequences of seven or more, and 15 percent of loan sequences involved 10 or more loans.670 The scope of that injury is quite substantial across the entire market for these loans. The Bureau concludes that this aggregate injury to many “re-borrowers” outweighs the countervailing access-to-credit benefits that other “re-borrowers” may receive as a result of lenders not reasonably assessing the borrower’s ability to repay the loan according to its terms, in light of all the provisions of the final rule, including the effect that § 1041.6 will have in reducing the magnitude of those benefits. As for the commenters who cited studies purporting to show that payday loans improved financial outcomes, the Bureau notes that all of the studies varied in their empirical rigor and the connection of their causal inferences to their documented findings. Based on its experience and expertise, the Bureau finds some studies to be more compelling than others. For example, several of these studies predicated their conclusions on comparisons of financial outcomes for consumers with and without access to payday loans, relying on access to payday loans based on 670 And again, the research shows those in longer sequences are less likely to anticipate the extent of re-borrowing. 500 geographic location as a proxy for actual use.671 Others that reached conclusions about better or similar financial outcomes for these groups relied on changes in credit scores, a narrow measure of financial well-being for the population of payday loan borrowers, whose credit scores are already strongly skewed toward the bottom of the customary ranges.672 The Bureau discussed many of these studies in the proposal; additional studies are mentioned here in light of comments received and are also discussed in further depth in the Section 1022(b)(2) Analysis in part VII below. Findings based on the access proxy, which are possible largely due to State-level variation in payday lending laws, do not demonstrate better financial outcomes for actual payday loan borrowers. While certainly instructive, the Bureau finds these studies are generally less compelling than those based on individual-level data that can identify actual payday borrowers and their use. Further, this research has focused almost exclusively on the question of what happens when all access to a given form of credit is eliminated, as opposed to when it is merely restricted (or, as in this rule, restricted only as to borrowers who cannot demonstrate an ability to repay). The evidence available from States that have imposed strong restrictions on lending, but not outright or de facto bans, suggests that, even after large contractions in this industry, loans remain widely available, and access to physical locations is not unduly limited.673 671 Adair Morse, Payday Lenders: Heroes or Villains?, 102 J. of Fin. Econ. 28 (2011); Jonathan Zinman, Restricting Consumer Credit Access: Household Survey Evidence on Effects Around the Oregon Rate Cap, at 5 (2008); Kelly D. Edmiston, Could Restrictions on Payday Lending Hurt Consumers?, Fed. Reserve Bank of K.C., Econ. Rev. 31, 37– 38 (1st Qtr. 2011). 672 Ronald Mann, Do Defaults on Payday Loans Matter?, (working paper Dec. 2014); Neil Bhutta, Payday Loans and Consumer Financial Health (April 27, 2014), Journal of Banking and Finance, Vol. 47, No. 1; Jennifer Priestley, Payday Loan Rollovers and Consumer Welfare (Dec. 4, 2014). 673 See Section 1022(b)(2) Analysis, part VII below. 501 One such study cited by commenters attempted to determine how households in North Carolina and Georgia fared following State actions to restrict payday lending. They reported an increase in the rate of bounced checks, Chapter 7 bankruptcy filings, and complaints against debt collectors and creditors.674 In an update to that paper, the authors expanded the time frame, analyzed more State-level payday bans, and considered the effects of enabling payday lending as well.675 They again found evidence that in response to limits on payday borrowing, bounced checks increased, as did complaints about debt collectors to the FTC, whereas Chapter 13 bankruptcy filings decreased. Numerous industry comments cited these studies, along with a related study that is no longer available.676 However, these studies each rely on a methodology that severely undermines their conclusions. Specifically, the original study’s assertion that checks are returned more frequently from States without payday lending—notably Georgia and North Carolina—relies on data that intermingles data from those States with data from numerous authorizing States (such as Louisiana, Alabama, Tennessee, and others), which makes the conclusions dubious at best. Indeed, in the original paper, more than half of the checks processed at the Charlotte, North Carolina check processing center actually came from States with payday loans. Additionally, the complaint data they cited are limited by the fact that the FTC is unlikely to receive complaints about payday lending (at the time, State regulators were more likely to receive such complaints). As such, the measure of complaints that the authors employ may not indicate the actual rate of credit-related complaints, let alone overall consumer 674 Donald P. Morgan & Michael R. Strain, Payday Holiday: How Households Fare after Payday Credit Bans, FRB of New York Staff Reports, No. 309, (Revised Feb. 2008). 675 Donald P. Morgan, Michael R. Strain & Ihab Selani, How Payday Credit Access Affects Overdrafts and Other Outcomes, Journal of Money, Credit, and Banking, 44(2-3): 519-531 (2012). 676 Donald P. Morgan, Defining and Detecting Predatory Lending,” Federal Reserve Bank of New York Staff Reports No 273 (2007). FRBNY webpage indicates report was “removed at the request of the author.” 502 satisfaction. While the later study improves on the previous studies by including more States, a longer period of analysis, and additional outcome measures, they still do not adequately address the shortcomings of their previous studies. This study also relies on data sources that commingle returned checks from States with payday bans with those from States that permit payday lending, which undermines its conclusions, and again relies on the simplistic measure of complaints received by the FTC. Other studies, rather than using differences across States in the availability of payday loans, have used data on the actual borrowers who apply for loans and are either offered loans or are rejected. One study used this approach to find that taking out a payday loan increases the likelihood that the borrower will file for Chapter 13 bankruptcy.677 The authors found that initial approval for a payday loan essentially doubled the bankruptcy rate of borrowers. Another study used a similar approach to measure the causal effects of storefront borrowing on borrowers’ credit scores.678 The authors found that obtaining a loan had no impact on how the consumers’ credit scores evolved over the following months. The authors noted, however, that applicants generally had very poor credit scores both prior to and after borrowing (or being rejected for) a payday loan. In each of these studies, the authors were unable to determine whether borrowers who were rejected by the lender from which they had data were able to take out a loan from another lender.679 677 Paige Skiba and Jeremy Tobacman. Do Payday Loans Cause Bankruptcy?, Working Paper (2015). Neil Bhutta, Paige Marta Skiba & Jeremy Tobacman, Payday Loan Choices and Consequences, Journal of Money, Credit and Banking, 47(2-3): 223-260 (2015). doi: 10.1111/jmcb.1275 679 As noted, some commenters had made dire predictions that the proposed rule might cause borrowers to turn to illegal lenders or “loan sharks.” As noted below in part VII, the Bureau is unaware of any data on the current prevalence of illegal lending in the United States, nor of data suggesting that such illegal lending is more prevalent in States where payday lending is not permitted than in States which permit it. 678 503 Two other studies have used data on payday borrowing and repayment behavior to compare changes over time in credit scores for different groups of borrowers. One measured changes over time in credit scores for borrowers who re-borrowed different numbers of times, and found that in some cases it appeared that borrowers who re-borrowed more times had slightly more positive changes in their credit scores.680 These differences were not economically meaningful, however, with each additional loan being associated with less than one point in credit score increase.681 The other compared the changes in credit scores of borrowers who defaulted on their loans with borrowers who did not, and also found no difference.682 Neither study found a meaningful effect of payday loan borrowing behavior on credit scores. Commenters also cited a laboratory experiment in which undergraduate students completed a novel computer exercise designed to test whether access to payday loans increased or decreased the likelihood of financial survival in the face of expense shocks. The experiment found that while subjects who used payday loans sparingly were more likely to survive the simulated 30-month period than those with no payday loans, heavy users that took out 10 payday loans or more over the course of the 30 months were less likely to survive than those who had no access to payday loans.683 One comment described the lender’s use of the Bureau’s financial well-being scale to compare the scores of its borrowers to those of consumers deemed by the commenter to be 680 Jennifer Priestly, Payday Loan Rollovers and Consumer Welfare (Dec. 5, 2014). Available at SSRN. The Priestley study also compared changes over time in credit scores of payday borrowers in different States, and attributed those differences to differences in the States’ payday regulations. This ignores differences in who chooses to take out payday loans in different States, given both the regulatory and broader economic differences across States, and ignores the different changes over time in the broader economic conditions in different States. 682 Mann, Ronald, Do Defaults on Payday Loans Matter, December 2014, Working Paper. 683 Bart J.Wilson and David W. Findlay, and James W. Meehan, and Charissa P.Wellford, and Karl Schurter, An Experimental Analysis of the Demand for Payday Loans (April 28, 2010). 681 504 “similarly situated” who did not use payday loans or did not have access to payday loans due to their State prohibiting the product. However, the commenter’s analytic methods cannot be used to determine causality, and their findings do not appear fully consistent with their conclusions. Furthermore, the comment noted that customers were more likely than non-customers to have incomplete surveys. It is unclear whether the survey may therefore have been affected by nonresponse bias by customers in greater financial distress. Non-customers may also have had characteristics that make them ineligible for a payday loan despite being “similarly-situated” based on other metrics. These factors, such as being unbanked or not having documented income, may also have influenced well-being scores. In the commenter’s first analysis, they report the median and mean financial well-being scale scores by State and overall for its payday customers and non-customer population and found that, in 11 States in which a high response rate was achieved, its median customer scored one point lower than a non-customer, and that the average customer scored 2.3 points lower than the average non-customer. The lender concluded this result showed no real negative effect of payday borrowing. However, the commenter also highlighted the findings from Texas, where customers had a higher score than non-customers, although the differences were the same or smaller than those reported nationally where the commenter surmised there was no significant effect. The Bureau recently conducted a national study of American consumers which found that the adults who reported using products such as payday, non-recourse pawn, and vehicle title loans in the previous 12 months had an average financial well-being score of 42, which was 13 505 points lower than adults who did not report using these products.684 Additionally, there is little overlap in the distribution of financial well-being scores among those consumers who have and have not used payday, non-recourse, and vehicle title loans.685 A second analysis conducted by the lender compared the scores of customers across different levels of payday loan usage and borrowing outcomes. Customers within the last year were grouped into five categories by the number of transactions they had, and grouped into four categories based on the outcome they experienced. Based on the median scores for each of the 20 categories a customer could be placed in given their borrowing and outcome status, the commenter concluded that there is no correlation between borrowers’ financial well-being score and the number of transactions. However, the commenter also acknowledged finding lower scores for those that have their balances written off. Despite this finding, the lender still concluded that there is no evidence to support a theory that payday loan use has a negative effect on financial well-being. More generally, the Bureau notes that all of these studies sought to measure the impact of payday loans, or eliminating payday lending, on all consumers generally. The Bureau is not opining on whether the payday industry, generally, is beneficial to consumers taken as a whole. Rather, the Bureau is assessing the impact of the identified practice of making payday loans (and other covered short-term loans and covered longer-term balloon-payment loans) to borrowers without making reasonable determinations that the borrowers have the ability to repay the loans according to their terms. In fact, the Bureau believes that covered short-term loans will still be 684 CFPB, Financial Well-Being in America, 57 (Sept. 2017), available at http://files.consumerfinance.gov/f/documents/201709_cfpb_financial-well-being-in-America.pdf. 685 Financial Well-Being in America, 57-58. 506 available to consumers facing a truly short-term need for credit in those States that allow them. More specifically, the Bureau believes the vast majority of consumers would be able to get at least six covered short-term loans in any 12-month period, with those borrowers who are able to satisfy an ability-to-repay assessment being able to get some number of additional loans. Notably, however, none of these studies was focused on the impact that payday lending has on the welfare of the sub-population of borrowers who do not have the ability to repay their loans. Industry commenters also suggested that consumers seem to be satisfied with covered short-term loan products, as shown by low numbers of complaints and the submission of positive stories about them to the “Tell Your Story” function on the Bureau’s website. In response, as noted earlier, the Bureau observed from its consumer complaint data that from November 2013 through December 2016 more than 31,000 debt collection complaints cited payday loans as the underlying debt, and over 11 percent of the complaints the Bureau has handled about debt collection stem directly from payday loans.686 And when complaints about payday loans are normalized in comparison to other credit products, the numbers do not turn out to be low at all. For example, in 2016, the Bureau received approximately 4,400 complaints in which consumers reported “payday loan” as the complaint product and about 26,600 complaints about credit cards.687 Yet there are only about 12 million payday loan borrowers annually, and about 156 million consumers have one or more credit cards.688 Therefore, by way of comparison, for every 686 Bureau of Consumer Fin. Prot., Monthly Complaint Report, at 12 (Dec. 2016), https://www.consumerfinance.gov/data-research/research-reports/monthly-complaint-report-vol-18/. 687 Bureau of Consumer Fin. Prot., Consumer Response Annual Report, Jan. 1 – Dec. 31, 2016, at 27, 33, (March 2017), available at https://www.consumerfinance.gov/documents/3368/201703_cfpb_Consumer-Response-AnnualReport-2016.PDF. 688 Bureau staff estimate based on finding that 63 percent of American adults hold an open credit card and Census population estimates. Bureau of Consumer Fin. Prot., Consumer Credit Card Market Report, at 36 (Dec. 2015), 507 10,000 payday loan borrowers, the Bureau received about 3.7 complaints, while for every 10,000 credit cardholders, the Bureau received about 1.7 complaints. In addition, faith leaders and faith groups of many denominations from around the country collected and submitted comments, which suggested that many borrowers may direct their personal complaints or dissatisfactions with their experiences elsewhere than to government officials. In addition, though the Bureau did receive a large number of comments from individual consumers relating their general satisfaction with these loan products, it also received a sizable number of comments to the contrary, where consumers or persons writing on their behalf detailed that many consumers experience negative effects with extended loan sequences. Furthermore, based on the analysis set forth above in Market Concerns—Underwriting, the Bureau did not overstate the extent of the injury to “re-borrowers” who receive singlepayment vehicle title loans, which were found to pose similar harms to consumers. Even though such loans may be non-recourse, which limits the extent of some harms, the injury to consumers of the risks of vehicle repossession often are extremely consequential on top of the other harms that flow from the structure and term of these loans, all of which leads to similar conclusions about the risks and harms of these loans. available at http://files.consumerfinance.gov/f/201512_cfpb_report-the-consumer-credit-card-market.pdf; U.S. Census Bureau, Annual Estimates of Resident Population for Selected Age Groups by Sex for the United States, States, Counties, and Puerto Rico Commonwealth and Municipios: April 1, 2010 to July 1, 2016 (Jun. 2017), available at https://factfinder.census.gov/bkmk/table/1.0/en/PEP/2016/PEPAGESEX. Other estimates of the number of credit card holders have been higher, meaning that 1.7 complaints per 10,000 credit card holders would be a high estimate. The U.S. Census Bureau estimated there were 160 million credit card holders in 2012, and researchers at the Federal Reserve Bank of Boston estimated that 72.1 percent of U.S. consumers held at least one credit card in 2014. U.S. Census Bureau, Statistical Abstract of the United States: 2012, at 740 tbl.1188, (Aug. 2011), available at https://www.census.gov/library/publications/2011/compendia/statab/131ed.html; Claire Greene, Scott Schuh, and Joanna Stavins, The 2014 Survey of Consumer Payment Choice: Summary Results, at 18 (Aug. 15, 2016), available at https://www.bostonfed.org/-/media/Documents/researchdatareport/pdf/rdr1603.pdf. And as noted above in the text, additional complaints related to both payday loans and credit cards are submitted as debt collection complaints with “payday loan” or “credit card” listed as the type of debt. 508 In the proposal, the Bureau did not address one countervailing benefit to consumers resulting from the identified practice—some commenters noted that some borrowers, even ones with an ability to repay, are currently able to obtain a non-underwritten loan without inquiries showing up on the borrower’s credit report. The Bureau acknowledges this can be a benefit to some consumers. However, the Bureau notes that the impact that a credit check will have on a borrower’s overall credit profile is limited and uncertain, given that every consumer’s consumer report differs and different creditors use different credit scoring models. One of the most experienced scoring companies, FICO, says the following about the impact of credit inquiries on a consumer’s score: “The impact from applying for credit will vary from person to person based on their unique credit histories. In general, credit inquiries have a small impact on one’s FICO Scores. For most people, one additional credit inquiry will take less than five points off their FICO Scores. For perspective, the full range of FICO Scores is 300-850.”689 Thus this minor effect has little bearing on the Bureau’s overall assessment of benefits and injury to consumers, especially in light of the adjustments made to the underwriting criteria in § 1041.5 of the final rule. Substitute Products. The Bureau has several responses to the commenters asserting that the proposed rule’s restrictions would make consumers worse off by forcing them to substitute more expensive and harmful credit products, and that the Bureau failed to account—or at least fully account—for the countervailing benefit that borrowers of covered loans do not incur the harms caused by these substitute products. 689 http://www.myfico.com/credit-education/credit-checks/credit-report-inquiries/. 509 As noted above, the Bureau has decided not to finalize proposed §§ 1041.8 to 1041.10. These proposed sections would have required lenders making covered longer-term loans, including both high-cost installment loans and loans with balloon-payment features, to comply with the ability-to-repay requirements. The proposed rules as applied to longer-term installment loans were one focus of the comments described above. Accordingly, to the extent those comments were predicated on such restrictions applying to covered longer-term installment loans, they have been rendered largely moot by the Bureau’s decision. The following discussion is thus limited to comments about the effects of the proposed restrictions on the making of covered short-term loans and covered longer-term balloon-payment loans. As a threshold matter, it is important to put the effects of the final rule’s restriction on borrowing in the proper context. A consumer would be denied an additional covered short-term or longer-term balloon-payment loan only if the consumer was neither able to demonstrate an ability to repay the loan nor eligible for a conditionally exempt covered short-term loan. Bureau simulations described in the Section 1022(b)(2) Analysis indicate the final rule would restrict only six percent of borrowers from initiating a sequence they would have started absent the rule. Furthermore, even if the impact of the decline in lending results in the closure of a substantial number of storefronts offering covered short-term or longer-term balloon-payment loans, the Bureau expects that the vast majority of consumers will not see a sizable increase in the distance to the nearest storefront. As discussed in more detail in the Section 1022(b)(2) Analysis, the Bureau’s analysis of the impact of storefront closures in several States after the imposition of State restrictions on payday lending found that over 90 percent of payday borrowers had to travel 510 no more than five additional miles to access their nearest payday lending storefront.690 This is in addition to the option of obtaining a covered loan online. It is equally important to note that predicting how this relatively limited number of consumers will react to a particular restriction on covered loans in a particular circumstance is an imprecise matter given that, as noted above, the particular suite of restrictions imposed by the final rule has not been imposed by any State. The best that can be done is to make reasonable predictions about how consumers will react to these restrictions based on research concerning similar restrictions imposed by various States and other types of research, and the Bureau accordingly relies on such research in this discussion to the extent possible. In addition, even assuming that each of the alternatives identified by the commenters is in fact more expensive or harmful than covered short-term or longer-term balloon-payment loans, to the extent that a given consumer who cannot obtain a loan under § 1041.5 or § 1041.6 has access to other alternatives that are as or less expensive than other alternatives, that consumer could use those less expensive substitutes rather than one or more of the allegedly worse alternatives. In this regard, it is important to note that the Bureau’s decision not to finalize proposed §§ 1041.8 to 1041.10 means that covered longer-term installment loans will be at least as available after the rule goes into effect as they are in current market. Thus consumers who cannot obtain a covered short-term or longer-term balloon-payment loan may be able to turn to a longer-term installment loan which, in the view of the commenters who were concerned about inferior alternatives, is not injurious. The Bureau emphasizes, however, that it remains 690 CFPB Report on Supplemental Findings, at 79. 511 concerned about potential consumer harms from longer-term installment loans where loan pricing and structure may reduce the incentive for lenders to engage in careful underwriting, and the Bureau will monitor evolution of the market and take action under its supervisory and enforcement authorities as necessary to address identified consumer harms. In addition, the Bureau observes that some consumers may have access to some forms of credit that are typically less harmful than covered short-term loans and covered longer-term balloon-payment loans. These include some of the types of loans excluded from the final rule, including non-recourse pawn loans (discussed further below), no-cost advances, and advances made under wage advance programs that enable employees to access earned and accrued wages ahead of their payday. These options also include loans made by lenders who choose to comply with the conditional exemptions for alternative loans (akin to the PAL products administered by the NCUA) and accommodation loans. The Bureau now turns to a consideration of evidence and arguments concerning each of the alleged inferior alternatives identified by industry commenters. Non-recourse pawn loans. As noted in the section-by-section analysis for § 1041.3(d)(5), which excludes non-recourse pawn loans from the scope of coverage of the final rule, the Bureau believes that non-recourse pawn loans do not pose the same risks to consumers as covered loans because consumers are more likely to understand and appreciate the risks associated with nonrecourse pawn loans, and the loss of a pawned item that the lender has physical possession of is less likely to affect the consumer’s other finances. In addition, a consumer who cannot afford to repay a non-recourse pawn loan at the end of the loan term has the option not to return for the previously-surrendered household item, thus ending his indebtedness to the lender without 512 defaulting, re-borrowing, or impacting his ability to meet other financial obligations. A study described in the Section 1022(b)(2) Analysis found that non-recourse pawn lending increased in States that banned payday lending; a similar substitution effect may occur to some degree for consumers who are unable to obtain additional covered loans.691 Overdraft. Industry commenters and some individual consumer commenters expressed concern that consumers who are unable to access additional covered loans after exhausting the options permitted under the proposal will overdraw their bank accounts more frequently. Before considering whether there is likely to be a substitution effect towards overdraft, the Bureau notes that because many lenders of covered loans obtain access to a consumer’s bank account for repayment, these loans are often the cause of overdrafts for consumers who are unable to repay, and they contribute to account closures. See Market Concerns—Payments and the section-bysection analysis for §§ 1041.7 and 1041.8 for more details. Thus, even if overdrafts and bounced checks were to serve as a substitute for covered loans among some consumers, there still might be a net reduction in overdraft usage as a result of the rule. Further, Bureau research discussed in the proposal and the Supplemental Report calls into question certain commenters’ assumptions that consumers who cannot obtain covered short-term or longer-term balloon-payment loans will overdraw their bank accounts more frequently. The Bureau analyzed substitution patterns among former users of the deposit advance product (DAP) offered by several depository institutions when the offering of this product was discontinued in 691 Brian Baugh, “What Happens When Payday Borrowers Are Cut Off from Payday Lending? A Natural Experiment,” (Fisher College of Bus., Ohio State U. 2015). 513 the wake of the prudential regulator guidance.692 With discontinuation of DAP, consumers who had previously taken DAP advances did not discernably substitute towards other credit products or exhibit sustained negative outcomes compared to their non-user counterparts. Specifically, the former DAP users did not overdraw their bank accounts more frequently relative to non-users after the discontinuance of DAP, nor did they experience long-term increases in bank account charge-off rates following DAP’s discontinuation. In addition, the analysis also found that former DAP users did not change their use of payday loans offered by non-depository institutions in any meaningful way relative to those that did not use DAP. Additionally, an academic paper exploring the relationship between payday loan access and overdrafts shows that reduced access to payday loans leads to a decrease in the number of days a household experiences overdrafts or bounced checks.693 The Bureau notes, however, that if demand for short-term liquidity is inelastic and outside options were limited, a decrease in access to one option will necessarily increase the demand for its substitutes.694 The Bureau also notes the 2008 Morgan and Strain study discussed in the Section 1022(b)(2) Analysis and cited by several commenters, updated in 2012, which found that bounced checks and complaints about debt collectors to the FTC increase, and Chapter 13 bankruptcy filings decrease, in response to limits on payday lending. The updated study found that the service fees received on deposit accounts by banks operating in a single State tend to increase with limits on payday lending, and the authors interpreted this as an 692 CFPB Report on Supplemental Findings, part 2. Brian Baugh, “What Happens When Payday Borrowers Are Cut Off From Payday Lending? A Natural Experiment,) Payday Lending? A Natural Experiment,) (Ph.D. dissertation, Ohio State University, 2015), available at http://fisher.osu.edu/supplements/10/16174/Baugh.pdf. 694 See Romeo, Charles. 2017. “Estimating the Change in Surplus from the Elimination of Deposit Advance Products.” Working Paper, Office of Research, Consumer Financial Protection Bureau. 693 514 indication that payday loans help to avoid overdraft fees. The Bureau reiterates its critiques of the Morgan and Strain study as described the Section 1022(b)(2) Analysis. Unregulated Loans. As noted, some commenters argued that limiting the number of covered loans a consumer could obtain may result in a consumer who cannot obtain a loan under § 1041.5 or § 1041.6 using unregulated or illegal loans. Evidence does not suggest that additional regulation of covered loans leads to more borrowing of these loans. The Bureau notes that the comments often conflate two distinct things. The first is unregulated loans made over the Internet (sometimes from Tribal lands or offshore locations) to consumers who may live in States where payday loans are prohibited by usury restrictions. The second loans made by individuals associated with local criminal enterprises (i.e., neighborhood loan sharks). For instance, commenters sometimes describe in vivid terms the possibility of the rule resulting in criminal loan sharking accompanied by violent behavior, but then go on to present as evidence for that possibility some data or anecdotes about unregulated lenders operating online. The Bureau treats these cases differently in turn below. One study compared usage of online payday loans in States with restrictive payday lending regulations to usage in States with permissive payday lending regulations, since some unlicensed lenders of online payday loans may offer such loans without regard to the law of the State in which the consumer resided.695 The study concludes that usage rates of online payday loans do not significantly differ between States with restrictive and permissive payday loan laws, calling into question the notion that more consumers would turn to illegal lending sources if 695 PEW, Payday Lending in America: Who Borrows, Where They Borrow, and Why, p. 19-24, available at http://www.pewtrusts.org/~/media/legacy/uploadedfiles/pcs_assets/2012/pewpaydaylendingreportpdf.pdf. 515 covered loans offered by compliant lenders were curtailed. Similarly, another analysis examined the market penetration of non-licensed lending in States with varying payday lending regulations and found that the presence of non-licensed lenders was relatively minimal in all States, though somewhat higher in States with restrictive payday lending regulations overall in some years and somewhat lower in States with restrictive regulations in other years. However, States with restrictive payday lending regulations that also vigorously enforced those laws consistently had very low market penetration for non-licensed payday lending.696 A trade group critical of the proposal submitted a comment referencing a study that it stated “confirms that where payday credit has been restricted, consumers turn to online and unlicensed lenders.”697 The Bureau has reviewed the underlying study and does not believe that it confirms the commenter’s premise. The analysis posits that after Texas enacted its payday and vehicle title regulations in 2012, there was an increase in online payday lending applications and at the same time a subsequent decrease in storefront payday lending applications—which the author takes to mean that borrowers turned to online lenders when storefront loans became less available. However, the Texas regulations involved a licensing and disclosure regime that did not limit access to payday lending. An alternative explanation may be that these developments reflect the general market trends of storefront payday lending decreasing relative to online lending, which was experiencing large national growth during this period. Relatedly, the study’s finding that non-licensed lenders increased their online lending market share in Texas between 696 See NonPrime101, Report 2, Does State Regulation of Small-Dollar Lending Displace Demand to Internet Lenders?, p. 7 (2015), available at https://www.nonprime101.com/wp-content/uploads/2013/10/Does-StateRegulation-of-Small-Dollar-Lending-Displace-Demand-to-Internet-Lenders2.pdf. 697 Anna Ellison, Policis. The Outcomes for Consumers of Differing Approaches to the Regulation of Small Dollar Lending. See https://www.nonprime101.com/wp-content/uploads/2016/05/A_Ellison_nonPrime101_051016.pdf. 516 2011 and 2012 is likely similar to what happened nationally and was not caused by Texas law. The author also found that payday lending occurs to some degree in all States, regardless of how intensely it is regulated. If the author’s hypothesis held true that payday demand is inelastic and non-licensed lenders would step in to fill a void that licensed lenders could not, the Bureau would expect the usage rates to be fairly similar in each of these groups of States, since they are all indexed to the subprime population. But it should be noted that use in restrictive and banned States is lower than in permissive States. Illegal lenders/loan sharks. Finally, the Bureau believes the risk that consumers will be denied access to credit due to the impacts of the final rule and will be forced to turn to illegal lenders such as loan sharks is not supported by available evidence. Although a number of commenters made this argument, they offered little to no specific evidence about the prevalence of loan sharking in States that restricted payday and vehicle title lending. The Bureau notes the receipt of a comment letter from a trade group referencing a paper that discusses, among other issues, analyses of loan sharking activity in other countries. The Bureau does not find this analysis to be persuasive, since the regulatory context, access to credit for subprime populations, and characteristics of unlicensed lending are quite different in those jurisdictions than in the United States, as the author of the study acknowledges.698 In addition, as noted above, under the final rule credit-impaired borrowers could still obtain credit through 698 The trade group letter cites Todd J. Zywicki, “The Case Against New Restrictions on Payday Lending,” Mercatus Center, George Mason Univ., No. 09-28 (July, 2009), available at https://www.mercatus.org/system/files/WP0928_Payday-Lending.pdf. The author of the study wrote that, “The flexibility of consumer credit markets in the United States has substantially reduced the importance of illegal loanshark lending,” and goes on to describe unregulated internet lending—rather than neighborhood loan sharking—as where credit-constrained consumers would turn. Id. at 20. The Bureau discusses issues relating to unregulated loans above. Moreover, the author notes that Japan and Germany both had strict price caps, which the Bureau is not authorized to impose. See id. at 18-19. 517 various alternatives discussed above (including conditionally exempt loans provided for in the rule and longer-term installment loans which are not subject to the ability-to-repay requirements of the final rule). Similarly, a State trade group commenter argued that the Bureau had not properly accounted for the possibility of loan sharking in its assessment of costs and benefits, arguing that racketeering actions related to lending are more highly concentrated in jurisdictions that do not allow alternative forms of credit such as Pennsylvania, New York, and New Jersey. However, the Bureau views what was cited as supposed support to be anecdotal, non-specific, and lacking evidentiary weight.699 Even if the Bureau assumed the commenter was correct that loan sharking activities are prevalent in those jurisdictions, the Bureau believes the evidence cited fails to establish even a basic correlation between loan sharking and State differences in authorizing small-dollar lending, let alone a causal link. The Bureau also notes receipt of a comment letter attaching a law review article analyzing the history of loan sharking in the consumer credit markets and the relationship between loan sharking and usury caps in the United States. The article argues that the “loanshark thesis” offered by proponents of deregulating the credit markets is “seriously flawed.” Among the evidence cited was that in Vermont, which has one of the lowest interest rate caps in the nation, no Federal indictments have been recorded in the State during the 20-year period 699 The commenter asserts a mere search of FBI or DOJ records or Google Scholar cases, or a general internet search, “all demonstrate the prevalence of loan shark and racketeering actions related to lending more highly concentrated in jurisdictions that do not allow alternative forms of credit.” However, the commenter then provides an example of a single case of loan sharking in Philadelphia in 2013, without citation to news articles, court records, or any other evidence. The commenter also mentions “other examples” in New York and New Jersey without any specification. 518 prior to 2012 (when the article was published) for engaging in an extortionate credit transaction, nor had the local press published a single story in that time about local black-market lending.700 The Bureau further notes that the U.K. Financial Conduct Authority (the FCA) recently issued a report summarizing feedback it had received in assessing the impacts of the FCA’s 2015 price cap on high-cost short-term credit.701 The FCA wrote, “We do not see strong evidence of a rise in illegal money lending because of the price cap.” The report explains the basis for the prediction it had made, in imposing the price cap, that less than 5 percent of declined applicants would consider turning to illegal money sources, and in the recent report the FCA stated that the results from their recent survey confirmed this prediction. The FCA cautioned that the individuals who use illegal money lenders are difficult to reach and reluctant to talk about their experience, but noted that they gleaned information through discussions with social service organizations and other individuals who could speak with authority on the prevalence of illegal lending behavior in the United Kingdom. If the hypothesis was that regulating payday and vehicle title lending in ways that restrict access would lead to an increase in illegal lending, then a nationwide price cap is the type of broad, substantive restriction on small-dollar lending that one may surmise would cause such a rise. Given the difficulty in generalizing across different legal systems and credit markets, the Bureau does not view such findings as dispositive, but does view them as instructive. At the very least, they cast doubt on the assertions made by the trade group that had cited the study about illegal lending in Germany and Japan discussed above. 700 Robert Mayer, “Loan Sharks, Interest-Rate Caps, and Deregulation,” 69 Wash. & Lee L. Rev. 807, 841 (2012). Financial Conduct Authority, High-cost credit: Including review of the high-cost short-term credit price cap, Feedback Statement FS17/2 (July 2017) at 5 https://www.fca.org.uk/publication/feedback/fs17-02.pdf. 701 519 Finally, the Bureau reemphasizes that the various types of alternatives described above will remain available. Thus, the Bureau concludes that the number of consumers who would seek these illegal options as a first resort is next to zero, and as a last resort is still quite low.702 Assessing Benefits to Competition. In the proposal, the Bureau concluded that the rule would not have a significant impact on competition, in part because the Bureau had observed, as discussed above, that when lenders make covered short-term loans they typically charge the maximum price permitted under State law. Many lenders objected to that claim in their comments, and some provided examples of how prices can differ—including statistics on the difference between State-regulated lender prices and online lender prices, and differences between nationwide average prices versus industry medians. Other commenters noted that lenders compete on non-price terms. The Bureau acknowledges that a certain amount of market consolidation may impact the competition involved in non-price terms, meaning consumers may be presented with fewer choices as to where to go to obtain a loan. The impact that market consolidation has on pricing, however, is generally capped by existing State law requirements. Another point made by industry commenters was that the Bureau’s own analysis showed that the proposed rule would lead to increased concentration in the market for covered short-term loans, thereby undermining competition. Indeed, these commenters asserted that the Bureau had understated the amount of decline in revenue that would follow from its proposal and thus had underestimated the impact of the proposal in reducing competition. These comments, however, 702 The Bureau notes that other government entities have the authority to prosecute such actors under applicable criminal statutes at the State and Federal level. 520 largely misunderstood the Bureau’s analysis of the actual effects on competition. The Bureau did believe that the requirement to underwrite covered loans by making a reasonable assessment of the borrower’s ability to repay the loan according to its terms would cause consolidation in the market, which the Bureau attempted to estimate to the extent feasible. Yet the Bureau presented preliminary findings, based on its observed experience of the markets in States that had adopted modifications to their own payday lending regulations, which indicated that market consolidation would not reduce meaningful access to credit among consumers. As discussed above, the upshot of such consolidation was that lenders remained almost as proximate and available to consumers as before. To the extent the industry commenters present different estimates, the Bureau is not persuaded of their likely accuracy, and these issues are addressed further in part VII, which presents the Bureau’s consideration of the benefits, costs, and impacts of the final rule on consumers and covered persons pursuant to section 1022(b)(2)(A) of the Dodd-Frank Act.703 Moreover, as discussed above, in light of the comments received, the Bureau has adjusted certain parameters of the proposed rule to simplify its scope, streamline the underwriting process, and add more flexibility within the existing framework, as described more fully below in the explanation of § 1041.5 of the final rule. The effect of these adjustments is to reduce the costs associated with complying with the rule, which likely will reduce the estimated amount of consolidation in the market for covered short-term loans. For all of these reasons, the Bureau concludes, based on its judgment and expertise, the comments it received on all sides of these issues, and the data on injury and the effects of the identified practice set forth above in Market Concerns—Underwriting and the analysis in part 703 12 U.S.C. 5512(b)(2)(A). 521 VII below, which presents the Bureau’s consideration of the benefits, costs, and impacts of the final rule on consumers and covered persons, that the practice of making covered loans without reasonably assessing the borrower’s ability to repay the loan according to its terms is injurious to consumers, on net and in the aggregate, taking into consideration the countervailing benefits of the identified practice. Consideration of public policy The Bureau’s Proposal Section 1031(c)(2) of the Dodd-Frank Act allows the Bureau to “consider established public policies as evidence to be considered with all other evidence” in determining whether a practice is unfair, as long as the public policy considerations are not the primary basis of the determination. In the proposal, the Bureau stated that public policy supports the proposed finding that it is an unfair practice for lenders to make covered loans without determining that the consumer will have the ability to repay the loan according to its terms. Specifically, as noted in the proposal, several consumer financial statutes, regulations, and guidance documents require or recommend that covered lenders must assess the customer’s ability to repay before extending credit. These include the Dodd-Frank Act provisions on closed-end mortgage loans,704 the CARD Act provisions on credit cards,705 guidance from the 704 Dodd-Frank Act section 1411, codified at 15 U.S.C. 1639c(a)(1) (“no creditor may make a residential mortgage loan unless the creditor makes a reasonable and good faith determination based on verified and documented information that, at the time the loan is consummated, the consumer has a reasonable ability to repay the loan, according to its terms, and all applicable taxes, insurance (including mortgage guarantee insurance), and assessments.”). 705 15 U.S.C. 1665e (credit card issuer must “consider[] the ability of the consumer to make the required payments”). 522 OCC on abusive lending practices,706 guidance from the FDIC on small-dollar lending,707 and guidance from the OCC708 and FDIC709 on deposit advance products. In addition, the Federal Reserve Board promulgated a rule requiring an ability-to-repay determination for higher-priced mortgages, although that rule has since been superseded by the Dodd-Frank Act’s ability-torepay requirement and the Bureau’s implementing regulations, which apply generally to mortgages regardless of price.710 In short, the Bureau stated in the proposal that Congress, State legislatures,711 and other agencies have found consumer harm to result from lenders failing to determine that consumers have the ability to repay before extending credit to them. The Bureau stated that these established policies provide support for its preliminary finding that it is unfair for a lender to make covered loans without determining that the consumer will have the ability to repay; and they likewise were seen as supporting the Bureau’s proposed imposition of the consumer protections in the proposed rule. The Bureau gave weight to the policy contained in these Federal consumer laws, and based its preliminary finding that the identified practice is 706 OCC Advisory Letter 2003-3, Avoiding Predatory and Abusive Lending Practices in Brokered and Purchased Loans (Feb. 21, 2003), available at http://www.occ.gov/static/news-issuances/memos-advisoryletters/2003/advisory-letter-2003-3.pdf (cautioning banks not to extend credit without first determining that the consumer has the ability to repay the loan). 707 FDIC Financial Institution Letter FIL-50-2007, Affordable Small-Dollar Loan Guidelines (June 19, 2007). 708 OCC, Guidance on Supervisory Concerns and Expectations Regarding Deposit Advance Products, 78 FR 70624, 70629 (Nov. 26, 2013) (“Deposit advance loans often have weaknesses that may jeopardize the liquidation of the debt. Customers often have limited repayment capacity. A bank should adequately review repayment capacity to assess whether a customer will be able to repay the loan without needing to incur further deposit advance borrowing.”). 709 FDIC, Guidance on Supervisory Concerns and Expectations Regarding Deposit Advance Products, 78 FR 70552 (Nov. 26, 2013) (same as OCC guidance). 710 Higher-Priced Mortgage Loan Rule, 73 FR 44522, 44543 (July 30, 2008) (“the Board finds extending higherpriced mortgage loans or HOEPA loans based on the collateral without regard to the consumer’s repayment ability to be an unfair practice. The final rule prohibits this practice.”). 711 See, e.g., 815 Ill. Comp. Stat. Ann. 137/20 (lender must assess ATR in making “high risk home loan”); Nev. Rev. Stat. Ann. § 598D.100 (it is unfair practice to make home loan without determining ATR); Tex. Educ. Code Ann. § 52.321 (State board will set standards for student-loan applicants based in part on ATR). 523 unfair, in part, on that significant body of public policy. Yet the Bureau did not make this consideration the primary basis for its preliminary determination of unfairness. Comments Received The Bureau received comments relating to the public policy implications of the proposed rule. One industry commenter argued that because the Bureau lacked substantial evidence for its other determinations, it was essentially basing the unfairness determination primarily on public policy, which is prohibited by the Dodd-Frank Act. Other industry commenters contended that public policy considerations militate against promulgating a rule that restricts access to credit to the extent described in the proposal. For example, some commenters claimed that restricting access to credit for certain borrower populations conflicts with public policy considerations underlying fair lending laws. Industry commenters also cited perceived conflicts with other sources of law as contravening public policy. One commenter made a similar argument about the proposal’s coverage of the furnishing and review of credit information, which it viewed as inconsistent with the Fair Credit Reporting Act and thus as inconsistent with public policy. Other commenters more simply argued that in addressing the perceived issues with covered loans, the Bureau should be required to defer to existing State regulatory approaches. Some commenters stated quite different views, as discussed previously. One trade association, in particular, stated that Congress plainly recognized the problems created by unregulated and less regulated lenders, and for that reason conferred on the Bureau new authority to supervise and write rules for the payday lending industry for the first time ever at the Federal level. More generally, consumer groups were strongly supportive of the Bureau’s legal authority 524 to develop and finalize the proposed rule. Rather than viewing other ability-to-repay provisions in Federal consumer law as implied negative restrictions on the Bureau’s authority, these commenters pointed to them and others (such as the Military Lending Act) as embodying a considerable trend of expanding public policy now supporting the principle that consumer lending generally should be premised on the borrower’s ability to repay. They also noted that some States now embody this principle in statute, and many more do so by judicial precedent. They noted that general statements of this principle in Federal and State law tend to define this approach as requiring the lender to establish the borrower’s ability to repay the loan while meeting basic living expenses and without re-borrowing. One commenter argued that the proposed rule contradicts other recent Federal policy that authorizes and even promotes mortgages, auto loans, and other types of long-term lending. Several commenters argued that the rule violates the public policy of federalism because it would prohibit certain lending practices that are otherwise allowed and regulated by State laws, which reinforce the structure of such loans and mitigate harms to consumers. On the other side of the issue, commenters argued that the Bureau’s rule is increasingly consistent with the evolving direction of State law. The Final Rule As an initial matter, the Bureau notes that public policy is only one factor that it uses to inform its unfairness assessments; it is not a prerequisite or an element of the legal determination or its primary basis. The Bureau has concluded that this rule is consistent with public policy, but commenters’ argument that the rule is primarily based on public policy is inaccurate. As stated in the proposal, the identified practice of making covered loans without reasonably assessing the 525 borrower’s ability to repay the loan according to its terms is unfair because it meets the three legal elements of unfairness, and the rule is also supported by public policy. The rule does not conflict with Federal fair lending laws. The Bureau will continue to expect creditors to treat borrowers of protected classes equally. Additionally, the rule does not conflict with the Fair Credit Reporting Act. Lenders can comply with the provisions of both this rule and the FCRA and will be expected to do so. To the extent that Federal policy is intended to promote long-term lending, this rule does not conflict with that objective. First, the Bureau is unaware of any Federal policy that specifically prefers long-term lending simply for the sake of long-term indebtedness. Certain Federal policies may allow longer-term installment lending in order to reduce payment amounts, but the covered short-term loans at issue in this rule do not involve reduced payment amounts as a result of re-borrowing. The Bureau does not agree with the commenters who claimed that this rule conflicts with general principles of federalism, even though some loans that would not be permissible under the rule would currently be permissible under State law. If the commenters’ argument were to be accepted, then any Federal regulation (other than rules prohibiting only the exact conduct already prohibited by the States) would create an impermissible conflict with principles of federalism. Yet that is not how our system of federalism works. Under the Constitution, both the States and the Federal government have coexisting, overlapping authority. This rule preserves that settled framework by stating explicitly that it does not preempt any State law that is more restrictive in its effects than the provisions of this rule. Existing State regulatory frameworks will continue to exist alongside this rule, in a version of cooperative federalism that is analytically similar to the 526 way parallel State and Federal laws have long operated in such fields as securities law, antitrust law, environmental law, and many others. The Bureau is unaware of any State laws that a lender of covered short-term loans cannot comply with as a consequence of this rule. Indeed, the making of covered short-term loans pursuant to State regulatory frameworks is already subject to significant Federal laws and regulations, as many commenters acknowledge. Those Federal laws include the Truth in Lending Act, the Equal Credit Opportunity Act, the Fair Credit Reporting Act, and others. To the extent those laws control or modify various aspects of the covered loans made pursuant to State law, they do not thereby contravene the principles of federalism. In fact, the final rule adopted by the Bureau also provides support for those States that effectively prohibit the making of certain types of covered loans by imposing a hard usury cap on such lending, insofar as the rule will restrict lenders from offering non-underwritten covered loans on-line or by other avenues of cross-border lending into those States, which are also empowered to enforce their usury caps against cross-border loans that violate those caps. The Bureau disagrees with the contention that it only has the authority to issue rules based on unfairness that incorporate an ability-to-repay standard if Congress expressly specified the use of such a standard. On the contrary, Congress created the Bureau and chartered it with the responsibility to identify and prevent unfair practices, employing general statutory definitional criteria as set forth in the Dodd-Frank Act. Congress did not explicitly preclude the issuance of rules based on unfairness that incorporate an ability-to-repay standard, and the Bureau has not found in the statute, its legislative history, or other authoritative sources any implied preclusion of rules based on unfairness that incorporate an ability-to-repay standard. And the Bureau is authorized to adopt appropriate rules when it has determined that an ability-to527 repay standard is appropriate to address a practice that it has identified as meeting the definition of “unfair” under the criteria enunciated by Congress in the statute. Indeed, Congress reinforced the Bureau’s authority to engage in rulemaking in this particular market by providing in section 1024(a)(1)(E) of the Dodd-Frank Act that this was one of three specified markets (along with mortgages and private student loans) where the Bureau had broad authority to adopt regulations that apply to “any covered person who . . . offers or provides to a consumer a payday loan.”712 As for those commenters who stated that the Bureau is obliged to consider and defer to State-law regimes for regulating covered loans, it suffices to note that this approach does not square with the terms of Federal law as prescribed in the Dodd-Frank Act. It also fails to recognize that even in light of varying State regulatory structures, the injury caused by covered loans persists in those States where it is permitted to exist. And those States, of course, are the sources of all the data that the Bureau has compiled on the harms of covered loans in the United States (since the so-called “prohibition States” cannot, by definition, be the source of any current data on the making or effects of those loans). Finally, commenters who criticized the Bureau as violating some version of public policy by acting too aggressively to limit or even eliminate covered short-term loans altogether were overstating their point while at the same time missing the point. Again, the approach proposed by the Bureau and now adopted in the final rule does not eliminate such loans. Rather, it merely imposes a requirement that they be underwritten by the lender making a reasonable assessment that the borrower will be able to repay the loan according to its terms. And especially in light of various adjustments the Bureau has now made to simplify and streamline the underwriting 712 12 U.S.C. 5514(a)(1)(E). 528 provisions in § 1041.5 of the final rule, along with some ability to make covered loans under the alternative provisions of § 1041.6, the notion that the final rule will eliminate these loans altogether is not well grounded in any factual analysis. Abusiveness Under sections 1031(d)(2)(A) and (B) of the Dodd-Frank Act,713 the Bureau may find an act or practice to be abusive in connection with a consumer financial product or service if the act or practice takes unreasonable advantage of: (A) a lack of understanding on the part of the consumer of the material risks, costs, or conditions of the product or service or of (B) the inability of the consumer to protect the interests of the consumer in selecting or using a consumer financial product or service. In the proposal, the Bureau stated that it appeared that a significant population of consumers does not understand the often-hidden risks and costs of taking out payday, single-payment vehicle title, or other covered loans, and further lack the ability to protect their interests in selecting or using such loans. It also stated that it appeared that lenders take unreasonable advantage of these consumer vulnerabilities by making loans of this type without reasonably determining that the consumer will have the ability to repay the loan. After considering the comments received, for the reasons described below, the Bureau concludes that it is an abusive practice to make covered short-term loans without reasonably assessing that the borrower will have the ability to repay the loan according to its terms. The Bureau concludes that many borrowers lack an understanding of the material risks and costs of these loans, based on evidence that many borrowers do not seem to understand the likelihood or the severity of the harms that can result from such unaffordable loans. The Bureau concludes 713 12 U.S.C. 5531(d)(2)(A)&(B). 529 that borrowers are unable to protect their interests based on the circumstances of many borrowers, such as their typically urgent need of credit, their perception that they often lack a realistic ability to shop for alternatives, and above all the difficulties they face after origination of the first unaffordable loan based on various features of the loan product that create and magnify the potential risks and harms. And finally, by making covered short-term loans without reasonably assessing the borrower’s ability to repay the loan according to its terms, and based on various features of the structure of such loans, lenders are taking unreasonable advantage of these vulnerabilities. General Comments Before turning to its analysis of the statutory prongs of the abusiveness standard, the Bureau can first address a small set of general comments on its use of the abusiveness standard generally. Some commenters asserted that the proposed rule improperly amounts to a “ban” on certain products, instead of focusing on the identified practice of making covered loans without reasonably assessing consumers’ ability to repay. Other commenters asserted that when a practice is expressly permitted by some applicable law, including State law, it cannot also be abusive. One commenter pointed to statements made in the Bureau’s own exam manual as ostensible support for opposing the Bureau’s use of its abusive authority to impose this rule. The suggestions that the rule effectuates a “ban” on products rather than a prohibition against acts or practices are inaccurate. The Bureau did not propose, and this final rule does not provide, that any covered short-term loans are prohibited. The practice of failing to make such an assessment has been identified by the Bureau as the practice that is both unfair and abusive. In response, the rule simply requires that such loans must be underwritten with a reasonable 530 assessment of the borrower’s ability to repay the loan according to its terms. Further analysis on the effect of this rule on the market for such loans can be found above in the discussion of the statutory unfairness prong, as well as in part VII, where the Bureau presents its assessment of the costs, benefits, and impacts of the final rule on consumers and covered persons pursuant to section 1022(b)(2)(A) of the Dodd-Frank Act.714 As to the assertion that a practice cannot be abusive when it is expressly permitted by some applicable law, this statement seems overbroad and inaccurate, for when a new rule is promulgated, it would often be the case that the conduct it now addresses would previously have been permitted, and perhaps even explicitly permitted, before the law was changed by the new rule.715 By the same token, the observation made about the Bureau’s examination manual is irrelevant, because the manual would only have been describing the existing state of the law prior to the promulgation of this new rule. Many if not most new rules adopted by the Bureau that add new substantive requirements may not be anticipated by examination manuals written to guide examiners in applying the pre-existing legal landscape before the rule was adopted. As is common when new rules are adopted, the Bureau plans to produce new examination procedures to reflect the new substantive requirements of this final rule. Moreover, the Bureau may properly exercise its statutory authority at any time to consider whether an identified practice meets the definitional prongs of unfairness or abusiveness, based on substantial evidence and research. When it does so, it reaches an 714 12 U.S.C. 5512(b)(2)(A). Without undermining this general point, it should be noted that where, as here, the Bureau is adopting rules pursuant to its authority to identify and prevent unfair and abusive practices, such rules are not necessarily creating new law so much as clarifying that these practices, which could have been addressed previously by the Bureau pursuant to its supervision and enforcement authority, are now addressed independently by essentially codifying them in the terms of the new rule. 715 531 appropriate conclusion that the identified practice is illegal under the provisions of the Federal statute, regardless of whether lenders had been engaging in the practice prior to the time the Bureau completed its new analysis. Furthermore, the fact that State laws on the same subject may be less restrictive in some respects than Federal law does not prohibit the promulgation of a regulation that is authorized by Federal statute, even though it may be more restrictive in some respects than those State laws. This is typical of how federalism traditionally works in other areas of parallel Federal and State law, such as securities, antitrust, environmental law, and many other areas. Consumers lack an understanding of material risks and costs The Bureau’s Proposal As discussed in the proposal, covered short-term loans, including payday and title loans, can and frequently do lead to a number of negative consequences that can pose serious financial problems for consumers. These effects flow from the identified practice of failing to underwrite such loans by making a reasonable assessment of the borrower’s ability to repay the loan according to its terms. The harms that borrowers tend to experience once they have taken out an unaffordable loan of this kind include default, delinquency, re-borrowing, and the collateral consequences of making unaffordable payments, including forgoing basic living expenses and major financial obligations to avoid the other injuries. All of these potentially harmful effects— including the direct costs that the borrower has to pay to the lender, as well as other costs that often are incurred as well—are among the “material risks and costs” of these loans, as the Bureau understood and reasonably interpreted that phrase. 532 In the proposal, the Bureau recognized that borrowers who take out a payday, title, or other covered short-term loan typically understand that they are incurring a debt which must be repaid within a prescribed period of time and that if they are unable to do so, they will either have to make other arrangements or suffer adverse consequences. The Bureau stated, however, that it did not believe that such a generalized understanding suffices to establish that consumers actually understand the material risks and costs of these products, and in particular the magnitude and severity of the risks and harms. Rather, the Bureau stated that it believed it was reasonable to interpret “understanding” in this context to mean more than a mere awareness that it was within the realm of possibility that a negative consequence could be experienced as a result of using the product. For example, consumers may not understand that a certain risk is very likely to materialize or that—even though relatively rare—the impact of a particular risk would be severe. As discussed in the proposal, the single largest risk to a consumer of taking out an initial covered short-term loan is that it will lead to an extended cycle of indebtedness that poses material risks and costs to the consumer. This occurs in part because of the identified practice, which can lead to lenders making unaffordable loans. It also occurs, in large part, because the term and structure of the loan generally require the consumer to make a lump-sum balloon payment within a short period, typically two weeks or a month after the loan is made, often absorbing such a large share of the consumer’s disposable income as to leave the consumer unable to pay basic living expenses and major financial obligations. As the Bureau stated in the proposal, in States where it is permitted, lenders often offer borrowers the enticing—but ultimately costly—alternative of paying a fee and rolling over the 533 loan or taking out a new loan to pay off the previous one, leaving the principal amount intact. Many borrowers choose this option, and a substantial population of them ends up in extended loan sequences because when the loan next comes due, they are in exactly the same situation all over again. Alternatively, borrowers may repay the loan in full when it comes due, but find it necessary to take out another loan over the course of the ensuing expense cycle because the large amount of money needed to repay the first loan, relative to their income, leaves them without sufficient funds to meet their other obligations and expenses. This also can often lead to an extended cycle of debt, posing material risks and costs to the consumer’s financial situation. This cycle of indebtedness affects a large segment of borrowers: as described above in Market Concerns—Underwriting, half of all storefront payday loan sequences contain at least four loans.716 One-third contain seven loans or more, by which point consumers will have paid charges equal to 100 percent of the original amount borrowed and still owe the full amount of the principal.717 Almost one-quarter of loan sequences contain at least 10 loans in a row.718 And looking just at loans made to borrowers who are paid weekly, bi-weekly, or semi-monthly, more than one-fifth (21 percent) of those loans are in sequences consisting of at least 20 loans.719 For loans made to borrowers who are paid monthly, almost half (46 percent) of the loans are in sequences consisting of at least 10 loans.720 The evidence summarized in the proposal and reinforced above in Market Concerns— Underwriting and again in the section on unfairness also shows that many consumers who take 716 CFPB Report on Supplemental Findings. Id. 718 Id. 719 Id. 720 Id. 717 534 out these loans appear not to understand, when they first take out the loan, how long they are likely to remain in debt and how costly and harmful that situation could be for them. Many borrowers tend to overestimate their likelihood of repaying the loan without re-borrowing and do not understand the likelihood that they will end up in an extended loan sequence. As the Bureau stated in the proposal, empirical evidence shows that a substantial population of borrowers, and especially those who end up in extended loan sequences, are not able to predict accurately how likely they are to re-borrow and thus how much they will end up paying over time. One study, in particular, found that consumers who end up re-borrowing numerous times—which are the consumers who suffer the most harm—are particularly bad at predicting the number of times they will need to re-borrow. Thus, many consumers who find themselves in a months-long cycle of indebtedness do not understand the material risks and costs of that consequence, and end up paying hundreds of dollars in fees above what they expected, while struggling to meet their other financial obligations. As recounted in the same sections identified above and in the proposal, the Bureau has observed similar outcomes for borrowers of single-payment vehicle title loans. For example, 83 percent of title loans are re-borrowed on the same day that a prior loan was due, and 85 percent of vehicle title loans are re-borrowed within 30 days of a previous vehicle title loan.721 Fifty-six percent of vehicle title loan sequences consist of more than three loans, 36 percent consist of at least seven loans, and almost one quarter—23 percent— consist of more than 10 loans.722 While there is no comparable research on the subjective expectations of title borrowers, the Bureau 721 722 CFPB Single-Payment Vehicle Title Lending. Id. 535 preliminarily found that the research in the payday context can be extrapolated to these other single-payment short-term products, given the significant similarities in the product structures, the characteristics of the borrowers, and the outcomes that many borrowers experience, as detailed above in part II and in Market Concerns—Underwriting. Consumers are also exposed to other material risks and costs in connection with these kinds of loans. As discussed in more detail in Market Concerns—Underwriting, the unaffordability of the payments creates, for many consumers, a substantial risk of default. Indeed, 20 percent of payday loan sequences and 33 percent of title loan sequences end in default.723 And 69 percent of payday loan defaults occur in loan sequences in which the consumer re-borrows at least once.724 For a payday borrower, the cost of default generally includes the cost of at least one, and often multiple, NSF fees assessed by the borrower’s bank when the lender attempts to cash the borrower’s postdated check or debit the consumer’s account via ACH transfer and the attempt fails. It is also known that NSFs on on-line payday loans are associated with a high rate of bank account closures, further jeopardizing the financial health and stability of these consumers. Defaults often expose consumers to other adverse consequences, such as aggressive debt collection activities. The consequences of default can be even more dire for a title borrower, including repossession of the consumer’s vehicle—which is the result in 20 percent of single-payment vehicle title loan sequences and can greatly complicate the borrower’s ability to earn the funds needed to repay such loans.725 723 CFPB Report on Supplemental Findings; CFPB Single-Payment Vehicle Title Lending. CFPB Report on Supplemental Findings. 725 CFPB Single-Payment Vehicle Title Lending. 724 536 The Bureau stated in the proposal that it believed a substantial population of consumers who take out payday, title, or other covered short-term loans do not understand the magnitude of these additional risks. The proposal also stated that borrowers—at least at the point where they are first deciding whether to take out the loan—are not likely to factor into their decision-making the severity of the harms they may suffer from default, delinquency, re-borrowing, and the collateral consequences of making unaffordable payments in an attempt to avoid these other injuries. Further adverse effects can include expensive bank fees, the potential loss of their bank account, aggressive debt collection efforts, and, with title loans, the risks and costs of losing their basic transportation to get to work or conduct their ordinary personal affairs. As discussed in the proposal, several factors can impede consumers’ understanding of the material risks and costs of these loans. At the outset, as discussed above, there is a mismatch between how payday and single-payment vehicle title loans are structured and marketed to consumers and how they operate in practice to support a business model based on repeated reborrowing. Although the loans are presented and marketed as stand-alone short-term products, lenders know and rely on the fact that only a minority of payday loans are repaid without any reborrowing. As discussed above, these loans often, instead, produce lengthy cycles of indebtedness through extended loan sequences of repeat re-borrowing. This is influenced by the term and the balloon-payment structure of the loans, which offer the limited options of either reborrowing by paying additional fees without paying down the principal amount or requiring a large payment to be made all at once, which can lead to severe consumer harm if the lender makes an unaffordable loan without reasonably determining that the borrower has the ability to repay the loan according to its terms. 537 In addition, consumers in extreme financial distress tend to focus on their immediate liquidity needs, rather than potential future costs, in a way that makes them highly susceptible to lender marketing. Payday and title lenders are generally aware of this vulnerability and often advertise the speed with which the lender will provide funds to the consumer, which may further cloud consumers’ ability to understand the risks and costs.726 But while covered short-term loans are marketed as being intended for short-term or emergency use,727 a substantial percentage of consumers do not repay the loan quickly and thus confront the harms of default, delinquency, reborrowing, and the collateral consequences of making unaffordable payments in an attempt to avoid these other injuries. Many consumers find themselves caught in a cycle of re-borrowing that is both very costly and very difficult to escape. Comments Received The Bureau received many comments relating to this prong of the abusiveness definition concerning consumers’ lack of understanding of material risks and costs associated with the kinds of loans covered by the rule. Industry participants, trade associations, and others who criticized the Bureau for proposing the rule in response to this concern maintained that consumers do understand the terms of the loans and the possible outcomes, making a more detailed understanding of the risks unnecessary, and making the rule unnecessary as well. They argue that it is unrealistic to require, as they believed the Bureau’s proposed rule did, that consumers develop an expert understanding of the characteristics of covered short-term loans. 726 In fact, during the SBREFA process for this rulemaking, numerous SERs commented that the Bureau’s contemplated proposal would slow the loan origination process and thus negatively impact their business model, though these points may be addressed by the development of automated underwriting, as discussed earlier. 727 For example, as noted above, the website for a national trade association representing storefront payday lenders analogizes a payday loan to a “cost-efficient ‘financial taxi’ to get from one payday to another when a consumer is faced with a small, short-term cash need.” 538 Those commenters who maintained that the risks and costs are sufficiently understood by consumers claimed that the proposed rule improperly substitutes the Bureau’s own judgments for those of consumers, denying them the ability to make a free choice to purchase products about which they do, in fact, know and appreciate how they work. Many commenters, including individual users of covered short-term loans, asserted that consumers use them effectively to cope with unexpected temporary expenses or shortfalls in income, to manage uneven income and cash flow challenges, and to avoid more expensive alternatives for handling other debt. They cited various studies to support the proposition that consumers understand the challenges and disadvantages of these loans, but opt for them as the best choice available among unappealing alternatives. Other commenters stated that no evidence suggests borrowers of covered loans generally suffer from infirmity or ignorance, but rather are well-educated and sophisticated in how they use financial services. Several commenters pointed to the relatively small number of consumer complaints submitted to the Bureau about these kinds of loans, and to the high volume of positive comments submitted about such loans in response to the proposal, which were viewed as showing that consumers who use these loans understand them. Many individual users of such loans likewise commented that they use these products advisedly to meet their particular needs. In the alternative, industry commenters contended that the Bureau’s method for determining that consumers do not understand these risks is flawed, such as by relying too heavily on concepts of behavioral economics, which would leave an essential premise for the rule unproven. Other commenters argued that consumers are generally accurate in predicting the duration of their borrowing, citing the Mann study and Professor Mann’s response to the 539 Bureau’s proposal, a point that was raised and discussed earlier in Market Concerns— Underwriting, as well as in the section on unfairness. Other commenters such as consumer groups agreed with the Bureau’s assessment in the proposal that many consumers do not understand the material risks and costs associated with these kinds of loans, which they viewed as resting on sound underpinnings of the facts and data marshaled by the Bureau. Once again, the commenters said this was especially true of borrowers who end up in extended loan sequences, and the financial circumstances of these consumers are materially undermined by their experience with such loans. They are unable to repay the loans when they come due, which leads them to re-borrow repeatedly and, in many instances, to suffer the injuries associated with being trapped in extended loan sequences. Consumer groups expressly agreed with the weight placed by the Bureau on concepts from behavioral economics such as “tunneling risk” and “optimism bias,” which they stated are well-established phenomena. Another commenter noted that their experiences with legal assistance clients showed consistent confusion about the risks, costs, and conditions of these loans, as well as the excessive optimism many consumers have about their expected ability to pay off the loans as they come due. This perspective was supported by many comments by and about individual users of such loans, whose experiences contrasted sharply with other cohorts of borrowers who commented on the proposal in more critical terms. Some industry commenters argued that lack of understanding must be evaluated at the level of each consumer and thus cannot serve as the basis for a broad rulemaking of general applicability. Some commenters pointed to prior statements by the Bureau’s Director, who stated that abusiveness cases are “unavoidably situational” and depend on an individualized 540 inquiry of the facts and circumstances presented. Other commenters noted that the abusiveness standard is worded in the singular—“a lack of understanding on the part of the consumer”—to support this assertion.728 Another commenter suggested that measures should be taken to combat advertising and marketing problems rather than accepting the restrictions on access to credit that would result from the proposed rule. Yet another industry commenter took a different approach, objecting that there was no evidence that the proposed rule could prevent the harms to consumers that it purported to address. The Final Rule After careful consideration of the comments received, the Bureau has concluded that when lenders make covered short-term loans without reasonably assessing whether borrowers have the ability to repay the loans according to their terms, consumers often lack understanding of the material risks and costs of these loans, which are often unaffordable and lead to the risks and harms of default, delinquency, re-borrowing, and the negative collateral consequences of forgoing basic living expenses and major financial obligations in order to avoid defaulting on their loans. Many of the points made by commenters objecting to whether the rule satisfies this prong of the definition of abusive practices rely on arguments that conflict with credible evidence cited by the Bureau in support of the proposed rule. That evidence is discussed more thoroughly in Market Concerns—Underwriting, the Section 1022(b)(2) Analysis, and the preceding section on unfairness. After consideration of the evidence and perspectives propounded by commenters, the 728 12 U.S.C. 5531(d). 541 Bureau generally adopts the evidentiary basis it had preliminarily set forth in the proposed rule as the basis for meeting this prong of the definition of abusiveness for purposes of the final rule. As stated in the proposal, the section on unfairness, Market Concerns—Underwriting, and the Section 1022(b)(2) Analysis, the Bureau has evidence showing that a significant proportion of consumers do not understand the kinds of harms that flow from unaffordable loans, including those imposed by default, delinquency, re-borrowing, and the collateral consequences of making unaffordable payments to attempt to avoid these other injuries. As noted above, the adverse effects for many consumers who find themselves caught up in extended loan sequences constitute severe harm, the likelihood of which is not understood by many consumers in advance. The Bureau thus concludes that a substantial population of borrowers lacks understanding of the material risks or costs of these loans. The Bureau does not dispute that many consumers may be knowledgeable about covered short-term loans and use them effectively, including making accurate predictions about their duration of borrowing. Yet for all the reasons discussed previously, the Bureau concludes that a significant population of consumers does not understand the material risks and costs of unaffordable loans that are made without reasonably assessing the borrower’s ability to repay the loan according to its terms. This does not mean that consumers are required to be experts in all aspects of how such loans function as a practical matter. But it does mean that if borrowers do not understand either their likelihood of being exposed to the risks of these loans or the severity of the kinds of costs and harms that may occur, then it is quite difficult to maintain the position that those same borrowers in fact understand the material risks and costs associated with unaffordable short-term loans. And the kinds of harms involved in the risks of default, 542 delinquency, re-borrowing, and the collateral consequences of making unaffordable payments to avoid these other injuries—including the interrelations among these injuries—can pose complex dynamics that are not likely to be well understood by many consumers. A number of commenters supported this view as well. Some noted that while some consumers might have a generalized understanding of how the debt associated with a covered loan can affect their economic circumstances, that understanding cannot be presumed to include an understanding of the broader risks and harms of such loans. These commenters also agreed with the Bureau that behavioral issues such as “tunneling” and “optimism bias” could have effects on decision-making that may affect consumers’ ability to use and manage covered loans successfully. Although some commenters criticized this approach as “novel” and relying too heavily on behavioral economics, the Bureau has no reason to believe that these theories and methodologies are particularly unconventional at this point of their development in the field of economics. Regardless, however, the Bureau concludes that these behavioral phenomena are equally consistent with economic analyses that would rest on models of rational behavior, given the particular circumstances of the consumers of these kinds of loans. The claim made here by industry commenters that payday loans have generated few consumer complaints, which mirrors the same claim made elsewhere by these commenters, is unpersuasive for reasons that have already been laid out in Market Concerns—Underwriting and the section on unfairness. When payday complaints are normalized, for example, in comparison to credit card complaints in view of the user population for each product, payday complaints 543 occurred more than twice as frequently.729 In any event, the volume of consumer complaints received by the Bureau is by no means an effective measure, by itself, to establish the presence or absence of consumer understanding. The Bureau believes there are a number of reasons why borrowers who find themselves in extended loan sequences do not submit a complaint to the Bureau about their negative experience with such loans. First, some borrowers may be embarrassed and thus less likely to submit complaints about their situation. Second, they may blame themselves for having gotten themselves caught up in a cycle of debt authorized by State law. Third, as some commenters indicated and the Bureau has observed around the country, faith leaders and faith groups may seem a more natural audience for some borrowers to appeal in relating their dissatisfactions with these experiences. The claim that abusiveness claims are “unavoidably situational,” and therefore the Bureau must make an individualized determination of abusiveness for each consumer, is unfounded. All decisions consumers make are individualized, but that fact does not preclude the Bureau from developing a general rule based on the statutory definitions of unfairness or abusiveness, as Congress clearly contemplated in section 1031(b) of the Dodd-Frank Act. It is true that the abusiveness standard is expressed in the statute in the singular. However, the Bureau also notes that it has the authority to declare “acts or practices” abusive, and it would be a reasonable interpretation of the statute to assume that Congress would not label abusive conduct aimed at a single consumer a “practice.” Further, it is true that each practice must be assessed based on the specific facts and circumstances before coming to an abusiveness 729 Bureau of Consumer Fin. Prot., Consumer Response Annual Report, Jan. 1 – Dec. 31, 2016, at 27, 33, (March 2017), available at https://www.consumerfinance.gov/documents/3368/201703_cfpb_Consumer-Response-AnnualReport-2016.PDF. 544 conclusion, yet the Bureau has done so here, and this does not mean it must assess the facts and circumstances as to each consumer. Comments suggesting that the Bureau did not prove borrowers were either infirm or ignorant are beside the point. The Bureau did not reach that conclusion, nor is it relevant under the terms of the statute applicable here. Rather, this prong of abusiveness only requires a lack of understanding. The final point raised by many industry and trade association commenters was that any lack of consumer understanding could be addressed by improved disclosures. They reinforced this point by asserting that the Bureau is obligated to seek reformed disclosures as a more modest intervention than requiring new underwriting criteria. These comments urging that the rule should mandate disclosures rather than adopt ability-to-repay requirements are addressed in more detail below in the section-by-section analysis of § 1041.5. For these reasons, the Bureau finds that many consumers lack an understanding of the material risks and costs associated with covered short-term loans made according to the identified practice of failing reasonably to assess the borrower’s ability to repay the loan according to its terms. Consumer inability to protect interests The Bureau’s Proposal Under section 1031(d)(2)(B) of the Dodd-Frank Act, an act or practice is abusive if it takes unreasonable advantage of the inability of the consumer to protect the interests of the 545 consumer in selecting or using a consumer financial product or service.730 As the Bureau stated in the proposal, consumers who lack an understanding of the material risks and costs of a consumer financial product or service often will also lack the ability to protect their interests in selecting or using that product. Nonetheless, if a consumer lacks understanding of the risks and costs of taking out such loans and yet could still find it easy to protect against them, then the consumer might be judged able to protect her interests. The Bureau also noted in the proposal that the structure of section 1031(d) is in the disjunctive, separately declaring it to be abusive to take unreasonable advantage either of consumers’ lack of understanding of material risks and costs or of their inability to protect their interests in using or selecting a product or service. As a matter of logic, then, Congress has determined that there could be situations where consumers do understand the material risks and costs of covered short-term loans yet are nonetheless unable to protect their interests in selecting or using these products. In particular, the Bureau stated in the proposal that consumers who take out covered short-term loans may be unable to protect their interests in selecting or using such loans, given their immediate need for credit and their inability in the moment to search out or develop alternatives that would enable them either to avoid the need to borrow or to borrow on terms within their ability to repay. As discussed in Market Concerns—Underwriting, consumers who take out these loans typically are financially vulnerable and have very limited access to other sources of credit. Their need is often acute. And consumers facing an immediate liquidity shortfall may believe that a covered loan is their only choice; a Pew study found that 37 percent of borrowers say they have been in such a difficult financial situation that they would take a 730 12 U.S.C. 5531(d)(2)(B). 546 payday loan on almost any terms offered.731 They may not have the time or resources to seek out, develop, or take advantage of alternatives. These factors may place them in such a vulnerable position when taking out these loans that they are unable to protect their interests. The Bureau also stated in the proposal that once consumers have commenced a loan sequence by taking out an unaffordable loan, they are likely to be unable to protect their interests in selecting or using subsequent loans. After they take out the initial loan, consumers are no longer able to protect their interests as a practical matter because they are already face to face with the competing injuries of default, delinquency, re-borrowing, or the collateral consequences of making unaffordable payments, with no other way to opt out of the situation. An unaffordable first loan can thus ensnare consumers in a cycle of debt from which they cannot extricate themselves without incurring some form of injury, rendering them unable to protect their interests in selecting or using these kinds of loans. Comments Received One commenter began by making a linguistic point that questioned whether the Bureau had conflated this prong of the abusive standard with the prior prong, suggesting that it was simply assuming that consumers taking out covered short-term loans inherently demonstrate an inability to protect their own interests, whereas many other consumers adequately protect their interests by deciding not to take out covered loans. More generally, commenters argued that lack of understanding is not enough to prove that a borrower has an inability to protect his interests. 731 Pew Charitable Trusts, How Borrowers Choose and Repay Payday Loans, at 20 (2013), http://www.pewtrusts.org/~/media/assets/2013/02/20/pew_choosing_borrowing_payday_feb2013-(1).pdf. It bears note that commenters correctly pointed out that the Bureau overstated the results of the Pew study by recounting a question as asking consumers whether they would take out a payday loan on “any terms,” rather than on “almost any terms.” Yet the Bureau does not find that this changes the fundamental point made in the Pew study. 547 Rather, these commenters asserted that the Bureau must show that it is actually impossible for consumers to protect their interests. In the same vein, an industry commenter argued that the Bureau’s claim in the proposal that consumers believe there are no better alternatives or that it would be too costly to shop for them fails to show inability to protect where such alternatives actually exist. Others repeated points they had made about the prior prong, observing that users of covered loans are not vulnerable or unsophisticated or irrational, but rather they do understand the terms and costs of those loans. One commenter analogized the language of this prong to the prohibition against unconscionable contracts in the Uniform Consumer Sales Practices Act, and asserted that the Bureau must therefore find consumers to be infirm, illiterate, or ignorant in order to satisfy this prong. Industry commenters also repeated their arguments that consumers tend to be accurate in their estimates of the duration of borrowing, and contended that re-borrowing is simply a preference for many consumers, rather than indicating an inability to protect their interests. These commenters also questioned the Pew study relied on by the Bureau, noting that the fact that 37 percent of short-term borrowers acknowledge they have been in an “immediate liquidity shortfall,” which they would pay off with payday loans on almost any terms offered, does not demonstrate consumers’ inability to protect their own interests. On the contrary, they argued that both competition and State laws protect consumers against problematic loan features and the study showed that the other 63 percent of consumers seek alternatives to covered loans when they perceived such loans to be harmful or problematic to them. 548 Commenters also asserted that no “seller behavior” occurs in making covered loans that deprives consumers of their ability to make informed decisions about their use of such loans. By contrast, consumer groups commented that covered loan borrowers are faced with an array of bad options, none of which provides them with the ability to protect their own interests. They described the significant difficulties that consumers regularly face when they are using covered short-term loans, which are traceable directly to the initial decision to take out loans that may prove to be unaffordable. And they urged that this consistent pattern is a reasonable demonstration of the proposition that a substantial portion of consumers using covered shortterm loans are unable to protect their own interests. The Final Rule After consideration of the comments received, the Bureau now concludes that when borrowers of covered loans are subjected to the identified lender practice of making such loans without reasonably assessing the borrower’s ability to repay, they are unable to protect their interests in selecting or using the loan product given the dynamics of this market and the structure and terms of these loans as described above and in Market Concerns—Underwriting. Once again, under section 1031(d)(2)(B) of the Dodd-Frank Act, an act or practice is abusive if it takes unreasonable advantage of the inability of the consumer to protect the interests of the consumer in selecting or using a consumer financial product or service.732 Consumers who lack an understanding of the material risks and costs of a consumer financial product or service often will be unable to protect their interests in selecting or using covered short-term loans because if they misunderstand the likelihood and extent of those material risks, they may 732 12 U.S.C. 5531(d)(2)(B). 549 not be aware that they should undertake efforts to protect their interests against those risks. And if they cannot reasonably estimate the nature and magnitude of the costs they could incur from unaffordable loans made in accordance with the identified practice, then they may not, as a practical matter, have the ability to protect their interests in the face of those material costs. To this extent, the provisions of section 1031(d)(2)(B) of the Dodd-Frank Act flow from the provisions of section 1031(d)(2)(A) on consumers who lack understanding, as noted in the proposal. But there are further reasons why consumers may be unable to protect their interests in using these loan products even if they largely understand the risks and costs involved. As discussed in the proposal and above in the section on unfairness, consumers who take out covered short-term loans may be unable to protect their interests in selecting or using such loans because many of them typically have an immediate need for credit and they cannot, in the moment, effectively identify or develop alternatives that would vitiate the need to borrow, allow them to borrow on terms within their ability to repay, or even allow them to borrow on terms not within their ability to repay but nonetheless on terms more favorable than those of a covered short-term loan. And as discussed in Market Concerns—Underwriting, many borrowers of these loans are financially vulnerable and have very limited access to other sources of credit. Confronted with an immediate liquidity problem, they may determine that a covered loan is the only option they have, as shown by the Pew study cited in the proposal, which found that 37 percent of borrowers say they have been in such a difficult financial situation that they would 550 take a payday loan on almost any terms offered.733 Because they find themselves in such vulnerable circumstances when they are deciding whether to take out an initial covered shortterm loan, they are unable, as a practical matter, to protect their interests. At this point, moreover, the dynamic changes even more dramatically, as described earlier in Market Concerns—Underwriting. Borrowers who take out an initial loan on unaffordable terms are generally unable to protect their interests in selecting or using further loans. After the first loan in a sequence has been consummated, the borrower is legally obligated to repay the debt. Consumers who lack the ability to repay that initial loan are faced with making a choice among competing injuries: default, delinquency, re-borrowing, or making unaffordable payments in an effort to avoid these other injuries while forgoing basic living expenses or major financial obligations in order to repay the loan. At this juncture, the consumer has no way out of the situation other than by deciding among competing harms. Having taken out the unaffordable first loan, borrowers generally will be not be able to protect their interests in selecting or using these kinds of loans. But the Bureau acknowledges that there are exceptions to this rule. For example, there may be consumers who encounter a windfall after taking out the loan but before repaying, such that none of the injuries occurs even though at the time the loan was originated the borrower would not have had an ability to repay. In addition, the set of problems faced by consumers who have already taken out an unaffordable loan can result in increased costs to consumers—often very high and unexpected 733 Pew Charitable Trusts, How Borrowers Choose and Repay Payday Loans, at 20 (2013), http://www.pewtrusts.org/~/media/assets/2013/02/20/pew_choosing_borrowing_payday_feb2013-(1).pdf. It bears note that commenters correctly pointed out that the Bureau overstated the results of the Pew study by recounting a question as asking consumers whether they would take out a payday loan on “any terms,” rather than on “almost any terms.” Yet the Bureau does not find that this changes the fundamental point made in the Pew study. 551 costs—that harm their interests. Sometimes these harms can occur in combination at different points in a single loan sequence, and the dynamics of how they interact with one another in their effects on the consumer can be complex. An unaffordable first loan can thus ensnare consumers in a cycle of debt with no reasonable means to extricate themselves without incurring further harm, rendering them unable to protect their interests in selecting or using these kinds of loans. The Bureau disagrees with the commenters who suggested that its determination that consumers taking out these loans are very often unable to protect their interests relied on the proposition that taking out such a loan is inherently demonstrative of an inability to protect oneself. Instead, the Bureau based its conclusions on the evidence that borrowers of these loans often have an urgent need and do not perceive any other options, especially once they have taken out an unaffordable loan and must confront the types of injury that they face when the next unaffordable payment comes due on their loan. A stark example of how consumers are unable to protect their interests by avoiding the injuries to which they are exposed by the identified practice is the substantial number of consumers who re-borrow—many of them repeatedly, and then eventually default—an outcome that is not in the interests of such consumers and thus one from which they would protect themselves if they were able. Other factors also hinder consumers in being able to protect their interests, such as the mismatch between how these loans are presented to consumers—as short-term, liquidityenhancing products that they can use to bridge an income shortfall until their next paycheck— and how they are actually designed and intended by lenders, as part of their business model, to function in long sequences of re-borrowing for a substantial population of consumers. Lenders offer a product whose term and balloon-payment structure, along with the common use of 552 leveraged payment mechanisms or vehicle security all tend to magnify the risks and harms to the borrower who fails to avoid the injuries that occur with extended loan sequences. Many consumers are unlikely to be able to protect their interests if they are extended an unaffordable loan and are rigidly confined within the limited options of repaying in full or re-borrowing, with no low-cost repayment or amortization options being extended. Consumers in this situation have the ability to make choices among the competing harms of default, delinquency, re-borrowing, or the collateral consequences of making unaffordable payments—though even the dynamics of these interrelated harms can become complex—but they are unable to protect their interests in avoiding those harms. The Bureau thus takes strong exception to the comment that re-borrowing is simply a preference for many consumers. If each loan in an extended loan sequence was itself an initial loan, such that it could be entered into simply with a view to the considerations moving the borrower to decide to take out a new credit obligation, then the comment would have more force. But a large volume of covered short-term loans is not at all of that kind: many of these loans are repeat re-borrowing that occurs in a setting where consumers generally face an unavoidable choice among different harms, including potentially severe harms from unaffordable loans and thus are unable to protect their interests. Therefore, the Bureau concludes that though borrowers of covered loans are not irrational and may generally understand their basic terms, these facts does not put borrowers in a position to protect their interests, given the nature of these loans if they are made on unaffordable terms. The Bureau again finds the comment that consumers accurately estimate their duration of borrowing to be a misleading account of the evidence it relies on here and elsewhere, which in 553 fact shows that consumers who are best able to predict accurately the duration of their borrowing are those who repay after little or no re-borrowing, and borrowers who end up in extended loan sequences are especially likely to err in estimating how long their loan sequences will last, though they are least able to protect their interests. Here as elsewhere, the key point is not that all consumers are unable to protect their interests, but that a substantial population of borrowers is unable to protect their interests in these circumstances. The Bureau does not agree that the language in the Dodd-Frank Act should be construed in light of the very different language of the Uniform Consumer Sales Practices Act, which one commenter urged should be interpreted as synonymous. The Dodd-Frank Act does not limit the instances in which a lender can take advantage of consumers’ inability to protect their interests to those where that inability is caused by infirmity, ignorance, illiteracy, or inability to understand the language of an agreement. Nor does the Bureau agree with commenters that asserted, in effect, that to satisfy the inability to protect condition, the Bureau must show there is no possible way for consumers to protect their interests. Rather, the Bureau reasonably interprets “inability to protect” in a practical manner under the circumstances. Thus, as the Bureau explained in the proposal and above, consumers who take out a covered short-term loan in the circumstance of their urgent need for funds, lack of awareness or availability of better alternatives, and no time to shop for such alternatives, are unable to protect their interests in selecting and using such a loan. The claim that no “seller behavior” occurs in making covered short-term loans that causes consumers to be unable to protect their interests is both incorrect and beside the point. First, it is incorrect because the identified practice of making these loans without reasonably 554 assessing the borrower’s ability to repay the loan according to its terms is itself seller behavior that causes some consumers—those who have been extended a loan—to be unable to protect their interests when the loan comes due and the consumer is unable to repay. Second, though seller behavior does bear on the “takes unreasonable advantage” prong of the definition and will be discussed further below, it has no relevance to the question of whether consumers lack the ability to protect their interests in the selection or use of the product. The Bureau does not find anything in the comments that undermines the soundness of the Pew study, which demonstrates that, by their own admission, consumers who take out these loans often find themselves in circumstances where they are not able to protect their interests. Moreover, the Bureau disagrees with the commenter that interpreted the negative answer to the survey question as meaning that 63 percent of respondents would seek alternatives to payday loans if the terms were perceived by them as harmful. This is pure speculation. One could likewise speculate that a negative response meant that the respondent would not seek an alternative loan and address their dire situation in some other manner. Moreover, there are many other reasons why a substantial majority of consumers may have opted not to utilize a covered loan, including that some do not need a loan at all. In contrast, there is only one plausible interpretation of an affirmative answer to the survey question, which is the one the Bureau has provided. The suggestion that consumers are adequately protected from the risks and consequences of covered short-term loans by industry competition and State laws is inaccurate in light of the data and analysis the Bureau has presented about the substantial risks and costs of these loans, which exist despite industry competition and the existing provisions of State laws. 555 Having considered the comments submitted, the Bureau has concluded that many consumers are unable to protect their interests in selecting or using covered short-term loans made in accordance with the identified practice of failing to make a reasonable assessment of the borrower’s ability to repay the loan according to its terms. Practice takes unreasonable advantage of consumer vulnerabilities The Bureau’s Proposal Under section 1031(d)(2) of the Dodd-Frank Act, a practice is abusive if it takes unreasonable advantage of any of several consumer vulnerabilities, including lack of understanding of the material risks, costs, or conditions of such loans or inability to protect their interests in selecting or using these loans.734 The Bureau stated in the proposal that the lender practice of making these loans without reasonably assessing that the consumer will have the ability to repay may take unreasonable advantage of both types of consumer vulnerabilities, though either would suffice to meet this prong of the abusiveness definition. The Bureau recognized that in any transaction involving a consumer financial product or service there is likely to be some information asymmetry between the consumer and the financial institution. Often the financial institution will have superior bargaining power as well. Section 1031(d) of the Dodd-Frank Act does not prohibit financial institutions from taking advantage of their superior knowledge or bargaining power to maximize their profit. Indeed, in a market economy, market participants with such advantages generally pursue their self-interests. However, section 1031 of the Dodd-Frank Act makes plain that there comes a point at which a 734 12 U.S.C. 5531(d)(2). 556 financial institution’s conduct in leveraging its superior information or bargaining power becomes unreasonable advantage-taking and thus is abusive.735 The Dodd-Frank Act delegates to the Bureau the responsibility for determining when that line has been crossed. Several interrelated considerations led the Bureau to believe that the practice of making payday, vehicle title, and other covered short-term loans without regard to the ability to repay may cross the line and take unreasonable advantage of consumers’ lack of understanding and inability to protect their interests. First, the Bureau noted in the proposal that the practice of making loans without regard to the consumer’s ability to repay the loan according to its terms stands in stark contrast to the practice of lenders in virtually every other credit market, and upends traditional notions of responsible lending enshrined in safety-and-soundness principles as well as in a number of other laws.736 The general principle of credit markets is that the interests of lenders and borrowers are closely aligned: lenders succeed (i.e., profit) only when consumers succeed (i.e., repay the loan according to its terms). For example, lenders in other markets, including other subprime lenders, typically do not make loans without first making an assessment that consumers have the capacity to repay the loan according to the loan terms. Indeed, “capacity” is one of the traditional three 735 A covered person taking unreasonable advantage of one or more of the three consumer vulnerabilities identified in section 1031(d) of the Dodd-Frank Act in circumstances in which the covered person lacks such superior knowledge or bargaining power may still be engaging in an abusive act or practice. 736 Dodd-Frank Act section 1411, codified at 15 U.S.C. 1639c(a)(1); CARD Act, 15 U.S.C. 1665e; HPML Rule, 73 FR 44522, 44543 (July 30, 2008); OCC Advisory Letter 2003-3, Avoiding Predatory and Abusive Lending Practices in Brokered and Purchased Loans (Feb. 21, 2003), available at http://www.occ.gov/static/news-issuances/memosadvisory-letters/2003/advisory-letter-2003-3.pdf; OCC, Guidance on Supervisory Concerns and Expectations Regarding Deposit Advance Products, 78 FR 70624 (Nov. 26, 2013); FDIC Guidance on Supervisory Concerns and Expectations Regarding Deposit Advance Products, 78 FR 70552 (Nov. 26, 2013). 557 “Cs” of lending and is often embodied in tests that look at debt as a proportion of the consumer’s income or at the consumer’s residual income after repaying the debt. In the markets for covered loans, however, lenders have built a business model that— unbeknownst to borrowers—depends on repeated re-borrowing, and thus on the consumer’s lack of capacity to repay such loans without needing to re-borrow. As explained in the proposal and in part II and Market Concerns—Underwriting above, the costs of maintaining business operations (which include customer acquisition costs and overhead expenses) often exceed the revenue that could be generated from making individual short-term loans that would be repaid without re-borrowing. Thus, in this market the business model of the lenders depends on a substantial percentage of consumers not being able to repay their loans when they come due and, instead, taking out multiple additional loans in quick succession. Indeed, upwards of half of all payday and single-payment vehicle title loans are made to—and an even higher percentage of revenue is derived from—borrowers in a sequence of 10 loans or more. This dependency on revenue from long-term cycles of debt has been acknowledged by industry stakeholders. For example, as noted in Market Concerns—Underwriting, an attorney for a national trade association representing storefront payday lenders asserted in a letter to the Bureau that “[i]n any large, mature payday loan portfolio, loans to repeat borrowers generally constitute between 70 and 90 percent of the portfolio, and for some lenders, even more.”737 Also relevant in assessing whether the practice identified here—of making covered shortterm loans without reasonably assessing the borrower’s ability to repay the loan according to its terms—involves unreasonable advantage-taking is the vulnerability of the consumers seeking 737 See Miller letter, cited in footnote 53, supra. 558 these types of loans. As discussed above in Market Concerns—Underwriting, payday and vehicle title borrowers—and by extension borrowers of similar covered short-term loans— generally have modest incomes, little or no savings, and have tried and failed to obtain other forms of credit. They generally turn to these products in times of need as a “last resort,” and when the loan comes due and threatens to take a large portion of their disposable income, their situation becomes, if anything, even more desperate. In addition, the evidence described above in Market Concerns—Underwriting suggests that lenders engage in practices that further exacerbate the risks and costs to the interests of consumers. In addition to the identified practice of making such loans without any underwriting to gauge their affordability, lenders rely on the term and balloon-payment structure of these loans to yield the intended result of extensive re-borrowing. Lenders market these loans as being for use “until next payday” or to “tide over” consumers until they receive income, thus encouraging overly optimistic thinking about how the consumer is likely to use the product. Lenders also make this re-borrowing option easy and salient to consumers in comparison to repayment of the full loan principal. Moreover, lenders typically limit the options available to borrowers by not offering or not encouraging borrowers to make use of alternatives that would reduce the outstanding principal over the course of a loan sequence, which would help consumers extricate themselves from the cycle of indebtedness more quickly and reduce their costs from reborrowing. Storefront lenders, in particular, encourage extended loan sequences by encouraging or requiring consumers to repay in person in an effort to frame the consumer’s experience in such a way to promote re-borrowing. Lenders often give financial incentives to employees to produce this outcome and thus reward them for maximizing loan volume. 559 Comments Received One trade association commented that lenders are allowed to take advantage of their superior knowledge and bargaining power and doing so is not contrary to law. In their view, the Bureau’s perspective that the re-borrowing model undergirding the market for covered loans stands in contrast to other markets is attributable to the restrictions imposed by State laws rather than by borrower needs and expectations. They also maintained that lenders have little incentive to take advantage of borrowers who they hope will return to them for subsequent loans after repaying those which are outstanding. By contrast, although consumer groups agreed with the general proposition that lenders can take advantage of superior knowledge and bargaining power, they emphasized that the proposed rule would prevent lenders from taking unreasonable advantage of consumers. They also noted that the financial vulnerability of many consumers who are likely to seek covered short-term loans is relevant to this inquiry. One commenter noted that a lender cannot take unreasonable advantage of a borrower through “acts of omission,” such as by failing to ask for pay stubs or other verification evidence or failing to check with consumer reporting agencies for information about the borrower’s credit history. Others asserted that an unreasonable advantage is not taken when lenders make loans to consumers with damaged credit or in need of cash, or advertise their loans as “quick” or “speedy” to cater to borrower needs, or offer terms that are readily and easily understood by borrowers. Some argued that the rule simply substitutes the Bureau’s judgment and risk tolerance for that of consumers. Still others argued that a lender cannot take unreasonable advantage of a consumer when the benefits of a loan exceed its costs. 560 The Final Rule The Bureau now concludes, after consideration of the comments received, that when lenders make covered short-term loans without reasonably assessing whether the borrower has the ability to repay the loan according to its terms, lenders take unreasonable advantage of consumers’ lack of understanding of the material risks, costs, and conditions of these loans, and also take advantage of their inability to protect their interests in selecting or using these loans. The Bureau does not dispute the proposition that lenders may take reasonable advantage of their superior knowledge and bargaining power. Nonetheless, in the proposal the Bureau preliminarily found that many lenders who make such loans have crossed the threshold to take impermissible and unreasonable advantage of those to whom they lend. The suggestion that these lenders have little incentive to take advantage of borrowers who are likely to be repeat customers is unfounded—there is an enormous difference between a scenario in which a borrower successfully repays a loan and later returns to apply for another loan (i.e., a true “repeat customer”), as compared to a scenario in which a borrower is forced to re-borrow again and again to cope with the problems posed by an unaffordable loan. Given that such a large majority of covered loans (over 80 percent) consist of loans procured through re-borrowing, and given that this is the core of the business model, it is evident that lenders have very significant incentives to take advantage of consumers’ lack of understanding of the material risks and their inability to protect themselves in the choice of the product. And once a consumer has taken a loan, lenders have at least equally significant incentives to take advantage of their inability to protect themselves with respect to the choice of the next loan in order to encourage re-borrowing. The factual background for the core elements of the Bureau’s conclusion that the “taking 561 unreasonable advantage” prong is met in these circumstances have been discussed at length in the section on unfairness and above in Market Concerns—Underwriting. For the sake of convenience, however, much of that analysis will be restated here. First, many consumers may not be able to protect their interests or to understand either the likelihood or the extent of the risks and costs of loans made in accordance with the identified practice of failing to make a reasonable assessment of the borrower’s ability to repay the loan according to its terms. In the face of these vulnerabilities, the general practice in this market is that lenders nonetheless make it their practice not to assess the borrower’s ability to repay. As a result, they typically have a significant volume of loans that are unaffordable from the outset in accordance with their terms. As discussed above in part II and in Market Concerns—Underwriting, this approach is in fact the core of the business model for most such lenders and reflects a deliberate decision on their part. Nothing in State or Federal law prohibits these lenders from engaging in meaningful underwriting on the loans they make. In this respect, the direction taken in this market is, in fact, out of step with traditional lender-borrower relationships in other loan markets, where the success of the lender is intertwined with the success of the borrower and determinations about loans that will be offered and accepted are preceded by underwriting assessments and determinations of this kind. Instead, the profitability of these lenders is built on, and depends upon, repeat re-borrowing by consumers. This model of lending premised on very minimal underwriting—often limited to screening only for potential fraud—is exacerbated by another common practice of these lenders once the initial loan, often unaffordable according to its terms, has been made. At this point, 562 these lenders typically provide the borrower with few or no repayment options other than either full repayment all at once or continued re-borrowing (which incurs another set of fees but provides no reduction of the loan principal). The array of repayment options provided in many other lending markets are virtually nonexistent here. Low-cost repayment or amortization options are typically not presented at all or are minimized or obscured in various ways. This again is a deliberate choice made by lenders in this market, not compelled by either State or Federal law. Indeed, the Bureau’s close experience over the past five years from exercising its supervision and enforcement authority over this market indicates that, even when such options are supposed to be afforded under provisions of some State laws, lenders often find ways to mask or obscure them or otherwise impede borrowers from availing themselves of them. Indeed, even consumers who are delinquent and have further demonstrated their inability to repay the loan according to its terms are encouraged to re-borrow, which leads many consumers to engage in extensive re-borrowing even where they eventually wind up in default. For many reborrowers, the upshot is that they end up making repeated payments that become increasingly unaffordable in the aggregate over time, even though a substantial number of them still will sustain the harms associated with default. The Bureau also has observed other lender conduct that greatly increases the risks and harms to consumers in these circumstances. Covered short-term loans, in particular, involve a basic mismatch between how they appear to function as short-term credit and how they are actually designed and intended by lenders, as part of their business model, to function in long sequences of re-borrowing for a substantial population of consumers. Lenders present these loans as short-term, liquidity-enhancing products that consumers can use to bridge an income 563 shortfall until their next paycheck. But in practice, across the universe of borrowers, these loans often do not operate that way. Lenders have designed the term of the loan, its balloon-payment structure, and the common use of leveraged payment mechanisms, including vehicle security, so as to magnify the risks and harms to the borrower. The disparity between how these loans appear to function and how they actually function increases the difficulties that consumers experience with these loans. Once consumers have taken out a loan, they have no practical means to avoid the injuries that will occur if the loan proves to be unaffordable. Consumers who obtain a covered shortterm loan that is beyond their ability to repay confront the harms of default, delinquency, reborrowing, or the collateral consequences of making unaffordable payments that would cause them to forgo basic living expenses or major financial obligations. They can make choices among these competing harms but not avoid them. And as discussed above in Market Concerns—Underwriting, and below in Market Concerns—Payments, lenders engage in other practices that further increase the likelihood and degree of harm, for instance by encouraging additional re-borrowing and its attendant costs even for consumers who are already experiencing substantial difficulties as they are mired in extended loan sequences, and by engaging in payment collection practices that are likely to cause consumers to incur substantial additional fees beyond what they already owe on the terms of the existing loan. Further adverse effects can include expensive bank fees, the potential loss of their bank account, aggressive debt collection efforts, and, with title loans, the risks and costs of having their vehicle repossessed, causing them to lose their transportation to work or conduct their ordinary personal affairs. 564 As discussed earlier, this practice of making loans without regard to the consumer’s ability to repay contrasts sharply with the regular practice of lenders in virtually every other credit market, and upends traditional notions of responsible lending enshrined in safety-andsoundness principles as well as in a number of other laws.738 The general principle of credit markets is that the interests of lenders and borrowers are aligned and lenders benefit only when their customers are successful in repaying their loans in accordance with the terms. For this reason, lenders in other markets, including other subprime lenders, typically do not make loans without first making an assessment that consumers have the capacity to repay the loan according to the loan terms. Yet the set of effects found in the market for covered short-term loans has the cycle of indebtedness at its core, as intended and effectuated by lenders in this market. And it affects a large segment of borrowers: as described above in Market Concerns—Underwriting, half of all storefront payday loan sequences contain at least four loans.739 One-third contain seven loans or more, by which point consumers will have paid charges equal to 100 percent of the original amount borrowed and still owe the full amount of the principal.740 Almost one-quarter of loan sequences contain at least 10 loans in a row, and looking just at loans made to borrowers who are paid weekly, biweekly, or semi-monthly, more than one-fifth (21 percent) of those loans are in 738 Dodd-Frank Act section 1411, codified at 15 U.S.C. 1639c(a)(1); CARD Act, 15 U.S.C. 1665e; HPML Rule, 73 FR 44522, 44543 (July 30, 2008); OCC Advisory Letter 2003-3, Avoiding Predatory and Abusive Lending Practices in Brokered and Purchased Loans (Feb. 21, 2003), available at http://www.occ.gov/static/news-issuances/memosadvisory-letters/2003/advisory-letter-2003-3.pdf; OCC, Guidance on Supervisory Concerns and Expectations Regarding Deposit Advance Products, 78 FR 70624 (Nov. 26, 2013); FDIC Guidance on Supervisory Concerns and Expectations Regarding Deposit Advance Products, 78 FR 70552 (Nov. 26, 2013). 739 CFPB Report on Supplemental Findings. 740 Id. 565 sequences consisting of at least 20 loans.741 For loans made to borrowers who are paid monthly, almost half (46 percent) of the loans are in sequences consisting of at least 10 loans.742 The figures for title loans are similar, and also are premised on a business model built around repeated re-borrowing: 56 percent of vehicle title loan sequences consist of more than three loans, 36 percent consist of at least seven loans, and almost one quarter—23 percent—consist of more than 10 loans.743 Regardless of what the outer bounds of “taking unreasonable advantage” may be, the Bureau concludes that the ways lenders have structured their lending practices here fall well within any reasonable definition of that concept. Here the identified practice of making loans without reasonably assessing the borrower’s ability to repay the loan according to its terms leads to unaffordable loans and all the harms that follow upon them. At a minimum, lenders take unreasonable advantage of borrowers when they develop lending practices that are atypical in the broader consumer financial marketplace, take advantage of particular consumer vulnerabilities, rely on a business model that is directly inconsistent with the manner in which the product is marketed to consumers, and eliminate or sharply limit feasible conditions on the offering of the product (such as underwriting and amortization, for example) that would reduce or mitigate harm for a substantial population of consumers. The Bureau now affirms that lenders take such unreasonable advantage in circumstances where they make covered short-term loans or covered longer-term balloon-payment loans without reasonably assessing the consumer’s ability to repay the loan according to its terms. 741 Id. Id. 743 Id. 742 566 The Bureau does not disagree with the commenters who noted that lenders do not take unreasonable advantage of consumers when they make loans to consumers with damaged credit or in need of cash, or they advertise their loans as quick or speedy to cater to borrower needs, or they offer terms that are readily and easily understood by borrowers. Neither in isolation nor taken together do these particular acts or practices constitute abusive behavior. The Bureau concludes instead that, by engaging in the identified practice, lenders take unreasonable advantage of consumer vulnerabilities. Moreover, the rule does not substitute the Bureau’s judgment and risk tolerance for those of consumers. Instead, it simply seeks to assure that lenders do not take unreasonable advantage of consumers’ lack of understanding or inability to protect their interests through use of the identified practice. Even well-educated and sophisticated consumers can lack understanding of a loan product whose structural effects are complex and opaque, leading many of them to the negative consequences that flow from an extended cycle of indebtedness. The Bureau disagrees with the commenters who noted that a lender cannot take unreasonable advantage of a borrower by failing to underwrite appropriately, such as by failing to ask for pay stubs or other verification evidence or failing to check with consumer reporting agencies for information about the borrower’s credit history. The thrust of these comments is that the lender cannot “take unreasonable advantage” by seeking to reduce burdens and make life easier for consumers and, in particular, cannot do so by “acts of omission.” On the contrary, the Bureau has shown that lenders utilize these and related practices to position a substantial population of borrowers to take out unaffordable loans that lead directly to debt cycles of longterm re-borrowing. And as the law has long recognized in various contexts, there is no material 567 distinction to be made between acts of omission and acts of commission, particularly here where these aspects of the identified practice take unreasonable advantage of consumer vulnerabilities. With respect to the comments that a lender cannot take unreasonable advantage of a consumer when the benefits of a loan exceed its costs, as stated above in the unfairness section, the Bureau has concluded that the countervailing benefits of the identified practice, rather than of the product itself, do not outweigh the substantial injury. In determining whether the lender takes unreasonable advantage, the Bureau’s focus is not on the variable experiences of the entire heterogeneous borrower universe, but rather on the adverse effects that the identified practice has on a substantial population of consumers where lenders are taking unreasonable advantage of their vulnerabilities by making unaffordable loans to them. Thus, for the sake of argument, even if it were true that a practice that is net beneficial for consumers cannot be found to take unreasonable advantage, that would not stand as an impediment to finding the practice at issue here to be abusive. Further, nothing in the final rule prevents any lender from offering loans whose benefits exceed their costs, regardless of the specific population for which that judgment is being made, as long as the lender does not engage in the identified practice of failing to make a reasonable assessment of ability to repay when making such loans. In sum, the Bureau concludes that where a borrower lacks understanding of the material risks and costs of covered short-term loans, or where the borrower lacks an ability to protect his own interests by using or selecting these loans, the lender takes unreasonable advantage of these consumer vulnerabilities by making a covered short-term loan without reasonably assessing the borrower’s ability to repay the loan according to its terms, where the natural result of that practice is that a substantial number of consumers will be caught up in extended loan sequences, 568 with the adverse consequences that have been amply canvassed above and in Market Concerns— Underwriting. The Bureau does not take issue with the comment that it should take into consideration the array of State laws governing covered short-term loans. The Bureau has carefully considered the effects of those laws and concludes that the laws in those States that authorize such loans do not adequately protect consumers, because the negative effects for consumers that are described at length in Market Concerns—Underwriting continue to exist despite those State laws. Having considered the comments submitted, the Bureau has concluded that there is substantial evidence and a sufficient basis to determine that the identified practice of making covered short-term and longer-term balloon-payment loans, without reasonably assessing the borrower’s ability to repay the loan according to its terms, takes unreasonable advantage either of the borrower’s lack of understanding of the material risks and costs of these loans or of the borrower’s inability to protect his own interests by using or selecting these loans. Section 1041.5 Ability-to-Repay Determination Required General Approach in Proposed Rule As discussed in the section-by-section analysis of § 1041.4 above, the Bureau tentatively concluded in the proposed rule that it is an unfair and abusive act or practice to make a covered short-term loan without reasonably determining that the consumer will have the ability to repay the loan. Section 1031(b) of the Dodd-Frank Act provides that the Bureau’s rules may include requirements for the purpose of preventing unfair or abusive acts or practices. The Bureau thus proposed to prevent the abusive and unfair practice by including in proposed §§ 1041.5 and 569 1041.6 certain minimum requirements for how a lender may reasonably determine that a consumer has the ability to repay a covered short-term loan. Proposed § 1041.5 set forth the prohibition against making a covered short-term loan (other than a loan that satisfies the protective conditions in proposed § 1041.7) without first making a reasonable determination that the consumer will have the ability to repay the covered short term loan. It also, in combination with proposed § 1041.6, specified the minimum elements of a baseline methodology that would be required for determining a consumer’s ability to repay, using a residual-income analysis and an assessment of the consumer’s prior borrowing history. In particular, proposed § 1041.6 would have required that a presumption of unaffordability applied if a consumer sought a new covered short-term loan within 30 days of a prior outstanding covered short-term loan, and applied a mandatory 30-day cooling-off period after the third such loan in a sequence. The Bureau proposed similar ability-to-repay requirements for covered longer-term loans, including covered longer-term balloon-payment loans, in proposed §§ 1041.9 and 1041.10. Given the parallel nature of proposed §§ 1041.5 and 1041.6 for covered short-term loans and proposed §§ 1041.9 and 1041.10 for covered longer-term loans, the Bureau will generally refer just to proposed §§ 1041.5 and 1041.6 to describe the proposed ability-to-repay framework, but will note where proposed §§ 1041.9 and 1041.10 differed from the framework for covered shortterm loans. The baseline methodology in proposed § 1041.5 rested on a residual-income analysis— that is, an analysis of whether, given the consumer’s projected income and major financial obligations, the consumer will have sufficient remaining (i.e., residual) income to cover the 570 payments on the proposed loan and still meet basic living expenses. The proposal also would have required lenders to track the timing of inflows and outflows of funds to determine whether there would be periods of shortfall that might prompt consumers to re-borrow soon after a previous covered short-term loan. In the proposal, the Bureau recognized that, in other markets and under other regulatory regimes, financial capacity is more typically measured by establishing a maximum debt-to-income (DTI) ratio.744 DTI tests generally rest on the assumption that as long as a consumer’s debt burden does not exceed a certain threshold percentage of the consumer’s income, the remaining share of income will be sufficient for a consumer to be able to meet non-debt obligations and other expenses. By its nature, DTI must be calculated by dividing total income and total expenses for the relevant time period, and does not permit the tracking of a consumer’s individual income inflows and major financial obligation outflows on a continuous basis over a period of time. For low- and moderate-income consumers, the Bureau expressed concern in the proposal that a DTI ratio would not be sufficiently sensitive to determine re-borrowing risk in the markets for covered loans. In particular, the Bureau noted that a DTI ratio that might seem quite reasonable for the “average” consumer could be quite unmanageable for a consumer at the lower end of the income spectrum and the higher end of the debt burden range.745 Ultimately, the 744 The Bureau noted in the proposal that, for example, DTI is an important component of the Bureau’s ability-torepay rule for mortgages in 12 CFR 1026.43. It is a factor that a creditor must consider in determining a consumer’s ability to repay and also is a component of the standards that a residential mortgage loan must meet to be a qualified mortgage under that rule. 745 The Bureau stated in the proposal that, for example, under the Bureau’s ability-to-repay requirements for residential mortgage loans, a qualified mortgage results in a DTI ratio of 43 percent or less. But for a consumer with a DTI ratio of 43 percent and low income, the 57 percent of income not consumed by payments under debt obligations is unlikely to indicate the same capacity to handle a new loan payment of a given dollar amount, compared to consumers with the same DTI and higher income. The Bureau further stated in the proposal that this is 571 Bureau posited in the proposal, whether a particular loan is affordable will depend upon how much money the consumer will have left after paying existing obligations and whether that amount is sufficient to cover the proposed new obligation while still meeting basic living expenses. The Bureau additionally stated in the proposal that, in contrast with other markets in which there are long-established norms for DTI levels that are consistent with sustainable indebtedness, the Bureau did not believe that there existed analogous norms for sustainable DTI levels for consumers taking covered short-term loans. The Bureau stated in the proposal that it thus believed that residual income was a more direct test of ability to repay than DTI and a more appropriate test with respect to the types of products covered in this rulemaking and the types of consumers to whom these loans are made. The Bureau emphasized in the proposal that it had attempted to design the residual income methodology specified in proposed §§ 1041.5 and 1041.6 to ensure that ability-to-repay determinations can be made through scalable underwriting models. While it was proposing that the most critical inputs into the determination rest on documentation, the Bureau noted that its proposed methodology would allow for various means of documenting major financial obligations and also permit alternatives to documentation where appropriate. The Bureau recognized in particular that rent often cannot be readily documented and therefore would have allowed for estimation of rental expense based on the housing expenses of consumers with households in the locality of the consumer. The Bureau’s proposed methodology also would not especially true if the low-income consumer also faces significant non-debt expenses, such as high rent payments, that may consume significant portions of the remaining 57 percent of her income. 572 have mandated verification or detailed analysis of consumers’ expenditures for basic living expenses. The Bureau stated in the proposal that it believed that such detailed analysis may not be the only method to prevent unaffordable loans and was concerned that it would substantially increase costs to lenders and consumers. Finally, the Bureau emphasized that the proposed methodology would not dictate a formulaic answer to whether, in a particular case, a consumer’s residual income is sufficient to make a particular loan affordable. For instance, the Bureau did not propose a specific minimum dollar threshold for adequate residual income. Instead, the proposed methodology would have allowed lenders to exercise discretion in arriving at a reasonable determination with respect to that question. Proposed § 1041.5 outlined the methodology for assessing the consumer’s residual income as part of the assessment of ability to repay. Proposed § 1041.5(a) set forth definitions used throughout proposed §§ 1041.5 and 1041.6. Proposed § 1041.5(b) set forth the proposed requirement for a lender to determine that a consumer will have the ability to repay a covered short-term loan and set forth minimum standards for a reasonable determination that a consumer will have the ability to repay such a covered loan. In the standards in proposed § 1041.5(b), the Bureau generally proposed to require a lender to determine that the consumer’s income will be sufficient for the consumer to make payments under a covered short-term loan while accounting for the consumer’s payments for basic living expenses and major financial obligations. 573 Proposed § 1041.5(c) set forth standards for verification and projections of a consumer’s income and major financial obligations on which the lender would be required to base its determination under proposed § 1041.5. Proposed § 1041.6 would have augmented the basic ability-to-repay determination required by proposed § 1041.5 in circumstances in which the consumer’s recent borrowing history or current difficulty in repaying an outstanding loan provides important evidence with respect to the consumer’s financial capacity to afford a new covered short-term loan. For example, proposed § 1041.6 would have imposed a presumption of unaffordability in various circumstances suggesting that a consumer lacked the ability to repay a current or recent loan, so that a lender would have been permitted to extend a new covered short-term loan under proposed § 1041.5 only if there was particular evidence of a sufficient improvement in financial capacity. In addition, where a consumer took out a sequence of three covered short-term loans, each within 30 days of the prior outstanding loan, proposed § 1041.6 would have imposed a mandatory 30day cooling-off period. The Bureau believed that these requirements would help consumers to avoid getting stuck in long cycles of debt. See section-by-section analysis for § 1041.5(d), below, for further discussion of proposed § 1041.6. The Bureau explained in the proposal that as an alternative to the proposed ability-torepay requirement, it had considered whether lenders should be required to provide disclosures to consumers warning them of the costs and risks of re-borrowing, default, and collateral harms from unaffordable payments associated with taking out covered short-term loans. However, the Bureau stated in the proposal that it believed that such a disclosure remedy would be significantly less effective in preventing the identified consumer harms, for three reasons. First, 574 the Bureau stated that disclosures would not address the underlying incentives in the market for lenders to encourage consumers to re-borrow and take out long sequences of loans. As discussed in the proposal’s section on Market Concerns—Short-Term Loans, the prevailing business model involves lenders deriving a very high percentage of their revenues from extended loan sequences. The Bureau stated in the proposal that while enhanced disclosures would provide additional information to consumers, the loans would remain unaffordable for consumers, lenders would have no greater incentive to underwrite more rigorously, and lenders would remain dependent for revenue on extended loan sequences of repeat re-borrowing by many consumers. Second, the Bureau stated in the proposal that empirical evidence had led it to believe that disclosures would have only modest impacts on consumer borrowing patterns for short-term loans generally and negligible impacts on whether consumers re-borrow. In the proposal, the Bureau discussed evidence from a field trial of several disclosures designed specifically to warn of the risks of re-borrowing and the costs of re-borrowing that showed that these disclosures had a marginal effect on the total volume of payday borrowing.746 Further, the Bureau discussed in the proposal its analysis of the impact of a change in Texas law (effective January 1, 2012) requiring payday lenders and short-term vehicle title lenders to provide a new disclosure to prospective consumers before each payday loan transaction.747 The Bureau observed in the proposal that, using the Bureau’s supervisory data, it had found that, with respect to payday loan transactions, there was an overall 13 percent decline in loan volume in Texas after the disclosure requirement went into effect, relative to the loan volume changes for the study period in 746 Marianne Bertrand and Adair Morse, “Information Disclosure, Cognitive Biases and Payday Borrowing,” 66 J. of Fin. 1865, at 1866 (2011). 747 See CFPB Report on Supplemental Findings, at Chapter 3. 575 comparison States.748 As discussed in the proposal, the Bureau noted that its analysis of the impacts of the Texas disclosures also showed that the probability of re-borrowing on a payday loan only declined by approximately 2 percent once the disclosure was put in place.749 The Bureau stated in the proposal that this finding indicates that high levels of reborrowing and long sequences of payday loans remain a significant source of consumer harm even with a disclosure regime in place.750 Further, the Bureau stated in the proposal that, as discussed in the proposal’s section on Market Concerns—Short-Term Loans, the Bureau has observed that consumers have a very high probability of winding up in very long loan sequences once they have taken out only a few loans in a row. The Bureau stated in the proposal that the extremely high likelihood that a consumer will wind up in a long-term debt cycle after taking out only a few loans contrasts sharply with the nearly negligible impact on consumer re-borrowing patterns of a required disclosure, which the Bureau viewed as providing further evidence that disclosures tend to be ineffective in addressing what the Bureau considered to be the core harms to consumers in this credit market. Third, the Bureau stated in the proposal that it believed that behavioral factors made it more likely that disclosures to consumers taking out covered short-term loans would be ineffective in warning consumers of the risks and preventing the harms that the Bureau sought to address with the proposal. The Bureau stated in the proposal that due to general optimism bias and the potential for tunneling in their decision-making, as discussed in more detail in the 748 See CFPB Report on Supplemental Findings, at 73. See CFPB Report on Supplemental Findings, 78-79. 750 The Bureau stated in the proposal that the empirical data suggests that the modest loan volume reductions are primarily attributable to reductions in originations; once a consumer has taken out the initial loan, the disclosure has very little impact on re-borrowing. 749 576 proposal’s section on Market Concerns—Short-Term Loans, consumers are likely to dismiss warnings of possible negative outcomes as not applying to them, and not to focus on disclosures of the possible harms associated with outcomes—re-borrowing and default—that they do not anticipate experiencing themselves. The Bureau stated in the proposal that to the extent consumers have thought about the likelihood that they themselves will re-borrow or default (or both) on a loan, a general warning about how often people re-borrow or default (or both) is unlikely to cause them to modify their approach by revising their own expectations about what the chances are that they themselves will re-borrow or default (or both). Legal Authority As noted above in the section-by-section analysis for § 1041.4, the Bureau has authority to prescribe rules applicable to a covered person or service provider identifying as unlawful unfair, deceptive, or abusive acts or practices in connection with any transaction with a consumer for a consumer financial product or service, or the offering of a consumer financial product or service.751 The Bureau has done so in § 1041.4. Additionally, the Bureau may include in such rules requirements for the purpose of preventing such acts or practices.752 It is based on that authority that the Bureau issues § 1041.5. A number of commenters, including several industry trade associations and lenders, challenged the Bureau’s authority to enact a prescriptive ability-to-repay requirement because Congress did not specifically authorize such a requirement with respect to payday loans and other loans the Bureau proposed to cover, in contrast to the mortgage and credit card markets. 751 752 12 U.S.C. 5531(b). Id. 577 Consumer advocates and some other commenters, however, argued that the Bureau had ample authority to impose the proposed ability-to-repay requirement under the UDAAP authority granted to the Bureau under the Dodd-Frank Act. These comments are addressed in the sectionby-section analysis for § 1041.4, above (“Identification of Unfair and Abusive Practice— Covered Loans”). More generally, the Bureau received a number of comments asserting that its proposed rule had exceeded its authority to prevent the unfair and abusive practice identified in § 1041.4, by prescribing more detailed underwriting requirements than would be required to avoid engaging in the identified unfair or abusive practice. By its terms, section 1031 of the Dodd-Frank Act authorizes the Bureau not only to “prescribe rules applicable to a covered person or service provider identifying as unlawful unfair, deceptive or abusive acts of practices” but also provides that “Rules under this section may include requirements for the purpose of preventing such acts or practices.” This latter phrase would be surplusage if the Bureau’s rulemaking authority were as circumscribed as these commenters suggest. Furthermore, as discussed above in part IV, courts have long held that rulemakings to remedy and prevent unfair acts and practices may include preventative requirements so long as those requirements have a “reasonable relation to the unlawful practices found to exist.”753 The Bureau believes that the final underwriting requirements as set forth in § 1041.5 are reasonably related to, and crafted adequately to prevent, the abusive and unfair practice identified in § 1041.4. The unfair and abusive practice is making covered short-term and longer-term balloon-payment loans without reasonably determining that consumers will have 753 AFSA, 767 F.2d at 988. 578 an ability to repay the loans according to their terms. Section 1041.5 sets forth a balanced approach, providing flexibility in some areas and bright-line guidance in others, that is aimed at ensuring that lenders account for net income, major financial obligations, and basic living expenses, and make a reasonable determination about whether a consumer will be able to repay the loan according to its terms, using those variables in a residual income or debt-to-income ratio calculation. And other provisions in § 1041.5, such as the cooling-off periods in paragraph (d), are likewise reasonably related to the identified practice in that they temporarily prohibit continued lending to consumers who have already received a sequence of three covered shortterm loans or covered longer-term balloon-payment loans in quick succession, to both protect them from further unaffordable loans and potentially enable them to escape from a cycle of indebtedness. General Comments Received In this general section, before describing the details of proposed § 1041.5, comments, and changes in the final rule on specific subparts of § 1041.5 below, the Bureau is addressing comments about the Bureau’s general proposed approach, including the overall burden of the proposed ability-to-repay requirements and general methodology proposed, the specificity of the rule, the comparison of the proposed approach to underwriting in other markets, the predictiveness of residual income methodologies, the decision not to adopt a disclosure-only remedy to the identified unfair and abusive practice, the decision not to permit a payment-toincome underwriting model and other alternatives suggested by commenters, and assertions that the rule will conflict with the interests of fair lending law. 579 The Bureau received a significant number of comments from a variety of stakeholders, including lenders of different types and sizes, industry trade associations, some service providers, some State and local elected officials, the SBA Office of Advocacy, a joint letter from five Members of Congress,754 and others asserting that the Bureau’s proposed ability-to-repay regime would, in the aggregate, be too burdensome, rigid, and complicated. One commenter stated that one of the chief virtues of payday and other covered loans is their lack of underwriting, and if underwriting were required, it is unlikely that businesses would make nearly as many covered short-term loans. Many commenters believed that the burden would be so high that it would significantly reduce access to credit, including even to consumers who do have the ability to repay. One commenter stated that some in the industry have estimated an increase in cost for each loan of about $30, and several commenters asserted that lenders would need to increase prices to cover the additional costs. Others argued that while the more burdensome underwriting requirements proposed in the rule may be common for banks making other types of loans; they would be new and quite difficult for non-bank lenders to implement. Relatedly, some commenters noted that the small balances of covered loans, particularly covered short-term loans which often are $500 or less, might not allow lenders to offset the additional costs required to comply with the underwriting requirements. Some commenters suggested that only large lenders would be able to survive the additional compliance cost. Several commenters, including a SER and five Members of Congress, cited a presentation by representatives of four specialty 754 In their letter, the Members made several critiques of the proposed ability-to-repay requirements along the lines of those made by other commenters as discussed below—that the proposed requirements would have been too complex, burdensome, and prescriptive; that they did not align with the underwriting rules in other credit markets; and that they would potentially constrict access to credit. However, unlike many of the other commenters who made similar arguments, the Members expressed general support for the proposal and expressed particular appreciation for the Bureau’s approach to addressing long-term re-borrowing. 580 consumer reporting agencies which appeared to suggest that the proposed ability-to-repay requirements would disqualify any consumer who earned under $40,000 per year, asserting that would effectively result in denial of credit access to 140 million Americans.755 Some commenters also suggested that the burdensome and complex underwriting requirements would significantly increase the time needed to underwrite a loan, and did not agree with the Bureau that lenders would be able to automate sufficiently to keep origination times short. The Bureau received a number of estimates on the time it would take to originate a loan. For example, one commenter asserted that it would take more than 10 minutes. Another said it would take 15-20 minutes to originate a loan manually. One estimated that it would increase transaction time by 15-45 minutes, while another said it would increase the time by 6-25 minutes. Another commenter wrote that origination already takes 20 minutes, and the new documentation requirements would add to that timing. And one trade association asserted that it would take three hours. Many of these commenters specifically focused on the Bureau’s proposal to require a residual income underwriting requirement, which they argued was overly burdensome and prescriptive. Commenters argued that prescribing such an underwriting methodology would be a novel approach that is not common in other credit markets, and would be inconsistent with the general merits of preserving flexibility in underwriting models. Several commenters cited the preamble discussion to the Bureau’s final ability-to-repay rule for mortgages as evidence of its 755 A comment letter by a SER attached the presentation from the specialty consumer reporting agency officials. The Bureau did not receive a copy of this presentation directly from the specialty consumer reporting agencies, three of whom submitted individual comment letters. Nor did any of them make the specific negative claims about the impacts of the proposal as had been made in the slides, although one indirectly alluded to similar statistics cited in the presentation. The presentation is undated, although it appears from the context to have been developed during the comment period. 581 novelty as an underwriting methodology.756 Several commenters asserted that the proposed residual income methodology would not prevent the default and re-borrowing injuries identified in the Bureau’s analysis, relying on studies that the commenters believed showed that residual income is not predictive of such outcomes. Commenters also stated that they believed that the proposed underwriting requirements were not specific enough with regard to such issues as estimates for basic living expenses, the general reasonableness standard for lenders’ ability-to-repay analyses, the lack of a numeric threshold or other guidance for what constitutes sufficient residual income, and what kinds of loan performance patterns would be evidence that a lender’s ability-to-repay analysis was inadequate. These commenters recognized that the Bureau had attempted to leave some amount of flexibility and discretion to lenders, but argued that more clarity was needed to reduce compliance risk associated with choices made in the “grey area.” One commenter noted that the underwriting model for mortgage loans from the U.S. Department of Veterans Affairs involves a more prescriptive methodology based on residual income that sets forth precise dollar figures for required residual income based on various variables,757 and that if a residual income approach was going to be adopted, the commenter believed this was a more workable model. Relatedly, a number of commenters, including several lenders and industry trade associations, suggested the Bureau permit use of a debt-to-income ratio as an alternative to residual income, citing the Bureau’s mortgage and credit card regulations (12 CFR 1026.43 and 756 Commenters cited a passage of the preamble from the mortgage ability-to-repay rule where the Bureau wrote that, “Except for one small creditor and the [U.S. Department of Veterans Affairs], the Bureau is not aware of any creditors that routinely use residual income for underwriting, other than as a compensating factor.” 78 FR 6407, 6486 (Jan. 30, 2013). 757 38 CFR 36.4340. 582 12 CFR 1026.51, respectively) as precedent for that approach. They also discussed how the DTI ratio is a more familiar and time-tested concept for lenders across other credit markets. Some of these commenters argued that the Bureau should permit, instead of require, a residual income underwriting model, and also allow lenders to use a more traditional method premised on a DTI ratio. A number of commenters, including several lenders and industry trade associations, argued that the proposed rule set forth ability-to-repay requirements that were more rigorous and burdensome than that set forth in the Bureau’s ability-to-pay rules for credit cards (12 CFR 1026.51) and ability-to-repay rules for mortgages (12 CFR 1026.43), and asserted that the inconsistency was unwarranted. The Bureau’s regulations under the CARD Act generally require underwriting that considers the consumer’s ability to make the required minimum periodic payments under the terms of the account based on the consumer’s income or assets and the consumer’s current obligations; provides that card issuers must establish and maintain reasonable written policies and procedures to consider the consumer’s ability to make the required minimum payments; and provides that reasonable policies and procedures include consideration of at least one of the following: the ratio of debt obligations to income; the ratio of debt obligations to assets; or the income the consumer will have after paying debt obligations.758 The Bureau’s regulation on mortgage underwriting requires that a lender of covered transactions must make a reasonable and good faith determination at or before consummation that the consumer will have a reasonable ability to repay the loan according to its terms, and allows lenders to use either the consumer’s monthly debt-to-income ratio or residual income in making 758 12 CFR 1026.51(a)(1). 583 that determination.759 These commenters argued that the Bureau’s underwriting regulations for these other markets were more flexible than the regulation proposed here. Some commenters believed it was illogical and unjustified to impose more prescriptive and restrictive underwriting and verification requirements for small-dollar loans when the Bureau imposes, in their view, less prescriptive and restrictive underwriting and verification requirements for other loans of much larger size (e.g., mortgages). Several commenters noted that the proposal would require a determination of the consumer’s ability to repay the entire principal amount while the credit card rules require a determination regarding the consumer’s ability to make minimum payments, stating or implying that this was a difference in legal standards for ability to repay and questioning the basis for it; one commenter suggested the Bureau was imposing a different standard because it did not “trust” consumers in this market to make decisions for themselves. On a similar note, some commenters stated that the underwriting requirements would be greater than those in the student loan and automobile loan (for purchase money) markets. Other commenters, including consumer advocates and at least some industry stakeholders (including several installment lenders), generally supported the underlying principle of the rule requiring lenders to make a reasonable determination that consumers have an ability to repay, noting that it is a fundamental, common-sense tenet of responsible lending in most loan markets. These commenters noted the precedent in the Bureau’s regulations relating to mortgages and credit cards, as well as the other Federal precedent noted above in Market Concerns— Underwriting. Some consumer groups agreed that an underwriting methodology based on residual income was the most appropriate underwriting model for determining whether 759 12 CFR 1026.43(c). 584 consumers have an ability to repay and asserted that alternative approaches were too permissive. Consumer advocates writing jointly suggested a number of specific changes to the proposal which in their view would strengthen elements of the ability-to-repay requirement, which are described in more detail below. Some commenters argued that the Bureau should allow an approach that would permit lenders to lend up to a prescribed payment-to-income ratio (generally suggested by commenters as 5 percent) as an alternative to a residual income underwriting approach, an approach the Bureau had contemplated in the Small Business Review Panel Outline and on which it specifically solicited comment in the proposal. During inter-agency consultations on the final rule, a fellow financial regulator also expressed support for this concept. These commenters argued that a payment-to-income approach would provide a streamlined compliance option for lower-cost lenders for whom the proposed ability-to-repay requirements would prove too cumbersome and expensive. These commenters cited positively the Bureau’s consideration of such a policy at the SBREFA process stage and criticized the Bureau’s failure to include the option as an alternative in the proposed rule. One research and public policy organization discussed in its comment letter potential additional policy suggestions that it believed would address criticisms of the approach raised by other stakeholders, including restricting lenders from using the payment-to-income approach if they experience high default rates (over 10 percent) and limiting the total loan cost to 50 percent of the amount borrowed. This commenter also sent a separate comment letter in conjunction with a number of large and mid-sized banks and other stakeholders endorsing the payment-to-income concept, arguing it would provide a streamlined and more cost-effective approach for depository institutions to make small-dollar loans. That 585 letter also provided a number of additional policy suggestions containing changes to the payment-to-income approach described in the Small Business Review Panel Outline, such as clarifying that evidence of regular deposits represents sufficient verification of income. The commenters also urged the Bureau to work with the federal prudential regulators to ensure sensible, streamlined regulatory oversight for small-dollar loans. In contrast, a number of consumer groups and other commenters strongly urged the Bureau not to adopt a payment-to-income approach and supported the Bureau’s decision not to propose it as an alternative. The consumer groups stated that they disagreed with a payment-toincome approach because it would not take into account consumer expenses, arguing that even a loan that is 5 percent of income could be unaffordable if the remaining income is allocated to expenses and emergency costs. One of these commenters noted that the Bureau’s study found that more than 40 percent of loans made under a 5 percent payment-to-income ratio would still default or be re-borrowed.760 The Bureau also received a number of comments objecting to its proposal to remedy the identified unfair and abusive practice through an underwriting requirement instead of disclosures alone. In particular, commenters stated that disclosure was a more appropriate remedy for any perceived lack of consumer understanding rather than complicated new underwriting requirements. They also argued that disclosures were a less restrictive alternative to the proposed ability-to-repay requirements and that the Bureau had not taken the disclosure option seriously. They pointed to model disclosures developed by industry trade associations as sufficient already to inform consumers of the high costs of using payday loans for an extended 760 CFPB Report on Supplemental Findings, at 25. 586 period. They also stated that the Bureau had not presented evidence that disclosures cannot adequately address the issue. One commenter specifically objected to the conclusions the Bureau derived from its analysis of the impact of the new Texas disclosures, which showed that following their introduction the disclosures decreased lending by 13 percent and the probability of re-borrowing by only 2 percent. The commenter argued that the appropriate conclusion is not that disclosure is ineffective, but rather, that consumers understand the costs and risks of payday loans and choose to take them out anyway. This commenter argued that the Bureau should have instead studied the impact the disclosures had on consumer understanding. Commenters raised other substantive and procedural arguments related to a disclosure alternative. An industry trade association argued that the Bureau had failed to respond to the trade association’s proposals to study and test enhanced disclosures, including a plan to partner with a firm that assisted the Bureau with the form design on the Bureau’s Know Before You Owe mortgage rulemaking. Several industry commenters argued that the Bureau’s discussion in the proposal of the marginal impacts of disclosures contradicted statements by the Bureau’s own researchers who had analyzed the impact of the Texas disclosures, noting that they had stated at a research conference in 2015 that enhanced disclosures can have economically meaningful impacts and that consumers who are more likely to end up in long-term debt cycles may be more responsive to disclosures.761 A large non-bank lender commenter cited the Bureau’s acknowledgment in a 2013 study that the Regulation E opt-in disclosures resulted in a majority of heavy over-drafters choosing not to opt-in to continued overdraft, as well as the lender’s own data indicating that its customers use extended payment plans at a higher rate (17.25% vs. 761 One lender commenter included a slide deck from this presentation in its comment letter as an attachment. 587 5.67%) in States that require disclosure, as evidence that disclosure produces successful outcomes. This comment also suggested that the Bureau should use TILA authority to create disclosures comparing the “all in” cost of credit to other alternatives and to apply the requirement across all consumer loan products including overdrafts. A trade group criticized the reliance on “dubious theories of behavioral economics” as a reason for rejecting the efficacy of disclosures. Finally, a separate trade group suggested that a disclosure requirement could be dynamic and require consumers to fill out a form that would demonstrate how much residual income they have each month based on projected income and expenses. Industry commenters, a joint letter from a number of State Attorneys General, letters from other attorneys general, SERs, and others argued that the Bureau had not considered as alternatives the less onerous approaches to regulating payday lending that many States have adopted. Commenters cited a variety of State laws, including laws about collection practices, disclosures, limits on the size and duration of loans, grace periods, limiting rollovers, principal repayment requirements, cooling-off periods, gross monthly income requirements, and even different ability-to-repay requirements. They also urged the Bureau to consider mixing and matching particular elements of the different State laws to find the right regulatory approach.762 Others argued that the Bureau should exempt entities operating in States that have payday laws. Other commenters urged the Bureau to consider additional less restrictive alternatives to the proposed ability-to-repay requirements, such as requiring lenders to offer extended payment 762 For example, one SER commenting proposed a hybrid of various State laws and other policy suggestions, calling for adoption of the Illinois gross monthly income requirement, a three-loan cap with provision of a fourth loan for emergencies with an off-ramp, and provision of reporting repayment of the off-ramp to nationwide consumer reporting agencies. An auto title lender suggested that the rule should permit the consumer to take advantage of all rollovers allowed by company policy and State law and require additional TILA disclosures. 588 plans, implementing a nationwide licensing and registration system, using existing enforcement authority to continue addressing “bad actors” or focus on unregulated or online lenders, or addressing consumer demand for payday loans by adopting measures to encourage consumer savings, similar to the Bureau’s “tools for saving.”763 Lastly, the Bureau received a number of comments asserting that the proposed rule conflicts with the Equal Credit Opportunity Act. They asserted that the proposal would have a disparate impact on women and minorities because they are more likely to be paid in cash, which is less documentable and would mean, as a result, that women and minority applicants for covered loans would be less likely to qualify for the loans under the ability-to-repay requirements. Additionally, some commenters argued that the proposal would prevent nonworking consumers, such as stay-at-home spouses, from receiving covered loans because they would not have their own individual income on which to rely for underwriting. They criticized the fact that the proposal did not permit consumers to rely on income from another person to which the consumer has a reasonable expectation of access, which may be considered under the Bureau’s credit card ability-to-pay rules. They noted, additionally, that the Bureau had amended those ability-to-pay rules in 2013 specifically to address a similar policy concern regarding access to credit for stay-at-home spouses, and questioned why the Bureau would apply a different standard in the proposal. Commenters further argued that the proposal’s allowance of estimates for rental housing expenses using locality-based data could create a disparate impact and look similar to more traditional “red-lining” discrimination. Commenters also argued that 763 Bureau of Consumer Fin. Prot., “Tools for saving: Using Prepaid Accounts to set aside funds; Innovation Insights,” (2016), available at https://www.consumerfinance.gov/data-research/research-reports/tools-saving-usingprepaid-accounts-set-aside-funds/. 589 the proposal’s definition of basic living expenses, which would have included expenses of any dependents of the consumer, would run afoul of Regulation B’s prohibition on seeking information about the consumer’s spouse. And more generally, some commenters argued that because covered loans are disproportionately used by minorities and women, the proposed rule would affect minority communities more significantly than other consumers. Final Rule As detailed below and in the discussion of specific parts of § 1041.5, the Bureau is finalizing the proposed ability-to-repay requirements for covered short-term loans and covered longer-term balloon-payment loans with substantial changes. These changes are designed to address various concerns raised by commenters, while still requiring lenders to engage in robust upfront underwriting procedures and providing targeted back-end protections to prevent consumers from getting stuck in long cycles of debt. In particular, the Bureau has made four substantial changes designed to make the final rule more flexible for both consumers and lenders, in order to facilitate efficient implementation and access to responsible credit: (1) the final rule permits use of a simplified underwriting calculation using either a residual income or debt-to-income methodology; (2) the final rule provides additional flexibility as to verification requirements, including permitting increased reliance by lenders on consumers’ written statements in appropriate circumstances; (3) the final rule permits consideration of situations in which the consumer has a reasonable expectation of access to others’ income or in which others regularly pay for certain of the consumer’s expenses; and (4) the final rule does not apply presumptions that a consumer will not be able to repay the second or third covered short-term loan or covered longer-term balloon-payment loan within a sequence. 590 The final rule thus consolidates, with modifications, parts of proposed §§ 1041.5 and 1041.6 for covered short-term loans and §§ 1041.9 and 1041.10 for covered longer-term balloonpayment loans in final § 1041.5. The conditional exemption for covered short-term loans originated under the separate requirements contained in proposed § 1041.7 is thus now renumbered as § 1041.6 in the final rule, and discussed separately below. The Bureau details its analysis for the individual elements of § 1041.5 below, after providing an overview of its response to the high-level issues summarized above and discussing the overall balance struck in the final rule. Burden, prescriptiveness, and complexity. As noted above, the Bureau received a significant number of comments from industry arguing that the underwriting requirements in the proposed rule would be too costly, take too much time to administer, be too restrictive, and require too much document verification. These commenters argued that the compliance burdens and underwriting restrictions would dramatically reduce loan origination volume, causing major impacts not only on lenders but on consumers as well through reduced access to credit, increased prices, and market consolidation. They also argued (as discussed separately further below) that the proposal unfairly imposed more rigorous underwriting requirements than the Bureau’s rules for other credit markets. As a general matter, the Bureau is sensitive to the concerns raised by many commenters regarding the burdens, prescriptiveness, and complexity of the proposal. The Bureau took some steps to address similar concerns that had been raised in response to the Small Business Review Panel Outline. For example, among the changes relative to the Outline, the proposal would have allowed lenders to use estimates of rental housing expenses instead of requiring verification of 591 lease documents, and included a 30-day, rather than a 60-day, definition of loan sequence and cooling-off period after a three-loan sequence. The Bureau also specifically sought comment in the proposal about automation and scalability, balancing the need for flexibility and innovation with the desire for regulatory certainty and related concerns. At the same time, the Bureau explained in the proposal that it believed that merely establishing a general requirement to make a reasonable determination that a consumer will have the ability to repay would provide insufficient protection for consumers and insufficient certainty for lenders. Rather, in light of stakeholder feedback to the Outline, Bureau experience, the experience with more general standards in some State laws, and the fact that lenders’ current screening is designed for more limited purposes, the Bureau believed that it was important to specify minimum elements of a baseline methodology for evaluating consumers’ individual financial situations. After careful consideration, the Bureau continues to believe that specifying a baseline underwriting methodology is not just reasonably related to preventing the unfair and abusive practices identified above, but also is necessary to a successful regulatory regime, as are targeted back-end protections to prevent consumers from becoming stuck in long cycles of debt. By requiring common-sense underwriting steps that incorporate both certain activities that are routine in other credit markets and tailored measures for the specific market, the Bureau believes that the baseline methodology substantially reduces the risk that consumers will obtain an initial unaffordable loan and provides greater regulatory certainty to lenders. At the same time, in light of the back-end protections, concerns about impacts on consumers who may have difficulty documenting certain income sources, and the need to leave room for lenders to innovate and 592 refine their methods over time, the Bureau believes that it possible to reduce the burdens, prescriptiveness, and complexity of the underwriting requirements in various ways relative to the proposal while still preserving the core of the essential consumer protections from the proposal. The four most significant changes to effectuate this revised framework, listed above, are summarized in the following discussion, with the section-by-section analysis of specific subparagraphs within § 1041.5 below providing further elaboration and detail. Beyond the four significant areas of change from the proposal, the Bureau has also taken a number of smaller steps to calibrate the ability-to-repay analysis in ways that differ from the proposal, which are described in the more detailed section-by-section analysis. First, as an initial matter, the Bureau agrees with commenters that the specific residual income methodology contained in the proposal for covered short-term loans would have been quite prescriptive in requiring lenders to track both the amount and timing of the consumer’s receipt of net income and payment of major financial obligations, as well as to project the consumer’s ability to cover major financial obligations and basic living expenses both during the loan term and for 30 days after the single highest payment.764 The proposal would not have required lenders to engage in detailed tracking of basic living expenses, but the analysis during 764 In contrast, the methodology for covered longer-term loans under proposed § 1041.9(b)(2) would have generally allowed lenders to calculate residual income on a monthly basis, although lenders making covered longer-term balloon-payment loans would also have had to evaluate consumers’ ability to cover major financial obligations and basic living expenses in the 30 days following the single highest payment on the loan. The proposal explained that for loans longer than 45 days, the Bureau generally believed that the particular number and amount of net income payments and payments for major financial obligations that will accrue between consummation and a payment due date were less instructive for determining a consumer’s residual income than for covered short-term loans. However, proposed comments 9(b)(2)(i)-1 and 9(b)(2)(ii)-1 emphasized that lenders would have been required to evaluate residual income for the month with the highest sum of payments in cases in which loan payments were not even, and to consider the amount and timing of major financial obligations in the period after the highest loan payment on a covered longer-term balloon-payment loan. 593 the 30 days after the highest loan payment in particular would have required specific attention to the timing of the consumer’s net income inflows and major financial obligation outflows.765 Upon further consideration, the Bureau believes it is appropriate to allow lenders a choice between residual income and debt-to-income methodologies, both of which would analyze the total amount of net income and major financial obligations during the month with the highest aggregate payments on the loan. Lenders can use this one-month snapshot to determine more generally whether the consumer has the ability to repay the loan without re-borrowing and can do so without having to track the specific timing of income receipts and major financial obligation payments. By simplifying the calculation to focus on the month in which the consumer is under the highest financial stress in connection with the covered short-term or covered longer-term balloon-payment loan, the final rule addresses concerns about compliance burden. The flexibility to use a debt-to-income methodology also allows lenders to use analyses that are more common in other credit markets, while maintaining appropriate tailoring in light of the variable payment structures and particular re-borrowing patterns evident in this market. See § 1041.5(a)(2) and § 1041.5(b)(2)(i) and the associated section-by-section analysis. Second, the Bureau has also made a number of modifications to the proposed requirements regarding verification evidence for consumer’s net income and major financial obligations. The final rule requires certain common-sense verification steps, such as requiring lenders generally to verify income, use a recent national consumer report to verify major 765 The proposed commentary examples in comment 5(b)(2)(i)-1.A and 5(b)(2)(ii)-1.i illustrate the granular focus that would have been required on the part of the lender to ascertain the timing of income receipts and expense payments as part of the broader ability-to-repay determination for covered short-term loans under proposed § 1041.5(b)(2). 594 financial obligations, and obtain a specialty consumer report from a registered information system in light of the fact that many covered loans are not reflected in national consumer reports. At the same time, the final rule reduces burden relative to the proposal and provides appropriate flexibility to consumers and lenders in cases in which verification is not reasonably available. For example, the final rule does not require income verification in all instances, as the proposed rule would have required. In those circumstances where a lender determines that a reliable income record is not reasonably available—as, for example, when a consumer receives some income in cash and spends that money in cash—the lender can reasonably rely on the consumer’s statements alone as evidence of income. See section-by-section analysis of § 1041.5(c)(2)(ii)(A) and associated commentary for further discussion. In addition, the final rule also no longer requires lenders to obtain a national consumer report for every single new loan. Rather, lenders may rely on a national consumer report that was obtained for a previous loan if the lender did so within the last 90 days, unless during the previous 90 days the consumer had taken out a sequence of three loans and thereby triggered a cooling-off period since the previous report was obtained. See section-by-section analysis of § 1041.5(c)(2)(ii)(D) and associated commentary for further discussion. And with respect to evidence of rental housing expenses, the final rule does not require a lender to verify them with a lease or with estimates based on data about general housing expenses in the locality of the consumer, as the proposed rule would have required. Instead, lenders are able to reasonably rely on consumers’ written statements for projecting rental housing expenses. See section-by-section analysis of § 1041.5(c)(2)(iii) and associated commentary for further discussion. 595 Third, unlike in the proposed rule, the final rule permits lenders and consumers to rely on income from third parties, such as spouses, to which the consumer has a reasonable expectation of access as part of the ability-to-repay analysis, as is generally true of the underwriting provisions for credit cards (although there are some distinctions described below, including that the lender must verify that the consumer has regular access to the funds). The final rule also permits the lender in certain circumstances to consider whether another person is regularly contributing to the payment of major financial obligations or basic living expenses. See sectionby-section analysis of § 1041.5(a)(5), § 1041.5(b)(1), and § 1041.5(c)(1) and associated commentary for further discussion. Fourth, the Bureau is not finalizing any of the presumptions of unaffordability from proposed §§ 1041.6 or 1041.10. The Bureau had proposed presumptions of unaffordability during the period in which a consumer had a covered loan outstanding, or for 30 days thereafter, under the theory that one can presume a consumer who returns within 30 days after paying off a prior loan was unable to repay that loan while still meeting other expenses (and hence likely would not be able to afford to repay a new loan). In light of the complexity associated with implementing that presumption, the Bureau is not finalizing these provisions, and is instead leaving the determination of whether a consumer has the ability to repay a second or third loan in a sequence to the reasonable discretion of the lender consistent with the requirements under § 1041.5. The Bureau will, however, view extensive re-borrowing, as observed through the lender’s performance metrics, as an indicator that the lender’s ability-to-repay determinations may not be reasonable. See section-by-section analysis of § 1041.5(b)(1), § 1041.5(d) and associated commentary for further discussion. 596 The Bureau has concluded that these significant changes will, collectively, reduce the upfront process burdens on lenders to underwrite these covered loans and provide more flexibility to consumers with regard to accounting for certain types of income, while maintaining the core elements of the proposal in reducing risks that consumers will become stuck in long cycles of unaffordable debt. The Bureau understands that any rule will impose some level of burden, especially for entities that have not previously had to comply with ability-to-repay standards. The Bureau is sensitive in particular to the concerns raised about the impacts on small lenders, by the SBA Office of Advocacy, the small entity representatives, and other stakeholders. The Bureau has analyzed these impacts in detail in the Regulatory Flexibility Analysis in part VIII, in addition to the compliance burdens on the industry in general in the Section 1022(b)(2) Analysis in part VII. As discussed in more detail in those sections, the Bureau has found that the compliance burdens of § 1041.5 will not impose undue costs, particularly as those burdens have been modified from the proposal in the final rule. For instance, the Bureau continues to expect that underwriting in accordance with the rule can largely be automated and that the market will evolve toward greater automation to manage operational costs and the time it takes consumers to obtain loans. Rather, the Bureau believes that the main impacts to the industry—including with regard to consolidation—are likely to be driven primarily by the question of how many consumers are reasonably determined to have the ability to repay covered short-term and longerterm balloon-payment loans and by the impact of the 30-day cooling-off period after the third loan in a sequence. As set forth in the Section 1022(b)(2) Analysis, the Bureau acknowledges that those impacts will be substantial and will likely drive significant consolidation and/or 597 product diversification, especially with respect to lenders who currently offer only short-term vehicle title loans. But putting limits on lending to consumers who lack the ability to repay is at the very heart of the rulemaking, as lenders’ failure to make reasonable ability-to-repay determinations in the market today is the crux of the unfair and abusive practice identified by the Bureau. As described above, the Bureau has concluded that it is necessary to proscribe that practice and adopt substantive regulatory measures reasonably designed to prevent it. The substantial changes in the final rule are intended to reduce the impact on lenders so that they are able to make reasonable ability-to-repay determinations without unnecessary cost. But the Bureau maintains its view expressed in the proposal that a robust ability-to-repay requirement is necessary or appropriate to prevent the unlawful practice identified by the Bureau, which leads to harms to many consumers. With regard to industry commenters who argued that the ability-to-repay requirements would have negative impacts on consumers in the form of increased time needed to obtain loans, increased prices, fewer lenders in close geographic proximity, and reduced access to credit in general, those issues are also addressed in greater detail in the Section 1022(b)(2) Analysis. As discussed in that section as well as with regard to specific elements of § 1041.5 below, the Bureau concludes that these impacts will generally be relatively modest. For example, as discussed above, the Bureau expects that the market will evolve toward automation in response to the rule, but for any lenders that choose to maintain an entirely manual system that loan processing time will be between 15 and 45 minutes.766 The Bureau also expects that compliance 766 As discussed in the Section 1022(b)(2) Analysis, the Bureau believes that changes from the proposal will facilitate automation under the final rule. While the Bureau has increased the estimate for purely manual 598 costs will not generally be passed through to consumers because many lenders are already charging the maximum amounts permitted by law, and that geographic impacts will be relatively modest in most areas. As described further below, the Bureau believes that a number of the modifications to final § 1041.5 will make it easier for consumers to access credit relative to the proposal, and consumers will also be able to access a limited number of covered short-term loans originated under § 1041.6 to deal with emergency situations or other needs. Indeed, the Bureau estimates that only six percent of current payday sequences would not be initiated due to the rule. Moreover, the Bureau disagrees with the commenters that argued that the proposal would preclude access to credit for any consumers who earn under $40,000 per year. As described in the Section 1022(b)(2) Analysis, the Bureau believes the analysis that underlies those comments rests on flawed assumptions and possible misunderstandings about the proposal.767 The Bureau notes that in making the changes to § 1041.5 to reduce the prescriptiveness of the upfront origination process requirements, it is not adopting many policy suggestions suggested by consumer groups that would have further increased verification requirements and other compliance burdens as well as further limiting re-borrowing. For example, consumer groups argued that lenders should never be permitted to rely on consumers’ written statements alone; that the Bureau should impose a cooling-off period after two loans in a sequence, rather than three; and that the final rule should impose an annual limit on all covered short-term loans underwriting relative to the proposal because a number of commenters had asserted that the original estimate was too low, the Bureau believes that the estimates for the final rule are lower than they would have been if all elements of the proposal had been adopted. Further, the Bureau believes that time for manual underwriting and the costs for lenders who choose to move toward a more automated model are not so concerning as to outweigh the benefits of preventing the identified unfair and abusive practice and the consequent risks and harms to consumers. 767 The Bureau also finds it significant that the undated presentation on which the commenters rely was not provided or discussed in individual comment letters submitted to the Bureau by three of the four specialty consumer reporting agencies that generated the analysis. 599 of six loans or 90 days of total indebtedness. The treatment of the consumer groups’ specific policy suggestions is discussed below in the relevant portions of the section-by-section for § 1041.5. At a broad level, however, the Bureau has concluded that the elements of the final rule as described further below will be sufficient to require lenders to engage in robust upfront underwriting and to provide targeted back-end protections to prevent consumers from getting stuck in long cycles of debt. In particular, the Bureau is finalizing a 30-day cooling-off period after a sequence of three covered short-term loans and applying it to sequences involving covered longer-term balloon-payment loans as well. The Bureau believes that the final rule as modified from the proposal will be sufficient to produce meaningful change in the incentives and practices of lenders in the affected markets, and that as long as those impacts are achieved it is appropriate to provide consumers and lenders with appropriate flexibility to meet individual circumstances under the rule. Furthermore, the Bureau acknowledges that in some cases the final rule provides more flexibility with respect to the ability-to-repay requirements than the Bureau indicated in the proposal that it was comfortable providing. For example, the Bureau is permitting lenders to reasonably rely on consumers’ written statements of net income if verification evidence is not reasonably available, in contrast to the proposal where it expressed concern about permitting loans to be made based on consumers’ written statements of income alone. The Bureau remains concerned about the same policy issues expressed in the proposal, but also sees merit in the arguments made by many commenters about the challenges of documenting certain types of income or obligations. The Bureau concludes that it has been able to calibrate this exception in the final rule appropriately to apply to those limited circumstances. As discussed further below, 600 the Bureau has also specifically emphasized that the ultimate reasonableness of lenders’ abilityto-repay determinations in such cases will be determined primarily by the pattern of outcomes for consumers. The Bureau has taken a similar approach with regard to other places where it has relaxed certain elements of the final rule relative to the proposal. The Bureau has judged that these changes strike an appropriate balance to ensure that the final rule provides core consumer protections that are necessary to address the identified harms in these markets, while at the same time reducing the burdens, complexity, and prescriptiveness of the proposed ability-to-repay requirements. Comparison to other markets. The changes described above in the final rule mean that relative to the proposal the rule is more consistent with underwriting practices in other consumer credit markets—whether specifically mandated by Federal law or as a matter of standard industry practice—while maintaining appropriately tailored requirements where the Bureau finds it appropriate to do so in light of the characteristics of the consumers who rely on covered shortterm and longer-term balloon-payment loans, the product structures used in these markets, and the particular patterns of re-borrowing seen in these markets. The Bureau notes that different markets warrant different regulatory interventions, as demonstrated by the fact that Congress itself has established very different regimes for underwriting mortgages and credit cards, and believes that calibration is appropriate to address particular consumer risks, industry practices, and product structures.768 768 With regard to student and automobile purchase-money loans, the Bureau notes that neither Federal consumer financial statutes nor regulations establish underwriting requirements for such loans. As the Bureau noted in proposing to exclude them from the scope of the final rule, both are quite distinct product markets that raise issues that are not present in the markets for covered short-term and longer-term balloon-payment loans. The Bureau 601 At a basic conceptual level, the final rule requires lenders to assess both consumer income and expenses using either a residual income or debt-to-income analysis. This is broadly consistent with the Federal underwriting requirements for both mortgages and credit cards, although the three regimes vary as to certain details in light of the products’ structure and the history of particular problems in their respective markets. For example, Congress specified a detailed regime for consideration of consumers’ ability to repay mortgage loans, including verification of both income and current obligations, after substantial evidence that “no-doc” loans helped to fuel a crisis in that market.769 In the credit card market, Congress imposed an obligation to consider consumers’ ability to make required payments on a credit card account, including heightened standards for consumers under the age of 21, in light of particular concerns that college students were being provided with amounts of debt that substantially exceeded their ability to make even minimum payments on their accounts.770 However, neither Congress nor the Federal Reserve Board, which was charged with implementing those requirements, chose to require specific verification requirements concerning income and expenses; the Board specifically noted that there had not been a record of the kinds of problems seen in the mortgage therefore disagrees with commenters that suggested that the proposal was somehow improper for failing to account for underwriting practices in these separate markets. As for check and ACH overdraft, the alternative to those fees is usually an NSF fee. For debit overdraft, the Federal Reserve Board created an opt-in regime which took effect in 2010 and which the Bureau is responsible for administering and enforcing. The Bureau has been studying the effects of that still-recent regime and opportunities to improve it. The Bureau also has been studying consumer outcomes with a particular focus on frequent overdrafters and is continuing to study the extent to which overdrafts occur in sequences that may suggest that repaying a prior overdraft led to a subsequent overdraft. 769 15 U.S.C 1639c(a)(1), (3), (4) (requiring assessment of consumer’s ability to repay a mortgage loan based on “verified and documented information,” including the consumer’s credit history, current income, current obligations, and various other factors). 770 15 U.S.C. 1637(c)(8), 1665e (requiring consideration of consumer’s ability to make required payments on a credit card account, but not verification). 602 market and that certain market conditions created strong incentives for lenders to exercise appropriate diligence even in the absence of specific Federal requirements.771 Similarly, the Bureau has tailored the details of the verification requirements and underwriting methodology in § 1041.5 based on the particular product structures and history of specific problems in the markets for covered short-term and longer-term loans. These include such factors as the frequency of lump-sum and irregular payment structures, the fact that many covered short-term and longer-term balloon-payment loans do not appear on national consumer reports, concerns that consumers who are in financial distress may tend to overestimate income or underestimate expenses, and lenders’ strong incentives to encourage mistaken estimates to the extent that doing so tends to result in more re-borrowing. The resulting final rule takes a common-sense approach by generally requiring lenders to obtain what verification evidence is reasonably available, while allowing reliance on consumer statements where other evidence is not. In their details, the income and expense verification requirements of the final rule are somewhat less onerous than the Bureau’s mortgage rules in 12 CFR 1026.43 and more onerous than the credit card rules for various groups of consumers in 12 CFR 1026.51.772 The final rule 771 The Board was also concerned about particular logistical problems where consumers wanted to open a credit card account at the point of sale with a retailer. 75 FR 7658,7721 (Feb. 22, 2010); 74 FR 54124, 54161 (Oct. 21, 2009). The rules therefore require creditors to consider information about income and current obligations, but not specifically to verify information supplied by a consumer. 12 CFR 1026.51(a)(1)(i). For a current description of industry’s routine reliance on consumer reports, see Bureau of Consumer Fin. Protection, “The Consumer Credit Card Market,” at 140-141 (2015), available at http://files.consumerfinance.gov/f/201512_cfpb_report-the-consumercredit-card-market.pdf. 772 To the extent that commenters asserted that the proposal’s verification and other requirements were disproportionate simply because covered short-term and longer-term balloon-payment loans have smaller balances than other credit products and mortgages in particular, the Bureau believes that there are certain fixed costs involved in responsible lending that do not vary much with size and that reducing below those minimums is unlawful. More generally as to overall processing times and burden, the Bureau concludes as summarized above and discussed in more detail in the Section 1022(b)(2) Analysis that a purely manual underwriting process for covered short-term and longer-term balloon-payment loans would still be quite modest, particularly compared to mortgage originations. 603 also has been modified in response to comments, discussed below, to allow consideration of situations in which consumers have a reasonable expectation of access to the income of other people and where another person regularly pays for certain expenses of a consumer, which is somewhat similar to the credit card rules but with more tailoring in light of the overall structure of § 1041.5 and general concerns about incentives to inflate income in the affected markets. As noted above, the Bureau received many comments from industry stakeholders suggesting that it apply the same rules as for credit card ability-to-pay rules under Regulation Z. The Bureau believes the response to these comments merits more extensive discussion. First, the Bureau disagrees with commenters that stated or implied that the proposed ability-to-repay requirement reflected a different legal standard for underwriting than the credit card ability-to-pay rule and questioned the basis for that difference, including the one commenter’s argument that the Bureau was imposing a different standard because it did not “trust” consumers in this market to make decisions for themselves. It is true that the credit card rules focus only on a consumer’s ability to make “required minimum payments,” which under credit card contracts are typically minimum monthly payments—typically finance charges, fees, and a small amount of principal—for however long it takes to pay off the principal.773 The ability-to-repay test set forth in the final rule requires the lender to determine whether the consumer can make “all payments on the loan.” As a legal standard, however, that is no different than the test under the CARD Act. That is, in both cases the rule requires that the 773 As the Board noted in issuing rules to implement the CARD Act standard, “Because credit card accounts typically require consumers to make a minimum monthly payment that is a percentage of the total balance (plus, in some cases, accrued interest and fees), the final rule requires card issuers to consider the consumer’s ability to make the required minimum payments.” 75 FR 7658, 7660 (Feb. 22, 2010). 604 lender assess the consumer’s ability to repay the payments required under the contract. What differs in the two contexts is the structure of the loan and thus the size of the required payments under the contract. Consumers under the typical covered short-term or longer-term balloon-payment loan have a legal obligation to repay the full amount of the loan when due in a single or large balloon payment, and the loans are presented to consumers as having a definite term. Consumers do not have the right to roll over or re-borrow; that is up to the discretion of the lender. Thus, to the extent that commenters implied that the Bureau should require that lenders inquire only about consumers’ ability to pay finance charges, such an approach would be fundamentally inconsistent with the structure of these loans and would ignore the fact that at some point the principal must be repaid in a single or large balloon payment. Indeed, to apply the ability-torepay test only to the finance charges would perpetuate one of the core concerns underlying this rule: that, as discussed in Market Concerns—Underwriting, these loans are presented to consumers as short-term loans to bridge until the next paycheck whereas in practice the loans operate quite differently.774 As discussed below in the 1022(b)(2) Analysis in more detail, there is substantial evidence that many consumers end up re-borrowing more than they expect and that consumers who end up in very long loan sequences in particular do not predict their usage patterns accurately.775 774 As discussed in Market Concerns—Underwriting and the section-by-section analysis for § 1041.4, the Bureau’s extensive research on the small-dollar lending market has focused to a large degree on the problem of consumers rolling over their loans on the due date or re-borrowing within 14 to 30 days of repayment of the prior loan. The product structure typically associated with covered short-term loans—a lump-sum payment due within 14 or 30 days of consummation and tied to the consumer’s payday—leads to the re-borrowing problem. 775 In contrast, credit cards are commonly understood to be an ongoing product. The Bureau further notes that the final rule does not cover open-end credit which amortizes over a period of more than 45 days without a balloon 605 The Bureau, furthermore, disagrees with commenters who asserted that the Bureau should follow the model of the credit card rules and not require verification of income. The Bureau believes that in view of the particular concerns about reliance on stated income in the market for covered short-term loans and covered longer-term balloon-payment loans, it is appropriate to include a baseline verification requirement in the final rule. Under the final rule, in § 1041.5(c)(2)(ii)(A), the lender must verify the consumer’s net income amount if verification evidence is reasonably available. If verification evidence as to some or all of the net income is not reasonably available, the lender may reasonably rely on the consumer’s statement of the amount. As described in the section-by-section analysis for § 1041.5(c)(2)(ii)(A) below, permitting lenders to reasonably rely on consumer statements of income in absence of verification evidence is a change from the proposal that addresses commenters’ concerns that consumers paid in cash will not be able to receive a loan if they otherwise would pass the abilityto-repay requirements. The Bureau does not believe, however, that merely requiring consideration of consumers’ stated amounts for net income and debt obligations as a baseline rule would provide sufficient consumer protections in this market. The Bureau notes that the income verification requirement in the final rule is generally aligned with current practices in the market for covered short-term loans (other than with regard to some vehicle title loans), where lenders typically request the consumer provide evidence of one pay cycle of income. Moreover, as discussed above, the Bureau understands that credit card issuers typically obtain a national consumer report for card applicants to ascertain “current obligations” under the credit card payment. Thus the rule does not restrict lenders from offering open-end credit plans with affordable minimum payments which amortize a loan over time. 606 ability-to-repay rules, which is similar to the obligation under the final rule for lenders making covered short-term loans and covered longer-term balloon-payment loans to obtain a national consumer report to verify debt obligations.776 Specificity. As discussed above in connection with the proposal and with regard to the final rule, the Bureau has attempted to balance the interests of specificity, which reduces uncertainty, with the interests of flexibility, which allows for innovation, competition, and diversification in business models. The Bureau received incompatible comments requesting that it shift further in both directions on the specificity-flexibility spectrum—sometimes from the same commenter when addressing different issues. Ultimately, as compared to the proposed rule, the Bureau found the commenters requesting more flexibility rather than additional prescriptiveness with regard to upfront underwriting procedures to raise the more compelling arguments, and decided to add more flexibility to the final rule as discussed generally above and with regard to individual elements below. At the same time, as discussed below, the Bureau has also refined the regulation text and commentary as appropriate in specific areas, for instance to provide clearer guidance on particular elements of the ability-to-repay analysis such as net income and estimation of basic living expenses and to discuss various fact patterns in examples. 776 Finally, a few commenters noted that the credit card rules allow lenders to consider the consumer’s debt-to-assets ratio as a means of satisfying the ability-to-pay requirement. The Bureau notes that this highlights the differences in the markets being regulated. While that approach might make sense in the context of credit cards, in the context of the markets at issue in this rule, many consumers will have exhausted their cash assets before seeking a covered loan. Moreover, as discussed in Market Concerns—Underwriting and the section-by-section analysis for §§ 1041.4 and 1041.6, the Bureau has concluded that vehicle title loans pose substantial harm to consumers in absence of robust underwriting that is tied to a consumer’s income and expenses, not the value of the vehicle. The Bureau is concerned that permitting lenders to rely on a debt-to-asset ratio for underwriting would potentially validate current practices by vehicle title lenders and fail to result in a meaningful change in current practices to remedy the identified harms. 607 With regard to commenters who criticized the general reasonableness standard, sought numerical thresholds or guidance on what constitutes sufficient residual income (or a specific debt-to-income ratio), or urged the Bureau to provide per se rules regarding what types of loan performance patterns indicate that a lender’s ability-to-repay analysis was unreasonable, those issues are discussed in more detail below in connection with § 1041.5(b)(1). While this rule provides substantial specificity as to upfront procedures, the Bureau does not provide a formulaic residual-income threshold or debt-to-income ratio to answer the question of whether a consumer has the ability to repay. The same is true for the Bureau’s mortgage and credit card ability-torepay rules.777 The Bureau does not believe it is possible to eliminate lender judgment in making these determinations, and thus believes that the general reasonableness standard is a critical element of the rule. Reasonableness is a widely used legal concept in both State and Federal law, and is what Congress required with respect to the underwriting of mortgages. The Bureau believes the standard in the final rule—which has been revised to include a substantial amount of new commentary clarifying how the reasonableness of ability-to-repay determinations will be evaluated—should provide a sufficiently discernible standard. As for loan performance, as discussed in final comments 5(b)-2.iii and 5(b)-2.iv, the Bureau will, among other things, use various outcome metrics on an aggregate basis to assess 777 The Bureau did adopt a 43 percent debt-to-income threshold for one type of “qualified mortgage,” which is subject to either a conclusive or rebuttable presumption of compliance with ability-to-repay requirements under the mortgage rules depending on particular loan terms. 12 CFR 1026.43(e)(2)(vi). However, the Bureau emphasized in adopting this threshold that it was based on longstanding benchmarks in the mortgage market (which do not exist in the markets for covered short-term loans and covered longer-term balloon-payment loans), that other types of qualified mortgages would allow lending to consumers with ratios in excess of 43 percent, and that the Bureau did not believe it was appropriate to set an across-the-board threshold for determining consumers’ ability to repay mortgage loans for similar reasons to those discussed here. See generally 78 FR 6408, 6460-62, 6470, 6526-28, 6533-35 (Jan. 30, 2013). 608 whether various underwriting models are indeed working as a practical matter to yield reasonable determinations of consumers’ ability to repay. However, such metrics must also be evaluated in their specific context, particularly given that the harms that arise from unaffordable loans may play out in different ways depending on lender practices and other variables. For example, lenders might have higher patterns of re-borrowing relative to defaults depending on their particular sales and collection practices, so establishing a single set of thresholds for all situations would be difficult. As discussed below, the Bureau has provided more specific guidance on the types of potentially relevant loan performance metrics and more examples discussing particular fact patterns, but believes that it is not practicable to establish numeric performance thresholds that would definitively demarcate whether a lender’s ability-to-pay determinations meet the reasonableness standard. See the discussion below regarding § 1041.5(b)(1) for more details. Using Residual-Income Analysis to Predict and Prevent Harms. As described above, several industry commenters asserted that the proposed requirement to determine consumers’ ability to repay is arbitrary because it will not actually predict and prevent the harms identified in the Bureau’s UDAAP analysis, particularly default and re-borrowing. For example, an industry commenter cited a study that uses what the researchers said was the residual income methodology specified in the proposed rule to examine the relationship between such residual income and default. Applying the residual income methodology to a large sample of storefront payday loan borrowers, the study compares consumers deemed to have positive residual income to consumers deemed to have negative residual income with respect to whether they repaid or defaulted on a particular test loan. In one such analysis using the borrower’s income most 609 recently observed by the lender, loans in which the borrower had positive residual income had a default rate of 11 percent, compared with a default rate of 14.7 percent for loans in which the borrower had negative residual income. The study concluded that little difference in default rates exists between these two populations, and that the residual-income analysis is not highly predictive of default. On the basis of these results, the industry commenter inferred that the proposed rule’s ability-to-repay requirement will not prevent consumers from defaulting. Setting aside the issue of whether the difference in default rates among loans for which the borrowers did and did not have residual income was meaningful,778 the Bureau does not agree with the commenter’s inference that an ability-to-repay requirement will not reduce the harms identified in the Bureau’s unfairness and abusiveness analyses above. The study focuses only on defaults in isolation, despite the fact that as the Bureau has explained numerous times (both in the proposed rule and elsewhere in the final rule), when consumers are faced with an unaffordable covered short-term loan, their most frequent response is to roll over short-term loans (in States where doing so is permitted) or nominally repay the loans, only to have to reborrow shortly thereafter. In its analysis, the Bureau found that only 28 percent of loan sequences consisted of single loans, with the remaining 72 percent of loan sequences consisted of at least one re-borrowing. For that 28 percent, 22 percent were repaid without re-borrowing, and only 6 percent defaulted.779 Where the lender has account access, such repayment is accomplished by a debit of the consumer’s account. Where the lender has obtained a postdated 778 The Bureau notes that the residual income test performed using the consumer’s recently-documented income as observed by the lender indicated that consumers with negative residual income defaulted on their loans 34 percent more often than consumers with some amount of positive residual income. 779 CFPB Supplemental Report, at 120. 610 check, such repayment is made either by way of that check or in light of the fact that the lender may deposit the check at any time. All of this explains why the default rate of covered shortterm loans for which the consumer does not have the ability to repay is relatively low. Indeed, the commenter effectively conceded this point when it claimed that by imposing a cooling-off period after the third loan in a sequence, the proposed rule will drive default rates higher. In addition, even if looking solely at default rates were a relevant metric, the study itself identifies a number of possible explanations for its finding of similar default rates for the two populations, including that account access may incentivize borrowers to prioritize paying the loan notwithstanding cash flow shortages affecting other expenses, which is one of the factors noted in the preceding paragraph.780 A specialty consumer reporting agency commenter made a similar argument based on a study it conducted using its own borrowing data. At a high level of generality, the study found very similar default rates for loans made to consumers with positive residual income compared to consumers with negative or zero residual income (with default rates of 16.1 percent and 16.2 percent, respectively). However, a more detailed analysis that disaggregates these consumers into varying degrees of residual income, ranging from those with negative residual income of negative $2,500 or less to those with more than $2,500 in positive residual income, showed higher default rates among consumers who have the most negative residual income (20.0 percent) compared to those with far less negative or positive residual income (15-16 percent). 780 The study does not, however, include in its list of explanations the main factor identified above, namely, that default rates for borrowers who lack the ability to repay are relatively low because their re-borrowing rates are so high. More generally, given that (i) re-borrowing rates are significantly higher than default rates, and (ii) it appears that the data used for this study could have been used to conduct a similar study of the re-borrowing rates for the two population, it is not clear why the researcher chose to conduct a study solely on default rates rather than a study on re-borrowing rates (or rather than a study that included both). 611 Relatedly, the study reported that first-time borrowers with positive residual incomes had slightly lower default rates than first-time borrowers with residual incomes that were zero or negative. In addition, the study found that consumers who triggered any of the proposed 30-day cooling-off periods had markedly lower default rates than consumers that did not trigger the criteria. Like the industry commenter, this commenter concludes that residual income is not a good predictor of default.781 The commenter likewise forecasted that the proposed rule’s restrictions on reborrowing will drive up default rates. In addition, citing the study results, the commenter urged the Bureau to modify the rule in three respects: (1) replace the ability-to-repay requirement with a propensity-to-repay requirement; (2) limit such an ability-to-repay requirement to first-time borrowers and those with low propensity to repay; and (3) eliminate all of the 30-day cooling-off periods. Given the close similarity of this commenter’s argument regarding the relationship between residual income and default to the argument of the industry commenter discussed above, the Bureau believes its response above to that argument applies equally to this one. For essentially the same reasons, the Bureau believes that the commenter’s proposed modifications of the rule are unwarranted and would, in fact, result in perpetuating most of the harm experienced by consumers in the current market.782 In addition to making the comment discussed above about default, the same industry commenter made a similar argument about re-borrowing: the commenter argued that ability to 781 This commenter also argued that it would therefore be inappropriate for the Bureau to base assessments of lenders’ compliance with the ability-to-repay requirements on their default rates. 782 The Bureau also addresses the recommendations to replace ability to repay with propensity to repay, and to remove all cooling-off periods, in the discussion of § 1041.5(d) below. 612 repay is no more predictive of re-borrowing than it is of default. In support of this claim, the commenter cited two studies. The first is the same study it cited in support of the “default” argument. In this instance, instead of describing the study as finding that there is a weak correlation between residual income and default, the commenter described it as finding that there is a weak correlation between ability to repay and repayment. The Bureau is not persuaded that this study provides such support. To be sure, if a study considers only default and repayment, its findings about default could be presented as findings about repayment, which is the mirror image of default in such a study. By the same token, however, given that such a study does not consider re-borrowing rates at all, it is unclear how findings about such rates can be derived from findings about default, or from mirror-image findings about repayment. The second study, which predated the proposed rule, contained a number of slides that reference ability to repay, the most pertinent of which appears to be one that includes the claim that consumers with large amounts of residual income are as likely to roll over their loans as consumers with limited residual income. Just below that is what appears to be a screen shot of a portion of a database or spreadsheet with various numbers and percentages. On its face, the statement does not appear to provide support for the commenter’s assertion. Nor does the commenter make any attempt to explain this page of the presentation.783 Disclosure alternative. The Bureau disagrees with commenters that asserted that a disclosure remedy would be sufficient to prevent either the unfair or abusive practice itself or the 783 Without citing any studies about either default or re-borrowing, another industry commenter argued that the Bureau had assumed without evidence that satisfaction of the proposed residual income test would predict and prevent injury from re-borrowing and default, and thus that it would be inappropriate for the Bureau to assess a lender’s compliance with that test based on performance metrics. The Bureau disagrees, as it has based the abilityto-repay requirement on a substantial body of evidence, including the evidence of re-borrowing rates cited above. ` 613 risks and harms to consumers from such practice, that the Bureau is compelled as a matter of law to adopt disclosure remedies to address any unfair or abusive practices that involve a lack of understanding by consumers, and that the Bureau erred in proceeding with the rulemaking instead of delaying it to conduct further disclosure research. The Bureau notes that consumer disclosures can be an important and effective tool in different circumstances and indeed has adopted disclosures to communicate various pieces of information to consumers in connection with this final rule. But for the reasons discussed in the proposal and below, the Bureau concludes that disclosures would not be sufficient to prevent the unfair and abusive practices identified in this rule. More generally, the Bureau concludes that it is not required to mandate disclosures to address any unfair or abusive practices that involve a lack of understanding by consumers, as opposed to adopting other approaches, such as the ability-to-repay provisions here, to prevent the unfair or abusive practices. Neither Congress784 nor other agencies785 nor the courts786 have 784 For instance, in the Dodd-Frank Act, Congress authorized the Bureau both to take action to identify and prevent unfair or abusive acts or practices and to impose disclosure requirements regarding any consumer financial product or service. If Congress had determined that disclosures were adequate and in fact required to address any unfair or abusive act or practice that involves consumer misunderstanding, then Congress could have directed the Bureau to adopt disclosures in such circumstances. Congress did not do so. 785 For example, the Federal Reserve Board promulgated a rule in 2010 prohibiting mortgage loan originator compensation from varying based on loan terms due to concerns about the steering of mortgage borrowers into less favorable terms than those for which they otherwise qualified. 75 FR 58509 (Sept. 24, 2010). The Board issued this rule under its TILA section 129(p)(2) authority to regulate unfair and deceptive practices in the mortgage market and had determined that a substantive approach was necessary. The Board found that, based on its experience with consumer testing, “disclosure alone is insufficient for most consumers to avoid the harm caused by this practice.” The Board also in its unfairness analysis discussed how a Regulation X disclosure promulgated by the Department of Housing and Urban Development similarly “is not likely by itself to prevent consumers from incurring substantial injury from the practice.” Id. at 58514-15. 786 See AFSA, 767 F.2d at 989 (upholding a rulemaking that “reasonably concluded” that the most effective way to eliminate an unfair practice concerning adoption of certain contractual remedies was to proscribe the contract clauses outright because “‘[d]isclosure alternatives would deal only partially with limited seller incentives to promote alternative remedies ... and would not address at all consumers’ limited incentives to search for information about remedies.’”). 614 adopted such a position. The Bureau is authorized by section 1031(b) of the Dodd-Frank Act to prescribe rules to identify unfair, deceptive, or abusive acts or practices and to include in such rules requirements for the purpose of preventing such acts or practices. The unfair and abusive practice the Bureau has identified in § 1041.4 is making covered short-term or longer-term balloon-payment loans without reasonably determining that consumers will have the ability to repay the loans according to their terms. No commenter claims that providing disclosures will prevent that practice. At most, effective disclosures could mitigate some of the harms from the failure to underwrite. In theory at least, disclosures could be so effective that any harms would be reasonably avoidable by the consumer and that consumers would no longer lack understanding of the material costs and risks of the product. However, as discussed below, the Bureau concludes that disclosures here would not have any such effect. The Bureau agrees that informing consumers that covered short-term loans or covered longer-term balloon-payment loans have high risks of default, re-borrowing, or default avoidance harms or that lenders are not underwriting such loans using the same sorts of practices that are common to other credit markets may cause some consumers to be more generally cautious in taking out such loans. Indeed, the Bureau’s analysis of the response by consumers to the new disclosure in Texas is consistent with this outcome. The Bureau finds it likely that the marginal difference in lending (around a 13 percent decrease in loan volumes) in fact resulted from consumers whose decisions were affected by the disclosures and decided not to borrow after better understanding the risks. However, generalized or abstract information does not inform the consumer of the risks of the particular loan in light of the consumer’s particular financial situation. Lenders would still 615 have strong incentives, given their overall business models, to make loans to consumers who cannot in fact afford to repay them according to their terms, as long as such consumers do not default early in their loan sequences. Because consumers using these loans—or at least those who end up in extended loan sequences—are not good predictors of how long it will take them to repay their loans, generalized disclosures are particularly unlikely to position consumers effectively to appreciate the risks they themselves would face from their loans and to make their decisions accordingly. In light of these circumstances, the Bureau finds that generalized disclosures to consumers will not prevent the unfair and abusive practice identified above or equip consumers to avoid the harms it causes as effectively as prohibiting lenders from engaging in the unfair and abusive practice in the first instance. The only disclosure that the Bureau could envision that could come close to positioning consumers to mitigate the unfair and abusive practice effectively would be an individualized forecast of whether the consumer could afford to repay the loan according to its term, and if not, a forecast of how long such repayment would be reasonably expected to take. While consumers are most familiar with their particular financial situations, lenders have the most information about their business models and the performance of their credit products over hundreds or thousands of individual cases. The Bureau notes, however, that no commenter has suggested such an approach, which would be unprecedented as a matter of mandatory disclosures under federal consumer financial law. Moreover, if anything, an individualized disclosure might require more compliance burden than the final rule to the extent that it would require a lender to 616 forecast how many rollovers or re-borrowing might be required in the event that a consumer is not likely to repay the entire balance during the initial loan term.787 Further, with disclosures in this specific context, the only option for a consumer warned about the risks of an unaffordable loan is simply not to take out the loan at all, since once a consumer takes out a loan that in fact turns out to be unaffordable the consumer’s only options are to choose between the harms associated with default, re-borrowing, or forgoing other major financial obligations or basic living expenses. Thus, the Bureau believes that it is telling that while the Texas disclosures appear to have caused some consumers to seek different options altogether, in the first instance, once they had already taken out a loan, there was only a 2 percent decrease in the probability of re-borrowing.788 The Bureau also addresses three other arguments commenters raised about disclosures. First, as to the specific trade group commenter’s argument that the Bureau was wrong to reject a formal invitation to engage in a study to test enhanced disclosures, the Bureau notes that this commenter had engaged in outreach with the Bureau for several years during the course of the rulemaking, yet did not present the disclosure trial proposal until less than two weeks before the 787 As noted earlier, one commenter suggested that a disclosure requirement could be dynamic and require consumers to fill out a form that would demonstrate how much residual income they have each month based on projected income and expenses. The Bureau notes that this suggestion bears some conceptual similarity to traditional installment lenders who, as noted in the proposal, work with their customers to prepare a budget itemizing income and expenses. However, in that case the lenders use the information to conduct an ability-to-repay analysis, which would not happen under the commenter’s suggested regime. As such, the Bureau believes this type of approach would not sufficiently address the identified harms. 788 For these reasons, the Bureau disagrees with the commenter that asserted that the Bureau’s economists made statements at a conference undermining the Bureau’s statements in the NPRM regarding the effectiveness of disclosures. The Bureau views those statements as compatible with its statements on this issue in the proposal and in this final rule. Specifically, the presentation asserted “borrowers more likely to end up in long-term debt cycles may be more responsive to disclosures” (emphasis added). The Bureau also notes that, even if these borrowers are relatively more responsive to disclosures, that fact would not equate to such disclosures being an effective means to reduce these sequences, let alone a viable substitute for the ability-to-repay approach set forth by the rule. 617 proposal was released and requested that the Bureau delay issuing a proposal or hold the comment period open during the pendency of the proposed study.789 Thus, in addition to the substantive reasons discussed above for why the Bureau concludes that generalized disclosures are insufficient to prevent the practice or harms identified, the Bureau rejected the request to delay the proposal in light of this strategic procedural posturing. The Bureau did indicate that it would be open to considering the results of any new research as part of the comment process, but no such evidence has been forthcoming. Second, the Bureau finds that commenters overstate the degree to which the Bureau is relying on behavioral economics in rejecting a disclosure alternative. As discussed above, there are both theoretical and data-driven explanations for why the Bureau does not share the view that disclosures will sufficiently remedy the observed harms. Lastly, the Bureau does not view as a viable option one commenter’s suggestion of requiring a new TILA disclosure that would potentially capture the “all-in” cost of credit. The Bureau finds that this disclosure would not be 789 The Bureau notes that the commenter presented the disclosure trial proposal to the Bureau at a meeting shortly after numerous press reports had already indicated that the proposal release was imminent. See, e.g., “CFPB to Propose Payday-Loan Rule on June 2,” Wall St. J. (May 18, 2016), available at https://www.wsj.com/articles/cfpbto-propose-payday-loan-rule-on-june-2-1463615308; “CFPB Set to Release Payday Lending Proposal on June 2,” Am. Banker, May 18, 2016, available at https://www.americanbanker.com/news/cfpb-set-to-release-paydaylending-proposal-on-june-2. The Bureau also notes receipt of a comment from an executive at a large lender who stated that he had sent correspondence to the Bureau in June 2015 following the Small Business Review Panel Outline release and the Small Business Review Panel meeting, which offered to make the commenter’s company available to conduct a controlled field trial to measure consumer outcomes relating to the proposals under consideration. The commenter noted that he had raised the idea again when he met with Bureau officials, along with trade groups and other lenders, in July of 2015. The commenter argued further that, at the meeting, Bureau officials were dismissive of the idea because it was “not a test and learn environment” and that the Bureau had not spoken to consumers and did not think it necessary to do so. The Bureau does not agree with the commenter’s assertions. To the extent any statements were made referring to a “test and learn” environment, Bureau officials were referring to the difficulty of incorporating a sandbox approach to testing policy ideas into an ongoing formal Federal rulemaking process, which was well underway at the time (see discussion elsewhere regarding other commenters’ ideas about sandbox approaches). Moreover, the Bureau has heard from consumers during the rulemaking process and views such feedback as meaningful, including its review of more than one million comments from individual commenters. See part III. 618 effective at preventing the unfair and abusive practice or rectifying the identified harms for the same reasons as described above. Payment-to-income alternative. While the Bureau is now allowing lenders to choose between underwriting approaches based either on a debt-to-income ratio or on residual income, the Bureau is not adopting an alternative approach centered on a payment-to-income ratio. The Bureau recognizes that many commenters have expressed strong support for this approach, including depository institutions interested in making lower-cost small-dollar loans. However, the Bureau notes that the particular proposal under consideration at the SBREFA stage and which these commenters have elaborated upon in their comments—namely a safe harbor for loans with a payment that takes up 5 percent or less of a consumer’s income—is far more relevant to the market for longer-term installment loans than for the loans covered by § 1041.4 and 1041.5, as those loans generally have lump-sum or other large irregular payments that far exceed a 5 percent payment-to-income ratio for the vast majority of consumers. Consider, for example, a consumer making $2,000 per month. A 5 percent payment-toincome ratio safe harbor would mean the consumer is only eligible for a $100 loan, assuming all payments on the loan would be due in one month; for loans due in two weeks—as is common for payday loans—the maximum loan amount would be only $50. Accordingly, the Bureau does not believe that lenders or consumers would be likely to use a 5 percent payment-to-income option in the short-term space, particularly where it is permissible to make loans under § 1041.6 in amounts of up to $500.790 To the extent the Bureau engages in further study and potential future 790 The Bureau also has some skepticism that a consumer’s ability to repay a covered short-term loan or covered longer-term balloon-payment loan can be evaluated without some consideration of major financial obligations and 619 rulemaking on longer-term installment products, the Bureau will continue to consider whether a payment-to-income approach either in the specific form suggested by the commenters or in other forms would be a reasonable alternative to an ability-to-repay requirement. State law regulatory approaches. As discussed above, many commenters argued that the Bureau failed to rigorously study existing State laws regulating small-dollar loans and consider more seriously whether one or more existing regulatory approaches in the States would be sufficient to address the concerns the Bureau identified in the market rather than the ability-torepay requirements. The Bureau also notes that in some cases, State Attorneys General or other State or local officials in the States cited by the aforementioned commenters as having model State regulatory approaches wrote in support of the proposed ability-to-repay requirements and of the proposal in general, reflecting a diversity of opinion about the sufficiency of the laws in those States to address the identified harms at the Federal level. The Bureau has over the past several years studied the regulatory approaches of many States carefully and, as discussed in part III, has engaged in outreach with a wide variety of stakeholders including elected officials and regulators in States that permit covered lending. The development of the proposal framework and the final rule has been informed by this understanding of these State laws. The Bureau provides more detail on State laws in part II, but some examples follow. basic living expenses, particularly in light of their lump sum or irregular payment features. For example, the Bureau notes that some States have limited short-term loans to 25 percent of income, but such limitations do not appear to have produced any substantial improvement in re-borrowing rates. See, e.g., Nev. Rev. Stat. sec. 604A.425.1(a); see also State Law Regulatory Approaches below. 620 A number of States set rollover thresholds that are higher than those in this final rule. Delaware permits four rollovers on payday loans, Missouri permits six on payday loans, and New Hampshire permits 10 rollovers on short-term title loans.791 Idaho, on the other hand, sets their rollover cap at three, similar to this rule.792 Other States, like California and Kentucky, impose fewer restrictions but cap payday loans at, for example, $500 (Kentucky) or $300 (California). Other commenters argued that States have imposed less onerous, but nonetheless effective, ability-to-repay frameworks that the Bureau should consider adopting instead of the proposed ability-to-repay requirements. For example, some commenters noted Utah as an example. Utah lenders must determine that a consumer has the ability to repay a loan based on one or more of the following sources: a consumer report from a consumer reporting agency, verification or proof of income, the borrower’s self-affirmation of ability to repay, or prior payment history with the lender from its own records.793 In addition, lenders may not roll over loans beyond 10 weeks, and once a year consumers may request extended repayment plans. It appears one significant difference between Utah law and this rule is in how that State treats reborrowing. In Utah a lender need only determine whether the consumer can repay the loan in the ordinary course, “which may include rollovers or extended payment plans,” and need not make a separate repayment determination on rollovers.794 To comply with § 1041.5(b), lenders will need to determine whether consumers have an ability to repay each loan according to its terms, 791 5 Del. Laws. Sec. 2235A(a)(2); Mo. Rev. Stat. sec. 408.500(6); N.H. Rev. Stat. sec. 399-A:19. Idaho Code Ann. Sec. 28-46-413(9). 793 Utah Code Ann sec. 7-23-401. 794 Id. 792 621 without re-borrowing. And Utah law allows 10 weeks of re-borrowing, as opposed to the Bureau’s cap of three loans in a sequence (under § 1041.5(d)), which would result in a shorter period for consumers taking out 14-day loans (approximately six weeks of re-borrowing), but a longer period for consumers taking out 30-day loans (approximately 12 weeks of re-borrowing). Of course, the Bureau’s approach is not more restrictive than that used by all the States. For example, only a minority of States, 19 by the Bureau’s count, permit vehicle title lending with lump-sum (typically short-term) structures, and 15 States and the District of Columbia either ban payday loans or set fee or interest caps that payday lenders find too low to sustain the business model (see part II). Even in States that do allow payday lending, certain parts of their payday lending laws may be more restrictive. For example, the cooling-off period imposed by Virginia in certain circumstances lasts 45 or 90 days,795 while the Bureau’s rule sets cooling-off periods, such as the one in § 1041.5(d), at 30 days. Commenters also raised Colorado’s laws as a model. However, following such an approach would involve banning covered short-term lending altogether since that State only allows loans of at least six months in term. To the extent the Bureau engages in further study and potential future rulemaking concerning longer-term installment products, the Bureau will continue to consider whether the Colorado model may provide additional insight.796 795 Va. Code Ann. sec. 6.2-1816. Specifically, the law requires a 45-day cooling-off period after a consumer has taken out five loans in 180 days and a 90-day cooling-off period after a consumer completes an extended payment plan. The Bureau received a comment letter from the State Attorney General in Virginia that urged the Bureau to finalize a 60-day cooling-off period or, at minimum, a 45-day cooling-off period, and discussed the above referenced 45-day cooling-off period under Virginia law as context for the request. See the discussion of § 1041.5(d) below for a more detailed description of the Bureau’s decision to adopt a 30-day cooling-off period in the final rule. 796 The Bureau also notes that Colorado does require lenders to obtain detailed information and credit histories from consumers for creditworthiness analysis in cases in which the loan exceeds a certain size threshold. 622 Though the Bureau closely studied the various States’ approaches as it developed this rule, the Bureau concludes that none of these State law frameworks, alone, would suffice to prevent the harms the Bureau has identified. As the Bureau noted in the proposal, above in Market Concerns—Underwriting and the section-by-section analysis for § 1041.4, and below in the Section 1022(b)(2) Analysis, the regulatory frameworks in most States do not appear to have had a significant impact on reducing re-borrowing and other harms that confront consumers of short-term loans. For example, the Bureau’s evidence shows that 24- and 48-hour cooling-off periods have a minimal impact on overall re-borrowing rates.797 As noted in the proposal, the Bureau studied re-borrowing rates from 2010-2011 in most of the States noted by commenters and found that, generally, over 80 percent of loans were re-borrowed regardless of the type of State restriction studied. This evidence suggests that the laws in those States at that time had not meaningfully prevented re-borrowing. Commenters have not rebutted these findings directly. Some instead challenge the premise that re-borrowing is an indicator of consumer harms. The Bureau addresses that issue above in Market Concerns—Underwriting and the section-by-section analysis for § 1041.4. Thus, the Bureau continues to believe that there is a need to adopt minimum Federal standards that apply consistently across all of these States. In setting the parameters of this final rule, the Bureau sought to prevent the harms identified in § 1041.4 from continuing. For that reason, the Bureau declines to exempt entities operating in any given State on the basis of the given State’s laws. The Bureau recognizes that States may wish to prevent more harms than are 797 CFPB Supplemental Findings, at 100-109. 623 prevented by this rule, and they are free to do so because, as noted earlier, this rule should be considered a floor and not a ceiling. See part IV (discussing preemption under the Dodd-Frank Act and noting that State usury caps are an example of State consumer protections that may extend beyond the floor of Federal law). Other alternatives. The Bureau does not believe that any of the other posited alternative approaches to regulating covered short-term or longer-term balloon-payment loans would be less onerous than, but as effective as, an ability-to-repay requirement. As noted in part II and Market Concerns—Underwriting sections and discussed at some length in the proposal, about 18 States require payday lenders to offer repayment plans to borrowers who encounter difficulty in repaying payday loans. The usage rate of these repayment plans varies widely, but in all cases it is relatively low.798 The Bureau believes the low take-up rate on these repayment plans may be 798 Washington permits borrowers to request a no-cost installment repayment schedule prior to default. In 2014, 14 percent of payday loans were converted to installment loans. Wash. Dep’t of Fin. Insts., “2014 Payday Lending Report,” at 7 (2014), available at http://www.dfi.wa.gov/sites/default/files/reports/2014-payday-lending-report.pdf. Illinois allows payday loan borrowers to request a repayment plan with 26 days after default. Between 2006 and 2013, the total number of repayment plans requested was less than 1 percent of the total number of loans made in the same period. Ill. Dep’t. of Fin. & Prof. Reg., “Illinois Trends Report All Consumer Loan Products Through December 2015,” at 19 (Apr. 14, 2016), available at http://www.idfpr.com/DFI/CCD/pdfs/IL_Trends_Report%202015-%20FINAL.pdf?ActID=1204&ChapterID=20. In Colorado, in 2009, 21 percent of eligible loans were converted to repayment plans before statutory changes repealed the repayment plan. State of Colorado, Dep’t of Law, Office of the Att’y Gen., “2009 Deferred Deposit Lenders Annual Report,” at 2 (2009), available at http://www.coloradoattorneygeneral.gov/sites/default/files/contentuploads/cp/ConsumerCreditUnit/UCCC/AnnualR eportComposites/2009_ddl_composite.pdf. In Utah, 6 percent of borrowers entered into an extended payment plan. G. Edward Leary, Comm’r of Fin. Insts. for the State of Utah to Hon. Gary R. Herbert, Governor, and the Legislature, (Report of the Commissioner of Financial Institutions for the Period July 1, 2013 to June 30, 2014), at 135, (Oct. 2, 2014), available at http://dfi.utah.gov/wp-content/uploads/sites/29/2015/06/Annual1.pdf. Florida law also requires lenders to extend the loan term on the outstanding loan by 60 days at no additional cost for borrowers who indicate that they are unable to repay the loan when due and agree to attend credit counseling. Although 84 percent of loans were made to borrowers with 7 or more loans in 2014, fewer than 0.5 percent of all loans were granted a cost-free term extension. See Brandon Coleman & Delvin Davis, “Perfect Storm: Payday Lenders Harm Florida Consumer Despite State Law,” at 4 (Ctr. for Responsible Lending, 2016), available at http://www.responsiblelending.org/sites/default/files/nodes/files/researchpublication/crl_perfect_storm_florida_mar2016_0.pdf. 624 due to lenders discouraging use of the plans or failing to promote their availability.799 At the very least, a rule that required only that lenders offer extended repayment plans would create significant evasion risk absent more complex provisions to try to prevent lenders from discouraging the use of repayment plans in order to make it more likely that such consumers will instead re-borrow. The Bureau is aware, from confidential information gathered in the course of statutory functions, that one or more payday lenders train their employees not to mention repayment plans until after the employees have offered renewals, and then only to mention repayment plans if borrowers specifically ask about them. Another alternative posited by commenters was increased or sustained enforcement attention focusing on the worst market actors, or focused on specific sub-markets like unregulated or offshore online lenders. As noted in part III, the Bureau has already engaged in extensive enforcement and supervisory activity in this market focused on a wide variety of practices. But, as noted in Market Concerns—Underwriting, the identified unfair and abusive practice in § 1041.4 is a market-wide practice. Continued enforcement and supervisory activity focused on the worst actors would simply not prevent the market-wide harms identified by the Bureau. In addition, the Bureau is sometimes criticized for “regulation through enforcement.” Thus, while the Bureau could bring enforcement actions against individual lenders for engaging in the practices identified here as unfair and abusive, the Bureau believes that it provides more 799 Colorado’s 2009 annual report of payday loan activity noted lenders’ self-reporting of practices to restrict borrowers from obtaining the number of loans needed to be eligible for a repayment plan or imposing cooling-off periods on borrowers who elect to take a repayment plan. State of Colorado, Dep’t of Law, Office of the Att’y Gen., “2009 Deferred Deposit Lenders Annual Report,” at 2 (2009), available at http://www.coloradoattorneygeneral.gov/sites/default/files/contentuploads/cp/ConsumerCreditUnit/UCCC/AnnualR eportComposites/2009_ddl_composite.pdf. This evidence was from Colorado under the state’s 2007 statute, which required lenders to offer borrowers a no-cost repayment plan after the third balloon loan. The law was changed in 2010 to prohibit balloon loans, as discussed in part II. 625 consistent protection for consumers and compliance guidance for industry to address marketwide harms through a detailed rulemaking that both defines the unfair and abusive practice, carefully outlines affirmative standards to prevent that practice, and provides a reasonable period for lenders to come into compliance with those standards. With regard to implementing a nationwide licensing and registration system, the Bureau has authority under the Dodd-Frank Act to prescribe rules regarding registration requirements applicable to covered persons, including those covered by this rule. The Bureau also has authority under 12 U.S.C. 5514(b)(7)(C) to prescribe rules to ensure that lenders under the Bureau’s nonbank supervision authority are legitimate entities and are able to perform their obligations to consumers, including by requiring background checks and bonding. Indeed, the Bureau has noted in its recent semi-annual regulatory agendas that it is evaluating stakeholder suggestions about creating such a system for these markets.800 But while such an action may assist with enforcement and supervision efforts (discussed above) and provide a better means of identifying lenders operating without State lending licenses, the Bureau does not believe that it would be effective in lieu of ability-to-repay requirements at remedying the identifying harms. A well-bonded lender with officers with a clean record, which is registered, would still be able to cause all of the identified harms noted in Market Concerns—Underwriting and the section-bysection analysis for § 1041.4 unless the Bureau took more substantive action (like adopting this rule). In response to the comment urging the Bureau to forgo rulemaking and instead focus on consumer education initiatives, the Bureau does not find that this would be a viable option for 800 See, e.g., 82 FR 40386, 40387 (Aug. 24, 2017). 626 significantly reducing the observed harms. While financial education is an important pillar of the Bureau’s work, and it will continue those efforts, it does not believe that its financial education efforts would impact saving rates broadly enough to have a substantial impact on the need to borrow to cover cash shortfalls across all consumers. Nor does the Bureau believe that generalized financial education, even if it succeeded in reaching all would-be-borrowers, could enable consumers to accurately predict their own likelihood of re-borrowing or defaulting. The Bureau recognizes that there will continue to be demand for credit from consumers who lack the ability to repay covered short-term or longer-term balloon-payment loans. See the discussion in the section-by-section analysis for § 1041.4 regarding substitution to alternative products. Fair lending. The Bureau expects that certain of the burden-reducing changes to the final rule will also address commenters’ concerns relating to fair lending. For example, under the final rule, when a reliable record to verify income is not reasonably available, a lender may now rely on a consumer’s statement of net income, provided such reliance is reasonable (see discussion of § 1041.5(c)(2)(ii)(A) and comment 5(c)(2)(ii)(A)-3 and -4, below). This change should reduce concern that members of protected classes would be denied access to credit solely because of the difficulty in verifying their income. Additionally, unlike the proposed rule, the final rule permits lenders to include in the consumer’s net income any income of another person to which the consumer has a reasonable expectation of access if the consumer documents that he or she has regular, verifiable access to such income (see 1041.5(a)(5) and comment 5(a)(5)-3). In the final rule, the lender is also permitted to rely on the consumer’s statement for rental housing expenses, provided such reliance is reasonable; this is a change from the proposal, which would have required a projection of rental housing expense using a reliable record or an estimate 627 based on survey or other data with respect to the consumer’s neighborhood (see § 1041.5(c)(2)(iii) and associated commentary). More generally, the Bureau notes that inquiries relating to dependents for purposes of estimating basic living expenses can be made consistent with Regulation B.801 For the foregoing reasons, the Bureau believes that the final rule is consistent with the requirements of ECOA and Regulation B. 5(a) Definitions Proposed § 1041.5(a) would have provided definitions of several terms used in proposed §§ 1041.5 and 1041.6. Virtually identical definitions and commentary appeared in proposed § 1041.9(a) for covered longer-term loans (including covered longer-term balloon-payment loans), with minor adjustments to account for the difference in the term of the products. In the final rule, the Bureau has revised several of the six proposed definitions for substance or clarity, made them applicable to both covered short-term loans and covered longer-term balloonpayment loans, and has added two more definitions in part to effectuate the new underwriting methodology based on debt-to-income ratio. A discussion of the proposed definitions, the comments received on those definitions, and the final definitions follows. 5(a)(1) Basic Living Expenses Proposed Rule Proposed § 1041.5(a)(1) would have defined basic living expenses as a component of the ability-to-repay determination as established in the proposed rule. The Bureau proposed to define basic living expenses as expenditures, other than payments for major financial obligations, 801 12 CFR 1002.5(d)(3) (“A creditor may inquire about the number and ages of an applicant’s dependents or about dependent-related financial obligations or expenditures, provided such information is requested without regard to sex, marital status, or other prohibited basis.”). 628 which a consumer makes for goods and services necessary to maintain the consumer’s health, welfare, and ability to produce income, and the health and welfare of members of the consumer’s household who are financially dependent on the consumer. Accordingly, the proposed definition of basic living expenses was a principle-based definition and did not provide a comprehensive list of all the expenses for which a lender must account. Proposed comment 5(a)(1)-1 provided illustrative examples of expenses that would be covered by the definition. It provided food and utilities as examples of goods and services that are necessary for maintaining health and welfare, and transportation to and from a place of employment and daycare for dependent children as examples of goods and services that are necessary for maintaining the ability to produce income. Proposed comment 5(b)-2.i.C would have clarified that as part of the reasonable abilityto-repay determination, the lender’s estimates of basic living expenses must be reasonable. Proposed comment 5(b)-4 would have provided examples of approaches to estimating basic living expenses that were reasonable or unreasonable. For discussion of how the final rule addresses the reasonableness of lender estimates of basic living expenses, see the section-bysection analysis of § 1041.5(b), where the commentary provisions relating to basic living expenses have been revised, as well as the immediately following discussion. The Bureau’s proposed definition gave lenders some flexibility in how lenders determine dollar amounts that meet the proposed definition, provided they do not rely on amounts that are so low that they are unreasonable for consumers to pay for the types and levels of expenses provided in the definition. The Bureau specifically noted in the proposal that a lender would not be required to verify or conduct a detailed analysis of every individual consumer expenditure. In contrast to major financial obligations, the Bureau explained that recent expenditures might not 629 reflect the amounts a consumer needs for basic living expenses during the term of a prospective loan. The Bureau expressed concern that such a requirement could substantially increase costs for lenders and consumers while adding little protection for consumers. The Bureau sought comment in the proposal on whether an alternative formulation focusing on expenses that are of the types that are likely to recur through the term of the loan and in amounts below which a consumer cannot realistically reduce them would be preferable; the Bureau had used similar concepts to define which expenses should be treated as major financial obligations as discussed further below in connection with § 1041.5(a)(3). The Bureau also sought comment on whether there are standards in other contexts that can be relied upon by the Bureau. The Bureau explained in the proposal that, for example, it was aware that the Internal Revenue Service (IRS) and bankruptcy courts have their own respective standards for calculating amounts an individual needs for expenses while making payments toward a tax delinquency or bankruptcy-related repayment plans. Comments Received The Bureau received many comments on the proposed definition of basic living expenses from a variety of stakeholders. In general, industry commenters criticized the proposed definition as overly vague and argued it would create uncertainty for lenders trying to comply with the proposed rule. A number of industry commenters asked for the Bureau to provide additional clarity on the definition. Some, including a trade association for payday lenders, suggested the Bureau include safe harbor amounts for basic living expenses due to the costs of having to establish a framework to estimate such expenses, particularly for smaller lenders. Some commenters argued that the standards were so vague that different lenders in good faith 630 could apply different definitions. One State Attorney General expressed concern that the vagueness in the proposed definition would lead to inconsistent interpretation of the rule. Industry commenters also raised a number of more discrete issues with the proposed definition. Some argued that the Bureau should let lenders assume that consumers could cut back on discretionary spending on items like restaurant meals, gym memberships, and the like, and that the proposed rule was not clear whether those types of expenses were included in the definition and whether lenders could assume that consumers would undertake some reductions in spending on those items for purposes of the basic living expenses estimates. Another commenter noted that the Bureau had not taken account of the fact that prices may change seasonally (as with back-to-school sales). Several commenters criticized the definition for including expenses for the health and welfare of the consumer’s dependents when, they argued, consumers may have spouses or other persons paying a portion of the household expenses, including those of dependents. (These issues are noted above in the discussion of general comments regarding ECOA and Regulation B.) They argued that the definition should be modified to account for such sharing of expenses. Most consumer advocates commenting on the rule expressed support for the concept of lenders having to estimate basic living expenses, but argued that the definition was underinclusive. For example, they questioned why the Bureau only included four examples of specific expenses. They also expressed support for including within the definition any expense that is likely to recur. They also criticized what they viewed as too permissive provisions in commentary regarding reasonable estimates of basic living expenses. Some of these commenters suggested specific expenses that should be explicitly added to the definition, such as alimony, 631 health insurance premiums other than those deducted from a consumer’s paycheck, cell phone payments, car insurance payments, and a number of other categories. Another suggestion was to change the definition to include typical expenses based on geography, income, and household size. In contrast, one organization generally supportive of the rule criticized the approach on this element of the financial analysis and argued that lenders should be expected to itemize basic living expenses because of the risk that estimates would be too low. The Bureau notes that it is responding to this comment in the discussion below of comment 5(b)-2.i.C.1. A public policy and research organization argued that childcare expenses, including diaper costs for new parents, could consume a large percentage of a consumer’s budget and therefore should be treated not as a basic living expense but as a major financial obligation to be verified. Several commenters urged the Bureau to use the IRS Collection Financial Standards to define the ambit of basic living expenses. They argued that the proposed definition was too ambiguous and could lead to confusion and potentially lender evasion; they argued that the IRS Collection Financial Standards would provide needed clarity for all parties involved. A lender commenter, a SER, took a different view, arguing that the IRS Collection Financial Standards should not be used for either estimating basic living expenses or rental housing and citing the average housing cost in Orange County, California, as an example of the Standards being “unrealistic.” Final Rule The Bureau has decided to finalize the proposal’s framing of the definition of basic living expenses as expenses that are “necessary” to maintain the consumer’s health, welfare, and ability 632 to produce income and the health and welfare of the members of the consumer’s household who are financially dependent on the consumer. As such, the regulatory text is being finalized with only minor wording changes from the proposal for clarity. However, the Bureau in response to comments is making a number of modifications to the commentary clarifying the definition, as described in more detail below. The Bureau concludes that the conceptual framework of the proposal remains the appropriate formulation for defining basic living expenses. The “necessary to maintain” language in the proposed definition is adapted largely from the IRS Collection Financial Standards, which set forth necessary expenses for repayment of tax delinquencies by taxpayers.802 The Bureau considered finalizing the alternative formulation on which it had sought comment (i.e., personal and household goods and services that are likely to recur and that are types of expenditures that the consumer cannot reasonably be expected to reduce or forgo during the term of the loan). However, while the focus on recurring obligations has been helpful in defining major financial obligations as discussed below, the Bureau is concerned about the complexity that would result from trying to differentiate recurring from non-recurring expenses and reducible from non-reducible expenses when it comes to more discretionary expenditures. 802 Internal Revenue Servs., “Collection Financial Standards,” https://www.irs.gov/businesses/small-businesses-selfemployed/collection-financial-standards (last revised Mar. 27, 2017) (providing that “Collection Financial Standards are used to help determine a taxpayer’s ability to pay a delinquent tax liability. Allowable living expenses include those expenses that meet the necessary expense test. The necessary expense test is defined as expenses that are necessary to provide for a taxpayer’s (and his or her family’s) health and welfare and/or production of income.”). The IRS Collection Financial Standards contain Local Standards for transportation expenses and housing expenses and utilities, and National Standards for other categories, such as food, clothing, out-of-pocket medical expenses, and miscellaneous items. The National and Local Standards are tied to different data sources, including the Bureau of Labor Statistics Consumer Expenditure Survey, U.S. Census Data, and the American Community Survey. The Standards are updated periodically. Both the categories and the amounts provided as estimates are found on the IRS website. 633 To give an example, newspaper and magazine subscriptions and health club memberships are not typically thought of as necessary expenses, but they generally are recurring. And whether such expenses are reducible during the term of the loan generally and the relevant monthly period that is the focus of the residual-income or debt-to-income analysis in particular may depend on such factors as the term of the relevant contracts (for both the loan and the product or service), the method by which payments are made (e.g., automatic debit versus monthly bill pay), and the applicable termination policies and penalties (e.g., advance notice of termination). The Bureau also is not aware of data sources that categorize the types and amounts of recurring expenses as distinguished from non-recurring expenses, in contrast to the “necessary” expense formulation which as noted above is derived from the IRS Collection Financial Standards. With regard to the commentary to § 1041.5(a)(1), the Bureau revised comment 5(a)(1)-1 and created a new comment 5(a)(1)-2. The revised comment 5(a)(1)-1 clarifies that estimating basic living expenses is part of the broader ability-to-repay determination under § 1041.5(b). The comment also clarifies that a lender may make a reasonable estimate of basic living expenses without making an individualized determination and includes a cross-reference to comment 5(b)-2.i.C. With regard to the amounts of basic living expenses, comment 5(b)-2.i.C has been revised in a number of ways to provide more guidance on how to reasonably estimate basic living expenses. Those changes are described below in the discussion of § 1041.5(b) and are to be read in tandem with the changes to commentary for the definition of basic living expenses in § 1041.5(a)(1). Comment 5(a)(1)-2 expands the examples of basic living expenses described in the proposal with some additional clarification to six items, which are: (1) food, (2) utilities not paid 634 as part of rental housing expenses, (3) transportation, (4) childcare, (5) phone and Internet service, and (6) out-of-pocket medical expenses (which would include insurance premiums to the extent not deducted from consumer’s paychecks as well as co-pays, prescriptions, and similar expenses). The comment also includes new language clarifying that basic living expenses do not include expenditures for discretionary personal and household goods or services and gives examples of newspaper subscriptions and vacation activities. Additionally, comment 5(a)(1)-2 notes that if the consumer is responsible for payment of household goods and services on behalf of the consumer’s dependents, those expenditures are included in basic living expenses. The comment further clarifies that the lender may reasonably consider whether another person is regularly contributing toward the consumer’s payment of basic living expenses when conducting a reasonable ability-to-repay determination (with a cross-reference to comment 5(b)-2.i.C.2). The Bureau agrees with the commenters who suggested that, when a lender estimates basic living expenses on an individualized basis, the Bureau should permit lenders to take this fact into account given that the proposed definition of basic living expenses included members of the consumer’s household who are financially dependent on the consumer. The inclusion of additional examples of basic living expenses in comment 5(a)(1)-2 and the new language describing examples of items that are not included in the definition are in response to comments asking for more specificity on what expenses are included in and what are excluded from the definition of basic living expenses. Commenters had specifically asked about the status of the items now addressed. The categories of out-of-pocket medical expenses and phone and Internet service have been added in view of comments urging the Bureau either to clarify the status of the items or to include them because of the view by the commenters that they 635 are necessary expenses. The category of utility payments also has been clarified to note that it includes utilities not paid as part of rental housing expense, in response to interagency comments from a Federal prudential regulator. The example of transportation as a basic living expense also has been broadened from the proposal, which included transportation to work as an example. The Bureau finds that transportation expenses for both personal and household use and for work is more consistent with the notion of “necessary” expenses for health, welfare, and the ability to work. The Bureau concludes that the six categories of expenses provided as examples are sufficient for estimating basic living expenses. To this end, the Bureau has included language in comment 5(b)-2.i.C.1 clarifying that a lender is not required to itemize the basic living expenses of each consumer but may instead arrive at estimates for the amount needed to cover the costs of food, utilities not paid as part of rental housing expenses, transportation, out-of-pocket medical expenses, phone and Internet services, and childcare. The comment also clarifies it would be reasonable for the lender to use data about these expenses from the Consumer Expenditure Survey of the Bureau of Labor Statistics or the IRS Collection Financial Standards, or a combination of the two data sources, to develop non-individualized estimates of basic living expenses for consumers seeking covered short-term or longer-term balloon-payment loans. The comment also clarifies that in using the data from those sources to estimate the amount spent on a particular category, the lender may make reasonable adjustments to arrive at an estimate of basic living expenses, for instance where a data source’s information on a particular type of basic living expenses overlaps with a type of major financial obligation as defined in § 1041.5(a)(3). 636 More explanation of the comment is provided in the section-by-section analysis for § 1041.5(b)(1), below. With regard to the comments requesting that the Bureau should provide safe harbor categories and amounts for basic living expenses, the Bureau believes that the IRS Collection Financial Standards are a useful source for developing estimates of basic living expenses. As explained earlier, the “necessary” expense concept at the heart of the definition in § 1041.5(a) is derived from the Standards. Lenders can use the Standards to estimate both the amounts and categories of expenses, and the Bureau would view such an approach as reasonable. As described above, comment 5(b)-2.i.C.1 now contains language recognizing that fact. At the same time, the Bureau recognizes that lenders may well want to make reasonable adjustments from that framework. The Bureau believes that in some cases the Standards may capture expenses that would not be relevant for a lender making a basic living expenses estimate for the relevant monthly period, which is the calendar month with the highest payments on the loan.803 And there also is overlap between some of the categories provided in the Standards and the items deemed in this rule as major financial obligations (such as automobile lease payments). A direct application of the Standards thus in some cases may create operational difficulty or result in an over-inclusive estimate for purposes of what is required under § 1041.5.804 803 For example, a consumer might not have transportation expenses such as licenses, registration, and maintenance, but the consumer presumably will have gas costs if she owns a car. 804 Regarding the comment by a SER who argued that using the IRS Collection Financial Standards would be unrealistic for estimating basic living expenses and rental housing expenses, as discussed below the Bureau is clarifying in comment 5(b)-2.i.C.1 that it would be reasonable to use the Standards to estimate the amounts or categories of basic living expenses, but the Bureau is not requiring use of the Standards, and the Bureau expects that lenders would have to make adjustments if they do use them. Moreover, the final rule no longer requires the lender to estimate housing expenses based on locality-based data. See § 1041.5(c)(2)(iii). 637 The Bureau also considered whether to use the Consumer Expenditure Survey of the Bureau of Labor Statistics (CEX) as a safe harbor. Like the Standards, the CEX is a useful source for information about consumers’ household expenditures which could inform estimates of basic living expenses. As with the IRS Collection Financial Standards, comment 5(b)-2.i.C.1 now clarifies that use of the CEX would be a reasonable method for estimating the categories and amounts of basic living expenses. However, because the CEX collects data at the household level, not the individual consumer level, and because of how it groups the categories of expenses, it too may be over-inclusive as to the amounts of the expenses, depending on whether the consumer has dependents or not match precisely the list of categories in comment 5(a)(1)-2. Put another way, the Bureau views both data sources as reliable and useful for the purposes of estimating various categories of basic living expenses, and believes it would be reasonable for lenders to draw on one or both of them or on their own experience (or on a combination of the lenders’ experience and these extrinsic data sources). But since the IRS Collection Financial Standards and the CEX each may be potentially over-inclusive or not match precisely the list of categories in comment 5(a)(1)-2, the Bureau expects that most lenders who use those sources will choose to make some reasonable adjustments or turn to supplemental sources. The Bureau finds that, cumulatively, the changes to comments 5(a)(1)-2 and 5(b)-2.i.C.1 described above will address commenters’ concerns. To recap, the commentary now contains: (1) additional examples of expense categories that are included in the definition; (2) clarification that it would suffice for lenders to estimate the six categories of expenses described in comment 5(a)(1)-2; (3) clarification around what is excluded from the definition; (4) new commentary 638 language clarifying that use of particular government data sources (IRS Collection Financial Standards and/or CEX) would be reasonable methods of estimating expenses; and (5) commentary explaining that lenders have flexibility to make adjustments based on the lender’s experiences and for other reasonable considerations. The Bureau recognizes that estimating basic living expenses will involve some complexity and burden, particularly initially while lenders are developing a system to comply with the rule’s requirements. (This is discussed in the Section 1022(b)(2) Analysis.) The Bureau does expect that, at least in some cases, service providers would be positioned to provide software to permit lenders to develop this capability. Indeed, some commenters appear to have developed their own methodologies in the course of researching the affected markets and commenting on the proposal. The Bureau does not want to unduly restrict the flexibility of lenders and service providers to develop innovative methods of estimating basic living expenses, which a more prescriptive approach might do. More generally, the Bureau emphasizes that at bottom the question will be whether the lender is acting reasonably in developing the estimates. The rule gives lenders substantial flexibility to develop estimates by consulting reliable data sources or developing reasonable estimates based on their own experience with similarly-situated consumers using at least the six categories of expenses provided as examples. Assuming a lender follows these procedural steps, the Bureau concludes that the strongest evidence of whether the estimations were in fact reasonable will be the performance of the loans in question; if a lender is consistently making unreasonable estimates of basic living expenses, the Bureau expects to see substantial reborrowing and default activity. 639 In response to one commenter, the Bureau has declined to modify the commentary to address specifically whether a lender should take account of the fact that prices may change seasonally. The Bureau finds this to be adequately covered by the general reasonableness standard, such that a lender could choose to do so if there were reason to believe, for example, that the monthly averages that a lender is using in estimating basic living expenses are not representative of expenses during a particular term. At least in certain regions, a lender could make a reasonable determination based on historical and local trends that the estimated expense allocation for utilities declines in the spring and fall, when electricity and gas bills are lower. The Bureau does not agree with the commenters who argued childcare expenses (including the costs of supplies for infant children) should not be basic living expenses and instead should be defined as major financial obligations and subject to verification. The Bureau believes that childcare expenses, particularly to the extent of including such items as diapers, could be difficult to verify and would not lend themselves to categorization as major financial obligations for which the primary source of verification is consumer reports from a nationwide consumer reporting agency. Therefore, the Bureau concludes that these expenses are better categorized as basic living expenses. The Bureau has determined that the changes to the basic living expenses definition described above, along with revisions to comment 5(b)-2.i.C described below, appropriately balance the weight of the comments. The Bureau acknowledges that it has left some flexibility in the definition, but believes this flexibility will permit lenders to develop methodologies that work best for them consistent with the requirement that the estimates are reasonable. 5(a)(2) Debt-to-Income Ratio 640 The Bureau has added a new definition at § 1041.5(a)(2) for debt-to-income ratio in light of its decision, in response to the criticisms of the proposed residual income approach, to permit lenders to choose to use that underwriting methodology. Due to the addition of this new definition, the remaining subsections of § 1041.5(a) are renumbered accordingly. The final rule defines debt-to-income ratio as the ratio, expressed as a percentage, of the sum of the amounts that the lender projects will be payable by the consumer for major financial obligations during the relevant monthly period and the payments under the covered short-term loan or covered longer-term balloon-payment loan during the relevant monthly period, to the net income that the lender projects the consumer will receive during the relevant monthly period, all of which projected amounts are determined in accordance with § 1041.5(c). The Bureau has also added a definition for relevant monthly period in § 1041.5(a)(7), which consists of the calendar month in which the highest sum of payments under the loan is due. The section-by-section analysis for § 1041.5(a)(7) below describes why the Bureau chose this particular time period as the relevant monthly period. The Bureau has added a new comment 5(a)(2)-1 to clarify aspects of the debt-to-income definition. Most notably, the comment clarifies that for covered longer-term balloon-payment loans, where the relevant monthly period may fall well into the future relative to the consummation of the loan, the lender must calculate the debt-to-income ratio using the projections made under § 1041.5(c) and in so doing must make reasonable assumptions about the consumer’s net income and major financial obligations during the relevant monthly period compared to the period covered by the verification evidence. The comment clarifies that, for example, the lender cannot assume, absent a reasonable basis, that there will be a substantial 641 increase in net income or decrease in major financial obligations between consummation and the relevant monthly period. The addition of this new definition ties to the broader revision of § 1041.5(b)(2) in the final rule. The changes to § 1041.5(b)(2) are described in more detail in the associated sectionby-section analysis below, but they bear some mention here given the interplay. As noted in the general § 1041.5 discussion above, under proposed § 1041.5(b)(2), the reasonable ability-torepay determination would have required the lender to project both the amount and timing of the consumer’s net income and major financial obligations, and to analyze the consumer’s finances during two distinct time periods: first for the shorter of the term of the loan or 45 days after consummation of the loan, and then also for 30 days after having made the highest payment under the loan. For covered longer-term loans (including covered longer-term balloon-payment loans), the two periods would have been the month with the highest payments on the loan and also for 30 days after having make the single highest payment on that loan. Upon further consideration of comments concerning the burdens involved in the proposed residual-income analysis and other factors, the Bureau has decided to streamline the calculations needed to support lenders’ determination of consumers’ ability to repay. Accordingly, the final rule simply requires lenders to make a projection about net income and major financial obligations and calculate the debt-to-income ratio or residual income, as applicable, during only a single monthly period, i.e., the relevant monthly period, which is the calendar month with the highest sum of payments on the loan. The debt-to-income ratio during this period is used as a snapshot of the consumer’s financial picture to draw conclusions about the consumer’s ability to pay, since it is the month in which the loan will cause the highest 642 amount of financial strain. Specifically, under § 1041.5(b)(2), the lender uses this information to reach a reasonable conclusion about whether the consumer has the ability to repay the loan while meeting basic living expenses and major financial obligations during: (1) the shorter of the term of the loan or 45 days after consummation of the loan, for covered short-term loans, and the relevant monthly period, for covered longer-term loans, and (2) for 30 days after having made the highest payment under the loan. This simplified approach—which also has been incorporated into the definition of residual income in § 1041.5(a)(8) for purposes of making the standards for both alternatives consistent—dovetails with the inclusion of the debt-to-income ratio methodology as an alternative to residual income. As discussed above, a debt-to-income methodology does not track a consumer’s individual income inflows and major financial obligation outflows on a continuous basis over a period of time. Section 1041.5(b) requires that lenders using debt-to-income ratios leave a sufficiently large percentage of income to cover basic living expenses. Commentary to § 1041.5(b) elaborates on this reasonableness standard in more detail. Comment 5(b)-2.ii.B clarifies that it would be unreasonable for the lender to assume that the consumer needs an implausibly low percentage of income to meet basic living expenses. The comment also clarifies in an example that a 90 percent debt-to-income ratio would leave an implausibly low percentage of income to meet basic living expenses. The Bureau does not intend to require lenders to set individualized thresholds for each consumer; instead, a lender could set its own internal thresholds in its policies and procedures, which would then be applied to individual loan applications. Whether a lender would be able to rely on one debt-to-income threshold for all borrowers, or enact multiple thresholds based on income tiers or other characteristics, would depend on whether application 643 of a single threshold or multiple thresholds resulted in reasonable ability-to-repay determinations in the run of cases, informed in part by the factors listed in comment 5(b)-2.iii. Lenders using a debt-to-income ratio will, in essence, be taking an individualized accounting of the consumer’s projected net income and major financial obligations within the relevant monthly period, which is the month in which a consumer will have to pay the most under the covered short-term or longer-term balloon-payment loan. The snapshot provided by the debt-to-income ratio, coupled with the lender’s estimate of the consumer’s basic living expenses during the relevant monthly period, will enable the lender to draw a reasonable conclusion about whether the consumer will be able to make payments for major financial obligations, make all payments under a covered longer-term balloon-payment loan, and meet basic living expenses during the loan term or 45 days following consummation (for covered short-term loans) or the relevant monthly period (for covered longer-term balloon-payment loans) and for 30 days after making the highest payment under the loan. This accounting of the consumer’s financial picture using a debt-to-income ratio is less granular than the proposed residual-income methodology, which would have required lenders to track the timing and amounts of net income and major financial obligations, and to analyze the consumer’s finances for two separate periods in proposed § 1041.5(b)(2). The Bureau had expressed concern in the proposal that a debt-to-income approach might be problematic in the context of the market for covered loans, due to the lack of long-established debt-to-income norms in this market, and noted that debt-to-income ratios which might seem quite reasonable for an “average” consumer might be quite unmanageable for a consumer at the lower end of the income spectrum and higher end of the debt burden range. Upon further consideration of the 644 comments focused on the complexity and burdens of the proposal, the Bureau concludes that it is appropriate to move to a simplified analysis that concentrates on the total inflows and outflows for the month in which the loan places the most financial strain on the consumer. In light of this change, the Bureau expects that lenders may be able to use either a debt-to-income ratio or a residual-income analysis, as long as they think carefully about the need for consumers to cover basic living expenses. For instance, lenders using a debt-to-income analysis may decide to set a more conservative ratio than lenders might use in other markets to account for the financial profiles of consumers in the markets for covered short-term loans or covered longer-term balloon-payment loans. Another option as referenced above may be to use different ratios for different subgroups of customers to account for differences in income, debt obligations, and other relevant factors.805 As described below, in the discussion of § 1041.5(b), the Bureau has not set particular debt-to-income ratios for lenders to use. As with other aspects of the ability-to-repay requirements, lenders would be expected to be reasonable. Section 1041.5(b) commentary, as described below, has been revised extensively to include additional clarification and examples of the reasonableness of ability-to-repay determinations, in response to many comments urging the Bureau to provide additional clarity. See discussion of 1041.5(b) for further elaboration. 5(a)(3) Major Financial Obligations Proposed Rule 805 The Bureau recognizes that the particular debt-to-income ratio approach in the final rule, while drawing inspiration from and sharing some similarities with the standard in credit card underwriting rules, has differences which the Bureau finds are justifiable as described in the general discussion of § 1041.5. 645 The Bureau proposed a definition for major financial obligations as a component of the ability-to-repay determination specified in proposed § 1041.5(b). Specifically, proposed § 1041.5(a)(2) would have defined the term to mean a consumer’s housing expense, minimum payments and any delinquent amounts due under debt obligations (including outstanding covered loans), and court- or government agency-ordered child support obligations. In comment 5(a)(2)1, the Bureau proposed to further clarify that housing expense includes the total periodic amount that the consumer applying for the loan is responsible for paying, such as the amount the consumer owes to a landlord for rent or to a creditor for a mortgage. It would have provided that minimum payments under debt obligations include periodic payments for automobile loan payments, student loan payments, other covered loan payments, and minimum required credit card payments. The Bureau explained in the proposal that the obligations that it included in the proposed definition were obligations that are typically recurring; that can be significant in the amount of a consumer’s income that they consume; and that a consumer has little or no ability to change, reduce, or eliminate in the short run, relative to their levels up until application for a covered short-term or longer-term balloon payment loan. The Bureau stated its belief that the extent to which a particular consumer’s net income is already committed to making such payments was highly relevant to determining whether that consumer has the ability to make payments under a prospective covered short-term loan. As a result, the Bureau believed that a lender should be required to inquire about such payments, that they should be subject to verification for accuracy and completeness to the extent feasible, and that a lender should not be permitted to rely on income already committed to such payments in determining the consumer’s ability to repay. The 646 Bureau further elaborated in the proposal that obligations included in the proposed definition are roughly analogous to those included in total monthly debt obligations for calculating monthly debt-to-income ratio and monthly residual income under the Bureau’s ability-to-repay requirements for certain residential mortgage loans, citing 12 CFR 1026.43(c)(7)(i)(A). In the proposal, the Bureau noted that it had adjusted its approach to major financial obligations based on feedback from SERs and other industry stakeholders in the Small Business Review Panel Outline. In the SBREFA process, the Bureau stated that it was considering including within the category of major financial obligations “other legally required payments,” such as alimony, and had considered an alternative approach that would have included utility payments and regular medical expenses. However, the Bureau noted in the proposal that it believed that it would be unduly burdensome to require lenders to make individualized projections of a consumer’s utility or medical expenses. With respect to alimony, the Bureau noted its belief that relatively few consumers seeking covered loans have readily verifiable alimony obligations and that, accordingly, inquiring about alimony obligations would impose unnecessary burden. The Bureau also noted that it did not include a category of “other legally required payments” because it believed that category, which was included in the Small Business Review Panel Outline, would leave too much ambiguity about what other payments are covered. The Bureau sought comment on whether to include alimony as a major financial obligation, as well as regarding other expenses such as telecommunication services. Comments Received The Bureau received a number of comments on its definition of major financial obligations. Some commenters argued that the proposal did not do enough to clarify the scope of 647 obligations that are included in major financial obligations. For example, commenters questioned whether a medical debt would be included. Consumer advocates and some other commenters urged the Bureau to include additional expenses in the definition, like taxes, childcare, medical expenses, telecommunications services, health insurance premiums, and homeowners insurance. Some commenters, including a Federal financial regulator during interagency consultation, asked for clarification on the treatment of alimony or questioned why it was excluded from the definition while child support obligations were included. Other commenters interpreted the proposed definition to mean that the definition of major financial obligations did not include the payments on non-covered loans and urged the Bureau to include them. Some industry commenters objected to the proposal to include delinquent amounts due on debt obligations in the definition of major financial obligations. They suggested that errors on credit reports often include defaulted debt, like medical debt, which could effectively halt the application process. Some commenters cited a Bureau report on the prevalence of consumers with outstanding medical debt in arguing that the proposal would impede credit access if medical debt was included as a major financial obligation. Others noted that, more generally, given how many consumers have accounts in collections on their credit reports, and the fact that the entire defaulted amount would need to be considered as a major financial obligation, this requirement would result in many consumers failing to demonstrate ability to repay and effectively being excluded from the market. Other commenters took a different view, arguing that delinquent amounts on non-covered loans should be part of the definition of major financial obligations. 648 Some commenters asked the Bureau to pinpoint the exact amount of time after which evidence on major financial obligations would become stale, and how long the calculations for major financial obligations remain valid. Similarly, commenters noted that it will be impossible to detect major financial obligations taken out the same day, or otherwise not reflected on national consumer reports because there is a delay between when a consumer takes out an obligation and when companies furnish to nationwide consumer reporting agencies. Some commenters argued that where basic living expenses or major financial obligations were deducted from a paycheck, they would be deducted twice from residual income because they would count as major financial obligations or basic living expenses but would not be counted in the definition of net income. The commenters cited examples of such “double deductions” where consumers sign up directly for bill-pay from a paycheck or if the deduction is required under State law in connection with payment of child support obligations. The Bureau received a comment suggesting that some of the categories of major financial obligations may not be able to be verified through national consumer reports, including escrowed amounts for property insurance and taxes. More broadly, industry commenters raised concerns about the accuracy of consumer reports, and being held accountable for inaccuracies in them. One commenter, a State trade association, criticized the proposal for not clarifying how lenders should treat debts of non-applicant spouses, as well as their income, in a community property State where debt obligations are considered equally owned and are split equally upon dissolution of the marriage. The commenter argued that the proposed ability-to-repay requirements were significantly flawed because did not take into account the interplay with State 649 community property laws. The commenter requested that the Bureau withdraw the proposal until it had adequately studied the issue. Finally, one consumer suggested that the Bureau’s identification of major financial obligations effectively prioritizes payment of other debts over covered loans. This commenter argued that the Bureau had not provided evidence that these debts were more important than covered loans. Final Rule The Bureau is finalizing the definition of major financial obligation at § 1041.5(a)(3) with certain substantive changes. The most significant change is that the Bureau has revised the reference to debt obligations to focus on “required payments under debt obligations (including, without limitation, outstanding covered loans).” In comment 5(a)(3)-1, the Bureau has provided further clarifications with regard to treatment of debt obligations to address commenters’ concerns about treatment of medical debt and other issues. First, comment 5(a)(3)-1 clarifies that the term “debt obligation” for purpose of § 1041.5(a)(3) does not include amounts due or past due for medical bills, utilities, and other items that are generally defined as basic living expenses under § 1041.5(a)(1). Second, the Bureau has provided a more robust definition of “required payments under debt obligations” drawing largely on language that was contained in proposed comments 5(a)(3)-1 and 5(c)(3)(ii)(B)-1. Third, the Bureau has added language to final comment 5(a)(3)-1 to include delinquent amounts on debt obligations within the concept of “required payments” only to the extent that such delinquent amounts are due as of the relevant monthly period, and not in cases in which an obligation on a covered short-term loan or a covered longer-term balloon-payment loan 650 is no longer outstanding or where the obligation is listed as charged off on a national consumer report. The Bureau has also included an example of a creditor adding delinquent amounts on periodic payments to the consumer’s next regularly scheduled periodic payment for an automobile loan payment. The Bureau believes that these changes cumulatively will address a significant portion of commenters’ concerns, particularly that resolving disputes about medical debts could effectively halt the application process. The Bureau has always intended that major financial obligations and basic living expenses be distinct categories, as evidenced by language in proposed and final § 1041.5(a)(1) defining the latter term, and the Bureau believes this further clarification will be helpful to reinforce the distinction. The Bureau recognizes that because of insurance and other factors, collections on medical bills can pose particular challenges for consumers. The Bureau believes that it may be appropriate for both consumers and lenders to exclude such irregular items from consideration. Because the general intent of the definition of major financial obligations generally is to capture recurring payments, the Bureau believes that a different rule is logical with regard to delinquent amounts on traditional consumer credit products by focusing on those amounts due in the relevant monthly period. The Bureau made a few other changes to § 1041.5(a)(3) and comment 5(a)(3)-1. The Bureau specified in the text of the regulation that required payments under debt obligations could include, but are not limited to, outstanding covered loans, to address the impression expressed by commenters that non-covered loans are not considered debt obligations. The Bureau also added language to the commentary to reflect the new underwriting approach regarding timing—namely that the projections and calculations lenders will need to conduct will be in relation to the 651 relevant monthly period, as defined in § 1041.5(a)(7). And the Bureau has clarified in comment 5(a)(3)-1 that the payments which must be included for a mortgage include principal, interest, and escrow if required. Second, the Bureau has revised § 1041.5(a)(3) to include child support obligations and alimony obligations in general, rather than focusing solely on court- or government agencyordered child support as in the proposal. As described above, at the SBREFA stage the Bureau had contemplated including both types of obligations generally within the definition of major financial obligations, but at the proposal stage decided to focus only on the obligations that were likely to be reflected in a national consumer report due to concerns that requiring lenders to verify other types of alimony or child support would be burdensome. Upon further consideration, the Bureau has concluded that the most reasonable approach is to include both types of expenses generally within the definition, and to permit lenders to rely on the information contained in consumers’ written statements about such obligations to the extent that they are not listed on national consumer reports. The Bureau has added associated regulatory text and commentary to § 1041.5(c) to effectuate this requirement. Finally, the Bureau has added a new comment 5(a)(3)-2 to specify that for purposes of the rule, motor vehicle leases shall be treated as a debt obligation. As explained in the Bureau’s separate rulemaking to define larger participants in the market for automobile financing, automobile leases often function similarly to automobile loans.806 In the Bureau’s experience, they are reported on national consumer reports—and, indeed, are often listed on such reports as 806 See generally 80 FR 37496, 37499 (June 30, 2015) (explaining that certain automobile leases are defined by statute as consumer financial products and services under the Dodd-Frank Act, and using the Bureau’s discretionary authority to define certain additional leasing arrangements as consumer financial products and services). 652 installment loans—and the Bureau believes that it will promote more effective determinations of consumers’ ability to repay a new covered short-term or covered longer-term balloon-payment loan to treat them as the equivalent of an automobile purchase loan. Regarding the comments asking for a broader definition of major financial obligations, the general theory behind the distinction between major financial obligations and basic living expenses under the final rule is that a major financial obligation is something a lender will need to calculate individually and generally to verify, while a lender will not need to do so for basic living expenses. The Bureau’s decision about what to include in the definition of major financial obligations has been influenced in part by considerations of administrability as well as size—all payments on debt obligations are included because they are generally both easily ascertained from a consumer report and tend to be large in amount. The other expenses that commenters recommended the Bureau include, such as childcare expenses, would not be ascertainable from a consumer report. Also, because housing is typically the largest recurring expense and is reflected on a credit report if the consumer has a mortgage, the Bureau thought it prudent for lenders to account specifically for that expense when performing their underwriting rather than including it in basic living expenses more generally. The Bureau does not agree that the definition for major financial obligations should be vaguer and more flexible. It includes rental housing payments and payments on debt obligations. The Bureau has generally provided flexibility in this rule, but where lenders are required to itemize specific obligations, the Bureau concludes that it is more reasonable to prescribe the specific obligations for which the Bureau will expect heightened attention. 653 As to commenters that expressed concerns about duplicative deductions, the Bureau has added comments 5(c)(2)(ii)(B)-2 and 5(c)(2)(ii)(C)-2 to address this issue, both of which clarify the provisions on verification evidence for debt obligations and child support and alimony obligations. The comments provide that if verification evidence shows that a debt obligation or child support or alimony obligation is deducted prior to the receipt of take-home pay, the lender does not include the obligation in the projection of major financial obligations under § 1041.5(c). The Bureau also added an example to comment 5(c)(1)-1 relating to similar facts. With regard to the comment that it would be difficult to verify some debt obligations on national consumer reports, the Bureau understands from its market monitoring that the nationwide consumer reporting agencies do in fact include most debt obligations in their national consumer reports, including payments necessary to cover escrowed items for mortgages. But, to the extent a consumer report does not include a debt obligation, lenders may reasonably rely on the information in the consumer’s written statement. As described in final § 1041.5(c)(1), a lender must consider major financial obligations that are listed in a consumer’s written statement even if they cannot be verified by the required sources. If the national consumer report does not show a consumer’s obligation because it is too recent or is not reported to a nationwide consumer reporting agency, and the consumer’s statement does not include the payment on the obligation in listing major financial obligations, a lender would be reasonable in not accounting for that obligation in the lender’s projection of major financial obligations and its residual income or debt-to-income calculation. Comment 5(c)(2)(ii)(B)-3 provides detailed guidance to lenders about how to reconcile inconsistent 654 information as between a consumer’s written statements and the verification evidence required under § 1041.5(c)(2)(ii)(B). With regard to the commenter writing about State community property laws, the Bureau does not believe there is a fundamental tension between the proposed ability-to-repay requirements and State community property laws and declines the request to withdraw the proposal based on this issue. As an initial response to this comment, the Bureau notes that it has revised the final rule based on other commenters’ input requesting that the final rule account for a consumer’s reasonable expectation of access to spousal or third-party income, as well as the payment by another person of a consumer’s major financial obligations or basic living expenses. Specifically, the Bureau has revised § 1041.5(a)(5), the definition of net income, and other provisions of § 1041.5 to provide that lenders may count as net income of the consumer any third party’s income to which the consumer has a reasonable expectation of access, which must be verified. The Bureau has also added a comment that clarifies that lenders may factor into the projections of major financial obligations the regular contributions of third parties to those obligations (comment 5(c)(1)-2). Similarly, the Bureau has clarified that if a lender is individually itemizing a consumer’s basic living expenses, the lender may consider whether other persons are regularly contributing to the consumer’s payment of basic living expenses (comment 5(b)-2.i.C.2). These changes are described in more detail in other parts of the sectionby-section analysis for § 1041.5. Thus, a consumer’s access to spousal income or the spouse’s contributions toward payment of a consumer’s major financial obligations or basic living expenses may be accounted for by the lender under the final rule, regardless of whether the consumer lives in a community 655 property State. The Bureau believes these changes would achieve for some consumers the same result as, for example, a rule that would permit a consumer to rely on the income of his spouse in a community property State.807 The Bureau does not find it sensible to create separate ability-to-repay requirements for community property States and common law property States.808 This would add complexity to the rule, pose challenges for examination and uniform enforcement of the rule, and add compliance burdens on providers operating in multiple States with different family law regimes. Furthermore, such an adjustment would not fit with the final rule’s orientation towards practical assessments of how much consumers pay in the short term for basic living expenses and major financial obligations, and practical access to income. For example, the final rule does not direct lenders to ascertain a consumer’s legal entitlement to income where the consumer does not have practical access to the funds. Nor did the commenter present any evidence that lenders in the market today have been taking into account State community property laws in making lending decisions. The Bureau disagrees with the commenter that argued that its identification of major financial obligations as obligations that must be itemized by category in underwriting suggests that the Bureau is prioritizing payment of other debt obligations over covered loans for which the 807 The Bureau acknowledges that the credit card ability-to-pay rules under Regulation Z discuss in commentary how reasonable expectation of access to the income of another person includes a legal entitlement to that income under a Federal or State regulation, including State community property laws. The Bureau is declining to adopt that standard in this final rule. See the section-by-section analysis of 1041.5(a)(5) (net income definition) and the general discussion above in § 1041.5 about why the Bureau is imposing different ability-to-repay standards for this market in contrast to the credit card market. 808 The commenter did not provide specific policy suggestions to address the issue, other than withdrawing the proposal which the Bureau is declining to do. The Bureau infers from the comment that this is one such policy option short of withdrawal. 656 lender is making an ability-to-repay determination. In fact, covered loans can also be major financial obligations (such as where a consumer has a concurrent loan outstanding). Rather, the Bureau is simply differentiating between major financial obligations that the consumer is already committed to and the obligation that would be incurred in connection with a new covered shortterm or longer-term balloon-payment loan. Finally, the Bureau declines the suggestion by commenters to include as major financial obligations property taxes and insurance that is not required to be paid in escrow to a mortgagee. The Bureau believes that the pool of consumers taking out covered short-term and longer-term balloon-payment loans who both own a home and who do not escrow their property taxes and insurance will be quite low.809 In the presumably small number of cases where consumers have a mortgage and do not pay taxes or insurance through a regular escrow arrangement, the Bureau also believes that the payments may be infrequent, particularly with regard to property taxes which, unless escrowed, are typically not paid monthly. Therefore, the Bureau believes it is unlikely in the vast majority of cases that these items would actually bear on the consumer’s financial balance sheet for purposes of the ability-to-repay requirement for a covered short-term loan,810 and thus these items should not be treated as a major financial obligation. The Bureau 809 As discussed in Market Concerns—Underwriting, for the population of payday borrowers, renting is twice as common as in the general U.S. population. See Skiba and Tobacman, “Payday Loans, Uncertainty, and Discounting: Explaining Patterns of Borrowing, Repayment, and Default,” at 5 (Apr. 1, 2008). Moreover, a recent analysis by CoreLogic shows that currently almost 80 percent of all mortgage borrowers are paying their taxes (and insurance) through escrow accounts. See Dominique Lalisse, Escrow vs. Non-escrow Mortgages: The Trend is Clear (June 21, 2017), available at http://www.corelogic.com/blog/authors/dominique-lalisse/2017/06/escrow-vsnon-escrow-mortgages-the-trend-is-clear.aspx#.WdRrL3IUns0. Finally, mortgage borrowers with higher loan-tovalue ratios are more prone to have required escrow arrangements, which could mean that payday borrowers are more likely to have escrow arrangements than the mortgage borrowing population at large. 810 Similarly, for a covered longer-term balloon-payment loan, it is relatively unlikely that such irregular expenses would come due in the relevant monthly period. 657 also is not treating them as a basic living expense for similar reasons, as well as the difficulty lenders would have in developing a non-individualized estimate of property taxes. 5(a)(4) National Consumer Report In proposed § 1041.5(a)(3), the Bureau defined national consumer report to mean a consumer report, as defined in section 603(d) of the Fair Credit Reporting Act (FCRA), 15 U.S.C. 1681a(d), obtained from a consumer reporting agency that compiles and maintains files on consumers on a nationwide basis, as defined in section 603(p) of the FCRA, 15 U.S.C. 1681a(p). In proposed § 1041.5(c)(3)(ii), the Bureau provided that a lender would have to obtain a national consumer report as verification evidence for a consumer’s required payments under debt obligations and under court- or government agency-ordered child support obligations. Reports that meet the proposed definition are often referred to informally as a credit report or credit history from one of the three major nationwide consumer reporting agencies or bureaus. A national consumer report may also be furnished to a lender from a consumer reporting agency that is not a nationwide agency, such as a consumer reporting agency that is a reseller. The Bureau did not receive comments on the specific definition of national consumer report, though it did receive comments on the requirement to obtain national consumer reports. The Bureau addresses those comments in the discussion regarding major financial obligations and § 1041.5(c). Therefore, the Bureau finalizes the definition as proposed, except renumbered as § 1041.5(a)(4). 5(a)(5) Net income Proposed Rule 658 In proposed § 1041.5(a)(4), the Bureau set forth a definition for net income as a component of the calculation for the ability-to-repay determination specified in proposed § 1041.5(b). Specifically, the Bureau proposed to define the term as the total amount that a consumer receives after the payer deducts amounts for taxes, other obligations, and voluntary contributions, but before deductions of any amounts for payments under a prospective covered short-term loan or for any major financial obligation. The Bureau explained in the proposal that the proposed definition was similar to what is commonly referred to as “take-home pay,” but is phrased broadly to apply to income received from employment, government benefits, or other sources. It would exclude virtually all amounts deducted by the payer of the income, whether deductions are required or voluntary, such as voluntary insurance premiums or union dues. The Bureau stated its belief that the total dollar amount that a consumer actually receives after all such deductions is the amount that is most instructive in determining a consumer’s ability to repay. Certain deductions (e.g., taxes) are beyond the consumer’s control. The Bureau further stated in the proposal that other deductions may not be revocable, at least for a significant period, as a result of contractual obligations into which the consumer has entered. Even with respect to purely voluntary deductions, most consumers are unlikely to be able to reduce or eliminate such deductions immediately—that is, between consummation of a loan and the time when payments under the loan would fall due. The Bureau also stated in the proposal that it believed that the net amount a consumer actually receives after all such deductions is likely to be the amount most readily known to consumers applying for a covered short-term loan (rather than, for example, periodic gross income) and is also the amount that is most readily verifiable by lenders through a variety of methods. The 659 Bureau stated in the proposal that the proposed definition would clarify, however, that net income is calculated before deductions of any amounts for payments under a prospective covered short-term loan or for any major financial obligations. The Bureau stated that it was proposing the clarification to prevent double-counting of any such amounts when making the ability-torepay determination. Comments Received The Bureau received a number of comments on its proposed definition of net income, raising a variety of issues. Several industry commenters argued that the Bureau should explicitly state that the definition includes a number of other sources of income that are paid at irregular times or in irregular amounts, including seasonal income, tips, bonuses, overtime pay, or commissions. Commenters also asked the Bureau to state explicitly that receipt of a number of other types of income should be included, such as child support, annuities, alimony, retirement, disability, prizes, jury awards, remittances, investment income, tax refunds, and legal settlements. A consumer advocate commenter took the opposite view, arguing that one-time lump-sum payments, tax refunds, legal settlements, or other income that is “not consistently reliable” should not be counted. This commenter argued that these income sources often are speculative and that consumers relying on them often take out payday or vehicle title loans in reliance on the expected funds only to see the payment delayed or to receive less funds than expected or not at all, leading to inability to repay and collateral consequences. Other commenters argued that the Bureau should allow lenders to include in net income any third-party income, like spousal income, because many individuals’ finances are managed on a household basis. Some suggested that the Bureau’s failure to do so was inconsistent with 660 CARD Act regulations, which permit card issuers to consider as the applicant’s income the income of another person if the applicant has a reasonable expectation of access to the other person’s income.811 The commenters argued that this created a disadvantage to stay-at-home spouses and would result in loss of credit access. They criticized the proposal for not addressing this issue in the same manner as the Bureau’s rulemaking in 2013 amending the CARD Act regulations. (Commenters raised related Regulation B issues addressed in the general § 1041.5 discussion above.) One commenter made arguments similar to those it made regarding major financial obligations, discussed with regard to § 1041.5(a)(3) above, arguing that the Bureau should have taken into account spousal income in community property States. Some commenters argued that the Bureau should use gross income instead of net income. One trade association argued that one of its members currently uses gross income, and that just this minor change would require training, systems updates, and changes to forms. Another noted that for Federal student loans, income-based repayment plans are assessed using adjusted gross income, and asserted that the Bureau’s proposal to use net income was merely a method of ensuring that fewer consumers would meet the standard. 811 The CARD Act regulations in commentary to 12 CFR 1026.51(a)(1)(i) clarify when card issuers may consider for purposes of the ability-to-pay test the income of another person to which the consumer has reasonable expectation of access. Two comments directly or indirectly reference community property laws. Comment 51(a)(1)(i)-4.iii clarifies that, “Consideration of the income or assets of authorized users, household members, or other persons who are not liable for debts incurred on the account does not satisfy the requirement to consider the consumer’s current or reasonably expected income or assets, unless a Federal or State statute or regulation grants a consumer who is liable for debts incurred on the account an ownership interest in such income and assets (e.g., joint ownership granted under State community property laws), such income is being deposited regularly into an account on which the consumer is an accountholder (e.g., an individual deposit account or a joint account), or the consumer has a reasonable expectation of access to such income or assets even though the consumer does not have a current or expected ownership interest in the income or assets.” Comment 51(a)(1)(i)-6.iv includes an example of where there is not reasonable expectation of access because, among other facts, “no Federal or State statute or regulation grants the applicant an ownership interest in that income.” 661 Some commenters argued that the Bureau should not require lenders to subtract voluntary deductions from the net income calculation, arguing that because these deductions are voluntary they thus could be diverted to cover basic living expenses, major financial obligations, or loan payments. Other commenters asked for further clarification of what “other obligations” and “voluntary contributions” would include. Still others argued that it would be very difficult in many instances to verify whether an employer was deducting for taxes or other items. Those commenters questioned whether lenders would be required to ascertain the consumer’s tax liability or be held responsible if the take-home pay figure used for the projection of net income was found to be based on erroneous information about tax deductions. A small rural lender commented that the proposed definition would create an inconsistent standard, positing that a loan applicant who withholds the maximum permitted amount would be less likely to pass the ability-to-repay requirement than another applicant who withholds the minimum amount, even if they work at the same job and earn the same salary. Another commenter asked for clarification on the situation where the verification evidence does not identify the payee or purpose of a deduction; the commenter noted this would likely occur with deposit account transaction history. Several industry commenters believed that the Bureau should allow lenders to include in net income the proceeds from the covered loan itself.812 These commenters argued that while it may make sense not to include proceeds in net income when a consumer is using those proceeds to pay for emergency expenses, it is conceptually inconsistent to exclude proceeds when they are being used to pay for basic living expenses or major financial obligations. For example, if a 812 Including proceeds in income, or deducting them from basic living expenses or major financial obligations, are mathematically and conceptually equivalent. Here, the Bureau addresses this line of argument as a request to include proceeds in income. But the Bureau’s response applies to both versions of the concept. 662 consumer uses proceeds to pay rent—which is a major financial obligation—commenters believed it would be unfair to have to treat rent as an obligation that the consumer would still have to pay in order to determine whether she would have the ability to repay the loan, unless the proceeds can be included in the net income calculation. They viewed this approach as improper “double-counting” of the major financial obligation or basic living expense paid with proceeds. Final Rule The Bureau is finalizing the definition of net income in § 1041.5(a)(5) with two changes from the proposal. The Bureau, moreover, has added three new comments to address various issues raised by the commenters. The first change is a technical change that aligns with the change in scope of the final rule. The proposal defined net income as the total amount the consumer receives after the payor deducts amounts for taxes, other obligations, and voluntary contributions, qualified with a parenthetical phrase reading “but before deductions of any amounts for payments under a prospective covered short-term loan or for any major financial obligation.” The definition of net income in proposed § 1041.9(a)(5) contained similar language referring to covered longer-term loans. In light of its decision not to finalize the ability-to-repay requirements as to all covered longer-term loans and to consolidate into § 1041.5 provisions from § 1041.9 relating to covered longer-term balloon-payment loans, the Bureau has changed the language to refer to “but before deductions of any amounts for payments under a prospective covered short-term loan or covered longer-term balloon-payment loan or for any major financial obligation.” Second, the Bureau agreed with commenters that it should allow lenders to include income from third parties where the consumer has a reasonable expectation of access to that 663 income, and § 1041.5(a)(5) of the final rule allows lenders to do so. In new comment 5(a)(5)-3, the Bureau clarifies that a consumer has a reasonable expectation of access to a third party’s income if the consumer has direct, practical access to those funds on a regular basis through a bank account in which the consumer is an accountholder. The Bureau also provided examples in comment 5(a)(5)-3 of what reasonable expectation of access would entail, including evidence of a joint bank account or of regular deposits from said third party into an account in the consumer’s name. A number of commenters had cited the Bureau’s CARD Act regulations as precedent for the request to include the income of another person in net income. The Bureau notes that the CARD Act regulations in 12 CFR 1026.51(a)(1)(i) contain commentary including a number of examples of whether an applicant had a reasonable expectation of access to the income of another person.813 This commentary was added in the Bureau’s amendments to the credit card ability-to-pay rules in 2013. The Bureau notes that it drew inspiration from this commentary in drafting the examples in comment 5(a)(5)-3, but the Bureau has not incorporated all of the examples. In particular, one example posited that the consumer has reasonable expectation of access where another person is regularly paying the consumer’s expenses, and another comment cited above includes an example of where there is not reasonable expectation of access because, among other facts, no Federal or State statute or regulation grants the applicant an ownership interest in that income. The former example, in the Bureau’s view, does not align well with the final rule insofar that the credit card example blends the distinction between income and expenses; as with the proposal, the final rule creates separate definitions for net income, major 813 See 12 CFR 1026.51(a)(1)(i), comment 51(a)(1)(i)-6. 664 financial obligations, and basic living expenses. Accordingly, the Bureau has dealt with contributions toward basic living expenses and major financial obligations in comment 5(b)2.i.C.2 and comment 5(c)(1)-2, respectively, of the final rule rather than in connection with the definition of net income. Also, the Bureau is not adapting the language referencing Federal or State statutes or regulations granting an ownership interest in income for similar reasons to those described above with regard to State community property laws in connection with major financial obligations. For further discussion on the differences more generally between the final rule and the CARD Act ability-to-pay regulations, see general § 1041.5 discussion above. Regarding the commenter that discussed State community property laws, similar to the treatment of this issue as applied to major financial obligations, the Bureau concludes that whether a consumer lives in a community property State does not change the consumer’s practical access to income, and thus the regulation does not need to distinguish between how lenders should treat net income from one State to another. However, as noted above, in response to other comments the Bureau has decided to allow lenders to rely on third-party income, including income from a spouse, if the consumer has a reasonable expectation of access to that income (see discussion of § 1041.5(a)(5) and comment 5(a)(5)-3). This is consistent with the Bureau’s general approach to whether a consumer has practical access to a spouse’s (or other third party’s) income. Given that this rule is closely focused on whether consumers will be able to meet their major financial obligations, make the payments on the loan, and pay basic living expenses in the near term, the Bureau determined that practical access to income was more important than legal entitlement to income. The Bureau also notes that attributing all community property to a consumer would not necessarily increase the odds that the consumer would be able 665 to meet the ability-to-repay requirement relative to the final rule, because in community property States, liabilities are also imputed to the spouse. The Bureau also noted in the earlier discussion that creating separate underwriting regimes depending on the family law of the State would create added complexity and also challenges for examination, enforcement, and compliance. The Bureau also agrees with commenters that the final rule should provide more clarity about and examples of what sources of income could be included in net income. The Bureau has added a detailed new comment, 5(a)(5)-1, addressing these issues. Specifically, the comment clarifies that net income includes income that is regularly received by the consumer as take-home pay, whether the consumer is treated as an employee or independent contractor, and also includes income regularly received by the consumer from other sources, such as court-ordered child support or alimony received by the consumer and any payments received by the consumer from retirement, social security, disability, or other government benefits, or annuity plans. Comment 5(a)(5)-1 further clarifies that lenders may include in net income irregular or seasonal income, such as tips, bonuses, and overtime pay, and that net income does not include one-time payments anticipated to be received in the future from non-standard sources, such as legal settlements, tax refunds, jury prizes, or remittances, unless there is verification evidence of the amount and expected timing of such income. The Bureau has included the verification requirement with regard to future one-time payments because they generally are uncertain as to timing or amount. Before basing an ability-to-repay determination on a projection of this sort, the Bureau believes it is important to be confident that income will be received during the relevant monthly period in the expected amount. Of course, lenders must always collect verification evidence about net income where it is reasonably available (see § 1041.5(c)(2)(ii)(A) 666 and comment 5(c)(2)(ii)(A)-3). Therefore, the effect of comment 5(a)(5)-1 is that when verification evidence is not reasonably available to project one-time income payment, then unlike with other sources of income, the lender cannot rely on the consumer’s statement of the amount alone. The Bureau does not agree with the commenter requesting the rule prohibit inclusion of these types of one-time income sources altogether, because if verification evidence as described is available, the Bureau believes it is appropriate to include such types of income in the definition of net income. The Bureau does not agree with commenters that it is more appropriate to calculate debtto-income or residual income based on gross income than net income. The ability-to-repay determination is intended to capture the amount of money the consumer will actually have available to pay for major financial obligations, basic living expenses, and loan payments in the month with the highest sum of payments on the loan. Income that is automatically diverted to taxes or other deductions would not be available to cover any of those expenses. While it is true, as one commenter noted, that student loan income-driven repayment plans are based on gross income, that is because an income-driven repayment plan is a flat percentage of income and does not account for basic living expenses or major financial obligations (see the discussion above about why a payment-to-income approach has not been adopted in this rule). At the same time, with regard to commenters that raised concerns about compliance burdens where they are relying on verification sources that do not clearly reflect whether deductions have been made from take-home pay, the Bureau believes it is not practicable to require lenders to engage in detailed inquiries and individual adjustments. Thus, the Bureau has clarified in comment 5(a)(5)-1 that the lender may draw reasonable conclusions from 667 information provided by the consumer and is not required to inquire further about deductions for the consumer’s taxes, other obligations, or voluntary contributions. This may mean that a lender could rely on gross income on a pay stub, if net income and/or deductions are not otherwise on the pay stub. Similarly, if a lender is verifying income via a bank statement, the lender may assume that the amount deposited is net of deductions. The Bureau also is adding commentary language to address the comments asking for clarification on the meaning of voluntary contributions and whether the lender must, or can, assume that voluntary contributions will be discontinued during the term. The Bureau has added comment 5(a)(5)-2 to provide further clarification about what would be included as a voluntary contribution deducted from income, giving an example of a consumer’s contribution to a defined contribution plan commonly referred to as 401K plans. In light of comments received, comment 5(a)(5)-2 also clarifies that a lender may inquire about and reasonably consider whether the voluntary contributions will be discontinued prior to the relevant monthly period, in which case deductions for those voluntary contributions would not need to be accounted for in the income calculation. New comment 5(a)(5)-2 also clarifies that an example of an “other obligation” is a consumer’s portion of payments for premiums for employer-sponsored health insurance plans. Treatment of loan proceeds. After careful consideration, the Bureau has decided not to include the loan proceeds in net income, or otherwise allow the lender to give a credit for or otherwise account for the proceeds in the estimation of basic living expenses or projection of major financial obligations. The Bureau acknowledges that some consumers use loan proceeds to cover basic living expenses or major financial obligations, but believes on balance that treating for loan proceeds as income is not appropriate for multiple reasons. 668 First, many consumers take out covered short-term or covered longer-term balloonpayment loans specifically to pay unusual, non-recurring or emergency expenses, or because covering such expenses in the recent past has left them without sufficient funds to cover basic living expenses or major financial obligations. The Bureau received many comments, including many from individual consumers, describing how consumers often use payday loans and other covered loans to cover emergency expenses. Payday lenders in their advertising also tend to cite this usage category as the primary purpose for using the product, and industry commenters noted it as a use case as well. Academic literature and surveys discussing usage patterns on payday loans have consistently found that a sizable number of consumers report using payday loans and other covered loans for non-recurring and emergency expenses. See part II and Market Concerns—Underwriting (citing a 2012 study by Center for Financial Services Innovation). Because money is fungible, the Bureau is concerned that disentangling the interplay between regular and irregular expenses would create significant compliance and examination challenges. Lenders would be expected to adhere to different rules depending on the stated intended use of the loan proceeds. This would put the lenders in the position of having to inquire in detail about consumers’ intended use for the loans, which consumers may feel is unduly intrusive. Such a provision would also be difficult to enforce given the fungible nature of the funds in question and raise questions about lender compliance burden and liability under the rule if they rely on a consumer’s statement of intended use that does not prove accurate. It also would create incentives for evasion. In addition, simply assuming that all consumers will use the loan proceeds to pay basic living expenses or major financial obligations would be as simple as the approach taken by the 669 Bureau, but is a problematic approach on policy grounds. Because many consumers use loan proceeds for reasons other than payment of major financial obligations or basic living expenses, such a rule would lead to lenders making loans to many consumers who plan to use the funds to cover a non-recurring or emergency expense, and thus the ability-to-repay determinations would be inaccurate in the opposite direction. As a result, the harms identified in Market Concerns— Underwriting and the section-by-section analysis for § 1041.4 would continue to exist and would likely be prevalent. Moreover, there is a question of timing. As referenced above and described in more detail below in connection with § 1041.5(a)(7) and § 1041.5(b)(2), the Bureau has decided to focus the calculation of debt-to-income or residual income on the relevant monthly period, which is the calendar month with the highest sum of loan payments. This snapshot is intended to focus on the month in which the loan places the greatest strain on the consumer’s finances, which is then used in turn by the lender to forecast the consumer’s ability to cover loan payments, major financial obligations, and basic living expenses both during the loan term and for 30 days after the single highest payment. To the extent that consumers use loan proceeds to cover major financial obligations or basic living expenses, that is likely to occur soon after consummation. Thus, except for loans with short terms made near the beginning of a calendar month, the Bureau believes that the proceeds will have been disbursed to cover expenses before the relevant monthly period and/or the 30 days after the single highest payment on the covered loan. Indeed, in light of the concern about high risk of re-borrowing in the markets for covered short-term and longer-term balloon-payments, this is precisely why the Bureau has focused the 670 analysis on the period of time in which the consumer is making the largest payment(s) on the loan and the major financial obligations and basic living expenses that are due soon thereafter. 5(a)(6) Payment under the covered short-term or longer-term balloon-payment loan Proposed Rule The Bureau proposed to define payment under the covered short-term loan, which was a component of the calculation for the ability-to-repay determination as specified in proposed § 1041.5(b). Specifically, the proposed definition of payment under the covered short-term loan in proposed § 1041.5(a)(5)(i) and (ii) would have included all costs payable by the consumer at a particular time after consummation, regardless of how the costs are described in an agreement or whether they are payable to the lender or a third party. Proposed § 1041.5(a)(5)(iii) would have set special rules for projecting payments on lines of credit if they are provided for under a covered short-term loan for purposes of the ability-to-repay test, since actual payments for lines of credit may vary depending on usage. Proposed § 1041.5(a)(5)(i) would have applied to all covered short-term loans. It defined payment under the covered short-term loan broadly to mean the combined dollar amount payable by the consumer in connection with the covered short-term loan at a particular time following consummation. The proposed definition further would have provided that, for short-term loans with multiple payments, in calculating each payment under the covered loan, the lender must assume that the consumer has made the preceding required payments and has not taken any affirmative act to extend or restructure the repayment schedule or to suspend, cancel, or delay payment for any product, service, or membership provided in connection with the covered loan. 671 Proposed § 1041.5(a)(5)(ii) similarly would have applied to all covered short-term loans and clarified that payment under the covered loan included all principal, interest, charges, and fees. The Bureau stated in the proposal that it believed that a broad definition was necessary to capture the full dollar amount payable by the consumer in connection with the covered shortterm loan, including amounts for voluntary insurance or memberships and regardless of whether amounts are due to the lender or another person. The Bureau noted that it is the total dollar amount due at each particular time that is relevant to determining whether or not a consumer has the ability to repay the loan based on the consumer’s projected net income and payments for major financial obligations. The amount of the payment is what is important, not whether the components of the payment include principal, interest, fees, insurance premiums, or other charges. In the proposal, the Bureau recognized, however, that under the terms of some covered short-term loans, a consumer may have options regarding how much the consumer must pay at any given time and that the consumer may in some cases be able to select a different payment option. The Bureau explained that the proposed definition would include any amount payable by a consumer in the absence of any affirmative act by the consumer to extend or restructure the repayment schedule, or to suspend, cancel, or delay payment for any product, service, or membership provided in connection with the covered short-term loan. Proposed comment 5(a)(5)(i) and 5(a)(5)(ii)-1 would have included three examples applying the proposed definition to scenarios in which the payment under the covered short-term loan includes several components, such as voluntary fees owed to a person other than the lender, as well as scenarios in which the consumer has the option of making different payment amounts. 672 Proposed § 1041.5(a)(5)(iii) included additional provisions for calculating the projected payment amount under a covered line of credit for purposes of assessing a consumer’s ability to repay the loan. As explained in proposed comment 5(a)(5)(iii)-1, the Bureau believed such rules were necessary because the amount and timing of the consumer’s actual payments on a line of credit after consummation may depend on the consumer’s utilization of the credit (i.e., the amount the consumer has drawn down) or on amounts that the consumer has repaid prior to the payments in question. As a result, if the definition of payment under the covered short-term loan did not specify assumptions about consumer utilization and repayment under a line of credit, there would be uncertainty as to the amounts and timing of payments to which the ability-torepay requirement applies. Proposed § 1041.5(a)(5)(iii) therefore prescribed assumptions that a lender must make in calculating the payment under the covered short-term loan. It would have required the lender to assume that the consumer will utilize the full amount of credit under the covered loan as soon as the credit is available to the consumer, and that the consumer will make only minimum required payments under the covered loan. The lender would then apply the ability-to-repay determination to that assumed repayment schedule. Proposed § 1041.9(a)(5)(iii) would have included parallel provisions, with a supplemental provision to account for the fact that it applied to longer-term loan structures. In addition to the same two assumptions that a lender must make in calculating the payment under proposed § 1041.5(a)(5)(iii), proposed § 1041.9(a)(5)(iii) also would have required the lender to assume that, if the terms of the covered longer-term loan would not provide for a termination of access to the credit line by a date certain and for full repayment of all amounts due by a date 673 certain, the consumer must repay any remaining balance in one payment on the date that is 180 days following the consummation date. Comments Received The Bureau received a number of comments that were generally supportive of the Bureau’s definition of payment under the covered short-term loan. A trade group representing open-end credit providers criticized this rule generally for reflecting what was, in the commenter’s view, the Bureau’s lack of understanding about openend credit provisions. They specifically criticized the proposal for, in the commenter’s view, not addressing how lines of credit with principal paydown requirements or with a specified duration would be treated. The Bureau also received a comment objecting to proposed § 1041.9(a)(5)(iii)(C), the parallel definition for the proposed underwriting section for covered longer-term loans, which would have provided that the whole balance of open-end longer-term credit should be considered to be due 180 days following the consummation date if there is not a date certain for termination of the line and repayment of any remaining balance. The commenter argued instead that the Bureau should use the maximum required payment under the terms of the agreement. Final Rule The Bureau has finalized the definition as proposed in § 1041.5(a)(6), with minor wording clarifications and the addition of references to payments for covered longer-term balloon-payment loans. The Bureau also has made minor adjustments to the examples in comment 5(a)(6)(i)-1 and 5(a)(6)(ii)-1 to reflect that the same definition applies to covered longer-term balloon-payment loans. 674 With regard to the rules for calculating payments on open-end loans, the Bureau has not imported the text from proposed § 1041.9(a)(5)(iii)(C) into this definition, which would have made a lender assume, for purposes of the ability-to-repay determination, that all advances under a longer-term open-end credit line would be due within 180 days of consummation if there is not a date certain for termination of the line and repayment of any remaining balance. Because the Bureau has decided to apply the ability-to-repay requirements only to covered longer-term balloon-payment loans that have the payment features as described in § 1041.3(b)(2), the Bureau does not believe that this provision is necessary to help lenders calculate potential loan payments. Put another way, if a loan without a date certain for termination of the line and repayment of any remaining balance qualifies as a covered longer-term balloon-payment loan under the rule, the Bureau believes the terms of the loan contract that create that balloon payment feature will be sufficient for lenders to calculate payments using the assumptions in § 1041.5(a)(6)(iii)(A) and (B). In comment 5(a)(6)(iii)-1, in addition to corresponding technical updates, the Bureau added a description of how a lender should calculate the payment amount for open-end credit when underwriting for a new advance, including when there is an outstanding balance. The comment states that lenders should use the same test with the same assumptions when they make a new ability-to-repay determination under § 1041.5(b)(1)(ii) prior to an advance under the line of credit that is more than 90 days after the date of a prior ability-to-repay determination for the line of credit, in order to determine whether the consumer still has the ability to repay the current credit line. The Bureau also disagrees with the commenter that argued the proposal reflects a lack of 675 understanding of open-end credit provisions. The commenter’s primary focus in asserting a lack of understanding appears to have been on certain assumptions about credit line usage and repayment that the proposal would have required lenders to use in periodically re-underwriting open-end loans. Those assumptions were admittedly complicated by the fact that the proposal would have applied to a broad range of product structures. However, the Bureau has since simplified and clarified those assumptions particularly in light of the narrowed scope of the final rule’s ability-to-repay requirements, which now apply only to covered short-term loans and covered longer-term balloon-payment loans. The Bureau believes the remaining assumptions— that consumers draw the maximum amount allowed on the loan and make minimum payments for as long as permitted under the loan contract—are logical for assessing consumers’ ability to repay and relatively simple to apply in conjunction with covered loans’ contractual terms governing principal pay-down and other matters. 5(a)(7) Relevant monthly period As described above, the Bureau has added a definition for relevant monthly period, which is the calendar month in which the highest sum of payments under the loan is due. This definition will be used as the period for which a lender will need to calculate residual income or a debt-to-income ratio. As noted in the discussion regarding debt-to-income ratio above, the concept of the relevant monthly period flows from the larger streamlining and reconceptualization of the requirements under § 1041.5(b)(2). The Bureau believes that instead of requiring lenders to make separate calculations to analyze consumers’ ability to cover major financial obligations, basic living expenses, and payments on the covered loan both during the term of the loan and for 30 days after the highest payment on the loan, it would be more 676 administrable to allow lenders to make a single monthly calculation that can then be used to evaluate more generally whether the consumer has the ability to cover all relevant expenses during the time periods described in § 1041.5(b)(2). Because the month with the highest sum of payments on the covered short-term or covered longer-term balloon-payment loan will be the month in which the loan places the greatest strain on the consumer’s finances, the Bureau believes that it is the logical period to use as a snapshot. Indeed, the Bureau had proposed to focus the underwriting analysis for covered longer-term balloon-payment loans on this specific period for this same reason.814 The Bureau considered starting the monthly clock on the date of the first of the loan payment(s), but ultimately concluded that a calendar month was easier to administer. Since billing cycles typically correspond to calendar months, the Bureau believes that it will be relatively straightforward for lenders to project income and major financial obligations based on consumer statements, income documentation, and national consumer reports. The Bureau also believes that calculating the residual income and debt-to-income ratio for a relevant monthly period defined by reference to a calendar month will generally give lenders a sense of total monthly inflows and outflows that can be projected to the time periods for which the lender must make a reasonable conclusion that, based on residual income or the debt-to-income ratio, the consumer can make payments for major financial obligations, make all payments under the loan, and meet 814 Specifically, for covered longer-term loans, proposed § 1041.9(b)(2) set out a two-part test. All lenders for all covered loans would have had to evaluate consumers’ residual income for the term of the loan under § 1041.9(b)(2)(i), which comment 9(b)(2)(i)-1.i explained could be satisfied by analyzing residual income for the month with the highest sum of payments (if applicable) under the loan. The second part of the test under proposed § 1041.9(b)(2)(ii) applied only to covered longer-term balloon-payment loans and would have required lenders to evaluate consumers’ ability to cover major financial obligations and basic living expenses for 30 days after the highest single payment. 677 basic living expenses. See discussion of § 1041.5(b)(2)(i) and (ii) and commentary for further information. The relevant monthly period is also the time period referenced under § 1041.5(b)(2)(i)(B) and (ii)(B). The Bureau considered alternative time periods for the relevant monthly period, such as the 30-day period starting at consummation, the 30-day period ending on the contractual due date, or the calendar month in which consummation occurred. The Bureau chose the specific calendar month in which the highest sum of payments under the loan will be due for the reasons discussed above, because it believes that the residual income and debt-to-income ratio will only be demonstrative of ability to repay if they reflect the calendar month in which the loan will strain the consumer’s monthly balance sheet the most. The Bureau notes that for covered longerterm balloon-payment loans, there may be challenges to projecting major financial obligations and net income, as the relevant monthly period may fall far into the future. Commentary in the definitions of debt-to-income ratio and residual income addresses this issue; see comments 5(a)(2)-1 and 5(a)(8)-1 which provide that for covered longer-term balloon-payment loans, lenders must make reasonable assumptions about that period compared to the period covered by the verification evidence, and gives examples. 5(a)(8) Residual income The Bureau proposed § 1041.5(a)(6) to define residual income as a component for the calculation of the ability-to-repay determination specified in proposed § 1041.5(b). It proposed to define the term as the sum of net income that the lender projects the consumer obligated under the loan will receive during a period, minus the sum of amounts that the lender projects will be payable by the consumer for major financial obligations during that same period. Proposed 678 § 1041.5(b) would have generally required a lender to determine that a consumer will have sufficient residual income to make payments under a covered short-term loan and to meet basic living expenses. The Bureau discussed above the comments that generally criticized its approach to requiring a residual-income analysis, which led the Bureau in the final rule to add the debt-toincome ratio as another option for lenders to use. Other comments about the Bureau’s general ability-to-repay framework were also listed above, and will be discussed further in addressing § 1041.5(b). The Bureau made a few changes to the definition of residual income as finalized in § 1041.5(a)(8). First, there were a number of technical edits, and the Bureau included “relevant monthly period” where appropriate to incorporate the revised approach to the timing of the underwriting calculations that must be made and thus parallel the definition of debt-to-income ratio. As discussed above, the Bureau has modified its approach to residual income calculations to allow lenders to calculate them on a net basis for the relevant monthly period, rather than focusing in detail on the timing of inflows and outflows within the time periods specified in § 1041.5(b)(2)(ii). The Bureau has also added into the residual income calculation the payments under the covered short-term or longer-term balloon-payment loan. This was a shift in structure from the proposal, but not substance. In the proposed rule, residual income was net income minus major financial obligations, and the result was used to make sure a consumer could afford the loan payments and basic living expenses. Now residual income is net income minus major financial obligations and loan payments, and the results will be used to determine whether consumers can 679 afford basic living expenses only. The Bureau thought it would be easier to reposition these variables so that the numbers for which the lender will need to make an individualized assessment—net income, major financial obligations, and loan payments—will all be used to come up with a single number. That will allow a lender to isolate the only estimated figure— basic living expenses. The Bureau notes that this “back-end” approach is consistent with the formulation in the Bureau’s mortgage ability-to-repay requirements and the definition of debt-toincome ratio in § 1041.5(a)(2). In addition, the Bureau added comment 5(a)(8)-1, which restates the definition of residual income and provides further clarification on how to project net income and major financial obligations for covered longer-term balloon-payment loans where the relevant monthly period may be well into the future. The Bureau states that the lender cannot assume, absent a reasonable basis, that there will be a substantial increase in income or decrease in major financial obligations between consummation and the relevant monthly period. As for all loans made under § 1041.5, lenders will generally be using figures verified by evidence of past payment amounts and income to project into the future. The Bureau recognizes that this projection will likely become somewhat less accurate as the time between verification evidence and the relevant monthly period lengthens, but notes that any further augmentations to amounts derived from verification evidence should be made only if a lender has a reasonable basis for doing so. 5(b) Reasonable determination required Overview The Bureau proposed to prohibit lenders from making covered short-term loans without first making a reasonable determination that the consumer will have the ability to repay the loan 680 according to its terms, unless the loans were made in accordance with the conditional exemption in proposed § 1041.7. Specifically, proposed § 1041.5(b)(1) would have required lenders to make a reasonable determination of ability to repay before making a new covered short-term loan, increasing the credit available under an existing loan, or before advancing additional credit under a covered line of credit if more than 180 days have expired since the last such determination. Proposed § 1041.5(b)(2) would have specified minimum elements of a baseline methodology that would be required for determining a consumer’s ability to repay, using a residual-income analysis and an assessment of the consumer’s prior borrowing history. It would have required the assessment to be based on projections of the consumer’s net income, basic living expenses, and major financial obligations that are made in accordance with proposed § 1041.5(c). It would have required that, using such projections, the lender must reasonably conclude that the consumer’s residual income will be sufficient for the consumer to make all payments under the loan and still meet basic living expenses during the shorter of 45 days or the term of the covered short-term loan. It would have further required that a lender must reasonably conclude that the consumer, after making the highest payment under the loan (typically, the last payment), will continue to be able to meet major financial obligations as they fall due, make any remaining payments on the loan, and meet basic living expenses for a period of 30 additional days.815 Finally, proposed § 1041.5(b)(2) would have required that, in situations in which the consumer’s recent borrowing history suggests that she may have difficulty repaying a new loan 815 Under proposed § 1041.9(b)(2) and comments 9(b)(2)(i)-1 and 9(b)(2)(ii)-1, the focus for analyzing covered longer-term balloon-payment loans would have been on two similar periods: (1) the month with the highest sum of loan payments; and (2) the 30 days after the single highest payment on the loan. 681 as specified in proposed § 1041.6, a lender must satisfy the requirements in proposed § 1041.6 before extending credit (i.e., the proposed presumptions of unaffordability and prohibitions on lending contained therein). As noted above in the general § 1041.5 discussion above, the Bureau received a significant number of comments asserting that the proposed ability-to-repay requirements were overly burdensome. Many commenters argued that they would lead to undue lost access to credit and excessive costs. The Bureau also received comments asserting that various aspects of the proposed ability-to-repay requirements were too restrictive and, on the other hand, too vague. Some commenters specifically argued that the reasonableness test animating the entirety of proposed § 1041.5 was overly vague and would lead to uncertainty about the Bureau’s expectations for compliance and potential challenges for examination and enforcement. These commenters included a wide spectrum of parties, including industry stakeholders, State banking supervisors, and some State Attorneys General. Consumer advocates, on the other hand, generally supported the proposed requirements while suggesting various means of strengthening them in their view. These comments are discussed in more detail in the discussion of individual sub-paragraphs within § 1041.5(b). As stated above, the Bureau has made a number of changes to § 1041.5(b) and its associated commentary in the final rule. As a general matter, these changes have been made in response to comments and have two primary purposes: to provide a streamlined set of requirements for evaluating the consumer’s ability to repay, which the Bureau believes will reduce burden, and to clarify the “reasonableness” standard for ability-to-repay determinations, which the Bureau believes will reduce uncertainty about the standards for compliance. The 682 specific changes to the rule and commentary to achieve these purposes are found in two areas: First, the Bureau has made substantial revisions to § 1041.5(b)(2), which sets forth the specific parameters of the general ability-to-repay determination in § 1041.5(b)(1), i.e., that the lender use the projections of net income and major financial obligations for the relevant monthly period and calculations of debt-to-income ratio or residual income for that same period to draw reasonable conclusions about the consumer’s ability to make the loan payments, pay for major financial obligations, and meet basic living expenses during specified time periods as described in final § 1041.5(b)(2)(i) and (ii). Second, the Bureau has substantially revised and expanded the commentary to 1041.5(b) to provide additional clarity on the expected components of a “reasonable” ability-to-repay determination and how reasonableness will be evaluated through the lender’s loan performance. Specifically, comment 5(b)-2.i has been revised to provide additional discussion of reasonable ability-to-repay determinations, in particular, additional clarification on reasonable estimates of basic living expenses. Comment 5(b)-2.ii now provides additional discussion of what constitutes an unreasonable ability-to-repay determination, including a new example involving a specific debt-to-income ratio. The final rule also significantly expands comment 5(b)-2.iii, which in the proposal described how evidence of the lender’s objective and comparative loan performance (i.e., rates of delinquency, re-borrowing, and default) may be evaluated to assess the reasonableness of ability-to-repay determinations. The comment now contains a broader list of indicators than the proposal (including default rates, re-borrowing rates, patterns of lending across loan sequences, evidence of delinquencies and collateral effects, and patterns of lenders “bridging” covered loans with non-covered loans) and provides more detail on how the Bureau 683 will use the loan performance metrics to evaluate lenders’ ability-to-repay determinations. The final rule also contains a new comment 5(b)-2.iv, which complements the expanded comment 5(b)-2.iii and provides four detailed examples of whether the lender is making reasonable or unreasonable ability-to-repay determinations. The Bureau also made several changes throughout § 1041.5(b) and its commentary to implement the decision to incorporate the part of proposed § 1041.9(b) that would have imposed similar ability-to-repay requirements for covered longer-term balloon-payment loans into § 1041.5. Thus, as finalized, at a high level, § 1041.5(b)(1) provides that lenders must make reasonable determinations that the consumer will have the ability to repay the loan according to its terms. Section 1041.5(b)(1)(i) applies to covered short-term loans and covered longer-term balloon-payment loans generally, while § 1041.5(b)(1)(ii) imposes requirements to determine consumers’ ability to repay periodically for open-end lines of credit. Finalized § 1041.5(b)(2) sets forth that a lender’s determination is reasonable only if it uses a debt-to-income ratio methodology as set forth in § 1041.5(b)(2)(i), or a residual income methodology as set forth in § 1041.5(b)(2)(ii). Under § 1041.5(b)(2), both the residual income and debt-to-income methodologies are used to project the consumer’s finances during the relevant monthly period so that the lender in turn can draw conclusions about the consumer’s ability to repay covered shortterm loans or covered longer-term balloon-payment loans without re-borrowing. This broader determination focuses for covered short-term loans on whether the consumer can make payments for major financial obligations, payments under the loan, and basic living expenses during the shorter of the loan term or 45 days following consummation, and for 30 days after the highest 684 payment under the loan, and for covered longer-term balloon-payment loans, on whether the consumer can make the same payments during the relevant monthly period and for 30 days after the highest payment under the loan. However, as described in the general § 1041.5 discussion and the discussion of the debt-to-income ratio definition in § 1041.5(a)(2), above, the debt-toincome ratio and residual income would not need to be calculated for all of those periods. Rather, the lender only needs to project net income and major financial obligations and calculate debt-to-income ratio or residual income, as applicable, for one calendar month—the relevant monthly period. The final rule reduces burden in at least two ways relative to the proposal, in addition to permitting use of a debt-to-income ratio as well as a residual-income analysis. Under proposed § 1041.5(b)(2), the reasonable ability-to-repay determination would have required the lender to project both the amount and timing of the consumer’s net income and major financial obligations, as well as to make separate calculations about the consumer’s finances during two distinct time periods: first for the shorter of the term of the loan or 45 days after consummation of the loan, and then also for 30 days after having made the highest payment under the loan. Under the final rule, however, lenders are instead required to make a projection about net income and major financial obligations and calculate the debt-to-income ratio or residual income, as applicable, during only the relevant monthly period, which is the calendar month with the highest payments on the loan. The debt-to-income ratio or residual income during this period is used as a snapshot of the consumer’s financial picture to draw conclusions about the consumer’s ability to pay. The lender then uses this information to make a reasonable conclusion that the consumer has the ability to repay the loan while meeting basic living expenses and major financial 685 obligations during the two specified time periods (which are not necessarily the same as the relevant monthly period, but may often overlap). The nature of the calculation has changed as well. While the proposal would have required lenders to pay particularly close attention to the timing of income and major financial obligations in the 30 days after the loan’s highest payment, the final rule requires that the calculations for the relevant monthly period focus on the total amount of net income and major financial obligations. The Bureau also notes that this simplified approach dovetails with the inclusion of the debt-to-income ratio methodology as an alternative to residual income. As discussed above, a debt-to-income methodology does not permit the tracking of a consumer’s individual income inflows and major financial obligation outflows on a continuous basis over a period of time. The same approach has also been incorporated into the definition of residual income in § 1041.5(a)(8) for purposes of making the standards for both alternatives consistent. As explained in more detail below, the Bureau believes that this approach will streamline the process for making the ability-to-repay determination required under 1041.5(b) because the lender will only be required to project net income and major financial obligations and make the calculation of debt-to-income ratio or residual income for one calendar month. The Bureau believes the revised approach will prove simpler for consumers as well.816 5(b)(1) Proposed Rule 816 For example, both consumers and lenders will not need to be as precise in tracking the timing of inflows and outflows within the periods in § 1041.5(b)(2)(i) and (ii). 686 In proposed § 1041.5(b)(1), the Bureau proposed generally that, except as provided in proposed § 1041.7, a lender must not make a covered short-term loan or increase the credit available under a covered short-term loan unless the lender first makes a reasonable determination of the consumer’s ability to repay the covered short-term loan. The proposed provision would also have imposed a requirement to determine a consumer’s ability to repay before advancing additional funds under a covered short-term loan that is a line of credit, if such advance would occur more than 180 days after the date of a prior required determination. Proposed § 1041.9(b)(1) would have included parallel provisions to proposed § 1041.5(b)(1) as applied to covered longer-term loans, except for certain conditional exemptions that are discussed above in connection with final § 1041.3(d)(7) and (8). Proposed § 1041.5(b)(1) would have required the ability-to-repay determination before a lender actually takes one of the triggering actions. The Bureau recognized in the proposal that lenders decline covered loan applications for a variety of reasons, including to prevent fraud, avoid possible losses, and to comply with State law or other regulatory requirements. Accordingly, the requirements of proposed § 1041.5(b)(1) would not have required a lender to make the ability-to-repay determination for every covered short-term loan application it receives, but rather only before taking one of the enumerated actions with respect to a covered short-term loan. Similarly, the Bureau explained in the proposal that nothing in proposed § 1041.5(b)(1) would have prohibited a lender from applying screening or underwriting approaches in addition to those required under proposed § 1041.5(b) prior to making a covered short-term loan. Proposed § 1041.5(b)(1)(ii) would have provided that, for a covered short-term loan that is a line of credit, a lender must not permit a consumer to obtain an advance under the line of 687 credit more than 180 days after the date of a prior required determination, unless the lender first makes a new reasonable determination that the consumer has the ability to repay the covered short-term loan. As the Bureau wrote in the proposal, under a line of credit, a consumer typically can obtain advances up to the maximum available credit at the consumer’s discretion, often long after the covered loan was consummated. Each time the consumer obtains an advance under a line of credit, the consumer becomes obligated to make a new payment or series of payments based on the terms of the covered loan. But when significant time has elapsed since the date of a lender’s prior required determination, the facts on which the lender relied in determining the consumer’s ability to repay may have changed significantly. As the Bureau explained in the proposal, during the Bureau’s outreach to industry, the Small Dollar Roundtable urged the Bureau to require a lender to periodically make a new reasonable determination of ability to repay in connection with a covered loan that is a line of credit. The Bureau stated in the proposal that it believed that the proposed requirement to make a new determination of ability to repay for a line of credit 180 days following a prior required determination appropriately balanced the burden on lenders and the protective benefit for consumers. 688 Reasonable determination. Under § 1041.5(b)(1) of the proposed rule, a lender would have to make a reasonable determination that a consumer will be able to repay a covered shortterm loan according to its terms. A consumer would have the ability to repay a covered shortterm loan according to its terms, under the proposed rule, only if the consumer is able to make all payments under the covered loan as they fall due while also making payments under the consumer’s major financial obligations as they fall due and continuing to meet basic living expenses during the shorter of the term of the loan or 45 days following consummation. The proposed rule would have also required that the lender determine if, for a period of 30 days after making the highest payment on the loan, the consumer will be able to pay major financial obligation as they fall due, make any remaining payments under the loan, and meet basic living expenses.817 Proposed comment 5(b)-1 would have provided an overview of the baseline methodology that would be required as part of a reasonable determination of a consumer’s ability to repay in proposed §§ 1041.5(b)(2) and (c) and § 1041.6. As noted in the general discussion of proposed § 1041.5(b), above, proposed comment 5(b)-2 would have identified standards for evaluating whether a lender’s ability-to-repay determinations under proposed § 1041.5 are reasonable. It would have clarified the minimum requirements of a reasonable ability-to-repay determination; identified assumptions that, if relied on by the lender, would render a determination not reasonable; and established that the overall 817 Under proposed § 1041.9(b)(2) and comments 9(b)(2)(i)-1 and 9(b)(2)(ii)-1, the focus for analyzing covered longer-term balloon-payment loans would have been on two similar periods: (1) the month with the highest sum of loan payments; and (2) the 30 days after the single highest payment on the loan. 689 performance of a lender’s covered short-term loans is evidence of whether the lender’s determinations for those loans are reasonable. The Bureau explained in the proposal that the proposed standards would not have imposed bright-line rules prohibiting covered short-term loans based on fixed mathematical ratios or similar criteria. Moreover, the Bureau stated that it did not anticipate that a lender would need to perform a manual analysis of each prospective loan to determine whether it meets all of the proposed standards. Instead, the Bureau explained that each lender would be required under proposed § 1041.18 to develop and implement policies and procedures for approving and making covered loans in compliance with the proposed standards and based on the types of covered loans that the lender makes. The Bureau noted in the proposal that a lender would then apply its own policies and procedures to its underwriting decisions, which the Bureau anticipated could be largely automated for the majority of consumers and covered loans. 690 Minimum requirements. Proposed comment 5(b)-2.i set out some of the specific respects in which a lender’s determination must be reasonable under the proposed rule with respect to covered short-term loans. For example, it noted that the determination must include the applicable determinations provided in proposed § 1041.5(b)(2), be based on reasonable projections of a consumer’s net income and major financial obligations in accordance with proposed § 1041.5(c) and be based on reasonable estimates of a consumer’s basic living expenses (which were further clarified under proposed comment 5(b)-4). It would also have to be consistent with the lender’s written policies and procedures required under proposed § 1041.18(b) and must be grounded in reasonable inferences and conclusions in light of information the lender is required to obtain or consider. Proposed comment 5(b)-2.i would have clarified that for a lender’s ability-to-repay determination to be reasonable, the lender must appropriately account for information known by the lender, whether or not the lender is required to obtain the information under proposed § 1041.5, that indicates that the consumer may not have the ability to repay a covered short-term loan according to its terms. For example, the Bureau explained, proposed § 1041.5 would not have required a lender to inquire about a consumer’s individual transportation or medical expenses, but if the lender learned that a particular consumer had a transportation or recurring medical expense that was dramatically in excess of the amount the lender used to estimate basic living expenses for consumers generally, proposed comment 5(b)-2.i would have clarified that the lender could not ignore that fact. The Bureau wrote in the proposal that, instead, it would have to consider the transportation or medical expense and then reach a reasonable determination that the expense did not negate the lender’s otherwise reasonable ability-to-repay determination. 691 For covered longer-term loans, proposed comment 9(b)-2.i would have paralleled comment 5(b)-2.i in all respects except for the addition of proposed comment 9(b)-2.i.F, would have provided that for covered longer-term loans, the reasonable determination must include appropriately accounting for the possibility of volatility in the consumer’s income and basic living expenses during the term of the loan, with a cross-reference to proposed comment 9(b)(2)(i)-2. Determinations that are not reasonable. Proposed comment 5(b)-2.ii would have provided an example of an ability-to-repay determination that is not reasonable for covered short-term loans. The example, in proposed comment 5(b)-2.ii.A, was a determination that relies on an assumption that the consumer will obtain additional consumer credit to be able to make payments under the covered short-term loan, to make payments under major financial obligations, or to meet basic living expenses. The Bureau stated in the proposal that it believed that a consumer whose net income would be sufficient to make payments under a prospective covered short-term loan, to make payments under major financial obligations, and to meet basic living expenses during the applicable period only if the consumer supplements that net income by borrowing additional consumer credit is a consumer who, by definition, lacks the ability to repay the prospective covered short-term loan. Similarly, proposed comment 9(b)-2.ii would have included two examples of unreasonable ability-to-repay determinations with respect to covered longer-term loans. The first example, proposed comment 9(b)-2.ii.A, was a parallel example to proposed comment 5(b)2.ii.A. The second example, in proposed comment 5(b)-2.ii.B, would have clarified that an unreasonable ability-to-repay determination is one that relies on an assumption that a consumer 692 will accumulate savings while making one or more payments under a covered longer-term loan and that, because of such assumed future savings, will be able to make a subsequent loan payment under a covered longer-term loan. The Bureau explained in the proposal that, like the prior comment, the Bureau is including this comment in an abundance of caution lest some lenders seek to justify a decision to make, for example, a multi-payment, interest-only loan with a balloon payment on the ground that during the interest-only period the consumer will be able to accumulate savings to cover the balloon payment when due. The Bureau explained further in the proposal that a consumer who finds it necessary to seek a covered longer-term loan typically does so because she has not been able to accumulate sufficient savings while meeting her existing obligations and expenses. The Bureau noted in the discussion in the proposal’s Market Concerns—Longer-Term Loans section regarding the high incidence of re-borrowing and refinancing coinciding with balloon payments under longer-term loans strongly and stated that it suggests that consumers are not, in fact, able to accumulate sufficient savings while making lower payments to then be able to make a balloon payment. The Bureau wrote in the proposal that a projection that a consumer will accumulate savings in the future is purely speculative, and basing an ability-to-repay determination on such speculation presents an unacceptable risk of an erroneous determination. The Bureau explained that believed that basing a determination of a consumer’s ability to repay on such speculative projections would not be reasonable. Performance of covered loans as evidence. The Bureau stated in the proposal that in determining whether a lender has complied with the requirements of proposed § 1041.5, there is a threshold question of whether the lender has carried out the required procedural steps, for example by obtaining consumer statements and verification evidence, projecting net income and 693 payments under major financial obligations, and making determinations about the sufficiency of a consumer’s residual income. The Bureau explained that in some cases, a lender might have carried out these steps but still have violated § 1041.5 by making determinations that are facially unreasonable, such as if a lender’s determinations assume that the amounts a consumer needs to meet basic living expenses are clearly insufficient for that purpose. The Bureau explained further in the proposal that, in other cases, the reasonableness or unreasonableness of a lender’s determinations might be less clear. Accordingly, proposed comment 5(b)-2.iii provided that evidence of whether a lender’s determinations of ability to repay for covered short-term loans are reasonable may include the extent to which the lender’s determinations subject to proposed § 1041.5 result in rates of default, delinquency, and re-borrowing for covered short-term loans that are low, equal to, or high, as compared to the rates of other lenders making similar covered loans to similarly situated consumers. The Bureau stated in the proposal that proposed comment 5(b)-2.iii would not mean that a lender’s compliance with the requirements of proposed § 1041.5 for a particular loan could be determined based solely on the performance of that loan. Nor, the Bureau stated in the proposal, would this proposed comment mean that comparison of the performance of a lender’s covered short-term loans with those of other lenders could be the sole basis for determining whether that lender’s underwriting complies with the requirements of proposed § 1041.5. The Bureau wrote in the proposal that, for example, one lender may have default rates that are much lower than the default rates of other lenders because it uses aggressive collection tactics, not because its determinations of ability to repay are reasonable. The Bureau wrote that similarly, the fact that one lender’s default rates are similar to the default rates of other lenders does not necessarily 694 indicate that their determinations of ability to repay are reasonable; the similar rates could instead reflect that their respective determinations of ability to repay are similarly unreasonable. The Bureau wrote in the proposal that it believed, however, that such comparisons would provide important evidence that, considered along with other evidence, would facilitate evaluation of whether a lender’s ability-to-repay determinations are reasonable. The Bureau elaborated in the proposal that for example, a lender may use estimates for a consumer’s basic living expenses that initially appear unrealistically low, but if the lender’s determinations otherwise comply with the requirements of proposed § 1041.5 and otherwise result in covered short-term loan performance that is materially better than that of peer lenders, the covered short-term loan performance may help show that the lender’s determinations are in fact reasonable. Similarly, the Bureau wrote, an online lender might experience default rates significantly in excess of those of peer lenders, but other evidence may show that the lender followed policies and procedures similar to those used by other lenders and that the high default rate resulted from a high number of fraudulent applications. The Bureau stated in the proposal that, on the other hand, if consumers experience systematically worse rates of default, delinquency, and re-borrowing on covered short-term loans made by one lender, compared to the rates of other lenders making similar loans, that fact may be important evidence of whether that lender’s estimates of basic living expenses are, in fact, unrealistically low and therefore whether the lender’s ability-to-repay determinations are reasonable. With respect to covered longer-term loans, the discussion in the proposal’s section-bysection analysis for proposed § 1041.9(b) and comment 9(b)-2.iii paralleled the discussion above. 695 Payments under the covered short-term loan. Proposed comment 5(b)-3 noted that a lender is responsible for calculating the timing and amount of all payments under the covered short-term loan. The Bureau explained in the proposal that the timing and amount of all loan payments under the covered short-term loan were essential components of the required reasonable determination of a consumer’s ability to repay under proposed § 1041.5(b)(2)(i), (ii), and (iii). Calculation of the timing and amount of all payments under a covered loan was also necessary to determine which component determinations under proposed § 1041.5(b)(2)(i), (ii), and (iii) apply to a particular prospective covered loan. Proposed comment 9(b)-3 mirrored the discussion in comment 5(b)-3 with regard to payments under the covered longer-term loan. Basic living expenses. A lender’s ability-to-repay determination under proposed § 1041.5(b) would have been required to account for a consumer’s need to meet basic living expenses during the applicable period, while also making payments for major financial obligations and payments under a covered short-term loan. The Bureau explained in the proposal that if a lender’s ability-to-repay determination did not account for a consumer’s need to meet basic living expenses, and instead merely determined that a consumer’s net income is sufficient to make payments for major financial obligations and for the covered short-term loan, the Bureau believed the determination would greatly overestimate a consumer’s ability to repay a covered short-term loan and would be unreasonable. The Bureau further explained that doing so would be the equivalent of determining, under the Bureau’s ability-to-repay rule for residential mortgage loans, that a consumer has the ability to repay a mortgage from income even if that mortgage would result in a debt-to-income ratio of 100 percent. The Bureau stated in the 696 proposal that it believed there would be nearly universal consensus that such a determination would be unreasonable. However, the Bureau recognized in the proposal that in contrast with payments under most major financial obligations, which the Bureau stated it believes a lender can usually ascertain and verify for each consumer without unreasonable burden, it would be extremely challenging to determine a complete and accurate itemization of each consumer’s basic living expenses. Moreover, the Bureau stated, a consumer may be somewhat more able, at least in the short-run, to reduce some expenditures that do not meet the proposed definition of major financial obligations. For example, the Bureau noted that a consumer may be able for a period of time to reduce commuting expenses by ride sharing. Accordingly, the Bureau did not propose to prescribe a particular method that a lender would be required to use for estimating an amount of funds that a consumer needs to meet basic living expenses for an applicable period. Instead, proposed comment 5(b)-4 stated the principle that whether a lender’s method complies with the proposed § 1041.5 requirement for a lender to make a reasonable ability-to-repay determination depends on whether it is reasonably designed to determine whether a consumer would likely be able to make the loan payments and meet basic living expenses without defaulting on major financial obligations or having to rely on new consumer credit during the applicable period. Proposed comment 5(b)-4 provided a non-exhaustive list of methods that may be reasonable ways to estimate basic living expenses. The first method was to set minimum percentages of income or dollar amounts based on a statistically valid survey of expenses of similarly situated consumers, taking into consideration the consumer’s income, location, and 697 household size. The Bureau explained in the proposal that this example was based on a method that several lenders had told the Bureau they use in determining whether a consumer will have the ability to repay a loan and is consistent with the recommendations of the Small Dollar Roundtable. The Bureau noted that the Bureau of Labor Statistics conducts a periodic survey of consumer expenditures that may be useful for this purpose. The second method was to obtain additional reliable information about a consumer’s expenses other than the information required to be obtained under proposed § 1041.5(c) to develop a reasonably accurate estimate of a consumer’s basic living expenses. The Bureau explained in the proposal that this example was not meant to suggest that a lender would be required to obtain this information, but was intended to clarify that doing so may be one effective method of estimating a consumer’s basic living expenses. The Bureau wrote that the method described in the second example may be more convenient for smaller lenders or lenders with no experience working with statistically valid surveys of consumer expenses, as described in the first example. The third example was any method that reliably predicts basic living expenses. The Bureau wrote that it was proposing to include this broadly phrased example to clarify that lenders may use innovative and data-driven methods that reliably estimate consumers’ basic living expenses, even if the methods are not as intuitive as the methods in the first two examples. The Bureau wrote that it expected to evaluate the reliability of such methods by taking into account the performance of the lender’s covered short-term loans in absolute terms and relative to other lenders, as discussed in proposed comment 5(b)-3.iii. Proposed comment 5(b)-4 also provided a non-exhaustive list of unreasonable methods of determining basic living expenses. The first example was a method that assumes that a 698 consumer needs no or implausibly low amounts of funds to meet basic living expenses during the applicable period and that, accordingly, substantially all of a consumer’s net income that is not required for payments for major financial obligations is available for loan payments. The second example was a method of setting minimum percentages of income or dollar amounts that, when used in ability-to-repay determinations for covered short-term loans, have yielded high rates of default and re-borrowing, in absolute terms or relative to rates of default and re-borrowing of other lenders making covered short-term loans to similarly situated consumers. Proposed comment 9(b)-4 would have paralleled the language of proposed comment 5(b)-4, and the relevant discussion in the proposal’s section-by-section analysis regarding this comment mirrored the discussion above. Comments Received The Bureau received a significant amount of comments on the standard set forth in § 1041.5(b)(1). The Bureau first addresses comments focused on the general ability-to-repay requirement itself, and then separately discusses comments received regarding the standards for assessing reasonableness of the ability-to-repay requirements, including proposed commentary in 5(b)-2. General ability-to-repay requirement. A wide spectrum of commenters wrote in support of the ability-to-repay requirement as a general matter, including a group of United States Senators, a number of State Attorneys General, many local and State elected officials, civil rights organizations, faith groups and individual clergy members, other advocacy organizations, numerous individual consumers writing as part of organized comment campaigns, and other commenters. Relatedly, consumer groups agreed with the Bureau’s basic premise in the 699 proposal that true ability to repay on a covered loan is not determined merely by whether a consumer repays the loan, but rather by whether the consumer has the ability to repay the loan, major financial obligations, and basic living expenses without the need to re-borrow. In fact, some consumer groups urged Bureau to revise the general ability-to-repay requirement in § 1041.5(b)(1) to read “ability to repay the loan according to its terms while meeting other obligations and expenses and without re-borrowing” to more expressly reflect that the standard was not just focused on lenders’ ability to collect payments from consumers no matter what the downstream consequences. These commenters cited statutory and regulatory language as precedent, such as language from HOEPA and the Federal Reserve Board’s higher-priced mortgage loan rule. Commenters who criticized the general reasonableness standard in proposed §§ 1041.5(b)(1) and 1041.9(b)(1) were split as to whether it was too vague, particularly as to the use of loan performance as a factor of the analysis, or too prescriptive, particularly in mandating specific upfront procedures. In one camp, several commenters objected generally to the use of a reasonableness standard, arguing that it is overly vague and would create uncertainty for compliance and examination. A group of State banking regulators commented that the proposed ability-to-repay requirement would be difficult to enforce because of the uncertain standards for making a reasonable determination. Other commenters criticized the proposal for not specifying the expected level of residual income that would be necessary for a determination to be reasonable. Some commenters referred to the lack of clarity on both front-end and performance standards as creating a “gotcha” regime. 700 On the other hand, some commenters argued that the final rule should be less prescriptive and designed to provide flexibility for innovation. A lender and a policy and research organization both argued that the Bureau’s rule should embrace a “sandbox” or pilot approach to the ability-to-repay requirements that would test policy interventions in the market before enshrining them into specific rules. One of these commenters suggested that a sandbox could, for example, be used to “test out and ‘right-size’” a payment-to-income or payment-to-deposits approach to underwriting. The other suggested that the Bureau establish a process for approving data sources used in underwriting. Relatedly, several commenters argued that the rule should embrace a principles-based approach to the ability-to-repay requirements which leaves more flexibility to lenders on the process and more closely scrutinizes the outcomes. One commenter cited its experience lending in the United Kingdom and discussed how the U.K. Financial Conduct Authority (FCA) in recent years has imposed regulations on small-dollar loans that are non-prescriptive. This lender described how it had successful implemented the FCA regulations and encouraged the Bureau to consider such an approach in this rulemaking. A number of commenters argued that the Bureau should create an exception or safe harbor to the rule for various scenarios, including for unusual, non-recurring, or emergency expenses. A group of State Attorneys General writing in opposition to the proposal questioned the Bureau’s reasoning for declining to create such an exemption. They argued that creating an exception for unusual circumstances—such as where a consumer has a documented medical emergency or a necessary furnace repair during the winter—would be no more difficult to implement than the proposal’s other requirements such as income and expense verification. 701 They argued that such an exemption would be invoked rarely, and also would provide States with more flexibility to impose their own requirements. They argued that failing to provide for an exception is “particularly incongruous” given that the proposal would require lenders to consider unusual expenses in determining a consumer’s ability to repay, citing the section-bysection analysis describing proposed comment 5(b)-2.i.E. Several commenters argued that the Bureau had failed to take into account a factor that lenders are currently using in their basic underwriting models—willingness to repay. These commenters argue that willingness to repay is often indicative of whether a consumer will default, and several commenters provided data regarding default rates. Several commenters discussed proposed comment 5(b)-2.i.E, which would have clarified that a reasonable determination includes the lender appropriately accounting for information known to the lender indicating the consumer may not have the ability to repay, even if the lender is not required to obtain the information. Consumer advocates urged that this language be included in the regulatory text. They also asked that the language be broadened to provide that “information known to the lender” include the following: (1) information on the national consumer report or registered information system reflecting delinquencies or defaults on covered loans, other forms of credit or debt obligations, basic living expenses within the past year; and (2) a pattern of re-borrowing known to the lender. A group of State Attorneys General commenting on the proposal interpreted this proposed comment to mean the rule would require lenders to consider unusual expenses in determining a consumer’s ability to repay. With regard to treatment of open-end lines of credit specifically under proposed § 1041.5(b)(1)(ii), consumer groups commenting on the rule also urged the Bureau to treat each 702 advance on a covered loan that is an open-end line of credit as a new loan for purposes of the ability-to-repay requirement. They expressed concern about the risks of open-end credit lines that are covered loans and believed the rule should have stricter requirements to prevent evasion and debt traps. One commenter, a State trade group representing open-end credit providers, took the opposite view. This commenter argued that the Bureau should exempt open-end lines of credit from the proposal and, in the alternative, that the Bureau should either address open-end lines of credit in a separate rulemaking along with credit cards or apply the requirements of the CARD Act in connection with open-end lines of credit that are covered in this rule. This commenter also argued that the condition under § 1041.5(b)(1)(ii) imposing a requirement to conduct an additional ability-to-repay determination after 180 days would contravene the definition of openend credit under Regulation Z, 12 CFR 1026.2(a)(20), which has a replenishment element. This commenter also argued that the proposal did not address the parameters for when the open-end credit provider can increase the amount of the line or when the consumer no longer has the ability to repay amounts outstanding after 180 days due to a deterioration of the consumer’s income or increase in expenses. Performance of a lender’s loans as evidence of ability to repay. As discussed briefly above, the Bureau received a substantial number of comments focusing specifically on proposed comment 5(b)-2.iii, which would have clarified that certain portfolio-wide backward-looking metrics of loan performance such as a lender’s re-borrowing and default rates, may be indicative of whether a lender’s determinations of ability to repay are reasonable. 703 Some commenters objected to the use of loan performance data, for instance by arguing that the use of performance metrics would unfairly penalize lenders for choices made by consumers. A number of commenters also argued that use of defaults or other metrics as measures of reasonableness could lead to unintended consequences, like creating a heightened incentive to aggressively collect delinquent loans. Several commenters also took particular issue with the Bureau’s use of defaults as a performance metric. Other commenters did not disagree that loan performance was potentially relevant to the question of whether a lender had made a reasonable determination of a consumer’s ability to repay the loan, but urged the Bureau to provide more concrete guidance. Several commenters encouraged the Bureau to set objective performance metric standards rather than relying on clarifying principles in commentary. For instance, a group of consumer advocates wrote that the Bureau should set a 5 percent default rate for vehicle title loans and payroll deduction loans and a 10 percent default rate for payday loans as thresholds that, if exceeded by the lender on a portfolio basis, would trigger heightened scrutiny of the lender’s practices to determine whether the ability-to-repay determinations are unreasonable.818 They also suggested that lenders whose loan performance exceeds those benchmarks would potentially be subject to enforcement actions or other required steps to mitigate such as refunding late fees, waiving back interest, or reducing loan principal. Another commenter similarly argued for the Bureau to treat lenders with a portfolio default rate on covered loans above 10 percent with heightened scrutiny. Other 818 In justifying the suggested default rate thresholds, consumer advocates made several arguments: that the 10 percent default rate threshold for payday loans was double the default rate chosen by the Bureau in the proposed conditional exemption for covered longer-term loans under proposed § 1041.12; that mainstream credit products have single-digit default rates; that the leveraged payment mechanism substantially lowers the default rate lenders otherwise would experience; and, that vehicle title loans present unique harms justifying an even lower threshold. 704 commenters argued that the Bureau should add more examples about the patterns of reborrowing that would be indicative of unreasonable ability-to-repay determinations. Some commenters actively advocated to use particular metrics. One commenter, a research and policy organization, generally supported the approach to use default data as a metric for evaluating ability to repay, stating that the clearest proof of effective underwriting processes should be found in consumer repayment outcome data rather than by assessing inputs into the product design alone. This commenter also argued that first-payment defaults would be a key indicator for the success of an underwriting model because absent fraud they clearly points to a mis-calibration in underwriting. Others argued that the Bureau should look to see whether consumers met expenses during the 30 or 60 days following the highest or last payment. Consumer groups also provided a list of additional performance metrics that they urged the Bureau to monitor as indicative of deficient ability-to-repay analyses, such as failed payments, late payments, requests for forbearance, aggressive collection practices, indications of consumers’ overdrafting or having trouble paying other expenses, and the extent of consumer injury (which they argued was influenced by a number of factors including late fees, debt collection practices, the interest rate and for how long interest was charged, and whether the lender sells or sues on the debt). In contrast, other commenters who generally supported the proposal and the reasonableness approach criticized the proposed comment 5(b)-2.iii for very different reasons and in particularly strong terms. These commenters objected to the language in the proposed comment suggesting that a review of the comparative performance metrics among lenders would be relevant to the evaluation of ability to repay. They suggested that this approach would 705 perpetuate high default or delinquency rates by incentivizing lenders to achieve only marginally better results than their competitors rather than meaningfully improved performance. A group of consumer advocates wrote that this provision was “among the most dangerous parts of the proposal” and “strongly impl[ies] that the metric for evaluating loan performance is simply not to be the worst of the worst.” The commenters noted the Bureau’s statements in the section-bysection analysis for the proposal that comparative performance metrics could not be the sole basis for a reasonableness determination and that factors such as aggressive collection efforts could be the reason for one lender’s default rates to appear lower than another, rather than ability to repay, but they argued that such statements were cautionary and would “be exploited.” Other commenters, including a large number of individual commenters writing as part of organized commenter campaigns, expressed concern that this provision would be a “business as usual loophole.” However, one commenter expressed support for the language regarding comparative performance metrics, arguing that such an analysis of comparative loan performance would help control for macroeconomic shifts that could affect large groups of consumers similarly. Final Rule The Bureau finalized the text of § 1041.5(b)(1) with adjustments to apply it to covered longer-term balloon-payment loans and a change to the time period for re-underwriting open-end lines of credit from every 180 days to every 90 days. The justification for this latter change is discussed below in the context of the Bureau’s response to comments asking for additional protections regarding open-end credit products covered by the proposal. The Bureau concluded that it was not necessary to further revise the regulation text in § 1041.5(b)(1) to refer expressly to consumers repaying the covered loan while meeting other obligations and expenses and 706 without re-borrowing, as these elements are expressly addressed in various other parts of the regulation text and commentary. The Bureau also made minor adjustments to the regulation text and commentary for clarity and conformity, such as to reflect policy decisions discussed elsewhere to permit lenders to analyze either a consumer’s debt-to-income ratio or residual income for the relevant monthly period and to cross reference other relevant commentary.819 In addition, the Bureau is making several substantive changes to the commentary to address various concerns raised in comments on the proposal. Specific elements of the ability-to-repay analysis. The Bureau made a number of substantive changes to the commentary for final § 1041.5(b)(1)(i) to address specific concerns about specific elements of the ability-to-repay test. First, with regard to basic living expenses, the Bureau has significantly revised comment 5(b)-2.i.C to elaborate on the estimation methods posited in the proposal. The Bureau did so in part in response to comments and also because of the Bureau’s decision to consolidate this comment with proposed comment 5(b)-4. The Bureau is not finalizing proposed comment 5(b)-4 because it believes it had some redundancy with other commentary language on basic living expenses, would have added complexity, and would have created some tension with comment 5(b)-2.i and -2.ii. The Bureau has chosen to harmonize the language regarding reasonable estimates of basic living expenses into one comment under § 1041.5(b). 819 For example, the Bureau revised final comment 5(b)-3 to reflect that the calculation of payments under the covered short-term loan or covered longer-term balloon loan focuses on the payments due during the relevant monthly period. 707 Specifically, comment 5(b)-2.i.C now has two subparagraphs. Comment 5(b)-2.i.C.1 emphasizes that the final rule does not specify a particular method that must be used to estimate basic living expenses, and that the lender is not required to itemize them for individual consumers. The comment goes on to clarify that a lender may instead arrive at estimates for the amount needed to cover the six categories of costs identified in § 1041.5(a)(1) based upon such sources as the lender’s own experience in making covered short-term loans or covered longerterm balloon-payment loans to similarly-situated consumers, reasonably reliable information available from government surveys or other publications about the basic living expenses of similarly-situated consumers, or some combination thereof. The Bureau disagrees with commenters who argued that the Bureau should require itemization, as that would create potentially substantial burdens for lenders and consumers and make automation harder. With regard to reliance on government sources, the comment also specifically clarifies that it would be reasonable for the lender to use data about the amounts spent on the six categories of basic living expenses identified in comment 5(a)(1)-2 from the IRS Collection Financial Standards or the CEX to develop non-individualized estimates of basic living expenses. However, the comment also notes that in using the data from those sources to estimate the amount spent on a particular category, the lender may make reasonable adjustments to arrive at an estimate of basic living expenses, for instance where a data source’s information on a particular type of basic living expenses overlaps with a type of major financial obligation as defined in § 1041.5(a)(3) or where a source groups expenses into different categories than comment 5(a)(1)-2. 708 As discussed above in connection with the final commentary to § 1041.5(a)(1), the Bureau intends to make clear that lenders have flexibility to make reasonable non-individualized estimates of basic living expenses and that, in doing so, they can rely on their own experience in estimating basic living expenses for similarly-situated consumers or upon governmental survey or data sources, some of which are now listed as examples. At the same time, for the reasons discussed above, while the Bureau believes that it would be reasonable for lenders to rely on either the IRS Collection Financial Standards or the Consumer Expenditure Survey, there is reason to believe that both may be over-inclusive or reflect some differences as to expense categorization. The Bureau believes it is therefore appropriate to emphasize that further reasonable adjustments are permitted to estimates that are primarily based on such sources. These changes are in part responsive to comments asserting that the standards in proposed comment 5(b)-4, which were consolidated with this comment, were too vague. The Bureau also has not finalized language in comment 5(b)-4 that would have referenced an example of reasonable basic living expense estimates being based on a survey taking into consideration a consumer’s income, household size, and location. The Bureau received a number of questions and comments about these categories, including those suggesting that consideration of location and household size would implicate fair lending law issues. As noted earlier, the Bureau does not believe estimates based on these categories would raise fair lending law issues, and the Bureau believes it will be difficult for lenders to arrive at reasonable estimates that apply without regard to household size or, for lenders operating in multiple States, without regard to differences in living costs. However, the Bureau believes including 709 commentary language of this sort might suggest that the final rule requires more precision in estimating than the Bureau intends. The Bureau has also added a comment 5(b)-2.i.C.2 regarding basic living expenses. This comment provides that if the lender is conducting an individualized estimate by itemizing the consumer’s basic living expenses (which earlier commentary clarifies the lender is not required to do), the lender may reasonably consider other factors specific to the consumer that are not required to be projected under § 1041.5(c). The comment clarifies that this could include whether other persons are regularly contributing toward the payment of basic living expenses. The comment clarifies that the lender can consider such consumer-specific factors only when it is reasonable to do so, and further notes that it is not reasonable for the lender to consider whether other persons are contributing toward the consumer’s payment of basic living expenses if the lender is also separately including in its projection of net income any income of another person to which the consumer has a reasonable expectation of access. As discussed above, the Bureau has made these changes to this comment based on comments to the proposal arguing that lenders should be permitted to account for the fact that other persons besides consumers themselves sometimes contribute to pay basic living expenses. The Bureau notes that it is permitting consideration of consumer-specific factors only if the lender is making an individualized determination. The Bureau believes it would be unworkable operationally and also potentially create a loophole if consumer-specific factors were permitted to be considered when the lender makes non-individualized estimates of basic living expenses. For example, the Bureau would be concerned if lenders developed a model for estimating basic living expenses that applied to all of their consumers or relevant subsets of them, and the model 710 assumed that a percentage of basic living expenses is always paid by persons other than the consumer. The comment also reflects the Bureau’s policy concern that if lenders were able to count both the income of another person to which the consumer has a reasonable expectation of access and assume that the consumer’s basic living expenses were being paid by that same person, it could result in a double-counting problem and an artificial inflation of net income (or deflation of basic living expenses); that is, the same income of another person to which the consumer claims access could be the income being used to pay for the consumer’s expenses. The Bureau believes it is a reasonable response to the comments asking for flexibility on this point to permit lenders to do one or the other—consider payment of basic living expenses by another person toward the estimate, or count as net income the other person’s income to which the consumer has a reasonable expectation of access. The Bureau also has decided not to finalize comment 5(b)-2.i.E, which would have stated that for a reasonable determination of ability to repay, the lender must appropriately account for information known by the lender whether or not the lender is required to obtain the information. The Bureau believes that this language created potential tension with other commentary indicating that lenders need not individually analyze basic living expenses because it would potentially have required substantial individual follow up that would negate the decision to allow lenders to rely on survey data and other generalized sources. The Bureau believes there is even more potential for this risk under the final rule, given that it now also allows lenders to rely on their historical experiences. The Bureau is therefore not finalizing the comment, but notes that it has had added other commentary as discussed separately below clarifying that lenders must, for example, take into account major financial obligations that consumers list on their written 711 statements even if those items are not reported on other sources. The Bureau believes that this more tailored guidance in particular circumstances will be more helpful to lenders in reconciling information from multiple sources. As such, the Bureau is declining the consumer groups’ suggestion to embed concepts into the rule that were discussed in the proposal’s section-bysection analysis for this proposed comment. General reasonableness standard. More generally, with regard to comments that expressed broader concerns about prescriptiveness, vagueness, and flexibility under § 1041.5(b)(1)’s reasonableness standard, the Bureau has made a number of adjustments to the commentary. First, the Bureau has expanded comment 5(b)-2.ii to provide more examples of front-end underwriting that would not meet the reasonableness standard. In addition, as discussed separately below, the Bureau added substantial additional text to comments 5(b)-2.iii regarding consideration of loan performance and added a new comment 5(b)-2.iv with illustrative examples of how the factors in 5(b)-2.iii would be used to evaluate the reasonableness of ability-to-repay determinations on the back end. The latter two comments are discussed separately below. With regard to comment 5(b)-2.ii, the Bureau has added a new subparagraph B to clarify that a lender’s determination would not be reasonable if it assumed a consumer needs implausibly low amounts or percentages of funds to meet basic living expenses. In the proposal, this language appeared in proposed comment 5(b)-4.ii.A, but the Bureau moved it for purposes of the final rule and revised it to address an example where a lender makes an unreasonable ability-to-repay determination by making a loan to consumer with a 90 percent debt-to-income ratio. The Bureau is adding this example in part to address the comments that the proposal did 712 not provide any indication of what thresholds would be considered sufficient for purposes of a reasonable ability-to-repay determination. The Bureau believes that a debt-to-income ratio in the range of 90 percent would not leave sufficient net income to cover consumers’ basic living expenses for purposes of this requirement. However, more generally, the Bureau is finalizing the general framework of considering whether an entity’s ability-to-repay determinations are reasonable. Reasonableness is a widely used legal concept in both State and Federal law, and is what Congress required with respect to the underwriting of mortgages, and so the Bureau believes the standard in the final rule—which, again, has been revised to include a substantial amount of new commentary clarifying how the reasonableness of ability-to-repay determinations will be evaluated—should provide a sufficiently discernible standard. The Bureau also declines to set more specific parameters about the level of residual income or debt-to-income ratio that would be considered reasonable or unreasonable for purposes of § 1041.5(b). Outside of extreme cases such as a 90 percent debt-to-income ratio, the Bureau believes that with regard to individual determinations of ability to repay, the acceptable level of residual income or debt-to-income ratio for a reasonable determination will depend on the circumstances. This question may also depend on whether lenders are using across-the-board DTI or residual income-thresholds or whether they are sorting their consumers into different categories and applying different thresholds for acceptable levels of DTI or residual income for consumers within those categories. There may be some debt-to-income thresholds that are sufficiently low that it would be reasonable to use a uniform debt-to-income threshold for all of the lender’s customers, whereas as thresholds get higher it may be reasonable to apply the 713 threshold to only subsets of the lender’s customers (such as customers in higher income tiers). The overarching principle, of course, is that the lender must make reasonable determinations of consumers’ ability to repay. Moreover, as discussed below, the Bureau believes that at least for lenders who follow the procedural requirements set forth in § 1041.5(c), the primary evidence with respect to the reasonableness of a lenders’ determinations will be the pattern of outcomes for consumers found to have the ability to repay. That is why the Bureau is adding detailed commentary to 5(b)-2.iii and a new comment 5(b)-2.iv clarifying the performance factors that would be reviewed for purposes of assessing reasonableness and giving examples. The Bureau declines the suggestion by some commenters to take a “sandbox” approach to components of the ability-to-repay requirement. The Bureau as a general matter supports innovation and policy experimentation through Project Catalyst and other initiatives. It simply does not believe this rulemaking is the best candidate for such an approach. Given the nature of the Federal rulemaking process and the particular history of this rulemaking—which has involved to date many years of study, outreach and deliberation, and where the effective date of the rule will not be for another 21 months after publication in the Federal Register—the Bureau is concerned that failing to finalize necessary components of the rule, such as the ability-to-repay requirement, and instead testing ideas in the market would not prove a fruitful value proposition in view of the further delays in finalizing the rule. Any policy ideas emanating from the sandbox would have to be reintegrated into a rulemaking process in any event, further forgoing valuable consumer protections in the Bureau’s view. With regard to the commenters suggesting a principles-based approach where outcomes are more important that procedures, the Bureau notes that the final rule strikes a balance between 714 a rules-based and an outcomes-based approach, with more emphasis than the proposal on the latter. First, the Bureau is taking a less prescriptive approach on certain key components of the ability-to-repay requirements, such as by permitting reasonable reliance on stated amounts for income in absence of reasonably available verification evidence. Second, as discussed below, the Bureau is expanding the discussion of how loan performance metrics will be used to evaluate ability-to-repay determinations. These changes reflect a greater emphasis on lender performance as a means of evaluating compliance with the ability-to-repay requirements. As to commenters asserting that the Bureau should allow for exceptions to the ability-topay framework for consumers who are seeking loans to pay for non-recurring, unusual, and emergency expenses, the Bureau declines this suggestion for several reasons. First, lenders will already have an alternative to § 1041.5 by lending under § 1041.6 of the final rule, which is not subject to the ability-to-repay requirements. That approach is available for consumers up to six times per year and can be used in any of the circumstances—including emergency situations— that the commenters noted, unless the consumer is in a cooling-off period. Second, the Bureau continues to believe that the policy challenges described in the proposal with crafting such an exception are profound, such as the difficulty of defining, by rule, unusual and emergency expenses, and disagrees that this would pose the same or less challenges as with the implementation of other aspects of the rule.820 Third, the Bureau believes that this type of exception would be extremely difficult to administer, for some the same reasons discussed in the section-by-section analysis for 820 While the proposal discussed the challenges to this exception in the context of alternatives considered to the presumption of unaffordability in proposed § 1041.6, the commenter referred to this language in the broader context of the ability-to-repay requirements. 715 § 1041.5(a)(5) in connection with suggestions made by other commenters to count the proceeds of the loan toward net income or as a credit against major financial obligations or basic living expenses. As discussed in the section-by-section analysis for § 1041.5(a)(5), the Bureau believes it is difficult if not impossible to construct a workable rule that would carve out from the requirement one type of usage case for a consumer—here, emergency expenses—but include other usage cases, such as payment of basic living expenses, given the fungibility of money, the potential intrusiveness of asking about why the consumer is taking out the loan, and the challenges of policing such a rule. Lastly, the Bureau does not agree that this exception would be used sparingly. This assertion contravenes empirical evidence, assertions by other commenters including many individual consumers, and lender advertising about the purpose of the loans.821 Moreover, the difficulty of enforcing this type of provision would create an incentive for evasion, where consumers simply state a reason that would fall under the exception and lenders accept that reason without further inquiry. Performance of a lender’s loans as ability to repay. As noted above, the Bureau received many comments asking for additional guidelines and clarity on what constitutes a reasonable ability-to-repay determination, including in some cases numerical thresholds above which would trigger heightened scrutiny or even consumer remedies. The Bureau appreciates the concerns raised by the commenters and has substantially expanded the language in comment 5(b)-2.iii and added new comment 5(b)-2.iv to further clarify how it will use loan performance metrics and 821 As noted in the section-by-section analysis for § 1041.5(a)(5) in discussion of the loan proceeds issue, the Bureau received many comments, including a large number from individual consumers, describing how consumers often use payday loans and other covered loans to cover their new needs or emergency expenses; payday lenders in their advertising tend to cite this usage category as the primary purpose for using the product; and academic literature and surveys discussing usage patterns on payday loans have consistently found that a sizable number of consumers report using payday loans and other covered loans for non-recurring and emergency expenses. 716 analysis in assessing whether a lender’s determinations of consumers’ ability to repay are reasonable. The specifics of the revised language are described in more detail below. The Bureau is declining, however, to provide a prescriptive standard or exhaustive list of factors that would show reasonableness, or a set of numerical thresholds tied to the factors such as a specific default rate that would constitute a per se violation or grounds for closer scrutiny. While the Bureau understands that reasonableness tests and multi-factor back-end performance metrics, without specific numerical thresholds, may not give lenders perfectly clear direction on how exactly to underwrite, the Bureau believes that on balance the more prudent option at this time is to preserve the principles-based approach of the proposal but add detail and illustrations. The Bureau believes it may be challenging to set thresholds that would apply across the board, given that lenders who make unaffordable loans may experience different rates of default, reborrowing, and other harms depending on collections practices and other factors. Furthermore, the Bureau also does not believe there is enough evidence at this time to codify specific numerical thresholds for default rates, re-borrowing rates, and the like, given that the practices identified in this rule are market-wide and that there is not currently a Federal ability-to-repay rule for this market. And the Bureau is concerned that setting particular benchmarks at this time would incentivize lenders to take steps to manage their rates aggressively through enhanced debt collection or even to manipulate the metrics to fall just beneath the threshold, neither of which would be a beneficial result. Further, to the extent that consumer group commenters urged the Bureau to establish numeric thresholds for enhanced scrutiny of particular lenders rather than outright thresholds for per se violations, such as 5 percent default rates for vehicle title loans and employer-based loans 717 and 10 percent threshold for payday loans, such a policy decision would not be made as part of a rulemaking, but rather, in the Bureau’s prioritization decisions regarding supervision or enforcement activity as the market evolves over time in response to the rule and other business developments. As noted above, comment 5(b)-2.iii does state that default rates can provide evidence that a lender’s ability-to-pay determinations were not reasonable.822 The Bureau also declines some commenters’ request to change the ability-to-repay standard to one focused on willingness or propensity to pay. The Bureau recognizes that many lenders today already employ predictive underwriting tools to screen out those with a propensity to default, a point noted in some comments. However, the Bureau’s core concern in this rulemaking is the determination of whether consumers have the ability to repay, i.e., the financial capacity to make the loan payments, pay for major financial obligations, and meet basic living expenses. The Bureau expects that lenders will continue to utilize in their underwriting models various methods for detecting fraud or willingness to repay, and nothing in the final rule precludes that from happening as long as they comply with the requirements of this rule. The assertion made by some commenters that default and re-borrowing are caused simply by consumer choice and not at all by lender practices—including the identified unfair and abusive practice that is the Bureau’s focus in this rule—runs counter to the analysis provided above in Market Concerns—Underwriting and seems to contradict their own comments that their customers are often living paycheck to paycheck. 822 The Bureau also notes that with regard to the specific thresholds suggested by the consumer groups, the Bureau does not find the justification compelling that the Bureau should designate a 10 percent portfolio default rate for payday loans because it is double the 5 percent rate included as part of a larger set of conditions for a proposed exemption for longer-term, and generally lower-cost, loans—an exemption which the Bureau is not finalizing. Nor does the Bureau believe commenters provided a compelling reason for why there should be a separate, and more stringent, 5 percent threshold for vehicle title and employer-based loans. 718 Regarding the comments about the use of comparative performance metrics and how that would create a “business as usual loophole,” as an initial matter the Bureau agrees with the concern voiced by consumer advocates, individual consumers, and others about a rule that would judge the reasonableness of ability-to-repay determinations based solely (or primarily) on a comparison of loan performance across lenders. The Bureau did not intend to promulgate a standard that would evaluate loan performance simply on not being “the worst of the worst.” The Bureau expressly noted in the proposal that comparative metrics are not the sole basis of judging compliance, that lenders cannot rely on comparative performance to excuse poor loan performance as measured more objectively, and that comparatively lower default rates could be caused by factors extrinsic to ability-to-repay determinations (such as aggressive debt collection). To further underscore and memorialize this intent, the Bureau has revised comment 5(b)2.iii to state specifically that evidence about comparative performance is not dispositive as to the evaluation of a lender’s ability-to-repay determinations. Additionally, this comment has been revised more generally to provide a more expansive discussion of the types of performance metrics used to evaluate the reasonableness of ability-to-repay determinations, along with several examples in comment 5(b)-2.iv showing lending patterns that indicate either reasonable or unreasonable ability-to-repay determinations. The combination of these changes provides more clarity that the reasonableness of ability-to-repay determinations are to be measured over a variety of dimensions (e.g., default rates, re-borrowing rates, patterns of lending across loan sequences, and delinquency-related harms such as late fees); non-comparative measures of loan 719 performance will be primary; and comparative performance metrics will be complementary. These changes are discussed in detail below. However, the Bureau has decided not to eliminate reference to comparative performance metrics altogether, as requested by the consumer advocates and other commenters. Although as noted above the fact that a lender’s outcomes are not among the worst of its peers is not sufficient to establish that the lender is making reasonable ability-to-repay determinations, outlier outcomes surely are probative of the unreasonableness of a particular lender’s ability-to-repay determinations. That is the import of comment 5(b)-2.iii and 5(b)-2.iv. The Bureau agrees with the consumer advocates that evaluating the ability-to-repay determinations should involve looking at indicators beyond default rates. Again, revised comment 5(b)-2.iii provides additional clarification on the types of performance metrics that will be evaluated. The list of factors has been expanded from the proposal. The commentary states that a variety of factors may be relevant, including rates of default, patterns of re-borrowing within loan sequences, patterns of re-borrowing across loan sequences, rates of delinquencyrelated harms (e.g., late fees and failed presentments), and patterns of lenders making noncovered loans that bridge gaps between sequences of covered loans. The Bureau has also clarified that loan performance may be evaluated across the lender’s entire portfolio of covered short-term or longer-term balloon-payment loans, as well as with respect to particular products, geographic regions, time periods during which the loans were made, or other relevant categorizations. Finally, the Bureau provides several new illustrative examples of lending patterns that would indicate reasonable or unreasonable ability-to-repay determinations in 720 comment 5(b)-2.iv. More discussion and explanation of these revised commentary provisions are found below. Comment 5(b)-2.iii has been revised and expanded in a number of important ways. First, it now states that evidence that a lender’s determinations of ability to repay are not reasonable may include, without limitation, the factors described under paragraphs (A) through (E) of the comment. This change refers to how the comment now lists the factors in separate paragraphs rather than the main body of the comment for organizational purposes and due to the additional level of detail provided. Second, comment 5(b)-2.iii now clarifies that these factors may be evaluated across a lender’s entire portfolio of covered short-term loans or covered longer-term balloon-payment loans or with respect to particular products, geographic regions, particular time periods during which the loans were made, or other relevant categorizations, and clarifies that other relevant categorizations would include, without limitation, loans made in reliance on consumer statements of income in the absence of verification evidence. The Bureau believes that this approach is important to identify potential troublesome patterns insofar as lenders could not simply blend the categories of covered loans evidencing poor performance with other types of covered loans made by the lender with better performance. Third, the comment now clarifies that the factors may be considered either individually or in combination with one another; that the factors are not absolute in their application and instead exist on a continuum and may apply to varying degrees; and that each of the factors is viewed in the context of the facts and circumstances relevant to whether the lender’s ability-to-repay determinations are reasonable. Finally, the comment clarifies that relevant evidence may also include a comparison of the factors listed in the comment on the part of the lender to that of other lenders making covered 721 short-term loans or covered longer-term balloon-payment loans to similarly situated consumers, but that such evidence about comparative performance is not dispositive as to the evaluation of a lender’s ability-to-repay determinations. This revised language above is a response to the criticisms of the proposed comment 5(b)-2.iii language regarding comparative performance metrics as evaluative tools, as discussed above. Comment 5(b)-2.iii is then organized into five sub-paragraphs elucidating the factors that will be evaluated. Comment 5(b)-2.iii.A addresses default rates, clarifying that this evidence includes defaults during and at the expiration of covered loan sequences as calculated on a per sequence or per consumer basis. The Bureau believes that a per-loan basis for calculating default rates would not be as accurate for purposes of evaluating whether reasonable ability-to-repay determinations are being made, because then a lender’s re-borrowing rate would substantially distort the metric. For example, on a per loan basis, a consumer who re-borrows twice and then defaults would have one-third the impact on the default rate that a consumer who defaults after the first loan would, even though both loan sequences end the same way. The Bureau also notes that the consumer advocates in their joint comment letter urged that any default rate metric that is used should be a per-customer or per-sequence default rate, for similar reasons. Comment 5(b)-2.iii.B addresses re-borrowing rates, which the comment clarifies as including the frequency with which the lender makes consumers multiple covered short-term loans or covered longer-term balloon-payment loans within a loan sequence as defined in § 1041.2(a)(14), i.e., consecutive or concurrent loans taken out within 30 days of a prior loan being outstanding. As discussed in many places in the final rule, including Market Concerns— Underwriting and the section-by-section analysis for § 1041.4, the Bureau has identified repeat 722 re-borrowing as a problem in this market meriting intervention and is requiring lenders to determine whether consumers have the ability to repay a covered short-term or longer-term balloon-payment loan without the repayment triggering a need to re-borrow over the ensuing 30 days. Thus, within-sequence re-borrowing rates will be critical in evaluating compliance with the ability-to-repay determination, as that is one of the core consumer harms that the requirements of the final rule are aiming to prevent. Comment 5(b)-2.iii.C lists patterns of lending across loan sequences as a third factor and clarifies that this evidence includes the frequency with which the lender makes multiple sequences of covered short-term loans or covered longer-term balloon-payment loans to consumers. The comment clarifies that this evidence also includes the frequency with which the lender makes new covered short-term loans or covered longer-term balloon-payment loans immediately or soon after the expiration of a cooling-off period under § 1041.5(d)(2) or the 30day period that separates one loan sequence from another, referencing the loan sequence definition in § 1041.2(a)(14). As noted in the section-by-section analysis for § 1041.4, while the Bureau has established a 30-day period as the measure for determining whether a consumer is likely to be re-borrowing the prior loan, there are circumstances in which new loans beyond the 30-day period would also be the result of the unaffordability of a prior loan rather than the result of a new borrowing need. For example, if a consumer does not have funds to pay major financial obligations or basic living expenses as they come due because the consumer used income that would pay those obligations to pay off a covered short-term loan, and the consumer falls behind on an obligation during the month after repaying a short-term loan and then returns to obtain a new loan 31 days after the prior loan was repaid, that would effectively mean that the prior loan 723 was not affordable. A pattern of consumers frequently returning to take out a new loan immediately after the end of a cooling-off period would thus be relevant in assessing whether the lender’s ability-to-repay determinations were reasonable. Comment 5(b)-2.iii.D lists a fourth factor, rates of delinquencies and collateral impacts. The comment clarifies that this evidence includes the proportion of consumers who incur late fees, failed presentments, delinquencies, and repossessions. The Bureau believes that evaluating the rates of late fees, failed presentments, delinquencies, and repossessions is highly relevant to the evaluation of ability-to-repay determinations because those metrics would indicate that consumers are struggling to repay their loans, even if they do not necessarily wind up in default. The Bureau discusses the consumer harms associated with failed presentments in § 1041.7. Comment 5(b)-2.E lists a fifth factor, patterns of non-covered lending. The comment clarifies that this evidence includes the frequency with which the lender makes non-covered loans shortly before or shortly after consumers repay a covered short-term loan or covered longer-term balloon-payment loan, and the non-covered loan bridges all or a substantial part of either the time period between two loans that otherwise would be part of a loan sequence or of a cooling-off period. The comment lists an example where the lender, its affiliate, or a service provider frequently makes 30-day pawn loans to consumers shortly before or soon after repayment of covered short-term loans made by the lender, and where the lender then makes additional covered short-term loans to the same consumers soon after repayment of the pawn loans. The Bureau included this factor as a way to address concerns, discussed by the Bureau in the proposal, about the possibility of lenders using non-covered loans as a way of “bridging” gaps between the making of covered loans in order to evade the cooling-off period and other 724 aspects of the proposal. The proposal attempted to address this issue more directly through rule provisions justified under the Bureau’s Dodd-Frank Act anti-evasion authority,823 but as described in the discussion below of §§ 1041.5(d) and 1041.6(d), the Bureau is not finalizing these provisions due to concerns about their efficacy and complexity and to the Bureau’s decision to significantly streamline the re-borrowing restrictions that had been in proposed § 1041.6 based on public comments. Upon further consideration, however, the Bureau has realized that if lenders are making these “bridge” loans on a frequent basis, it may be an indication that the consumers are struggling to repay the preceding covered short-term or covered longer-term balloon-payment loan and therefore the underlying ability-to-repay determination on the earlier loan may have been unreasonable. The Bureau believes that revised comment 5(b)-2.iii provides a relatively comprehensive list of factors that broadly capture the types of ascertainable outcomes that would be useful in evaluating the reasonableness of lenders’ ability-to-repay determinations. As such, the Bureau declines to include all of the factors urged to be added by the consumer advocates, including the loan’s interest rate and the “extent and aggressiveness of the lender’s debt collection practices.” At least some of the examples suggested by the consumer groups would be very difficult if not impossible to measure quantitatively; others may be more aptly described as potential examples 823 The proposal would have defined “non-covered bridge loan” in proposed § 1041.2(a)(13) and provided in proposed § 1041.6(h) that if the lender or an affiliate made a non-covered bridge loan while a covered short-term loan under proposed § 1041.5 or § 1041.7 or a covered longer-term balloon-payment loan under proposed § 1041.9 was outstanding or for 30 days thereafter, the days during which a non-covered bridge loan is outstanding would not have counted toward any of the time periods in proposed § 1041.6, including the proposed 30-day cooling-off period following a three-loan sequence. More explanation of this provision and the reasons for why the Bureau is not finalizing it are found in the discussion of § 1041.5(d), below. 725 of evasion rather than indicators of unreasonable ability-to-repay determinations; and still others in the Bureau’s view are overly restrictive, such as the suggestion regarding interest rates. Other commenters’ suggestions about which metrics would be most indicative of a failure to make a reasonable ability-to-repay determination, such as first-payment defaults absent those due to fraud, are helpful and may help inform Bureau analyses once the rule takes effect. However, the Bureau is not at this time rank-ordering the metrics because it believes that, depending on the facts and circumstances, any one of the factors, or multiple factors working in tandem, may be indicative of whether an ability-to-repay methodology is unreasonable. As a complement to revised comment 5(b)-2.iii, the Bureau has also added a new comment 5(b)-2.iv. This comment contains four detailed examples of fact scenarios illustrating how the factors in comment 5(b)-2.iii might constitute evidence about whether lenders’ abilityto-repay determinations are reasonable under § 1041.5(b). The Bureau is including these examples as a further response to criticisms that proposed comment 5(b)-2.iii, and § 1041.5(b) more broadly, did not provide sufficient guidance on how reasonableness on ability-to-repay determinations would be evaluated. These examples are non-exhaustive. The examples focus on fact scenarios where lenders’ portfolios include multiple factors from comment 5(b)-2.iii and where the factors are present to varying degrees, thus illustrating how the factors will be evaluated in combination. The first example, in comment 5(b)-2.iv.A, describes a scenario in which a significant percentage of consumers who obtain covered short-term loans from a lender under § 1041.5 reborrow within 30 days of repaying their initial loan, re-borrow within 30 days of repaying their second loan, and re-borrow shortly after the end of the cooling-off period that follows the initial 726 loan sequence of three loans, and how, based on the combination of these factors, this evidence suggests that the lender’s ability-to-repay determinations are not reasonable. This example illustrates a pattern where the lender’s consumers experience frequent re-borrowing— specifically, where a significant percentage of the lender’s consumers take out a full sequence of three covered short-term loans and then return to borrow shortly after the end of the cooling-off period, beginning another sequence. This would implicate the factors in both comment 5(b)2.iii.B and 5(b)-2.iii.C. The second example, in comment 5(b)-2.iv.B, describes a scenario in which a lender frequently makes at or near the maximum number of covered short-term loans permitted under the conditional exemption in § 1041.6 to consumers early within a 12-month period (i.e., the loans do not require ability-to-repay determinations) and then makes a large number of additional covered short-term loans to those same consumers under § 1041.5 (i.e., the loans require ability-to-repay determinations) later within the 12-month period. The example assumes that the loans made under § 1041.5 are part of multiple loan sequences of two or three loans each and the sequences begin soon after the expiration of applicable cooling-off periods or 30-day periods that separate one loan sequence from another. The example clarifies that this evidence suggests that the lender’s ability-to-repay determinations for the covered short-term loans made under § 1041.5 are not reasonable. The example notes further that the fact that some of the loans in the observed pattern were made under § 1041.6 and thus are conditionally exempted from the ability-to-repay requirements does not mitigate the potential unreasonableness of the ability-torepay determinations for the covered short-term loans that were later made under § 1041.5. 727 This example is intended to illustrate the potential interaction of the provisions under § 1041.5 and § 1041.6 and how the reasonableness of the lender’s ability-to-repay determinations for loans made under § 1041.5 would be evaluated if the lender makes a combination of loans under the different provisions to consumers during a given time period. Here, the lender is making loans to many consumers more or less continuously throughout the year (i.e., long loan sequences, borrowing shortly after cooling-off periods expire), with the § 1041.6 loans made toward the beginning of the year and § 1041.5 loans made later in the year. This pattern suggests that the lender is not making reasonable ability-to-repay determinations for the loans made under § 1041.5. This is the case even though some of the loans in the pattern did not require such an ability-to-repay determination. Put another way, the mere fact that the first set of loans in the pattern did not require an ability-to-repay determination does not insulate the lender from scrutiny if the subsequent loans show a pattern of long loan sequences and frequent borrowing shortly after cooling-off periods expire. The third example, in comment 5(b)-2.iv.C, is a variation of the preceding example. The facts are that a lender frequently makes at or near the maximum number of loans permitted under § 1041.6 to consumers early within a 12-month period and then only occasionally makes additional covered short-term loans to those same consumers under § 1041.5 later within the 12month period, and that very few of those additional loans are part of loans sequences longer than one loan. The example clarifies that absent other evidence that the ability-to-repay determination is unreasonable (i.e., presence of the factors in comment 5(b)-2.iii.A through E), this evidence suggests that the lender’s ability-to-repay determinations for the loans made under § 1041.5 are reasonable. In contrast to the preceding example where the lender made a large number of 728 § 1041.6 loans and a large number of § 1041.5 loans within a given time period and the latter loans were made in long sequences and close in time (broken up only by the cooling-off periods), under this example the vast majority of loans are made under § 1041.6, and there is little to no evidence of re-borrowing on the § 1041.5 loans. Therefore, this pattern reflects the permissible maximization of lending under § 1041.6 and the incidental making of additional § 1041.5 loans within the given time period, a pattern that is not suggestive of unreasonableness. Comment 5(b)-2.iv.D contains the final example. The pattern described is that within a lender’s portfolio of covered short-term loans, a small percentage of loans result in default; consumers generally have short loan sequences (fewer than three loans); the consumers who take out multiple loan sequences typically do not begin a new loan sequence until several months after the end of a prior loan sequence; and there is no evidence of the lender or an affiliate making non-covered loans to consumers to bridge cooling-off periods or the time periods between loan sequences. The example clarifies that this evidence suggests that the lender’s ability-to-repay determinations are reasonable. Although this example does indicate the presence of two factors from comment 5(b)-2.iii (i.e., defaults and re-borrowing), it illustrates that the degree to which these factors are present is germane to the overall evaluation. The re-borrowing is typically less than a full loan sequence, defaults are infrequent, and while there are some consumers who borrow multiple sequences, they are spread further apart, suggesting that new borrowing needs are driving the re-borrowing rather than the spillover effects of the prior loans. Therefore, this pattern does not indicate potentially unreasonable ability-to-repay determinations. Re-underwriting of open-end credit. Finally, with regard to the special rule requiring reunderwriting of open-end credit on a periodic basis under § 1041.5(b)(1)(ii), the Bureau is 729 concerned that the consumer group commenters’ suggestion to require lenders to underwrite each individual advance separately would be unduly burdensome particularly as to small advances. However, the Bureau has further considered the timeline it proposed, and decided to adjust the final rule to require in § 1041.5(b)(1)(ii) that the lender must make a new ability-to-repay determination prior to an advance on an open-end line of credit if more than 90 days has elapsed since the initial determination, rather than every 180 days as proposed. The Bureau believes it is reasonable to require a new ability-to-repay determination once a quarter for an open-end line of credit, which for example would mean that a consumer would be re-underwritten after taking a monthly advance three times in a row. This revised time period also aligns with the revised requirement in § 1041.5(c)(2)(ii)(D), which as discussed below generally exempts lenders from the requirement to obtain a new national consumer report to verify debt obligations, child support obligations, and alimony obligations if the lender or its affiliates has previously obtained such a report in the prior 90 days (unless the consumer had triggered a cooling-off period since the report was last obtained). The Bureau disagrees with the commenter that argued that the Bureau should exempt open-end lines of credit from the proposal or, in the alternative, should address open-end lines of credit in a separate rulemaking along with credit cards or apply the requirements of the CARD Act in connection with open-end lines of credit that are covered in this rule. The Bureau notes that while open-end products are not as common in the affected markets as closed-end products, the Bureau did conduct substantial research as part of this rulemaking concerning deposit advance products, which can be structured as open-end credit. The Bureau believes that consumers can be harmed just as much by unaffordable open-end credit as unaffordable closed730 end credit, and that both products are therefore appropriately subject to the final rule. With regard to why the Bureau is not imposing the same rules for open-end products as the CARD Act regulations—an alternative approach suggested by the commenter—see the general discussion above for § 1041.5 about the comparison between the two rules. The Bureau also disagrees with the more technical arguments made by the same commenter about the proposed requirement to assess consumers’ ability to repay an open-end line of credit where the consumer requests a new advance more than 180 days after the lender’s last assessment of the consumer’s ability to repay.824 5(b)(2) Proposed Rule Proposed § 1041.5(b)(2) set forth the Bureau’s specific proposed methodology for making a reasonable determination of a consumer’s ability to repay a covered short-term loan. Specifically, it would have provided that a lender’s determination of a consumer’s ability to repay is reasonable only if, based on projections in accordance with proposed § 1041.5(c), the lender reasonably makes the applicable determinations provided in proposed § 1041.5(b)(2)(i), 824 Specifically, the commenter argued that this provision would be inconsistent with the definition of open-end credit under Regulation Z. One element of that definition focuses on whether the amount of credit that may be extended to the consumer is generally made available to the extent that any outstanding balance is repaid. 12 CFR 1026.2(a)(20)(iii). The commentary to Regulation Z distinguishes open-end credit on this ground from situations in which the consumer has to apply for each advance individually under a closed-end credit feature. However, the Regulation Z commentary also emphasizes that this distinction does not prevent creditors offering open-end products from periodically adjusting their credit limits or refusing to make an individual extension of credit “due to changes in the creditor’s financial condition or the consumer’s creditworthiness.” Comment 1026.2(a)(20)-5. The Bureau believes that the final rule here is consistent with this Regulation Z commentary, in that the final rule periodically requires a lender to evaluate whether the consumer has the ability to repay the entire amount available under an open-end line of credit. With regard to how the lender would decide after such an assessment whether to increase the line or to take other action where the consumer’s credit has deteriorated such that she can no longer make the outstanding payments, the Bureau would expect lenders to make decisions in accordance with the updated ability-to-repay analysis as to whether a change in the credit line is appropriate in either direction. 731 (ii), and (iii). Proposed § 1041.5(b)(2)(i) would have required an assessment of the sufficiency of the consumer’s residual income during the term of the loan, and proposed § 1041.5(b)(2)(ii) would have required an assessment of an additional 30-day period after having made the highest payment on the loan in light of the harms from loans with short-term structures. In proposed § 1041.5(b)(2)(iii), the Bureau would have required compliance with further requirements in proposed § 1041.6 in situations where consumers’ borrowing history suggests that they may have difficulty repaying additional credit. Proposed § 1041.9(b)(2) would have imposed similar requirements on covered longer-term balloon-payment loans. More specifically, proposed § 1041.5(b)(2)(i) would have provided that for any covered short-term loan subject to the ability-to-repay requirement of proposed § 1041.5, a lender must reasonably conclude that the consumer’s residual income would be sufficient for the consumer to make all payments under the covered short-term loan and to meet basic living expenses during the shorter of the term of the loan or for 45 days following consummation. The Bureau believed that if the payments for a covered short-term loan would consume so much of a consumer’s residual income that the consumer would be unable to meet basic living expenses, then the consumer would likely suffer injury from default or re-borrowing, or suffer collateral harms from having to make unaffordable payments. The parallel provision in 1041.9(b)(2)(i) applicable to covered longer-term loans would have provided for a reasonable conclusion about the sufficiency of the residual income during the loan term. Proposed comment 9(b)(2)(i)-1.i would have clarified that for covered longer-term loans, a reasonable conclusion about the sufficiency of the residual income for the month in which the highest sum of payments were due on the loan would have satisfied this requirement. 732 In proposing § 1041.5(b)(2)(i), the Bureau recognized that, even when lenders determine at the time of consummation that consumers will have the ability to repay a covered short-term loan, some consumers may still face difficulty making payments on these loans because of changes that occur after consummation. The Bureau noted in the proposal that, for example, some consumers would experience unforeseen decreases in income or increases in expenses that would leave them unable to repay their loans. Thus, the fact that a consumer ended up in default is not, in and of itself, evidence that the lender failed to reasonably assess the consumer’s ability to repay the loan ex ante. The Bureau explained that proposed § 1041.5(b)(2)(i) would instead have looked to the facts that were reasonably knowable prior to consummation and prohibited a lender from making a covered short-term loan if the lender lacked a reasonable basis at consummation to conclude that the consumer would be able to repay the covered loan while also meeting basic living expenses and major financial obligations. The Bureau further explained in the proposal that while some consumers may have so little (or no) residual income as to be unable to afford any loan at all, for other consumers the ability to repay will depend on the amount and timing of the required repayments. Thus, the Bureau noted, even if a lender concludes there is no reasonable basis for believing that a consumer can pay a particular prospective loan, proposed § 1041.5(b)(2)(i) would have not prevented a lender from making a different covered loan with more affordable payments to such a consumer, provided that the loan is consistent with State law and that the more affordable payments would not consume so much of the consumer’s residual income that she would be unable to meet basic living expenses. Proposed comment 5(b)(2)(i)-1 would have provided more detailed guidance on the 733 calculations needed for the applicable period under § 1041.5(b)(2)(i), explaining that a lender complies with the requirement in § 1041.5(b)(2)(i) if it reasonably determines that the consumer’s projected residual income during the shorter of the term of the loan or the period ending 45 days after consummation of the loan will be greater than the sum of all payments under the covered short-term loan plus an amount the lender reasonably estimates will be needed for basic living expenses during the term of the covered short-term loan. The Bureau explained in the proposal that this method of compliance would have allowed the lender to make one determination based on the sum of all payments that would be due during the term of the covered short-term loan, rather than having to make a separate determination for each respective payment and payment period in isolation in cases where the short-term loan provide for multiple payments. Under the proposed rule, the lender would have had to make the determination for the actual term of the loan, accounting for residual income (i.e., net income minus payments for major financial obligations) that would actually accrue during the shorter of the term of the loan or the period ending 45 days after consummation of the loan. The Bureau wrote that it believed that for a covered loan with short duration, a lender should make the determination based on net income the consumer will actually receive during the term of the loan and payments for major financial obligations that will actually be payable during the term of the loan, rather than, for example, based on a monthly period that may or may not coincide with the loan term. The Bureau explained that when a covered loan period is under 45 days, determining whether the consumer’s residual income will be sufficient to make all payments and meet basic living expenses depends a great deal on, for example, how many paychecks the consumer will actually 734 receive during the term of the loan and whether the consumer will also have to make no rent payment, one rent payment, or two rent payments during that period. Proposed comment 9(b)(2)(i)-1 contained similar content but also emphasized that determination of whether residual income will be sufficient for the consumer to make all payments and to meet basic living expenses during the term of a covered longer-term loan (including covered longer-term balloon-payment loans) requires a lender to reasonably account for the possibility of volatility in the consumer’s residual income and basic living expenses over the term of the loan. The Bureau further stated in that proposed comment that a lender reasonably accounts for the possibility of volatility in income and basic living expenses by reasonably determining an amount (i.e., a cushion) by which the consumer’s residual income must exceed the sum of the loan payments under the loans and the amount needed for basic living expenses. Proposed comment 5(b)(2)(i)-2 clarified what constitutes “sufficient” residual income for a covered short-term loan, explaining that residual income is sufficient as long as it is greater than the sum of payments that would be due under the covered loan plus an amount the lender reasonably estimates will be needed for basic living expenses. Proposed comment 9(b)(2)(i)-2 was identical. The proposal also would have required lenders who make covered short-term loans and covered longer-term balloon-payment loans to assess consumers’ finances for a second, distinct time period under §§ 1041.5(b)(2)(ii) and 1014.9(b)(2)(ii), respectively. Specifically, those sections would have required that before making such loans, a lender must reasonably conclude that the consumer will be able to make payments required for major financial obligations as they 735 fall due, make any remaining payments under the loan, and meet basic living expenses for 30 days after having made the highest payment under the loan on its due date. Proposed comment 5(b)(2)(ii)-1 noted that a lender must include in its determination under proposed § 1041.5(b)(2)(ii) the amount and timing of net income that it projects the consumer will receive during the 30-day period following the highest payment, in accordance with proposed § 1041.5(c). Proposed comment 5(b)(2)(ii)-1 also included an example of a covered short-term loan for which a lender could not make a reasonable determination that the consumer would have the ability to repay under proposed § 1041.5(b)(2)(ii). The Bureau noted in the proposal that it proposed to include the requirement in § 1041.5(b)(2)(ii) for covered shortterm loans because research showed that these loan structures are particularly likely to result in re-borrowing shortly after the consumer repays an earlier loan. As discussed in the proposal, when a covered loan’s terms provide for it to be substantially repaid within 45 days following consummation, the fact that the consumer must repay so much within such a short period makes it especially likely that the consumer will be left with insufficient funds to make subsequent payments under major financial obligations and meet basic living expenses. The Bureau noted that the consumer may then end up falling behind in paying major financial obligations, being unable to meet basic living expenses, or borrowing additional consumer credit. Such consumers may be particularly likely to borrow new consumer credit in the form of a new covered loan. The Bureau further elaborated in the proposal that this shortfall in a consumer’s funds is most likely to occur following the highest payment under the covered short-term loan (which is typically but not necessarily the final payment) and before the consumer’s subsequent receipt of significant income. The Bureau noted, however, that depending on the regularity of a 736 consumer’s income payments and payment amounts, the point within a consumer’s monthly expense cycle when the problematic covered short-term loan payment falls due, and the distribution of a consumer’s expenses through the month, the resulting shortfall may not manifest itself until a consumer has attempted to meet all expenses in the monthly expense cycle, or even longer. The Bureau noted that indeed, many payday loan consumers who repay a first loan and do not re-borrow during the ensuing pay cycle (i.e., within 14 days) nonetheless find it necessary to re-borrow before the end of the expense cycle (i.e., within 30 days). The Bureau noted in the proposal that in the Small Business Review Panel Outline, the Bureau described a proposal under consideration to require lenders to determine that a consumer has the ability to repay a covered short-term loan without needing to re-borrow for 60 days, consistent with the proposal in the same document to treat as part of the same loan sequence a loan taken out within 60 days of having a prior covered short-term loan outstanding. The Bureau noted in the proposal that several consumer advocates had argued that consumers may be able to juggle expenses and financial obligations for a time, so that an unaffordable loan may not result in re-borrowing until after a 30-day period. The Bureau proposed a 30-day period for both purposes. The Bureau wrote that it believed that the incidence of re-borrowing caused by such loan structures would be somewhat ameliorated simply by determining that a consumer would have residual income during the term of the loan that exceeds the sum of covered loan payments plus an amount necessary to meet basic living expenses during that period. But if the loan payments consume all of the consumer’s residual income during the period beyond the amount needed to meet basic living expenses during the period, the Bureau wrote in the proposal, then the 737 consumer will have insufficient funds to make payments under major financial obligations and meet basic living expenses after the end of that period, unless the consumer receives sufficient net income shortly after the end of that period and before the next set of expenses fall due. The Bureau noted that often, though, the opposite is true: a lender schedules the due dates of loan payments under covered short-term loans so that the loan payment due date coincides with the consumer’s receipt of income. The Bureau noted that this practice maximizes the probability that the lender will timely receive the payment under the covered short-term loan, but it also means the term of the loan (as well as the relevant period for the lender’s determination that the consumer’s residual income will be sufficient under proposed § 1041.5(b)(2)(i)) ends on the date of the consumer’s receipt of income, with the result that the time between the end of the loan term and the consumer’s subsequent receipt of income is maximized. Thus, in the proposal, the Bureau wrote that even if a lender made a reasonable determination under proposed § 1041.5(b)(2)(i) that the consumer would have sufficient residual income during the loan term to make loan payments under the covered short-term loan and meet basic living expenses during the period, there would remain a significant risk that, as a result of an unaffordable highest payment (which may be the only payment, or the last of equal payments), the consumer would be forced to re-borrow or suffer collateral harms from unaffordable payments. The Bureau wrote that the example included in proposed comment 5(b)(2)(ii)-1 was intended to illustrate just such a result. In proposed § 1041.5(b)(2)(iii), the Bureau would have required the lender to determine that the requirements of proposed § 1041.6 are satisfied when making a covered short-term loan for which a presumption of unaffordability under proposed § 1041.6 applies. 738 Comments Received The Bureau received a number of comments on proposed § 1041.5(b)(2), and specifically the time period and sufficiency of the residual income model. Many of the comments pertaining to this section were already discussed above in the discussion of comments received pertaining to § 1041.5 more generally, § 1041.5(a), and § 1041.5(b)(1). On the time period, several consumer advocate commenters suggested that residual income should be assessed under § 1041.5(b)(2)(ii) for 60 days following the highest payment. Other commenters argued that the time period in question should run from the last payment instead of the highest payment, arguing that this would ensure that the consumer does not need to re-borrow throughout the entirety of the loan term and thereafter. As articulated by the commenters, if a consumer’s highest payment came more than 30 days before the end of the loan term, then under the Bureau’s proposed requirement, the lender would only need to make a reasonable conclusion about whether the consumer could repay until the end of the loan term (and there would not be a 30-day period after to assess for re-borrowing). Industry commenters asserted that forecasting for income and expenses as they come due, including the timing of those payment and expenses, during the various overlapping proposed time periods would be infeasible. Others made the opposite argument, asserting that lenders should at least be encouraged to assess actual basic living expenses during the two time periods specified in proposed § 1041.5(b)(2). As discussed above, a number of commenters asserted that the residual income model was unduly restrictive or otherwise inadequate for assessing whether a consumer has the ability to repay. Some argued that if the Bureau is using a residual income approach, it should model its 739 test after the Department of Veterans Affair’s residual income test, which includes objective numerical standards. Many other commenters, as noted in the general § 1041.5 discussion above, asserted that a debt-to-income ratio was a more well-accepted and time-tested underwriting model. Other commenters argued, as noted above, for a loan-to-income or payment-to-income approach instead. Others argued, also as noted above, that a residual income test would be too burdensome. Still other commenters pointed to data showing that residual income is not indicative of whether a consumer will default. These comments are discussed in more detail in the introduction to § 1041.5 and the summary of § 1041.5(b)(1) above. The Bureau also received a number of comments relating to how proposed § 1041.9(b) would have required lenders to include a cushion to account for income volatility over the course of a covered longer-term loan, arguing that to do so would be purely speculative. Final Rule As described in the general § 1041.5 discussion and in the discussion of the debt-toincome ratio definition in 1041.5(a)(2) above, the Bureau has made a substantial number of changes to § 1041.5(b)(2) of the final rule. To summarize, as described in the general § 1041.5 discussion above, under proposed § 1041.5(b)(2) the reasonable ability-to-repay determination would have required the lender to project both the amount and timing of the consumer’s net income and major financial obligations and draw reasonable conclusions about the consumer’s ability to repay during two distinct time periods: first for the shorter of the term of the loan or 45 days after consummation of the loan,825 and then also for 30 days after having made the highest payment under the loan. This 825 The proposal would have designated this time period to cover the term of the loan for covered longer-term loans. 740 requirement is being streamlined in the final rule. Lenders are instead required to make a projection about net income and major financial obligations and calculate the debt-to-income ratio or residual income, as applicable, during only a single monthly period, i.e., the relevant monthly period. The Bureau has defined that term in § 1041.5(a)(7) as the calendar month with the highest payments on the loan, which is generally consistent with the analysis that the Bureau proposed to use for covered longer-term balloon payment loans under proposed § 1041.9(b)(2)(i) and focuses on the time in which the loan places the greatest strain on the consumer’s finances. Lenders can use the debt-to-income ratio or residual income during this relevant monthly period as a snapshot of the consumer’s financial picture to draw conclusions about the consumer’s ability to repay the covered short-term loan or covered longer-term balloon-payment loan without re-borrowing. Specifically, under § 1041.5(b)(2), the lender uses this information to make a reasonable conclusion that the consumer has the ability to repay the loan while meeting basic living expenses and major financial obligations during: (1) the shorter of the term of the loan or 45 days after consummation of the loan, for covered short-term loans, and the relevant monthly period, for covered longer-term balloon-payment loans, and (2) for 30 days after having made the single highest payment under the loan. This simplified approach also dovetails with the inclusion of the debt-to-income ratio methodology as an alternative to residual income. As discussed above, a debt-to-income methodology does not permit the tracking of a consumer’s individual income inflows and major financial obligation outflows on a continuous basis over a period of time. In response to commenters arguing that forecasting the timing of income flow and payment obligations over the applicable period will be difficult, the Bureau has adjusted the rule. 741 While § 1041.5(b)(2) still requires the lender to generally make a reasonable conclusion about whether the consumer can pay major financial obligations, loan payment amounts, and basic living expenses for the loan term and 30 days after the largest payment, the Bureau has adjusted the rule such that the lender does not need to specifically project both the amount of the payments and the timing of the payments during those periods. Rather, the lender is required to account only for the amounts of such payments—and not the timing of them—during a single calendar month, the relevant monthly period. The relevant monthly period is defined in 1041.5(a)(7) as the calendar month in which the payments on the loan are highest. The Bureau has also revised commentary to § 1041.5(b)(2) to discuss how lenders are to use the projections of net income and major financial obligations during the relevant monthly period as a baseline of information to then make reasonable inferences and draw a reasonable conclusion about the time periods described in § 1041.5(b)(2). As noted above, § 1041.5(b)(2) has been revised and expanded largely as a way of accommodating the inclusion in the final rule of an option for lenders to use a debt-to-income methodology in lieu of a residual income methodology. Although some of the revisions are substantive and are described below, most of the changes reflect the creation of a parallel set of provisions to apply to the debt-to-income methodology. Thus § 1041.5(b)(2) of the final rule is now split so that subparagraph (i) addresses the debt-to-income ratio methodology, and subparagraph (ii) addresses the residual income methodology. Lenders will only have to comply with one or the other subparagraph depending on which methodology they choose. The Bureau described the debt-to-income ratio methodology above in the discussion of § 1041.5(a), but, to recap, a lender may determine whether a consumer will have a high enough 742 percentage of net income remaining to pay for basic living expenses after paying major financial obligations and the loan payments during the relevant monthly period. As discussed earlier, the Bureau has not set the threshold for how high a percentage would meet the test and will allow lenders to use their reasoned judgment. The Bureau believes that a lender may find that different thresholds are effective for consumers with different income levels and family sizes. However, a lender could conceivably use a single threshold, and lenders that choose to vary the thresholds will almost surely develop different approaches of doing so. The test will be whether the thresholds deployed by any given lender lead to reasonable determinations of whether consumers have the ability to repay their loans according to the loan terms. Of course, if lenders set thresholds based on reasoned judgment, but then find they do not work in practice, the Bureau will expect them to adjust accordingly. The Bureau has not imported the requirement under proposed comment 9(b)(2)(i)-2 (also cross-referenced in proposed comment 9(b)-2.i.F) that lenders must allow a cushion for income volatility. The proposal did not include a requirement to account for income volatility for covered short-term loans, and the Bureau sees no reason to add one in the final rule. The Bureau is skeptical that such a requirement is needed for covered short-term loans due to their shorter duration. Moreover, the Bureau is not finalizing this comment as to covered longer-term balloonpayment loans, which are included in the scope of § 1041.5(b)(2). The Bureau proposed the cushion requirement with respect to covered longer-term loans because installment loans would have predominated that category. For those loans, the proposed ability-to-repay requirement would have focused on the affordability of the regular periodic payment. The Bureau believed 743 that if a consumer had only just enough money to cover that payment in a “normal” month, the loan would prove unaffordable over its term due to income or expense volatility. The final rule, however, covers only longer-term loans with a balloon payment and requires underwriting such loans to assess whether the consumer will be able to make the payments in the month with the highest sum of payments. Therefore, the Bureau does not believe it is necessary to add a cushion to that calculation. In addition to substantially revising the text of § 1041.5(b)(2) in light of these major changes, the Bureau has also revised the comments. Comment 5(b)(2)-1 reiterates the general methodology, and notes that if there are two payments that are equal to each other in amount and higher than all other payments, the highest payment under the loan is considered the later in time of the two. Comments 5(b)(2)(i)-1 and -2 explain how the relevant monthly period for calculating the debt-to-income ratio is not identical to the periods for which a lender is assessing ability to repay in § 1041.5(b)(2)(i), and explains that in fact they may overlap. Comment 5(b)(2)(i)-2 explains that the lender uses the projections about the consumer’s net income and major financial obligations during the relevant monthly period and the calculation of the consumer’s debt-to-income ratio as a baseline of information on which to make reasonable inferences and draw a reasonable conclusion about whether the consumer will be able to pay major financial obligations, make the payments on the loan, and meet basic living expenses during the periods specified in § 1041.5(b)(2)(i). The comment further states that the lender cannot assume, for example, in making those reasonable inferences, that the consumer will defer payment on major financial obligations or basic living expenses until after the 30-day period that follows the date of the highest payment on the loan, or assume that the obligations and expenses 744 will be less than in the relevant monthly period. The Bureau provides examples of this dynamic in comment 5(b)(2)(i)-3. Comments 5(b)(2)(ii)-1 through -3 provide parallel guidance as to covered longer-term balloon-payment loans. Lastly, the Bureau did not finalize the content in proposed § 1041.5(b)(2)(iii), which would have required lenders to satisfy further requirements under proposed § 1041.6 before making a covered short-term loan in circumstances where the consumer’s recent borrowing or current difficulties paying off an existing loan suggested that they did not have the ability to repay a new loan. As discussed below, the Bureau has instead finalized certain elements of proposed § 1041.6 as final § 1041.5(d). 5(c) Projecting Consumer Net Income and Payments for Major Financial Obligations Overview Proposed § 1041.5(c) specified the requirements for obtaining information directly from consumers as well as various forms of verification evidence for use in projecting consumers’ net income and major financial obligations for purposes of the ability-to-repay requirements under proposed § 1041.5(b). Following the Bureau’s review and consideration of the comments to the proposal, the Bureau is finalizing § 1041.5(c) with substantial changes to provide more flexibility with regard to verification requirements and to provide more detailed guidance for how lenders should treat discrepancies between consumers’ written statements and verification evidence. The Bureau has carefully balanced the final rule to require substantial improvements in current industry verification practices, while providing appropriate flexibility for lenders and consumers in situations in which verification evidence is not reasonably available. 745 Specifically, the Bureau had proposed § 1041.5(c) in the following manner: subparagraph (1) set forth the general evidentiary standards for reasonably projecting net income and major financial obligations and the standards for addressing inconsistencies between the consumers’ stated amounts for such items and verification evidence; subparagraph (2) addressed one narrow way in which lenders could deviate from information in verification evidence; and subparagraph (3) governed how and when lenders must obtain verification evidence for net income and major financial obligations. The Bureau is not finalizing much of the content in subparagraph (2), as described below, and, for increased clarity, the Bureau is now placing the content from subparagraph (2), to the extent that content is being finalized or amended, into subparagraph (1). Accordingly, final § 1041.5(c)(1) describes the general evidentiary standards, the standards for addressing inconsistencies between the consumers’ stated amounts for net income and major financial obligations and the verification evidence, and the process for when lenders can deviate from the information in verification evidence; and § 1041.5(c)(2) governs how and when lenders must obtain verification evidence for net income and major financial obligations. 5(c)(1) General Proposed Rule With regard to covered short-term loans, in proposed § 1041.5(c)(1), the Bureau provided that for a lender’s projection of the amount and timing of net income or payments for major financial obligations to be reasonable, the lender must obtain both a written statement from the consumer as provided for in proposed § 1041.5(c)(3)(i) and verification evidence as provided for in proposed § 1041.5(c)(3)(ii). Proposed § 1041.5(c)(1) further provided that for a lender’s 746 projection of the amount and timing of net income or payments for major financial obligations to be reasonable, it may be based on a consumer’s statement of the amount and timing only to the extent the stated amounts and timing are consistent with the verification evidence. As the Bureau explained in the proposal, the Bureau believed verification of consumers’ net income and payments for major financial obligations was an important component of the reasonable ability-to-repay determination. Consumers seeking a loan may be in financial distress and inclined to overestimate net income or to underestimate payments for major financial obligations to improve their chances of being approved. Lenders have an incentive to encourage such misestimates to the extent that as a result consumers find it necessary to re-borrow. The Bureau further stated in the proposal that this result is especially likely if a consumer perceives that, for any given loan amount, lenders offer only a one-size-fits-all loan repayment structure and will not offer an alternative loan with payments that are structured to be within the consumer’s ability to repay. As the Bureau noted, an ability-to-repay determination that is based on unrealistic factual assumptions will yield unrealistic and unreliable results, leading to the very consumer harms that the Bureau’s proposal was intended to prevent. Accordingly, proposed § 1041.5(c)(1) would have permitted a lender to base its projection of the amount and timing of a consumer’s net income or payments for major financial obligations on a consumer’s written statement of amounts and timing under proposed § 1041.5(c)(3)(i) only to the extent the stated amounts and timing are consistent with verification evidence of the type specified in proposed § 1041.5(c)(3)(ii). Proposed § 1041.5(c)(1) also provided that in determining whether and the extent to which stated amounts and timing are consistent with verification evidence, a lender may reasonably consider other reliable evidence 747 the lender obtains from or about the consumer, including any explanations the lender obtains from the consumer. In the proposal, the Bureau stated its belief that the proposed approach would appropriately ensure that the projections of a consumer’s net income and payments for major financial obligations will generally be supported by objective, third-party documentation or other records. The Bureau further stated, however, that the proposed approach also recognized that reasonably available verification evidence may sometimes contain ambiguous, out-of-date, or missing information. For example, the net income of consumers who seek covered loans may vary over time, such as for a consumer who is paid an hourly wage and whose work hours vary from week to week. In fact, a consumer is more likely to experience financial distress, which may be a consumer’s reason for seeking a covered loan, immediately following a temporary decrease in net income from more typical levels. Accordingly, the Bureau stated that the proposed approach would not have required a lender to base its projections exclusively on the consumer’s most recent net income receipt shown in the verification evidence. Instead, it allowed the lender reasonable flexibility in the inferences the lender draws about, for example, a consumer’s net income during the term of the covered loan, based on the consumer’s net income payments shown in the verification evidence, including net income for periods earlier than the most recent net income receipt. At the same time, the proposed approach would not have allowed a lender to mechanically assume that a consumer’s immediate past income as shown in the verification evidence will continue into the future if, for example, the lender has reason to believe that the consumer has been laid off or is no longer employed. 748 The Bureau stated in the proposal, that in this regard, the proposed approach recognized that a consumer’s own statements, explanations, and other evidence can be important components of a reliable projection of future net income and payments for major financial obligations. Proposed comment 5(c)(1)-1 included several examples applying the proposed provisions to various scenarios, illustrating reliance on consumer statements to the extent they are consistent with verification evidence and how a lender may reasonably consider consumer explanations to resolve ambiguities in the verification evidence. It included examples of when a major financial obligation in a consumer report is greater than the amount stated by the consumer and of when a major financial obligation stated by the consumer does not appear in the consumer report at all. The Bureau stated in the proposal that it anticipated that lenders would develop policies and procedures, in accordance with proposed § 1041.18, for how they project consumer net income and payments for major financial obligations in compliance with proposed § 1041.5(c)(1) and that a lender’s policies and procedures would reflect its business model and practices, including the particular methods it uses to obtain consumer statements and verification evidence. The Bureau stated its belief that many lenders and vendors would develop methods of automating projections, so that for a typical consumer relatively little labor would be required. In proposed § 1041.5(c)(2), the Bureau proposed an exception to the requirement in proposed § 1041.5(c)(1) that projections must be consistent with the verification evidence that a lender would be required to obtain under proposed 1041.5(c)(3)(ii). As discussed below, the required verification evidence would have normally consisted of third-party documentation or other reliable records of recent historical transactions or of payment amounts. Proposed 749 § 1041.5(c)(2) would have permitted a lender to project a net income amount that is higher than an amount that would otherwise be supported under proposed § 1041.5(c)(1), or a payment amount for a major financial obligation that is lower than an amount that would otherwise be supported under proposed § 1041.5(c)(1), only to the extent and for such portion of the term of the loan that the lender obtains a written statement from the payer of the income or the payee of the consumer’s major financial obligation of the amount and timing of the new or changed net income or payment. As the Bureau explained in the proposal, the exception was intended to accommodate situations where a consumer’s net income or payment for a major financial obligation will differ from the amount supportable by the verification evidence. For example, a consumer who has been unemployed for an extended period of time, but who just accepted a new job, may not be able to provide the type of verification evidence of net income that generally would have been required under proposed § 1041.5(c)(3)(ii)(A). Proposed § 1041.5(c)(2) would have permitted a lender to project a net income amount based on, for example, an offer letter from the new employer stating the consumer’s wage, work hours per week, and frequency of pay. The lender would have been required to retain the statement in accordance with proposed § 1041.18. Proposed § 1041.9(c) included parallel requirements applicable to covered longer-term loans. Comments Received The Bureau received many comments on the proposed verification requirements from a variety of stakeholders. Many of these commenters argued that the verification requirements were overly burdensome, too prescriptive, and not appropriate to this credit market in contrast to 750 the mortgage and credit card markets. Other industry commenters asked the Bureau to provide more specificity around verification requirements to reduce uncertainty. These commenters included both industry stakeholders and other parties, such as several State Attorneys General and the SBA Office of Advocacy. Many individual consumers, often commenting as part of letter-writing campaigns, also criticized aspects of the verification requirements, particularly the requirement for lenders to obtain a national consumer report for each loan to verify debt obligations. Consumer advocates, on the other hand, generally argued that the verification requirements were calibrated appropriately or, in some places, were too permissive. Some of these arguments are described in the general § 1041.5 discussion at the outset of the section-bysection analysis for this section. These arguments are also described with more particularity in discussion below of sub-sections of this overall section, such as the requirements under § 1041.5(c)(2)(ii)(A) and (B) (verification evidence for net income and major financial obligations, respectively). Commenters generally argued that there are many consumers who have an ability to repay, but who cannot verify income, and that they would be harmed by the verification requirements. Specifically, many commenters cited consumers who work in the cash economy or who had seasonal or sporadic work as consumers who would be unable to access credit under the proposal because of the income verification requirements. One industry trade group representing community banks argued that some consumers use cash to pay for basic living expenses, so deposit account records would not provide accurate verification evidence. These comments are addressed in the discussion of § 1041.5(c)(2). 751 One commenter argued that the Bureau should not impose any verification evidence requirements until the Bureau could prove that consumers were harmed by lenders failing to collect evidence to verify consumer claims. A number of industry commenters asserted that the Bureau had failed to explain why it was applying more vigorous verification requirements to payday loans than to mortgages and credit cards. Some of these arguments are described in the general § 1041.5 discussion above. Some commenters argued that requiring similar verification requirements undermined the business model of payday and title loan companies, which they argued are built around speed, convenience, and lack of intrusive underwriting, and that consumers desire these features of the business model. Many individual consumers, often writing as part of organized letter-writing campaigns made similar comments. They described favorably their experience with payday loans based on the lack of a credit check requirement, the ease of the application process, and the respect they feel they receive from the origination process at payday lenders (in contrast to their experience at banks, which they argued was more intrusive and impersonal). Other commenters argued that the Bureau could and should provide safe harbors or exceptions for certain lenders who meet various criteria. For example, one commenter, an online lender, argued that the Bureau should not impose any income verification requirements on shortterm lenders with below market average charge-offs and that the Bureau should set a safe harbor loss rate of under 15 percent for first-time customers. A trade group representing vehicle title lenders commented that income verification is incompatible with the business model for the vehicle title loan product and its customer base. The commenter argued that vehicle title lenders would have difficulty obtaining the information 752 from consumers; that the time it would add to the process is disproportionate for this type of loan; and that it would undermine the value and competitiveness of the product. A number of commenters argued that the more rigorous underwriting requirements would involve personal questions that many consumers would believe violate their privacy and so would resist answering, or viewed such questions as too intrusive for a small-dollar loan as opposed to a much larger extension of credit. Similarly, many individual commenters expressed concerns about providing their personal information to lenders, and were concerned about their privacy and also the risk of data breach. Some industry commenters provided similar comments, stating that the need to create real-time, centralized databases for obtaining information on consumers during underwriting would increase consumers’ exposure to data breach risk. A number of commenters, including several lenders and consumer reporting agencies, argued that the Bureau should adopt a validation instead of a verification model, in which lenders could compare statements about income, basic living expenses, or major financial obligations to various third-party data sources or data models, and perform manual processing and verification only when the validation process identifies an anomaly. Some of these commenters noted that the U.K. Financial Conduct Authority guidelines on small-dollar lending permit such an approach. Another provided data comparing deviations from historical 12-month average stated income to default rates, finding that the further a consumer’s stated income deviated from that consumer’s historic average, the higher the default rate (with significantly higher default rates as consumers’ stated income is multiples higher than the historic average). More broadly, commenters argued that the proposed verification requirements did not take into consideration shared payment of major financial obligations by consumers and other 753 persons, such as expenses shared with spouses and cohabitants. Consumer advocates argued, alternatively, that claims of shared major financial obligations should be allowed only with verification evidence. The issues raised in these comments in some cases overlap with the issues discussed in the section-by-section analysis for § 1041.5(a)(1) (definition of basic living expenses) and § 1041.5(a)(3) (definition of major financial obligations). The Bureau received a number of comments relating to how proposed § 1041.5(c)(1) and (c)(2) would have addressed inconsistencies between the consumers’ stated amounts and the verification evidence, when deviation from the stated amounts would have been permitted, and what additional steps would have been required in those circumstances. Consumer advocates argued that lenders should not be allowed to rely on consumer statements that are inconsistent with verification evidence unless relying on the consumer statements would result in a projection of a lower income amount or a higher major financial obligations amount. Others expressed concern that the ability to deviate from amounts in the verification evidence based on explanations from the consumer would be an easy way to skirt the verification requirements in the proposal. On the other hand, industry commenters suggested that lenders should be able to deviate from amounts in verification evidence based on borrower statements. Specific to proposed § 1041.5(c)(2), a number of industry commenters argued that a requirement to procure statements from payors or payees would pose significant privacy concerns for consumers. Online lenders and their trade groups expressed concerns about the practicality and burdens on both the consumer and the lender with respect to the verification requirements. They argued that document verification disadvantages online lenders because documents submitted by fax, mobile image capture, or email scan are frequently illegible or easily misinterpreted; mobile 754 image capture does not work for pay stubs; and even if the customers could submit the documents via mobile app, lenders would need to manually process them on the back end. They also expressed concerns about the fraud and security risks related to consumers taking photos of sensitive documents to submit to online lenders via a smartphone. Lastly, some commenters noted concerns about potential double-deductions, where a national consumer report identifies a debt obligation or child support obligation that may have already been deducted from the consumer’s gross income prior to the consumer’s receipt of takehome pay. The concern was that the portion of the gross income deducted for this obligation would not be included in net income but would still be counted as a major financial obligation. Final Rule After carefully considering the comments received, the Bureau has finalized the core elements of § 1041.5(c)(1) to require lenders to obtain consumers’ written statements and various forms of verification evidence in order to reasonably project net income and major financial obligations for the relevant monthly period as required by § 1041.5(b). However, the Bureau has adopted a number of changes to the proposed approach to provide lenders with greater flexibility to rely on consumers’ written statements in appropriate circumstances and to clarify how lenders should address situations in which there are inconsistencies between consumers’ written statements and consumer reports or other verification evidence. The Bureau has also incorporated some elements of proposed § 1041.5(c)(2) into the commentary on § 1041.5(c)(1), but is not adopting a categorical requirement that lenders may only project increases in net income or decreases in major financial obligations if they obtain a written statement from the payer of the income or the payee of the obligation. 755 Specifically, final § 1041.5(c) specifies that a lender must obtain the consumer’s written statement in accordance with § 1041.5(c)(2)(i), obtain verification evidence as required by § 1041.5(c)(2)(ii), assess information about rental housing expense as required by § 1041.5(c)(2)(iii), and make a reasonable projection of the amount of a consumer’s net income and payments for major financial obligations during the relevant monthly period. As described in more detail in connection with final § 1041.5(c)(2) below, each of those provisions has been modified in turn to allow lenders more flexibility in reasonably relying on information in consumers’ written statements where particular income or major financial obligations cannot be verified through reasonably available sources. For example, § 1041.5(c)(2)(ii)(A) allows lenders to reasonably rely on consumers’ written statements with regard to income that cannot be verified through pay records, bank account records, or other reasonably available sources. Section 1041.5(c)(2)(iii) also allows lenders to reasonably rely on consumers’ written statements with regard to rental housing expense, but not with regard to mortgages that can be verified from a national consumer report. The Bureau also revised § 1041.5(c)(1) to address different types of potential inconsistencies between consumers’ written statements and verification evidence in more detail. Thus, final § 1041.5(c)(1) specifically requires lenders to consider major financial obligations that are listed in a consumer’s written statement, even if they cannot be verified by the sources provided for as verification evidence under § 1041.5(c)(2)(ii)(B). This requirement is consistent with various Bureau statements in the proposal and with proposed comment 5(c)(1)-1.G, which included an example in which a consumer’s child support payment did not appear on a national consumer report, but the Bureau has concluded that the requirement implicit in the example 756 should be reflected in a more direct statement in the regulation text. With regard to other types of inconsistencies between the consumer’s written statement and verification evidence, the final rule provides that a lender may base the amounts of net income or major financial obligations on the consumer’s written statement only as specifically permitted under § 1041.5(c)(2) or to the extent the stated amounts are consistent with the verification evidence. Consistent with the proposal, § 1041.5(c)(1) states that in determining consistency with verification evidence, the lender may reasonably consider other reliable evidence the lender obtains from or about the consumer, including any explanations the lender obtains from the consumer. While the basic elements of proposed § 1041.5(c)(1) remain intact in the final rule, the Bureau has made a number of significant changes to § 1041.5(c)(1). First, as discussed above in connection with § 1041.5(a) and (b), the Bureau is not requiring lenders to project the specific timing of major financial obligations or income. Thus, the Bureau has eliminated all references to the need to verify timing throughout this provision. Second, the Bureau is not finalizing proposed § 1041.5(c)(2). That section would have required a lender to obtain a written statement from a payor of income or a payee of major financial obligations in order to project income in a higher amount, or to project major financial obligations in a lower amount, than would otherwise have been supported by the verification evidence. The Bureau upon further consideration believes this requirement would be too onerous and inflexible, and may also raise privacy concerns if a consumer had to explicitly ask for a written statement from an employer. Because the Bureau is not finalizing proposed § 1041.5(c)(2), it is renumbering proposed § 1041.5(c)(3), which is being finalized (as described in further detail below), as § 1041.5(c)(2). 757 The Bureau believes that the final rule strikes an appropriate balance that will require substantial and reasonable improvements in current industry verification procedures while also addressing concerns that the proposal would be too burdensome to implement and would deny consumers access to credit in situations in which their finances are difficult to verify. The Bureau agrees with consumer advocates that verifying net income and major financial obligations is important to ensure the soundness of ability-to-repay determinations. But the Bureau also found the concerns raised by industry commenters regarding the burden of the verification requirements to be compelling in some instances, as noted below. In response to commenters asserting that the Bureau must first determine that lack of verification evidence is causing harms to consumers before imposing verification requirements, the Bureau notes that it has found harms associated with failing to make reasonable determinations that a consumer has the ability to repay the loan, and had identified the practice as unfair and abusive (as discussed in the section-by-section analysis for § 1041.4 of the final rule). To make a reasonable determination that a consumer has the ability to repay, lenders must satisfy certain reasonable verification requirements, which have been loosened somewhat in light of the concerns raised by commenters. In other words, the verification requirements are reasonably related to preventing the identified unfair and abusive practice in § 1041.4. As discussed above, this is the legal standard for exercise of the Bureau’s prevention authority under section 1031(b) of the Dodd-Frank Act. Moreover, as consumer groups noted and as the proposal stated, there are particular concerns in this market that that consumers who are in financial distress may tend to overestimate income or underestimate expenses, and lenders have strong incentives to encourage 758 misestimates to the extent that doing so tends to result in more re-borrowing. Thus, the Bureau believes that the practice of making loans without verification evidence is a contributing cause of the harms previously discussed. This premise was further validated by data submitted by a commenter, on 1.2 million covered loan applicants in 2014 to support arguments on a different issue. The analysis tracked the degree to which consumers’ stated income deviated from a 12month historical average for that consumer and compared it to default rates. The data showed that default rates increased as a consumer’s stated income deviated from that same consumer’s 12-month average. Some of this could be due of course to unexpected changes in income after the point of prediction, but it may also suggest that the stated income predictions were inaccurate in the first instance. Indeed, the commenter’s data suggests that 35 percent of the 1.2 million applicants studied provided stated income that was 1.5 or more times higher than their own 12month averages and that those borrowers saw significantly higher default rates than other applicants. The Bureau disagrees with arguments that the proposal would have imposed more rigorous verification requirements than it has in the mortgage market under Regulation Z, but in any event as discussed in detail in the introduction to § 1041.5 above, the Bureau believes that the final rule’s income and expense verification requirements are somewhat less onerous than the Bureau’s mortgage rules in 12 CFR 1026.43 and more onerous than the credit card rules for various groups of consumers in 12 CFR 1026.51.826 The Bureau recognizes that the Regulation 826 In determining whether a consumer will have a reasonable ability to repay the loan according to its terms, a mortgage lender must verify all information that the creditor relies upon, including income, assets, and debt obligations. 12 CFR 1026.43. The mortgage ability-to-repay rules under Regulation Z do not contain an exception that permits lenders to rely on a consumer’s statement of income if verification evidence is not reasonably available, 759 Z rules for credit cards do not impose similar verification requirements for income, although pulling consumer reports is a widespread industry practice. As noted above, each credit market is different and warrants different regulations. For further explication on this issue, see the discussion at the beginning of the section-by-section analysis for § 1041.5. The Bureau does not agree with comments requesting the Bureau grant safe harbors regarding the verification requirements for lenders meeting certain criteria such as below-market average charge-off rates. The Bureau does not believe the comments provided adequate data or justification for the particular safe harbors suggested. These changes also would add certain amounts of operational complexity. Additionally, the Bureau is not convinced that, as finalized, the verification requirements are so onerous as to warrant a safe harbor; see discussion elsewhere in this section of the various ways in which the requirements are being relaxed from the proposal. Allowing lenders to collect any less information, for example, through a safe harbor, would significantly undermine the lender’s ability-to-repay determinations under § 1041.5(b), which rely on a reasonable projection of net income and major financial obligations which is grounded in relevant evidence concerning the consumer’s current or recent income and obligations. The Bureau is not revising the final rule to allow lenders to rely on validation or modeling of income or expenses in lieu of verification, as suggested by a number of commenters. As described in the section-by-section analysis for § 1041.5(c)(2), the final rule relaxes the verification requirement in a variety of ways, such as not requiring verification of rental housing for example. Nor do those rules permit a lender to dispense with obtaining a consumer report if the lender has done so with respect to the consumer in the prior 90 days. 760 expenses and permitting reliance on stated amounts of income where verification evidence is not reasonably available. Thus, one of the reasons for expressly permitting the validation or modeling of income and expenses in the final rule as a broad alternative to verification—that it would permit lenders to make loans to consumers with undocumented cash income—has been addressed in a different manner. Furthermore, the rule permits income verification to be done electronically via transaction account data or payroll data, which may be particularly useful to online lenders. Additionally, the Bureau does not have reason to believe that income validation or modeling is a viable option in many contexts covered by § 1041.5, at least as an across-the-board substitute for income verification. The loans covered by § 1041.5 are, for the most part, shortterm loans and the rule requires the lender to project net income for the relevant monthly period. Whatever the reliability of income validation or income estimation modeling may be in assessing a consumer’s average monthly income or annual income, the Bureau does not believe that these techniques provide an adequate substitute for obtaining verification evidence, when reasonably available, of the consumer’s current income or income in the recent past. However, the Bureau has no objection to lenders using validation or modeling methods as a backstop in situations in which consumers’ income cannot be verified through traditional means or continuing to experiment with them in addition to traditional verification methods in order to develop a more complete picture of the strengths and weaknesses of those methods. The Bureau will continue to monitor developments in this area. As noted in Background and Market Concerns—Underwriting, the Bureau understands that obtaining verification evidence for income is a common practice in most of the covered 761 markets (except with regard to some vehicle title lending), and thus, the Bureau’s requirement to verify income is unlikely to upend current norms in those markets. The Bureau notes that the Small Dollar Roundtable, including several lenders, supported an income verification requirement. The Bureau agrees with commenters representing vehicle title lenders who argued that requiring income verification would present more of an adjustment for vehicle title lenders than payday lenders. However, the Bureau is not convinced that this is a compelling reason to not require income verification for vehicle title lenders. Commenters’ arguments are essentially that because a vehicle title lender has security for the loan, the lender’s business model is to forgo underwriting, and not obtain evidence of income, and that the Bureau’s rule should permit that business model to continue as is. But the Bureau has identified particular consumer harms associated with this business model (see Market Concerns—Underwriting), and that is precisely why the Bureau believes it is important that vehicle title lenders be required to underwrite the loans based on consumers’ ability to repay and not rely on the asset value as a substitute for underwriting. Were the Bureau to exclude vehicle title lenders from the verification provisions of the rule, it would be antithetical to one of the goals of this rule, which is to require reasonable determinations that consumers have the ability to repay loans according to their terms. More broadly, the Bureau added comment 5(c)(1)-2 as one of several steps taken to address commenters who urged the Bureau to allow lenders to recognize situations in which other persons regularly contribute to a consumer’s income or regularly pay a consumer’s expenses. Specifically, this comment clarifies that, when it is reasonable to do so, a lender may take into account consumer-specific factors, such as whether other persons are regularly 762 contributing toward paying the consumer’s major financial obligations. Comment 5(c)(1)-2 also notes, however, that it is not reasonable for the lender to consider whether other persons are regularly contributing toward the consumer’s payment of major financial obligations if the lender is separately including in its projection of net income any income of another person to which the consumer has a reasonable expectation of access. As discussed also in connection with § 1041.5(a)(5) and (a)(1) concerning others’ contributions to net income and basic living expenses, respectively, this clarification is intended to avoid double-counting. Regarding comments by online lenders and their representatives that the proposed verification requirements would disadvantage and prove impractical to online lenders and would raise fraud or security risks, the Bureau believes that these comments are largely overstated or mooted in view of the scope and substance of the final rule’s ability-to-repay requirements. First, the Bureau understands that online lenders generally fund the loans they make by depositing those loans into consumers’ checking accounts and collect payment by debiting those accounts. Thus, consumers obtaining online loans have transaction accounts that can be used to verify income electronically. As discussed below in the section-by-section analysis, comment 5(c)(2)(ii)(A)-3 has been added to clarify that the consumer’s recent transaction account deposit history is a reliable record (or records) that is reasonably available if the consumer has such an account and to note that that with regard to such bank account deposit history, the lender could obtain it directly from the consumer or, at its discretion, with the consumer’s permission via an account aggregator service that obtains and categorizes consumer deposit account and other 763 account transaction data.827 Furthermore, in the rare case in which a consumer without a transaction account seeks an online loan, the consumer may be able to provide verification evidence through online payroll records or by electronically transmitting a picture of a pay stub from her smart phone. Thus, the concern of commenters that the income verification requirement will require a scanner or fax machine, or will implicate widespread issues around data transmission or fraudulent documentation, seems misplaced. The Bureau also notes that the commenters’ concerns are moot to the extent that they were focused primarily on longer-term loans without balloon payments, given that such loans are not covered by the ability-to-repay requirements in the final rule. In light of the significant revisions it has made to the proposed rule, the Bureau has rewritten many of the examples in the commentary for § 1041.5(c)(1). In response to the comments received, the Bureau has also added commentary in both § 1041.5(c)(1) and (2) to clarify exactly when lenders can deviate from verification evidence. As discussed further below with regard to specific types of information, the Bureau recognizes that there is some risk of evasion, as consumer groups noted, but has decided to allow lenders to rely on consumers’ written statements in limited circumstances to augment the picture painted by verification evidence, as long as those statements are consistent and reasonable. The Bureau does so in recognition of the evident fact that many borrowers of covered loans have cash income that they spend in cash rather than deposit in a transaction account, and thus would be adversely affected by an overly rigid income verification requirement. For example, in comment 5(c)(1)-1.iii, the 827 As noted in the proposal, based on its market outreach the Bureau understands that at least some online lenders utilize account aggregator services. 764 Bureau notes that it would be reasonable to rely on consumers’ written statements to supplement verified income (by, for example, identifying and explaining a separate source of cash income in a reasonable amount), so long as there is no reasonably available evidence to verify that other source (like deposit account statements). Additionally, and consistent with the proposal, comment 5(c)(1)-1.iv states that a lender acts reasonably in relying on a consumer’s explanation to project income where there is inconsistent verification evidence such as, for example, where a consumer explains that she was sick and missed two days of work, and thus made less income than usual in the most recent period covered by the verification evidence and that the prior period covered by the evidence is more representative of the consumer’s income. Similarly, other examples in the commentary address inconsistencies between a consumer’s written statement and verification evidence with regard to major financial obligations. Specifically, comment 5(c)(1)-1.vi emphasizes that lenders must consider major financial obligations that are listed on the consumer’s written statement but not on a national consumer report or other verification sources, while comment 5(c)(1)-1.vii addresses a situation in which a national consumer report lists a debt obligation that does not appear on the consumer’s written statement. Lastly, the Bureau added comment 5(c)(1)-1.viii, to provide an example clarifying that a lender can deduct from major financial obligations the child support payments that a lender reasonably determines, based on a combination of verification evidence and an explanation from the consumer, have already been deducted from net income, a concept that is further described in § 1041.5(c)(2). 5(c)(2) Evidence of Net Income and Payments for Major Financial Obligations Overview 765 Proposed § 1041.5(c)(3) provided more detailed requirements for collection of a written statement from the consumer concerning the amount and timing of net income and required payments for various major financial obligations, as well as various types of verification evidence for particular categories of major financial obligations. As explained above in connection with proposed § 1041.5(c)(1) and (c)(2), proposed § 1041.5(c)(3) generally would have required lenders to base their projections on amounts shown in the verification evidence, with only limited reliance on the written statements. In light of the challenges in documenting housing expenses where a consumer does not have a formal mortgage or lease, however, proposed § 1041.5(c)(2)(ii)(D) would have permitted lenders to use a reliable estimate of rental housing expenses for consumers with households in the same locality as the consumer, based either on a source such as the American Community Survey of the United States Census Bureau or a lender’s own applicants, provided that the lender periodically reviewed the reasonableness of its estimates by comparing them to statistical survey data or other reliable sources. The Bureau had proposed that more permissive approach to rental housing expense following feedback during the SBREFA process and other outreach about a stricter verification approach to rental housing expense. The Bureau is finalizing proposed § 1041.5(c)(3) as § 1041.5(c)(2) of the final rule, with a number of modifications to the proposal that are intended to relieve unnecessary burdens of verification and to provide greater flexibility and clarity to lenders and consumers in situations in which a source of net income or a major financial obligation cannot be verified through the sources that lenders are required to obtain under the final rule. The Bureau has also modified the final rule to reflect policy decisions addressed in more detail above, including the decision to 766 relax proposed requirements for lenders to project the timing of individuals’ net income and major financial obligations as part of the broader ability-to-repay determination, the decision to include alimony as a major financial obligation, and the decision to allow lenders to account for situations in which the consumer has a reasonable expectation of access to others’ income or in which other parties regularly pay for a consumer’s major financial obligation. 5(c)(2)(i) Consumer Statements Proposed Rule Proposed § 1041.5(c)(3)(i)—which is being finalized, with adjustments, in § 1041.5(c)(2)(i) of the final rule—would have required a lender to obtain a consumer’s written statement of the amount and timing of net income, as well as of the amount and timing of payments required for categories of the consumer’s major financial obligations (e.g., credit card payments, automobile loan payments, housing expense payments, child support payments, and the like). The lender would then use the statements as an input in projecting the consumer’s net income and payments for major financial obligations during the term of the loan. The lender would also have been required to retain the statements in accordance with proposed § 1041.18. These statements were intended to supplement verification evidence because verification evidence may sometimes contain ambiguous, out-of-date, or missing information. Proposed comment 5(c)(3)(i)-1 would have clarified that a consumer’s written statement includes a statement that the consumer writes on a paper application or enters into an electronic record, or an oral consumer statement that the lender records and retains or memorializes in writing and retains. It further would have clarified that a lender complies with a requirement to obtain the consumer’s statement by obtaining information sufficient for the lender to project the 767 dates on which a payment will be received or will be paid through the period as required under proposed § 1041.5(b)(2). This proposed comment included the example that a lender’s receipt of a consumer’s statement that the consumer is required to pay rent every month on the first day of the month is sufficient for the lender to project when the consumer’s rent payments are due. Proposed § 1041.5(c)(3)(i) did not specify any particular form or even particular questions or particular words that a lender must use to obtain the required consumer statements. Comments Received and Final Rule The Bureau received few comments about the written statements in their own right, and is finalizing the proposed regulation and commentary as § 1041.5(c)(2)(i) in the final rule. The Bureau has revised the regulation text slightly for clarity and to reflect the decision to allow consideration of the amount of any income of another person to which the consumer has a reasonable expectation of access, as discussed above in connection with § 1041.5(a)(5) (definition of net income). The regulation text and commentary have also been edited to omit references to the timing of particular income and major financial obligation payments, in light of the final rule’s changes with regard to use of debt-to-income ratios and revisions to the residualincome analysis as discussed above in connection with § 1041.5(a)(2) (debt-to-income ratio definition) and § 1041.5(b)(2) (ability-to-repay determination methodologies). Comments concerning lenders’ ability to rely on written statements in the absence of verification evidence are discussed in more detail below. 5(c)(2)(ii) Verification Evidence In proposed § 1041.5(c)(3)(ii), the Bureau would have required a lender to obtain verification evidence for the amounts and timing of the consumer’s net income and payments for 768 major financial obligations for a fixed period prior to consummation. It separately specified the type of verification evidence required for net income and each component of major financial obligations. The Bureau explained in the proposal that the requirements were designed to provide reasonable assurance that lenders’ projections of consumers’ finances were based on accurate and objective information, while also allowing lenders to adopt innovative, automated, and less burdensome methods of compliance. 5(c)(2)(ii)(A) Proposed Rule In proposed § 1041.5(c)(3)(ii)(A), the Bureau specified that for a consumer’s net income, the applicable verification evidence would be a reliable record (or records) of an income payment (or payments) covering sufficient history to support the lender’s projection under proposed § 1041.5(c)(1). It did not specify a minimum look-back period or number of net income payments for which the lender must obtain verification evidence. The Bureau explained in the proposal that it did not believe it was necessary or appropriate to require verification evidence covering a look-back period of a prescribed length. Rather, the Bureau indicated that the sufficiency of the history for which a lender would obtain verification evidence may depend on the source or type of income, the length of the prospective covered longer-term loan, and the consistency of the income shown in the verification evidence that the lender initially obtains, if applicable. Proposed comment 5(c)(3)(ii)(A)-1 would have clarified that a reliable transaction record includes a facially genuine original, photocopy, or image of a document produced by or on behalf of the payer of income, or an electronic or paper compilation of data included in such a 769 document, stating the amount and date of the income paid to the consumer. It further would have clarified that a reliable transaction record also would include a facially genuine original, photocopy, or image of an electronic or paper record of depository account transactions, prepaid account transactions (including transactions on a general purpose reloadable prepaid card account, a payroll card account, or a government benefits card account), or money services business check-cashing transactions showing the amount and date of a consumer’s receipt of income. The Bureau explained in the proposal that the proposed requirement was designed to be sufficiently flexible to provide lenders with multiple options for obtaining verification evidence for a consumer’s net income. For example, the Bureau noted that a paper pay stub would generally satisfy the requirement, as would a photograph of the pay stub uploaded from a mobile phone to an online lender. In addition, the Bureau noted that the requirement would also be satisfied by use of a commercial service that collects payroll data from employers and provides it to creditors for purposes of verifying a consumer’s employment and income. Proposed comment 5(c)(3)(ii)(A)-1 would also have allowed verification evidence in the form of electronic or paper bank account statements or records showing deposits into the account, as well as electronic or paper records of deposits onto a prepaid card or of check-cashing transactions. Data derived from such sources, such as from account data aggregator services that obtain and categorize consumer deposit account and other account transaction data, would also generally satisfy the requirement. During outreach, service providers informed the Bureau that they currently provide such services to lenders. 770 The Bureau explained in the proposal that this approach was designed to address concerns that had been raised during the SBREFA process and other industry outreach prior to the proposal. In particular, some SERs and industry representatives had expressed concern that the Bureau would require outmoded or burdensome methods of obtaining verification evidence, such as always requiring a consumer to submit a paper pay stub or transmit it by facsimile (fax) to a lender. Others questioned requiring income verification at all, stating that many consumers are paid in cash and therefore have no employer-generated records of income. The Bureau explained in the proposal that proposed § 1041.5(c)(3)(ii)(A) was intended to respond to many of these concerns by providing a wide range of methods to obtain verification evidence for a consumer’s net income, including electronic methods that can be securely automated through third-party vendors with a consumer’s consent. The Bureau explained that in developing the proposal, Bureau staff met with more than 30 lenders, nearly all of which stated they already use some method—though not necessarily the precise methods the Bureau was proposing—to verify consumers’ income as a condition of making a covered loan. The Bureau stated that its proposed approach thus would accommodate most of the methods they described and that the Bureau was aware of from other research and outreach. It was also intended to provide some accommodation for making covered loans to many consumers who are paid in cash. For example, under the Bureau’s proposed approach, a lender would have been able to obtain verification evidence of net income for a consumer who is paid in cash by using deposit account records (or data derived from deposit account transactions), if the consumer deposits income payments into a deposit account. The Bureau explained in the proposal that lenders often require consumers to have deposit accounts as a condition of obtaining a covered loan, so the Bureau believed that lenders 771 would be able to obtain verification evidence for many consumers who are paid in cash in this manner. The Bureau recognized in the proposal that there would be some consumers who receive a portion of their income in cash and do not deposit it into a deposit account or prepaid card account. For such consumers, a lender may not be able to obtain verification evidence for that portion of a consumer’s net income, and therefore generally could not base its projections and ability-to-repay determinations on those amounts. The Bureau stated in the proposal that where there is no verification evidence for a consumer’s net income, the Bureau believed the risk would be too great that projections of net income would be overstated and that payments under a covered short-term loan consequently would exceed the consumer’s ability to repay, resulting in all the harms from unaffordable covered loans identified in the proposal. For similar reasons, the Bureau did not propose to permit the use of predictive models designed to estimate a consumer’s income or to validate the reasonableness of a consumer’s statement of her income. The Bureau noted that it had received recommendations from the Small Dollar Roundtable, comprising a number of lenders making the kinds of loans the Bureau was considering whether to cover in this rulemaking and a number of consumer advocates, urging the Bureau to require income verification. Comments Received Many commenters, particularly industry stakeholders, were generally concerned that the income verification requirements would create inaccurate portrayals of consumers’ income because many types of income would not be verifiable. These commenters specifically focused on consumers who are paid in cash, noting that these consumers would likely not have a way, 772 except account statements, to verify income. One trade group commenter said even then, some consumers use cash income directly to pay basic living expenses (without depositing it in an account). Commenters similarly argued that the Bureau’s verification regime had not accounted for consumers who have seasonal or irregular income, such as tips, bonuses, and overtime pay. Commenters also asked for clarity on how income earned in amounts and from sources other than regular payroll would be handled under the rule, and expressed concern with strict verification requirements that would make it difficult for consumers with these types of income to prove future income with past documentation. Commenters argued that these consumers who work in the “cash economy” make up a substantial portion of the customer base for covered lenders, and cited numerous examples of occupations such as restaurant workers, hair stylists, or day laborers who are routinely paid in cash. Others argued that the Bureau should allow stated income based on consumer statements, noting that credit card issuers do not need to verify income. Consumer groups generally supported the income verification requirements and urged the Bureau not to permit lenders to rely on stated income in any circumstances. They argued that variations from verification evidence based on the consumer statements should be permitted only if they result in a lower projection of income (i.e., a more conservative estimate). Also, as stated earlier, many commenters argued that the Bureau had not established a way to account for income from third parties to which a lender has a reasonable expectation of access (or even a legal right), like spousal income. These comments are described in the sectionby-section analysis for § 1041.5(a)(5). 773 Some commenters argued that consumers of online loans would need a fax machine or scanner to submit evidence of income, something that many of their customers do not own. These comments are described in more detail above in the section-by-section analysis for § 1041.5(c)(1). Commenters asked for further detail about what constitutes a “sufficient history” of net income for purposes of the verification requirement, a phrase appearing in the proposed regulation text without corresponding commentary. These commenters asked how long a lender should look back (e.g., for how many pay stubs) to establish a sufficient history. One trade group asked for a safe harbor of two pay cycles of verification evidence for covered longer-term loans, citing NCUA requirements. Other lenders asked whether they could look far back into the past, for example, at the bonus payment from last year, to help establish whether the borrower is likely to receive one this year. Consumer advocates argued that for longer-term loans with a duration of longer than six months, “sufficient history” should correspond to the length of the loan. Final Rule The Bureau has carefully considered the comments received and has concluded that it is appropriate to make two significant modifications to proposed § 1041.5(c)(3)(ii)(A). First, while the Bureau continues to believe that it is critical for lenders to obtain reliable records of net income if they are reasonably available, the Bureau has decided to permit lenders discretion to reasonably rely on consumers’ written statements of net income where such records cannot be obtained. Second, with regard to situations in which the consumer has a reasonable expectation of access to the income of another person, the Bureau has decided to permit lenders discretion to 774 reasonably rely on such income but only if they have obtained verification evidence of regular access to that income, such as documentation of a joint account. Specifically, in the final rule, § 1041.5(c)(2)(ii)(A)(1) has been revised to provide that the lender must obtain a reliable record (or records) of an income payment (or payments) directly to the consumer covering sufficient history to support the lender’s projection under § 1041.5(c)(1) if a reliable record (or records) is reasonably available. Section 1041.5(c)(2)(ii)(A)(1) has also been revised in the final rule to provide that if a lender determines that a reliable record (or records) of some or all of the consumer’s net income is not reasonably available, then the lender may reasonably rely on the consumer’s written statement described in § 1041.5(c)(2)(i)(A) for that portion of the consumer’s net income. The Bureau has added two comments in the final rule to accompany these changes in the regulation text. First, comment 5(c)(2)(ii)(A)-3 clarifies the meaning of “reasonably available” records. The comment clarifies that a reliable record of the consumer’s net income is reasonably available if, for example, the consumer’s source of income is from her employment and she possesses or can access a copy of her recent pay stub. The comment clarifies that the consumer’s recent transaction account deposit history is a reliable record (or records) that is reasonably available if the consumer has such an account. The comment further clarifies that with regard to such bank account deposit history, the lender could obtain it directly from the consumer or, at its discretion, with the consumer’s permission via an account aggregator service that obtains and categorizes consumer deposit account and other account transaction data. The comment also clarifies that in situations in which income is neither documented through pay stubs or transaction account records, the reasonably available standard requires the lender to act in good 775 faith and exercise due diligence as appropriate for the circumstances to determine whether another reliable record (or records) is reasonably available. Second, comment 5(c)(2)(ii)(A)-4 clarifies when a lender can reasonably rely on a consumer’s statement if a reliable record is not reasonably available. The comment clarifies that § 1041.5(c)(2)(ii)(A) does not permit a lender to rely on a consumer’s written statement that the consumer has a reasonable expectation of access to the income of another person. The comment further clarifies that a lender reasonably relies on the consumer’s written statement if such action is consistent with a lender’s written policies and procedures required under § 1041.12 and there is no indication that the consumer’s stated amount of net income on a particular loan is implausibly high or that the lender is engaged in a pattern of systematically overestimating consumers’ income. The comment clarifies that evidence of the lender’s systematic overestimation of consumers’ income could include evidence that the subset of the lender’s portfolio consisting of the loans where the lender relies on the consumers’ written statements to project income in absence of verification evidence perform worse, on a non-trivial level, than other covered loans made by the lender with respect to the factors noted in comment 5(b)-2.iii indicating poor loan performance (e.g., high rates of default, frequent re-borrowings). The comment also clarifies that if the lender periodically reviews the performance of covered shortterm loans or covered longer-term balloon-payment loans where the lender has relied on consumers’ written statements of income and uses the results of those reviews to make necessary adjustments to its policies and procedures and future lending decisions, such actions indicate that the lender is reasonably relying on consumers’ written statements. The comment provides an example of how such necessary adjustments could include, for example, the lender changing its 776 underwriting criteria for covered short-term loans to provide that the lender may not rely on the consumer’s statement of net income in absence of reasonably available verification evidence unless the consumer’s debt-to-income ratio is lower, on a non-trivial level, than that of similarly situated applicants who provide verification evidence of net income. Finally, the comment clarifies that a lender is not required to consider income that cannot be verified other than through the consumer’s written statement. The Bureau emphasizes four points relating to the changes in the final rule permitting lenders to reasonably rely on consumer statements of net income where reliable records for verification are not reasonably available. First, the test for whether a reliable record is reasonably available is not whether the consumer brings it with him to the store, but rather is akin to whether such records could have been brought because they do, in fact, exist. Pay stubs and transaction account history records documenting income are considered reliable records as clarified by comment 5(c)(2)(ii)(A)-3. If the consumer possesses or can access these types of records, the consumer has to provide them as needed to verify the consumer’s written statement and the lender cannot merely rely on the consumer’s written statement. Second, the Bureau expects that such reliance on consumers’ written statements will occur in relatively narrow circumstances. These would include situations where a consumer has a primary job where she receives a traditional pay stub but has a side business or job where the consumer is paid in cash and cannot document the income, and the small number of cases where a consumer is paid entirely in cash for her primary job and has no transaction account or deposits only a portion of cash wages in the account. In the vast majority of cases, the Bureau expects 777 that the consumer will have a pay stub or transaction account history that can serve as a reliable record to verify the relevant net income. Third, as stated in comment 5(c)(2)(ii)(A)-4, lenders are not required to consider income that cannot be verified other than through the consumer’s written statement (i.e., where a reliable record is not reasonably available). However, if they do so they are still subject to a reasonableness standard. The comment specifically notes that a lender reasonably relies on the consumer’s written statement only if such action is consistent with a lender’s written policies and procedures required under § 1041.12 and there is no indication that the consumer’s stated amount of net income on a particular loan is implausibly high or that the lender is engaged in a pattern of systematically overestimating consumers’ income. The comment also discusses what types of performance patterns might constitute evidence of a lender’s systematic overestimation of income and ways in which lenders could monitor and make adjustments to their policies and future lending decisions in the face of such evidence. The Bureau thus expects to monitor lenders for systematic overestimation of income where lenders are relying on consumers’ stated income amounts. The Bureau will look at whether lenders themselves are monitoring such loans and making appropriate adjustments to their underwriting policies and procedures and lending decisions. Fourth, the Bureau recognizes that, generally, the current practice among storefront payday lenders (but not vehicle title lenders) is to verify at least one pay stub of income for an initial loan. The Bureau thus believes that lenders have strong incentives to continue that practice rather than shift toward a widespread model of relying on stated income. With vehicle title lenders there are greater incentives for lenders to forgo verification and rely on the asset 778 value of the vehicle. But under the final rule, even for vehicle title lenders, the lender can only reasonably rely on the consumer’s statement of income when a reliable record is not reasonably available. The Bureau believes this approach responds appropriately to the comments from industry and other stakeholders about how the proposed verification requirements would not have accounted for, and potentially would have disadvantaged, individuals who are paid in cash and could afford to repay the loan but may not have the necessary documentation. At the same time, the Bureau believes that the final rule’s requirements that lenders’ reliance on consumers’ written statements of income must be reasonable and that lenders can only rely on such written statements when the records are not “reasonably available”—along with the detailed guidance in commentary about the meaning of those terms and the expectations around lender monitoring— will provide guardrails against lender overreliance on consumers’ written statements of income and the potential for abuse of this provision. For these reasons, as well as those noted in the several paragraphs above, the Bureau disagrees with the suggestion by the consumer group commenters that lenders should not be permitted to rely on consumers' written statements of income in any circumstances or that they should only be permitted to use a more conservative estimate. The other significant change is in response to statements by commenters that some consumers rely on income from third parties such as spouses or partners. The Bureau has added § 1041.5(c)(2)(ii)(A)(2) which permits consumers to rely on third party income, but only when the lender obtains verification evidence to support the fact that the consumer has a reasonable expectation of access to that income. The Bureau recognizes that many consumers either pool 779 their income in households or rely on third-party income, such as contributions from siblings or from parents to adult children. Given this fact, the Bureau in finalizing the rule is allowing lenders to rely on third-party income when calculating net income. However, the Bureau is adopting a different approach with regard to verification of such income relative to income received directly by the consumer. As described above, for a consumer’s income, a lender must obtain verification evidence unless it is not reasonably available. For third-party income, a lender must obtain verification evidence that the consumer has a reasonable expectation of access to that income for such income to be included in the ability-to-repay analysis. Comment 5(c)(2)(ii)(A)-1 clarifies that such evidence could consist of bank account statements indicating that the consumer has an account into which the other person’s income is regularly deposited. With regard to income that is not the consumer’s own income, the Bureau judges it is important for lenders to obtain objective evidence of regular access. The Bureau acknowledges that in this regard the rule imposes a more demanding verification requirement than applies under the CARD Act with respect to “accessible income” but notes again that, as explained earlier, differences between the credit card market and the market for short-term and balloon-payment loans warrant the differences in treatment; see the general discussion of § 1041.5 and the discussion of § 1041.5(a)(5) and comment 5(a)(5)-3 for further detail. In response to commenter requests for clarification about what constitutes “sufficient history” for purposes of projecting income, the Bureau has added a new comment 5(c)(2)(ii)(A)2 to provide general guidance. The comment states that: for covered short-term loans, one pay cycle would typically constitute sufficient history; and for longer-term balloon payment loans, two pay cycles generally would constitute sufficient history. However, the comment also 780 clarifies that additional verification evidence may be needed to resolve inconsistency between verification evidence and consumers’ written statements, and depending on the length of the loan. For covered longer-term balloon-payment loans, a national trade association for online lenders suggested a safe harbor for sufficient history of two pay stubs, citing National Credit Union Administration requirements for certain loans. In contrast, consumer groups argued that for covered longer-term loans greater than six months in duration, the final rule should require a look-back period of the length of the loan. The Bureau declines to adopt the consumer groups’ suggestion, because such a long look-back period would impose significant burdens on lenders and consumers to provide many months of pay stubs or bank statements, at least for loans of significant length. At the same time, the Bureau does not believe a safe harbor of two pay cycles would be appropriate, given that in some circumstances more income history might be necessary to project future income. The Bureau has structured comment 5(c)(2)(ii)(A)-2 to take into account these competing considerations. In response to a comment to the proposal seeking clarification on how far back lenders may look to make reasonable projections of future net income, specifically citing the issue of annual bonuses, § 1041.5(c)(2)(ii)(A), as clarified by new comment 5(c)(2)(ii)(A)-2, does not preclude the lender from requesting additional verification evidence dating back to earlier periods where needed to make the lender’s projection of income reasonable. 5(c)(2)(ii)(B), (C), and (D) Proposed Rule 781 The Bureau proposed separate provisions to detail the verification requirements for different types of major financial obligations in proposed § 1041.5(c)(3)(ii)(B) (debt obligations), § 1041.5(c)(3)(ii)(C) (child support), and § 1041.5(c)(3)(ii)(D) (rental housing expense), respectively. Specifically, in proposed § 1041.5(c)(3)(ii)(B) the Bureau specified that for a consumer’s required payments under debt obligations, the applicable verification evidence would be a national consumer report, the records of the lender and its affiliates, and a consumer report obtained from an information system currently registered pursuant to § 1041.17(c)(2) or (d)(2), if available. The Bureau believed that most typical consumer debt obligations other than covered loans would appear in a national consumer report. Many covered loans are not included in reports generated by the nationwide consumer reporting agencies, so the lender would also be required to obtain, as verification evidence, a consumer report from consumer reporting agency that specifically registers with the Bureau under part 1041. As discussed above, proposed § 1041.5(c)(1) would have permitted a lender to base its projections on consumer statements of amounts and timing of payments for major financial obligations (including debt obligations) only to the extent the statements are consistent with the verification evidence. Proposed comment 5(c)(1)-1 included examples applying that proposed requirement in scenarios where a major financial obligation shown in the verification evidence is greater than the amount stated by the consumer and where a major financial obligation stated by the consumer does not appear in the verification evidence at all. Proposed comment 5(c)(3)(ii)(B)-1 would have clarified that the amount and timing of a payment required under a debt obligation are the amount the consumer must pay and the time by which the consumer must pay it to avoid delinquency under the debt obligation in the absence of 782 any affirmative act by the consumer to extend, delay, or restructure the repayment schedule. To the extent the national consumer report and the consumer report from a registered information system omit information for a payment under a debt obligation stated by the consumer, the Bureau explained in the proposal that the lender would simply base its projections on the amount and timing stated by the consumer. The Bureau also emphasized in the proposal that proposed § 1041.5(c)(3)(ii)(B) would not have required a lender to obtain a consumer report unless the lender is otherwise prepared to make a loan to a particular consumer. Because obtaining a consumer report adds some cost, the Bureau assumed in the proposal that lenders would order such reports only after determining that a consumer otherwise satisfied the ability-to-repay test so as to avoid incurring costs for applicants who would be declined without regard to the contents of the report. Similarly, in proposed § 1041.5(c)(3)(ii)(C), the Bureau specified that for a consumer’s required payments under court- or government agency-ordered child support obligations, the applicable verification evidence would be the same national consumer report that serves as verification evidence for a consumer’s required payments under debt obligations under proposed § 1041.5(c)(3)(ii)(B). To the extent the national consumer report omitted information for a required payment, the Bureau explained in the proposal that the lender could simply base its projections on the amount and timing stated by the consumer, if any. Comments Received Many industry commenters and many individual consumer commenters objected broadly to the proposed requirements to collect verification evidence on major financial obligations on the grounds of burden, efficacy, and negative consequences for consumers. For example, many 783 individual consumer commenters and several lenders and industry trade groups argued that requiring a credit check for every loan will harm consumers’ credit by lowering their credit scores. Others stated that many consumers do not have a credit history, and so the credit check will not work. Still others claimed that the credit check and requirement to obtain a report from a registered information system would be costly for lenders. The SBA Office of Advocacy encouraged the Bureau to eliminate the credit check requirement because they argued it is an unnecessary hurdle based on feedback from small business roundtable participants. They also noted the costs to small businesses, citing the Bureau’s estimate in the proposal that a consumer report will cost approximately $2.00 for small lenders versus $0.55 for larger lenders. They also reported that SERs stated that the actual cost of a consumer report may be as high as $12.00. Commenters more specifically asked the Bureau to require that registered information systems only charge lenders a fee if a report is actually obtained (as opposed to an inquiry that generates no hits). Other commenters asked for a safe harbor when they rely on information from a consumer report, noting that the information in a consumer report may be inaccurate. A specialty consumer reporting agency commenting on the proposed provision requiring lenders to obtain a national consumer report to verify debt obligations and child support obligations wrote that it agreed with the Bureau’s assumption that lenders will stage the ordering of credit reports. The commenter wrote that it expected lenders will have a “two-step process” for obtaining national consumer reports—they would first order the separate required report from the registered information system to determine the borrowing history on covered loans and would conduct a preliminary underwriting assessment, and that only if the applicant passed that 784 first phase would the lender then order the national consumer report as part of the final ability-torepay determination. Commenters noted that credit report information is for the past, and not the future for which the lender would need to project major financial obligations. The commenters asked for clarification on whether in these instances a lender can trust a consumer’s statements regarding future payments, or how the lender will be able to project for any changes to the obligation in the future. Final Rule After careful consideration of the comments, the Bureau is finalizing proposed § 1041.5(c)(3)(ii)(B) and 5(c)(3)(ii)(C) as final § 1041.5(c)(2)(ii)(B) and 5(c)(2)(ii)(C), respectively, to address verification evidence for debt obligations and for alimony and child support. With regard to debt obligations, the final rule is consistent with the proposal in that it generally requires that lenders search their own records and those of affiliates and obtain consumer reports from both a nationwide consumer reporting agency and from an information system that has been registered for 180 days or more pursuant to § 1041.11(c)(2) or is registered pursuant to § 1041.11(d)(2), if available. However, in recognition of commenters’ concerns about the burdens on lenders with regard to the requirement to obtain a national consumer report (particularly on small lenders as described by the SBA Office of Advocacy) and the possibility of small negative impacts on some consumers’ credit scores as discussed further below, the Bureau has adopted new § 1041.5(c)(2)(ii)(D) to permit lenders and their affiliates to rely on a national consumer report that was obtained within the prior 90 days, provided that the consumer 785 did not complete a three-loan sequence and trigger the mandatory 30-day cooling-off period under § 1041.5(d)(2) since the prior report was obtained. Even with this change, the Bureau acknowledges that there will be some costs associated with obtaining consumer reports from nationwide consumer reporting agencies, and costs associated with obtaining a report from a registered information system. The Bureau has estimated, in its Section 1022(b)(2) Analysis below, that the cost of obtaining a report from a registered information system will likely be around $0.50 “per-hit,” and has estimated that the cost of pulling a consumer report from a nationwide consumer reporting agency will run somewhere between $0.50 and $2.00 each, depending on the report. The Bureau agrees that these are not small costs. However, they are essential to making sure that the lender can adequately determine that a borrower has an ability to repay, and are essential to the proper administration of the cooling-off period found in § 1041.5(d). In particular, given the importance of tracking consumers’ borrowing patterns with regard to covered short-term loans and covered longer-term balloon-payment loans under § 1041.5 in order to comply with § 1041.6 and with the cooling-off period provisions of § 1041.5(d), the Bureau believes it is important to require that lenders obtain new reports from registered information systems for each such loan where available. Further, as noted in the section-by-section analysis for § 1041.4, the Bureau believes that any impact on consumer’s credit scores will be minimal as a result of the requirements under § 1041.5(c)(2)(ii), for several reasons. First, as discussed above, the final rule in general only requires a credit check no more than once every 90 days, rather than for every loan. Second, as discussed in the proposal, the Bureau expects that lenders making loans under § 1041.5 will only 786 order national consumer reports after determining that the consumer otherwise satisfies the rule’s eligibility requirements and the ability-to-repay test using a consumer report from a registered information system so as to avoid incurring these costs for applicants who would be declined without regard to the contents of the national consumer report. In this regard, the Bureau notes the comment described earlier from a specialty consumer reporting agency which predicted that lenders would develop a “two-step process” for obtaining credit reports—they would first order the report from the registered information system and would determine the borrowing history on covered loans, along with a preliminary underwriting assessment, and that only if the applicant passed that first phase would the lender then order the national consumer report as part of the final ability-to-repay determination. Thus, the Bureau expects that many consumers who apply for loans but are denied based on information reflected in a report from a registered information system will have no negative impacts on their credit scores.828 Third, as discussed in the 1022(b)(2) Analysis, the Bureau is projecting that the majority of covered short-term loans that would be made under the final rule would be made under § 1041.6, not § 1041.5, so this particular requirement may affect only a small number of consumers. Moreover, as a more general matter, the impact that any credit check with a nationwide consumer reporting agency will have on a borrower’s overall credit profile is limited and uncertain, given that every consumer report differs and different creditors use different credit scoring models. One of the most experienced scoring companies, FICO, says the following 828 The Bureau notes that in the Section 1022(b)(2) Analysis, there is discussion of how lenders may potentially minimize the cost impacts of these requirements by obtaining both the consumer report from the registered information system and a national consumer report as part of a consolidated report. Even with the consolidated reports envisioned there, however, lenders and the providers for the registered information systems could stagger the delivery of such reports such as to minimize the negative scoring impacts on consumers. 787 about the impact of credit inquiries on a consumer’s score: “The impact from applying for credit will vary from person to person based on their unique credit histories. In general, credit inquiries have a small impact on one’s FICO Scores. For most people, one additional credit inquiry will take less than five points off their FICO Scores. For perspective, the full range of FICO Scores is 300-850.” Through the Bureau’s market monitoring and outreach it also understands that such a decrease in credit score may only be reflected on consumer reports for up to 12 months or could be fixed during that time period. For these reasons, the Bureau believes that the negative impacts claimed by commenters resulting from lenders having to obtain national consumer reports will be minimal. The Bureau recognizes, as commenters note, that consumer reports always include historical information. Thus, in projecting forward to the relevant monthly period, there will be times when lenders will have to make reasonable adjustments based on the consumer’s written statement and other sources as discussed in § 1041.5(c)(1) and (2) and related commentary. In addition, the Bureau reconsidered, as commenters noted, whether it was inconsistent to count child support but not alimony as a major financial obligation, especially where alimony is court- or government agency-ordered, and thus, likely reported on a consumer report. (Commenters also had questioned why receipt of alimony or child support was not included as net income, as discussed in the section-by-section analysis for § 1041.5(a)(5)). In light of the fact that, like other major financial obligations, alimony could potentially appear on a consumer report, or alternatively, a lender could rely on a written statement from the consumer about alimony, and the fact that alimony meets the general definition of a major financial obligation, the Bureau has decided to adjust § 1041.5(c)(2)(ii)(C) and the corresponding commentary to 788 state that both alimony and child support obligations should be verified where possible from a national consumer report and that lenders may otherwise reasonably rely on information provided in a consumer’s written statement for purposes of verification. In addition, in response to commenters asking for a safe harbor for instances where information in a consumer report is inaccurate, the Bureau has added comment 5(c)(2)(ii)(B)-3 to clarify more specifically how lenders should resolve conflicting information about major financial obligations as between a consumer’s written statement and various forms of verification evidence. The comment also clarifies that a lender is not responsible for information about a major financial obligation that is not owed to the lender, its affiliates, or its service providers if such obligation is not listed in a consumer’s written statement, a national consumer report, or a consumer report from an information system that has been registered for 180 days or more pursuant to § 1041.11(c)(2) or is registered pursuant to § 1041.11(d)(2). A similar provision addressing inaccurate or incomplete information in consumer reports from an information system that has been registered for 180 days or more pursuant to § 1041.11(c)(2) or is registered pursuant to § 1041.11(d)(2) has been included in the commentary for § 1041.6. With regard to the privacy concerns raised by commenters, the lender need only obtain information about the borrower’s individual income, information that is on consumer reports (including a report from a registered information system), and information contained in the borrower’s written statement. In the modern era, it is quite typical for creditors to have access to consumer reports, and many other parties, including employers, often do as well. The Bureau expects lenders to act in accordance with permissible use restrictions as prescribed in the Fair Credit Reporting Act and other privacy laws and regulations to the extent applicable. Lenders 789 will also ask consumers questions about, and receive verification evidence on, income. In the payday market, this will likely make only a marginal difference with respect to privacy because the payday market typically already collects this information. It will have a more significant impact for vehicle title lending borrowers, who would now have to obtain income verification. The Bureau recognizes that some consumers will be troubled by the increased scrutiny into borrowers’ private information, as noted by many individual commenters, but believes that these concerns have been somewhat reduced by changes to the final rule and in any event are worth the benefits of requiring income verification. The Bureau has made a number of technical and structural, as well as substantive, changes to § 1041.5(c)(2)(ii) and the related commentary to implement the policy changes discussed above, and the policy changes found throughout other sections in § 1041.5. Lastly, in response to commenters’ concerns that lenders may “double-count” certain major financial obligations if they are deducted from income, the Bureau notes that it has added comment 5(c)(2)(ii)(B)-2 and comment 5(c)(2)(ii)(C)-2, which specify that if verification evidence shows that a debt obligation, child support obligation, or alimony obligation is deducted from the consumer’s income, the lender does not include those amounts in the projection of major financial obligations. This change and the comments underlying the change are discussed in more detail in the section-by-section analysis for § 1041.5(a)(3) (definition of major financial obligations), above. 5(c)(2)(iii) Proposed Rule 790 The Bureau proposed a more flexible approach with regard to rental housing expenses in proposed § 1041.5(c)(3)(ii)(D) than with regard to other major financial obligations. Specifically, proposed § 1041.5(c)(3)(ii)(D) specified that for a consumer’s housing expense (other than a payment for a debt obligation that appears on a national consumer report obtained by the lender under proposed § 1041.5(c)(3)(ii)(B)), the applicable verification evidence would be either a reliable transaction record (or records) of recent housing expense payments or a lease, or an amount determined under a reliable method of estimating a consumer’s housing expense based on the housing expenses of consumers in the same locality. Proposed comment 5(c)(3)(ii)(D)-1 described each of the options for verification evidence in more detail. Most importantly, proposed comment 5(c)(3)(ii)(D)-1.iii provided examples of situations in which a lender used an amount determined under a reliable method of estimating a consumer’s share of housing expense based on the individual or household housing expenses of similarly situated consumers with households in the same locality, such as relying on the American Community Survey of the U.S. Census Bureau to estimate individual or household housing expense in the locality (e.g., in the same census tract ) where the consumer resides. In the alternative, the comment also provided that a lender may estimate individual or household housing expense based on housing expense and other data (e.g., residence location) reported by applicants to the lender, provided that it periodically reviews the reasonableness of the estimates that it relies on using this method by comparing the estimates to statistical survey data or by another method reasonably designed to avoid systematic underestimation of consumers’ shares of housing expense. It further explained that a lender may estimate a consumer’s share of household expense by reasonably apportioning the estimated household housing expense among 791 the people sharing the housing expense as stated by the consumer, or by another reasonable method. The Bureau explained in the proposal that this approach was designed to address concerns that had been raised in the SBREFA process and other industry outreach prior to the proposal. In particular, the Small Business Review Panel Outline had referred to lender verification of a consumer’s rent or mortgage payment using, for example, receipts, cancelled checks, a copy of a lease, and bank account records. As discussed in the proposal, some SERs and other lender representatives stated that many consumers would not have these types of documents readily available. Few consumers receive receipts or cancelled checks for rent or mortgage payments, they stated, and bank account statements may simply state the check number used to make a payment, providing no way of confirming the purpose or nature of the payment. Consumers with a lease would not typically have a copy of the lease with them when applying for a covered loan, they stated, and it would be unduly burdensome, if not impracticable for them to locate and transmit or deliver a copy of the lease to a lender. Comments Received Several commenters argued that the Bureau’s standards around verifying housing expenses were unfair and would lead to a significant number of “false negatives” (i.e., unintended denials of credit) for consumers who can, and do regularly, pay for rental housing expense but do not possess the requisite verification evidence. Commenters claimed that many consumers have non-traditional living arrangements where there is no documented lease, or live rent-free with a relative, and thus would not have any verification evidence of rent. Some commenters argued that asking consumers for verification of rental housing expense would be 792 considered intrusive, particularly for those consumers living in informal rooming arrangements. The Bureau also received a number of comments arguing that the proposal had not accounted for consumers who share rental housing expense and where the formal documentation does not reflect the arrangement. For example, if two roommates pay rent on the same lease, the verification evidence would indicate that only one of them may have to pay the full rent. Other commenters claimed that rental agreements were difficult to procure, and thus, it would be impractical to require one. Commenters also argued that bank statements would not be sufficient to verify housing expenses because they might not show the names of the recipients of the rental payments. A number of commenters raised issues with the provision in proposed § 1041.5(c)(3)(ii)(D) that lenders could estimate rental housing expenses by using a reliable method (either locality-based data or data on their own customers) as an alternative to collecting a reliable record of rental housing expense such a lease. As explained earlier, the Bureau had included the alternative in response to feedback during the SBREFA process and outreach that the Small Business Review Panel Outline took too strict of an approach to verification of rental housing expense. Commenters were critical of this proposed provision on a number of grounds. Some argued it would be burdensome for lenders, particularly small lenders, to develop statistical estimates. Other commenters argued that using census tract data, as given in a proposed commentary example, could substantially overstate housing expenses by failing to account for the greater amount of shared living arrangements among payday borrowers or the demographics of this borrowing population. One trade group argued that, at minimum, the Bureau should allow “validation” of housing expenses based on a consumer’s stated history and 793 circumstances. Some commenters also raised concerns that taking into account locality-based information on housing expenses in underwriting decisions could violate ECOA and Regulation B (see discussion of these issues in the more general § 1041.5 discussion above). Consumer groups, on the other hand, commented that the proposal’s treatment of rental housing expenses was too permissive. They argued that rental housing expense should be verified wherever possible, and that if verification evidence is not available, the lender should have to use the larger of the locality-based average or the consumer’s statement. They also argued that if there is a shared arrangement, lenders should obtain verification evidence (such as a lease or checking account activity) or reliable third-party evidence (like a co-tenant statement). They expressed concern about proposed commentary language permitting lenders to apportion household expense based solely on the consumer’s statement. And they argued that the Bureau should consider providing “portfolio-level guardrails” that indicate whether housing estimates not based on verification evidence and lender assertions of shared housing expense are more likely to be unreasonable, and subjecting lenders whose portfolios have those indicators to higher scrutiny. Final Rule The Bureau is finalizing proposed § 1041.5(c)(3)(ii)(D) as final § 1041.5(c)(2)(iii) with a number of significant changes as discussed below. In response to the many comments criticizing the Bureau for proposing to require, for rental housing, either a reliable record or an estimate based on the housing expenses of consumers with households in the same locality (including concerns about fair lending interests discussed above), the Bureau has adjusted § 1041.5(c)(2)(iii) to provide that lenders may reasonably rely on the consumer’s written 794 statement to project rental housing obligations. New comment 5(c)(2)(iii)-1 states that a lender reasonably relies on the consumer’s written statement if such actions are consistent with a lender’s policies and procedures, there is no evidence that the stated amount on a particular loan is implausibly low, and there is no pattern of the lender underestimating consumers’ rental housing expense. The Bureau views these clarifications as analogous to those in comment 5(c)(2)(ii)(A)-4 regarding reasonable reliance on stated amounts for net income, and refers to the explanatory explanation above. The Bureau recognizes that there are likely a significant number of consumers, as noted by commenters, that have non-traditional living situations, live with roommates without being on the lease, rent on a month-to-month basis without a lease or a current lease, sublet, or live with a third party (like parents). The Bureau also recognizes that requiring consumers with a lease to present those documents to obtain a loan could prove burdensome, especially for consumers applying online. For these reasons, the proposal did not require applicants to provide a lease even where one existed. Instead, the proposal allowed lenders instead to rely on verification evidence consisting of data that could be used to validate the reasonableness of a consumer’s statement of rental housing expenses. However, the Bureau is persuaded by the weight of the comments suggesting that the proposal’s approach to estimation of expenses raised a number of challenges. Specifically, the Bureau is persuaded by commenters who argued that data on the median or average rental expenses for households in the same locality may not accurately reflect the median or average demographic or housing expenses of customers of covered short-term and longer-term balloon-payment loans and would thus potentially overstate the amount of rental 795 housing expense for prospective borrowers. Furthermore, the Bureau is persuaded that even if it were possible to determine the average or median rental expense for these consumers, such data would not be useful in validating the reasonableness of any individual consumer’s statement of her rental housing expenses, which could be vastly different from the average or median consumer. Finally, the Bureau agrees with commenters who noted that small lenders would be at a disadvantage in obtaining statistical validation evidence. The Bureau continues to recognize the risks entailed in permitting lenders to rely on stated rental expenses—including the risk that consumers will misstate or be induced to misstate their expenses—which are concerns echoed by the consumer groups in their comment. But the Bureau nonetheless is persuaded that the available alternatives are not practical and therefore is permitting lenders to rely on consumers’ written statements of rental housing expense where it is reasonable to do so. 5(d) Additional limitations on lending—covered short-term loans and covered longer-term balloon-payment loans Proposed § 1041.6 would have augmented the basic ability-to-repay determination in proposed § 1041.5 in circumstances in which the consumer’s recent borrowing history or recent difficulty repaying an outstanding loan provides important evidence with respect to the consumer’s financial capacity to afford a new covered short-term loan. In particular, proposed § 1041.6 would have imposed a presumption of unaffordability when a consumer returned for a covered short-term loan within 30 days of a prior covered short-term or covered longer-term balloon-payment loan being outstanding. Presumptions would also have been imposed in particular circumstances indicating that a consumer was having difficulty repaying an outstanding covered or non-covered loan outstanding that was made or was being serviced by the 796 same lender or its affiliate. Under the proposed approach, lenders would have been able to overcome a presumption of unaffordability only in circumstances where there was a sufficient improvement in financial capacity. This would have applied, for instance, where there was evidence that the prior difficulty with repayment was due to an income shock that was not reasonably expected to recur or where there was a reasonable projected increase of income or decrease in major financial obligations during the term of the new loan. However, after the third covered short-term loan in a sequence, proposed § 1041.6 would have imposed a mandatory 30day cooling-off period. The proposed section also contained certain additional provisions that were designed to address concerns about potential evasion and confusion if consumers alternated in quick succession between covered short-term loans under proposed § 1041.5 and other types of credit products. Similarly, proposed § 1041.10 would have applied parallel presumptions of unaffordability to new covered longer-term loans based on consumers’ recent borrowing history on certain types of covered loans or difficulty repaying a current covered or non-covered loan, that was made or was being serviced by the same lender or its affiliate, although it would not have imposed a mandatory cooling-off period after a three loan sequence. Proposed § 1041.10 also would have imposed certain restrictions to address concerns about potential evasion and confusion if consumers alternated in quick succession between a covered short-term loan under proposed § 1041.7 and other types of credit products. After consideration of the comments received as discussed further below, the Bureau has decided to finalize only selected elements of proposed §§ 1041.6 and 1041.10, consolidated as § 1041.5(d) of the final rule. Specifically, the Bureau is finalizing a 30-day mandatory cooling797 off period after a consumer has completed a three-loan sequence of covered short-term loans, covered longer-term balloon-payment loans, or a combination thereof. It is also finalizing restrictions on certain re-borrowing within 30 days of a covered shorter-term loan made under final rule § 1041.6 (which was § 1041.7 in the proposal) being outstanding.829 As explained below in detail, the Bureau is not finalizing several provisions, including any of the proposed presumptions of unaffordability. Thus, the Bureau is finalizing adjusted portions of proposed § 1041.6(a)(2), (f), and (g) and proposed § 1041.10(a)(2) and (e) on a combined basis in § 1041.5(d) for both covered short-term and longer-term balloon-payment loans as discussed further below. Proposed Rule In proposed § 1041.6, the Bureau proposed to require the lender to factor evidence about the consumer’s recent borrowing history and difficulty in repaying an outstanding loan into the ability-to-repay determination and, in certain instances, to prohibit a lender from making a new covered short-term loan to the consumer under proposed § 1041.5 for 30 days. The Bureau proposed the additional requirements in § 1041.6 for the same basic reason that it proposed § 1041.5: to prevent the unfair and abusive practice identified in proposed § 1041.4 and the consumer injury that results from it. The Bureau explained in the proposal that it believed that these additional requirements would be needed in circumstances where proposed § 1041.5 alone might not suffice to prevent a lender from making a covered short-term loan that the consumer would lack the ability to repay. 829 See also the section-by-section analysis of final § 1041.6(d) below, which discusses a related provision limiting loans by lenders or their affiliates within 30 days of a prior outstanding loan under § 1041.6 by the same lender or its affiliates. 798 Proposed § 1041.6 would have generally imposed a presumption of unaffordability on continued lending where evidence suggested that the prior loan was not affordable for the consumer, indicating that the consumer could have particular difficulty repaying a new covered short-term loan. Specifically, such a presumption would have applied in three circumstances: (1) under proposed § 1041.6(b), when a consumer sought a covered short-term loan during the term of a covered short-term loan made under proposed § 1041.5 and for 30 days thereafter; (2) under proposed § 1041.6(c), when a consumer sought a covered short-term loan during the term of a covered longer-term balloon-payment loan made under proposed § 1041.9 and for 30 days thereafter; and (3) under proposed § 1041.6(d), when a consumer sought to take out a covered short-term loan when there were indicia that the consumer was already struggling to repay an outstanding loan of any type—covered or non-covered—with the same lender or its affiliate. The Bureau explained in the proposal that a central component of the preventive requirements in proposed § 1041.6 was the concept of a re-borrowing period—a period following the payment date of a prior loan during which a consumer’s borrowing of a covered short-term loan is deemed evidence that the consumer is seeking additional credit because the prior loan was unaffordable. When consumers have the ability to repay a covered short-term loan, the loan should not cause consumers to have the need to re-borrow shortly after repaying the loan. As discussed in the proposal, including in the proposal’s discussion of Market Concerns—Short-term Loans, however, the Bureau believed that the fact that covered short-term loans require repayment so quickly after consummation makes such loans more difficult for consumers to repay consistent with their basic living expenses and major financial obligations without needing to re-borrow. Moreover, as the Bureau explained, most covered short-term 799 loans—including payday and vehicle title loans—also require payment in a single lump sum, thus exacerbating the challenge of repaying the loan without needing to re-borrow. For these loans, the Bureau stated in the proposal that it believed that the fact that a consumer returns to take out another covered short-term loan shortly after having a previous covered short-term loan outstanding frequently indicates that the consumer did not have the ability to repay the prior loan and meet the consumer’s basic living expenses and major financial obligations. This also may provide strong evidence that the consumer will not be able to afford a new covered short-term loan. The Bureau further explained that a second covered short-term loan shortly following a prior covered short-term loan may result from a financial shortfall caused by repayment of the prior loan. The Bureau noted that evidence shows that re-borrowing for short-term loans often occurs on the same day that a loan is due, either in the form of a rollover of the existing loan (where permitted by State law) or in the form of a new loan taken out on the same day that the prior loan was repaid. Some States require a cooling-off period between loans, typically 24 hours, and the Bureau found that in those States, if consumers take out successive loans, they generally do so at the earliest time that is legally permitted.830 The Bureau interpreted these data to indicate that these consumers could not afford to repay the full amount of the loan when due and still meet their basic living expenses and major financial obligations. In the proposal, the Bureau stated that it is less facially evident whether a particular loan is a re-borrowing that was prompted by the unaffordability of a prior loan when that new loan is taken out after some time has elapsed since a consumer has repaid the prior loan (and after the 830 CFPB Report on Supplemental Findings, at Chapter 4. 800 expiration of any State-mandated cooling-off period). The fact that consumers may cite a particular income or expense shock is not dispositive, since a prior unaffordable loan may be the reason that the consumer cannot absorb the new change. The Bureau stated in the proposal that on balance, the Bureau believed that for new loans taken out within a short period after a prior loan ceases to be outstanding, the most likely explanation is the unaffordability of the prior loan—i.e., the fact that the size of the payment obligation on the prior loan left these consumers with insufficient income to make it through their monthly expense cycle. The Bureau explained in the proposal that to provide a structured process that accounts for the likelihood that the unaffordability of an existing or prior loan is driving re-borrowing and that ensures a more rigorous analysis of consumers’ individual circumstances, the Bureau believed that an appropriate approach would be to impose presumptions when new loans fall within a specified re-borrowing period, rather than engaging in an open-ended inquiry. The Bureau thus proposed to delineate a specific re-borrowing period, during which a new loan will be presumed to be a re-borrowing.831 In determining the appropriate length of the re-borrowing period, the Bureau described how it had considered several different possible periods. The Bureau proposed a 30-day period, but also considered periods of 14, 45, 60, or 90 days in length. 831 The Bureau explained in the proposal that re-borrowing takes several forms in the market for covered short-term loans. As used throughout the proposal, re-borrowing and the re-borrowing period include any rollovers or renewals of a loan, as well as new extensions of credit. The Bureau explained that a loan may be a “rollover” if, at the end of a loan term, a consumer only pays a fee or finance charge in order to “roll over” a loan rather than repaying the loan. The Bureau noted that similarly, the laws of some States permit a lender to “renew” a consumer’s outstanding loan with the payment of a finance charge, and that more generally, a consumer may repay a loan and then return to take out a new loan within a fairly short period. The Bureau stated in the proposal that it considers rollovers, renewals, and re-borrowing within a short period after repaying the prior loan to be functionally the same sort of transaction— and generally used the term re-borrowing in the proposal to cover all three scenarios, along with concurrent borrowing by a consumer whether from the same lender or its affiliate or from different, unaffiliated lenders. 801 The Bureau also considered an option that would tie the length of the re-borrowing period to the term of the preceding loan. In evaluating the alternative options for defining the re-borrowing period (and, in turn, the definition of a loan sequence), the Bureau described in the proposal how it was seeking to strike a balance between two alternatives. The first would be a re-borrowing period that is too short, thereby not capturing substantial numbers of subsequent loans that are in fact the result of the spillover effect of the unaffordability of the prior loan and inadequately preventing consumer injury. The second would be a re-borrowing period that is too long, thereby covering substantial numbers of subsequent loans that are in fact the result of a new need for credit, independent of such effects. The Bureau further described how this concept of a re-borrowing period is also intertwined with the definition of loan sequence. Under proposed § 1041.2(a)(12), the Bureau would have defined loan sequence as a series of consecutive or concurrent covered short-term loans in which each of the loans is made while the consumer currently has an outstanding covered short-term loan or within 30 days after the consumer ceased to have such a loan outstanding. The Bureau explained in the proposal that the Bureau’s 2014 Data Point analyzed repeated borrowing on payday loans using a 14-day re-borrowing period reflecting a bi-weekly pay cycle, the most common pay cycle for consumers in this market.832 For the purposes of the 2014 Data Point, a loan was considered part of a sequence if it was made within 14 days of the prior loan. The Bureau stated in the proposal that it had adopted this approach in its early research in order to obtain a relatively conservative measure of re-borrowing activity relative to 832 See generally CFPB Data Point: Payday Lending. 802 the most frequent date for the next receipt of income. However, the 14-day definition had certain disadvantages, including the fact that many consumers are paid on a monthly cycle, and a 14-day definition thus does not adequately reflect how different pay cycles can cause somewhat different re-borrowing patterns. The Bureau stated in the proposal that upon further consideration of what benchmarks would sufficiently protect consumers from re-borrowing harm, the Bureau turned to the typical consumer expense cycle, rather than the typical income cycle, as the most appropriate metric.833 The Bureau noted that consumer expense cycles are typically a month in length with housing expenses, utility payments, and other debt obligations generally paid on a monthly basis. Thus, where repaying a loan causes a shortfall, the consumer may seek to return during the same expense cycle to get funds to cover downstream expenses. The proposals under consideration in the Small Business Review Panel Outline relied on a 60-day re-borrowing period, based on the premise that consumers for whom repayment of a loan was unaffordable may nonetheless be able to juggle their expenses for some time so that the spillover effects of the loan may not manifest themselves until the second expense cycle following repayment. As explained in the proposal, upon additional analysis and extensive feedback from a broad range of stakeholders, the Bureau tentatively concluded that the 30-day definition incorporated into the proposal may strike a more appropriate balance between the competing considerations, chiefly because so many expenses are paid on a monthly basis. 833 The Bureau noted in the proposal that researchers in an industry-funded study also concluded that “an entire billing cycle of most bills—rent, other loans, utilities, etc.—and at least one paycheck” is the “appropriate measurement” for purposes of determining whether a payday loan leads to a “cycle of debt.” Marc Anthony Fusaro & Patricia J. Cirillo, “Do Payday Loans Trap Consumers in a Cycle of Debt?,” (2011), available at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=1960776. 803 The Bureau stated its belief that loans obtained during the same expense cycle are relatively likely to indicate that repayment of a prior loan may have caused a financial shortfall. Similarly, in analyzing supervisory data, the Bureau found that a considerable segment of consumers who repay a loan without an immediate rollover or re-borrowing nonetheless return within the ensuing 30 days to re-borrow.834 The Bureau stated in the proposal that accordingly, if the consumer returned to take out another covered short-term loan—or, as described in proposed § 1041.10, certain types of covered longer-term loans—within the same 30-day period, the Bureau believed that this pattern of re-borrowing indicated that the prior loan was unaffordable and that the following loan may likewise be unaffordable. On the other hand, the Bureau stated its belief that for loans obtained more than 30 days after a prior loan, there is a higher likelihood that the loan is prompted by a new need on the part of the borrower, and is not directly related to potential financial strain from repaying the prior loan. The Bureau further explained that while a prior loan’s unaffordability may cause some consumers to need to take out a new loan as many as 45 days or even 60 days later, the Bureau believed that the effects of the previous loan are more likely to dissipate once the consumer has completed a full expense cycle following the termination of a prior loan that has been fully repaid. For these reasons, the Bureau believed at the time it developed the proposed rule that a 45-day or 60-day definition would be too broad. The Bureau also stated in the proposal that a reborrowing period that would vary with the length of the preceding loan term would be operationally complex for lenders to implement and, for consumers who are paid either weekly or bi-weekly, may also be too narrow. 834 CFPB Report on Supplemental Findings, at Chapter 5. 804 Accordingly, using this 30-day re-borrowing window, the Bureau proposed a presumption of unaffordability for covered short-term loans made while a prior loan is outstanding or within a 30-day period after the end of the term of the prior loan. As proposed, however, the presumption could have been overcome in various circumstances suggesting that there is sufficient reason to believe the consumer would, in fact, be able to afford the new loan even though she is seeking to re-borrow during the term of or shortly after a prior loan. The Bureau recognized, for example, that there may be situations in which the prior loan would have been affordable but for some unforeseen disruption in income that occurred during the prior expense cycle and which is not reasonably expected to recur during the term of the new loan. The Bureau also recognized that there may be circumstances, albeit less common, in which even though the prior loan proved to be unaffordable, a new loan would be affordable because of a reasonably projected increase in net income or decrease in major financial obligations. To effectuate these policy decisions, proposed § 1041.6(a) would have set forth the general requirement for lenders to obtain and review information about a consumer’s borrowing history from the records of the lender and its affiliates, and from a consumer report obtained from an information system currently registered pursuant to proposed § 1041.17(c)(2) or (d)(2), if available, and to use this information to determine a potential loan’s compliance with the requirements of proposed § 1041.6. Proposed § 1041.6(b) to (d) would have defined the set of circumstances in which the Bureau believed that a consumer’s recent borrowing history makes it unlikely that the consumer 805 can afford a new covered short-term loan, including concurrent loans.835 In such circumstances, a consumer would be presumed not to have the ability to repay a covered short-term loan under proposed § 1041.5. Specifically, the presumption of unaffordability would have applied: (1) under proposed § 1041.6(b), when a consumer sought a covered short-term loan during the term of a covered short-term loan made under proposed § 1041.5 and for 30 days thereafter; (2) under proposed § 1041.6(c), when a consumer sought a covered short-term loan during the term of a covered longer-term balloon-payment loan made under proposed § 1041.9 and for 30 days thereafter; and (3) under proposed § 1041.6(d), when a consumer sought to take out a covered short-term loan when there are indicia that the consumer is already struggling to repay an outstanding loan of any type—covered or non-covered—with the same lender or its affiliate. Proposed § 1041.6(e) would have defined the additional determinations that a lender would be required to make in cases where the presumption applies in order for the lender’s ability-to-repay determination under proposed § 1041.5 to be reasonable despite the unaffordability of the prior loan. The presumption of unaffordability in proposed § 1041.6(c) would have provided that a consumer is presumed not to have the ability to repay a covered short-term loan under proposed 835 In the proposal, the Bureau noted that the proposed ability-to-repay requirements would have not prohibited a consumer from taking out a covered short-term loan when the consumer has one or more covered short-term loans outstanding, but instead accounted for the presence of concurrent loans in two ways: (1) a lender would have been required to obtain verification evidence about required payments on debt obligations, which were defined under proposed § 1041.5(a)(2) to include outstanding covered loans; and (2) any concurrent loans would have been counted as part of the loan sequence for purposes of applying the presumptions and prohibitions under proposed § 1041.6. The Bureau explained in the proposal that this approach differs from the conditional exemption for covered short-term loans under proposed § 1041.7, which generally would have prohibited the making of such a loan if the consumer has an outstanding covered loan. The Bureau noted that for further discussion, see the sectionby-section analysis of proposed § 1041.7(c)(1), including an explanation of the different approaches and notation of third-party data about the prevalence of concurrent borrowing in this market. 806 § 1041.5 during the time period in which the consumer has a covered longer-term balloonpayment loan made under proposed § 1041.9 outstanding and for 30 days thereafter. The Bureau stated in the proposal that it believed that when a consumer seeks to take out a new covered short-term loan that would be part of a loan sequence, there is substantial reason for concern that the need to re-borrow is being triggered by the unaffordability of the prior loan. Similarly, covered longer-term balloon-payment loans, by definition, require a large portion of the loan to be paid at one time. The Bureau described its research suggesting that the fact that a consumer seeks to take out another covered longer-term balloon-payment loan shortly after having a previous covered longer-term balloon-payment loan outstanding will frequently indicate that the consumer did not have the ability to repay the prior loan and meet the consumer’s other major financial obligations and basic living expenses. The Bureau stated that it had found that the approach of the balloon payment coming due is associated with significant re-borrowing. However, the need to re-borrow caused by an unaffordable covered longer-term balloon is not necessarily limited to taking out a new loan of the same type. The Bureau explained that if the borrower takes out a new covered short-term loan in such circumstances, it also is a reborrowing. Accordingly, in order to prevent the unfair and abusive practice identified in proposed § 1041.4, the Bureau proposed a presumption of unaffordability for a covered shortterm loan that would be concurrent with or shortly following a covered longer-term balloonpayment loan. In proposed § 1041.6(d), the Bureau would have established a presumption of unaffordability when a lender or its affiliate sought to make a covered short-term loan to an existing consumer in which there are indicia that the consumer cannot afford an outstanding loan 807 with that same lender or its affiliate. The triggering conditions would have been a delinquency of more than seven days within the preceding 30 days, expressions by the consumer within the preceding 30 days that he or she cannot afford the outstanding loan, certain circumstances indicating that the new loan is motivated by a desire to skip one or more payments on the outstanding loan, and certain circumstances indicating that the new loan is solely to obtain cash to cover upcoming payments or payments on the outstanding loan. The Bureau believed that the analysis required by proposed § 1041.6(d) would have provided greater protection to consumers and certainty to lenders than requiring that such transactions be analyzed under proposed § 1041.5 alone. Proposed § 1041.5 would have required generally that the lender make a reasonable determination that the consumer will have the ability to repay the contemplated covered short-term loan, taking into account existing major financial obligations that would include the outstanding loan from the same lender or its affiliate. However, the presumption in proposed § 1041.6(d) would have provided a more detailed roadmap as to when a new covered short-term loan would not meet the reasonable determination test. In proposed § 1041.6(f), the Bureau also would have established a mandatory cooling-off period prior to a lender making a fourth covered short-term loan in a sequence. As stated in the proposal, the Bureau believed that it would be extremely unlikely that a consumer who twice in succession returned to re-borrow during the re-borrowing period, and who seeks to re-borrow again within 30 days of having the third covered short-term loan outstanding, would be able to afford another covered short-term loan. Because of lenders’ strong incentives to facilitate reborrowing that is beyond the consumer’s ability to repay, the Bureau believed it appropriate, in proposed § 1041.6(f), to impose a mandatory 30-day cooling-off period after the third covered 808 short-term loan in a sequence, during which time the lender cannot make a new covered shortterm loan under proposed § 1041.5 to the consumer. This period was intended to ensure that after three consecutive ability-to-repay determinations have proven inconsistent with the consumer’s actual experience, the lender could not further worsen the consumer’s financial situation by extending additional unaffordable debt to the consumer. In its discussion of proposed § 1041.6(f), the Bureau stated that the ability-to-repay determination required by proposed § 1041.5 is intended to protect consumers from what the Bureau believes may be the unfair and abusive practice of making a covered short-term loan without making a reasonable determination of the consumer’s ability to repay the loan. If a consumer who obtains such a loan seeks a second loan when, or shortly after, the payment on the first loan is due, that suggests that the prior loan payments were not affordable and triggered the new loan application, and that a new covered short-term loan will lead to the same result. The Bureau stated that it believes that if a consumer has obtained three covered short-term loans in quick succession and seeks to obtain yet another covered short-term loan when or shortly after payment on the last loan is due, the fourth loan will almost surely be unaffordable for the consumer. In the proposal, the Bureau described how the Bureau’s research underscores the risk that consumers who reach the fourth loan in a sequence of covered short-term loans will wind up in a long cycle of debt. Most significantly, the Bureau found that 66 percent of loan sequences that reach a fourth loan end up having at least seven loans, and 47 percent of loan sequences that 809 reach a fourth loan end up having at least 10 loans.836 For consumers paid weekly, bi-weekly, or semimonthly, 12 percent of loan sequences that reach a fourth loan end up having at least 20 loans during a 10-month period.837 And for loans taken out by consumers who are paid monthly, more than 40 percent of all loans to these borrowers were in sequences that, once begun, persisted for the rest of the year for which data were available.838 The Bureau explained in the proposal, further, that the opportunity to overcome the presumption for the second and third loan in a sequence means that by the time that the mandatory cooling-off period in proposed § 1041.6(f) would apply, three prior ability-to-repay determinations will have proven inconsistent with the consumer’s actual experience. If the consumer continues re-borrowing during the term of or shortly after repayment of each loan, the pattern suggests that the consumer’s financial circumstances do not lend themselves to reliable determinations of ability to repay a covered short-term loan. After three loans in a sequence, the Bureau stated that it believes it would be all but impossible under the proposed framework for a lender to accurately determine that a fourth covered short-term loan in a sequence would be affordable for the consumer. The Bureau stated in the proposal that in light of the data described above, the Bureau believed that by the time a consumer reaches the fourth loan in a sequence of covered short-term loans, the likelihood of the consumer returning for additional covered short-term loans within a short period of time warrants additional measures to mitigate the risk that the lender is not furthering a cycle of debt on unaffordable covered short-term loans. To prevent the unfair and 836 Results calculated using data described in Chapter 5 of the CFPB Report on Supplemental Findings. Results calculated using data described in Chapter 5 of the CFPB Report on Supplemental Findings. 838 CFPB Report on Supplemental Findings, at Chapter 1. 837 810 abusive practice identified in proposed § 1041.4, the Bureau stated the belief that it may be appropriate to impose a mandatory cooling-off period for 30 days following the third covered short-term loan in a sequence. The Bureau’s overall approach to the re-borrowing restrictions in proposed § 1041.6 was fairly similar to the framework included in the Small Business Review Panel Outline, but contained some adjustments in response to feedback from the SERs, agency participants, and other stakeholders. For instance, the Bureau proposed a 30-day definition of loan sequence and a 30-day cooling-off period rather than a 60-day definition of loan sequence and a 60-day coolingoff period which was in the Small Business Review Panel Outline. The Bureau also proposed to provide greater specificity and flexibility about when a presumption of unaffordability would apply, for example, by proposing in § 1041.6(b)(2) certain exceptions to the presumption of unaffordability for a sequence of covered short-term loans where the consumer is seeking to reborrow no more than half the amount that the consumer has already paid on the loan. In those instances, the Bureau explained, the predicate for the presumption of unaffordability may no longer apply. The proposal also provided somewhat more flexibility about when a presumption of unaffordability could be overcome by permitting lenders to determine that there would be sufficient improvement in the consumer’s financial capacity for the new loan, under proposed § 1041.6(e). This standard would have included both documented increases in income or decreases in expenses since the prior borrowing (the Small Business Review Panel Outline standard of “changed circumstances”) plus where reliable evidence indicated that the need to reborrow was caused by a specific income decline that would not recur. The Bureau also continued to assess potential alternative approaches to the presumptions framework, as outlined 811 in the proposal,839 and specifically sought comment in response to the Small Business Review Panel Report on whether a loan sequence should be defined with reference to a period shorter or longer than 30 days. Proposed § 1041.10 would have applied a parallel set of presumptions of unaffordability to new covered longer-term loans where consumers had had a covered short-term or covered longer-term balloon-payment loan outstanding within the last 30 days or where there were indicia that consumers were having difficulty repaying a current loan of any type from the same lender or its affiliates. The Bureau’s logic in proposing to apply these presumptions was the same as described above with regard to proposed § 1041.6: Because covered longer-term balloon-payment loans also involve lump-sum or other large irregular payments that appear to exacerbate the challenge of repaying such loans without needing to re-borrow, there is substantial reason for concern that the need to re-borrow within a short time period is being triggered by the unaffordability of the prior loan. The Bureau did not specifically propose to impose a mandatory cooling-off period after a sequence of covered longer-term loans (whether they had a balloon payment or not), but sought comment on the general issue of whether a consumer’s intensity of use during a defined period of time warranted additional protections. 839 As discussed in the proposal, the Bureau had considered a number of alternative approaches to address reborrowing in circumstances indicating the consumer was unable to afford the prior loan. One alternative was to limit the overall number of covered short-term loans that a consumer could take out within a specified period, rather than using the loan sequence and presumption concepts, and when and if a mandatory cooling-off period should apply. Another was to simply identify circumstances that might be indicative of a consumer’s inability to repay that would be relevant to whether a lender’s determination under proposed § 1041.5 or § 1041.9 is reasonable. A third was whether there was a way to account for unusual expenses within the presumptions framework without creating an exception that would swallow the rule. The Bureau explained its concerns about each of these approaches in the proposal and broadly sought comment on alternative approaches to addressing the issue of repeat borrowing in a more flexible manner, including the alternatives described above, and on any other framework for assessing consumers’ borrowing history as part of an overall determination of ability-to-repay. 812 Finally, proposed §§ 1041.6 and 10 would also have established certain rules with regard to the prospect that consumers might switch back and forth between different types of covered or non-covered loans over time. In particular, proposed §§ 1041.6(g) and 1041.10(e) would have prohibited lenders under certain circumstances from making a covered short-term loan or a covered longer-term loan, respectively, while a prior covered short-term loan to the same consumer made under the conditional exemption in proposed § 1041.7 was outstanding or for 30 days thereafter. Because loans under that exemption are subject to certain principal reduction requirements over a sequence of three loans, the Bureau was concerned that the protections provided by that provision could be abrogated if a consumer were induced instead to take out a different kind of covered loan. Also, proposed §§ 1041.6(h) and 1041.10(f) would have suspended the 30-day count for purposes of determining whether a loan was subject to a presumption of unaffordability or the mandatory cooling-off period for short-term loans if a lender or its affiliate made a non-covered bridge loan within 30 days of a prior outstanding covered short-term loan or covered longer-term balloon-payment loan. The Bureau would have defined non-covered bridge loan in proposed § 1041.2(a)(13) as a non-resource pawn loan made by the same lender or its affiliate that is substantially repayable within 90 days of consummation. In the proposal, the Bureau described how this provision would address the concern that these types of loans could be used by lenders or their affiliates to bridge gaps between the making of covered loans, creating a continuous series of loans as a way of evading the proposed re-borrowing restrictions. Comments Received 813 The Bureau received numerous comments on the proposed re-borrowing restrictions. Stakeholders generally supportive of the rule criticized the restrictions for not going far enough, and stakeholders generally critical of the rule thought these restrictions went too far in a number of ways. Many consumer groups and other commenters argued that the Bureau should adopt a 45, 60, or 90 day cooling-off and re-borrowing period instead of a 30-day period, asserting that it takes longer than 30 days for a consumer to reach financial equilibrium. These arguments were based largely on arguments that had already been raised in response to the Small Business Review Panel Outline. The consumer advocates raised additional arguments for why the 30-day period was too short, including evidence from the U.S. Financial Diaries project and from national delinquency data on unsecured debt that they interpreted to suggest that consumers who take covered loans have monthly expense cycles greater than 30 days, and often in excess of 60 days. A State Attorney General urged that if the Bureau were not to adopt a 60-day cooling-off period, the Bureau should consider a 45-day cooling-off period as a more restrictive alternative to the proposed 30-day cooling-off period. Consumer groups and a broad spectrum of other commenters—including a group of U.S. Senators, several State Attorneys General, faith leaders, civil rights organizations, and other stakeholders generally supportive of the proposal—asked that the Bureau limit covered shortterm lending overall to 6 loans per year and 90 days per year. As discussed in the section-bysection analysis for § 1041.6, these stakeholders opposed the inclusion of the exemption for covered short-term loans, which contained these loan and time-in-debt limits as conditions of the proposed exemption. They argued that those same limits should apply to the making of all 814 covered short-term loans, in addition to the ability-to-repay requirements applicable to each loan and the various presumptions of unaffordability. The consumer groups asserted that these limits are “rooted in significant precedent” such as the FDIC’s 2005 guidelines on payday lending and State loan limits in Washington and Delaware. Consumer advocates also argued that that the Bureau should adopt a two-loan cap instead of a three-loan cap, because they believed that after two loans the ability-to-repay analysis already will have proven to be flawed. They argued that the rationale for imposing the threeloan limit in proposed § 1041.6(f) was equally applicable after two loans, i.e., it is extremely unlikely that a consumer attempting to borrow a third loan within a short period of time will be able to repay that loan given the prior re-borrowing. A number of commenters urged the Bureau to adopt additional restrictions under proposed § 1041.6. Several commenters raised concerns about the potential ability of consumers to take out multiple loans at a time, or to switch back-and-forth either between covered and noncovered loans or between short-term and longer-term loans, which could be ways of evading the proposed rule’s requirements. One commenter argued that the Bureau should consider any type of loan to be a non-covered bridge loan—rather than just non-recourse pawn loans of 90 days or fewer in duration—if it is used to bridge a gap between two sequences (or through a cooling-off period). Similarly, a number of other commenters argued that the Bureau should make the intrasequence presumptions stronger, arguing that lenders would likely still lend, and allow some amount of re-borrowing, unless there were stronger restrictions after the first and second loan. Consumer groups argued that tighter verification requirements should apply to loans being made that overcome the presumption of unaffordability. One State Attorney General expressed 815 concern about consumers taking out short-term and longer-term loans in quick succession as a way of evading the proposal and urged the Bureau to place greater restrictions on this type of lending pattern. Several commenters argued that the presumptions should apply in other scenarios, such as whenever a loan went delinquent, or when a consumer had repaid a loan made by an unaffiliated lender within 30 days. Others asked whether lenders can rely on consumer statements to determine whether a consumer had a prior loan with an unaffiliated lender. Consumer advocates also criticized the proposed exception to the presumption of unaffordability when the amount being borrowed was no more than half of the amount paid on the prior loan. They argued that this would incentivize lenders to make loans larger than consumers could initially afford at the outset and “then flip the clearly unaffordable portions, extracting excess costs each time.” Industry commenters, along with some other stakeholders, generally criticized the reborrowing restrictions in proposed § 1041.6. Many of them focused specifically on the proposed presumptions of unaffordability. Several industry commenters argued that the specific standards for overcoming presumptions provided too little flexibility or that they were vague and needed to be clarified. One trade group commenter argued that lenders essentially would have to become “financial planners” to determine whether a consumer had a “sufficient improvement in financial capacity”—the standard for overcoming the presumption—which the commenter viewed as untenable. Others asked for exceptions to the presumptions in various scenarios. Some commenters offered alternatives, such as off-ramps or exemptions for consumers who were taking out smaller or less expensive loans than they had previously. 816 A State trade association for lenders also criticized the exception to the presumption of unaffordability. The commenter argued that it would harm a more responsible consumer who borrowed a smaller amount initially but then developed a need for additional funds in excess of 50 percent of the initial loan amount. Several commenters argued that the Bureau should eliminate the presumptions against unaffordability and the cooling-off periods because consumers who have previously repaid are the most likely to repay in the future. One commenter, a specialty consumer reporting agency, discussed its analysis of data which it interpreted to show that a consumer who triggered the cooling-off period was more likely to repay than a consumer who had not, citing default rates. Similarly, commenters argued that consumers who pay off a loan have factually proven that they have an ability to repay, and thus there should be no limitation on future lending. Still other commenters argued that under the proposal consumers would be penalized twice for taking out a new loan while another loan remains outstanding, because the other loan would already be considered a major financial obligation. One lender commented that it was generally supportive of the proposed ability-to-repay requirements and viewed those requirements as sufficient, mitigating the need for additional re-borrowing restrictions. More broadly, many commenters argued that the cooling-off period should trigger after more loans have been made, or should be shorter, primarily arguing that the cooling-off periods as proposed would have a substantial impact on revenue, and would prevent consumers from obtaining credit when they need it. One commenter argued that the cooling-off period alone would reduce revenue by 71 to 76 percent. Others claimed that a cooling-off period would bar consumers from access to credit, and consumers cannot control when they might need it. A 817 small entity representative criticized the cooling-off period and the impacts it would have on this person’s small business. Several commenters argued that setting loan limits would cause consumers to over-borrow in order to tide themselves through the period when they would be restricted from borrowing. Commenters suggested a number of alternatives to the cooling-off period proposed, arguing that these alternatives would be less restrictive. Some commenters recommended that the Bureau create an off-ramp or repayment plan as an alternative to a cooling-off period, or alternatively, provide for exceptions where a consumer can prove that a new need has arisen. And some commenters asked the Bureau to take a more flexible approach when setting coolingoff periods, which would allow lenders to fluidly set their own thresholds based on outcomes, or give safe harbors while various industry participants try out different options. Some commenters called this a “sandbox” regulatory approach. A group of State Attorneys General opposed the proposed approach and asked the Bureau to allow the States to set their own restrictions, such as rollover caps, limits on the number of loans that may be taken out in a given timeframe, and cooling-off periods, to better reflect local conditions and allow for experimentation. They argued that States that impose rollover or annual limits, such as Washington and Missouri, should be allowed to continue that practice within a broader minimum Federal regulatory framework. The SBA Office of Advocacy encouraged the Bureau to reconsider the proposed coolingoff period and suggested that, if one were deemed necessary, it should be shortened from 30 days. The SBA Office of Advocacy noted that small entity representatives had criticized the cooling-off period based on its negative revenue impacts. It also passed along feedback from 818 small entities attending roundtables that some of their clients do not operate on a 30-day billing cycle, including some who pay their rent on a weekly basis; the 30-day cooling-off period would prevent these consumers from obtaining funds that may be needed for essential expenses. In its comment letter, the SBA Office of Advocacy acknowledged and expressed appreciation for the fact that the Bureau had shortened the period from the Small Business Review Panel Outline, which contemplated a 60-day period, but nonetheless argued that 30-days was too restrictive. An industry trade group criticized what it perceived as the proposal setting a blanket limit of six loans in a 12-month period for all covered short-term loans, not just exempt loans. The commenter argued the number was arbitrary and not backed by data. The commenter wrote that a more “appropriate limit that strikes the balance” between preventing consumers from relying too much on short-term loans and allowing the market for these loans to continue would be to limit covered short-term loans to eight loans during a 12-month consecutive period. The Bureau discusses substitutes and general considerations of access to credit in the Section 1022(b)(2) Analysis, as well as in the section-by-section analysis for § 1041.4. The Bureau received a significant number of comments from individual consumers who wrote as part of organized letter-writing campaigns. Among the more common themes in the letters was opposition to loan limits and cooling-off periods. Many individual consumers of such loans argued vehemently that these measures would intrude on consumer choice, would harm consumers who had no other credit options, and would cause consumers to turn to unsavory lending options. A number of them were concerned specifically about the burden and length of the 30-day cooling-off period, noting that it ignored the urgency of the need for immediate funds. Some were concerned that the re-borrowing limitations would result in loan denials and impede 819 their ability to access needed funds easily and quickly. These commenters specifically noted the need for funds for unexpected emergencies, like car repairs. Some simply declared these limits “unwarranted”. Many of these commenters believed the proposal to be setting firm annual limits on the making of all types of covered short-term loans. Lastly, the Bureau received some comments on the requirement to review borrowing history under proposed § 1041.6(a) by obtaining and reviewing information about a consumer’s borrowing history from a consumer report obtained from a registered information system. Consumer groups argued that a lender should have to check a State registry, if available, when a registered information system is unavailable. Others asked whether lenders would need to establish a backup registered information system in anticipation of potential periods in which the one the lender regularly uses may be unavailable. Final Rule After carefully considering the comments, the Bureau has decided to finalize only selected elements of §§ 1041.6 and 1014.10 in final § 1041.5(d).840 In particular, the Bureau has decided not to adopt the presumptions framework specified in the proposal, but rather rely primarily on the mandatory 30-day cooling-off period after the third loan in a sequence of covered short-term loans, covered longer-term balloon-payment loans, or a combination thereof. As specified below, the Bureau believes that this “circuit breaker,” when combined with the front-end ability-to-repay determination required under final § 1041.5(a) through (c), will protect consumers from long cycles of debt and strongly incentivize lenders to adopt more consumer- 840 As noted above, § 1041.6(d) is a related provision that restricts a lender and its affiliates from making loans within 30 days after a prior outstanding loan under § 1041.6 by the same lender or its affiliates. 820 friendly business models rather than relying on extensive consumer re-borrowing. At the same time, the Bureau believes that this shift will substantially simplify the final rule relative to the proposal, giving consumers more flexibility to manage their finances within short sequences and reducing burden on lenders. The Bureau is also adopting certain other parts of proposed §§ 1041.6 and 1014.10 concerning the basic obligation to review consumers’ borrowing history to determine whether a cooling-off period is triggered, and the restrictions on making covered short-term loans or covered longer-term balloon-payment loans under § 1041.5 within 30 days after an outstanding covered short-term loan under § 1041.6. The Bureau has made conforming changes to the commentary, as well as adding examples and other clarifications as discussed further below. Presumptions of unaffordability. The Bureau continues to believe the basic premise articulated in the proposal, as summarized above, that re-borrowing shortly after a previous covered short-term loan or covered longer-term balloon-payment loan can be important evidence that a consumer lacked the ability to repay the initial loan and that a consumer likely will not be able to afford a similar subsequent loan. When consumers have the ability to repay a covered short-term or covered longer-term balloon-payment loan, the loan should not cause consumers to have the need to re-borrow soon after repaying the balance, or when the prior loan is outstanding. Thus, the Bureau believes that the most likely explanation for a consumer returning to re-borrow shortly after paying off a previous covered short-term loan or covered longer-term balloonpayment loan is that the prior loan’s payment obligation left the consumer with insufficient income to make it through the balance of their expenses. 821 However, the Bureau also recognizes that there are occasional situations in which a consumer may experience an income or expense shock while a loan is already outstanding, and that the proposed presumptions framework did not provide a simple method of distinguishing such cases. In particular, the Bureau recognizes that defining the standard for overcoming the presumption would have either required extremely detailed inquiries of consumers, risked substantial evasion, or both. The Bureau agrees with the commenters who criticized the vagueness and workability of that standard contained in the proposal. As a result, the presumptions framework both would have imposed substantial compliance burdens on lenders and would have risked denying credit in some situations to consumers who had experienced an intervening borrowing need while a loan was already outstanding and would have been able to repay a second or third loan. Upon further consideration, the Bureau believes that the general ability-to-repay analysis under § 1041.5 in combination with a mandatory cooling-off period under § 1041.5(d)(2) provides a more appropriate way to balance the competing considerations with regard to reborrowing. The Bureau concludes that if a lender appropriately complies with § 1041.5(b) and (c) and makes a reasonable determination that the consumer will have the ability to repay the loan, the separate presumptions of unaffordability should be unnecessary to prevent re-borrowing in cases where the re-borrowing is attributable to the unaffordability of the prior loan. Of course, the presumptions were intended to be triggered in instances where it appeared that the lender was not making reasonable determinations of ability to repay. In the final rule, the Bureau has instead decided to rely on the reasonableness of ability-to-repay determinations. The determination of reasonableness will be based on whether a lender complies with the reasonable 822 determination and verification requirements in § 1041.5(b) and (c), including whether the outcome-related factors listed in comment 5(b)-2.iii indicate that the lender’s ability-to-repay determinations are reasonable as required in § 1041.5(b). Those factors include the frequency with which a lender makes multiple covered short-term or longer-term balloon-payment loans within a sequence. The Bureau believes that these requirements and measures will ensure that lenders shift their approach away from relying on extended loan sequences, and that lenders will appropriately factor in consumers’ prior borrowing history in making ability-to-pay determinations, especially with respect to loans that would constitute second or third loans in a sequence. If a lender fails to do so, the lender’s determinations would not be considered reasonable under § 1041.5(b). For the same reasons, the final rule does not include the presumptions framework of the proposal to address circumstances where there are indicia that consumers are struggling to repay a current loan—whether covered or non-covered or made by the same lender or its affiliate—as had been proposed in § 1041.6(d) and § 1041.10(c), respectively. Here, too, the Bureau believes that the combination of the ability-to-pay requirements coupled with a 30-day cooling-off period applied after the third covered short-term loan or covered longer-term balloon-payment loan in a sequence will be sufficient to prevent the unfair and abusive practice identified in § 1041.4. Cooling-off period. As noted above, a significant number of commenters objected to the cooling-off period, which the Bureau is finalizing largely as proposed for covered short-term loans and extending to covered longer-term balloon-payment loans in § 1041.5(d)(2). Thus, under the final rule, a lender cannot make a covered short-term loan or covered longer-term balloon-payment loan during the time period in which the consumer has one of those types of 823 loans outstanding or for 30 days thereafter if the new loan would be the fourth loan in a sequence of covered short-term loans, covered longer-term balloon-payment loans, or a combination thereof. Some commenters argued that consumers who have repaid a previous loan (or two or three loans in a sequence) and come back to borrow again within 30 days are consumers who are able to repay because they did not previously default, and thus, the Bureau should not impose cooling-off periods based on patterns of re-borrowing. But this ignores one of the central premises of §§ 1041.4 and 1041.5 of the final rule, which is that when a consumer avoids default by re-borrowing, it does not reflect that the consumer has the ability to repay the loan according to its terms. The industry’s current underwriting models do not account for re-borrowing risk because such re-borrowing helps to ensure that the lenders’ business model produces consistent revenue. But the very purpose of this rule is to ensure that lenders determine whether a consumer will be able to repay the loan and pay basic living expenses and major financial obligations without the need to re-borrow, thereby avoiding a significant harm identified above in Market Concerns—Underwriting and in the section-by-section analysis of § 1041.4. The Bureau’s decision to finalize the cooling-off period is also tied to its decision not to finalize the presumptions for the first or second loan in a sequence, as described above. The Bureau continues to believe that most consumers who return for a new loan within 30 days of paying off a previous loan had trouble meeting their obligations and needed to take out a new loan to cover the deficit left by paying off the old loan. For these consumers, such an “early return” suggests the consumer is beginning or continuing a cycle of re-borrowing, and the prior ability-to-repay determination was insufficient in some way. But there are other consumers who 824 did have an ability to repay, but who simply encountered an independent need for borrowing again within 30 days of paying off a prior loan, such as an unexpected car repair. The Bureau did not finalize the presumptions, in part, because the high bar for overcoming the presumptions would have prevented such consumers from obtaining additional credit that they can repay. But when a consumer returns to take out a fourth loan in a sequence, the Bureau concludes that is sufficient evidence to suggest that the consumer is not borrowing because of an independent need for funds, such as a non-recurring, unusual, or emergency expense. After all, at that point, the consumer would have had four such “new needs” during a relatively short period of time, each within 30 days of each other. Rather, it is much more likely that a cycle of re-borrowing has become manifest and the need for additional borrowing is due to the spillover effects of the prior borrowing. This conclusion is borne out in the Bureau’s data. The data show that consumers who take out more than three loans in a row are significantly more likely to be in a cycle of indebtedness that leads to 10 or more loans in a sequence than they are to repay that fourth loan and not re-borrow.841 Relatedly, the Bureau reiterates the data points noted in the proposal as support for this conclusion. The Bureau found that 66 percent of loan sequences that reach a fourth loan end up having at least seven loans, and 47 percent of loan sequences that reach a fourth loan end up having at least 10 loans.842 For consumers paid weekly, bi-weekly, or semimonthly, 12 percent of loan sequences that reach a fourth loan end up having at least 20 841 CFPB Report on Supplemental Findings, at Chapter 5. Specifically, approximately 22 percent of consumers repaid their first short-term loan without taking out another, and roughly 10 percent repaid the sequence with the second loan, but the percentage of consumers who repaid after the third, fourth, fifth, and sixth loans without reborrowing continued to drop, to approximately 5 percent and below, and more than 20 percent of consumers took longer than 10 loans to repay their loan sequence. 842 Results calculated using data described in Chapter 5 of the CFPB Report on Supplemental Findings. 825 loans during a 10-month period.843 And for loans taken out by consumers who are paid monthly, more than 40 percent of all loans to these consumers were in sequences that, once begun, persisted for the rest of the year for which data were available.844 The Bureau thus concludes that though it is not finalizing the presumptions, it is appropriate to finalize the cooling-off period after three loans in a sequence to prevent the unfair and abusive practice identified in in § 1041.4, and that doing so will still leave room for consumers who experience a new need to obtain credit via a second and even third loan in a sequence. Additionally, as the Bureau first stated in the proposal, if a lender’s ability-to-repay determinations resulted in re-borrowing three consecutive times in a sequence, the Bureau believes that is sufficient to suggest that either the lender’s ability-to-repay determinations are generally not reasonable, or the lender’s underwriting methodology does not work for the specific consumer’s circumstances. Of course, even well-underwritten credit includes some consumer defaults. But if a consumer returns to re-borrow three times in a sequence, that would likely suggest that the determinations are coming to erroneous results. Again, the Bureau believes that if a lender’s ability-to-repay determinations lead to the need to re-borrow three times in a row, it is unlikely that the fourth loan will produce a better outcome. The Bureau is finalizing a three-loan cap, instead of a different threshold such as a two-loan cap as suggested by certain consumer groups. As discussed above, a consumer’s taking three loans in a row is very strong evidence that the consumer did not have the ability to repay the prior loans and likely 843 844 Results calculated using data described in Chapter 5 of the CFPB Report on Supplemental Findings. CFPB Report on Supplemental Findings, at Chapter 1. 826 would not be able to repay another loan. It is not as apparent whether a consumer’s taking two loans in a row would provide such clear evidence. Furthermore, the Bureau notes that by including covered longer-term balloon-payment loans, it has also changed the additional limitations on lending for longer-term balloon-payment loans as compared to what was in proposed § 1041.10. Again, in proposed § 1041.10(b), the Bureau proposed a rebuttable presumption that a consumer would not have the ability to repay a longer-term loan (including a longer-term balloon-payment loan) if taken out while a covered short-term loan made under § 1041.5 or a longer-term balloon-payment loan made under § 1041.9 was outstanding and for 30-days thereafter. In the same way and for the same reasons that the Bureau is not finalizing the presumptions for covered short-term loans, the Bureau is not finalizing the presumptions for longer-term balloon-payment loans in proposed § 1041.10. However, after three longer-term balloon-payment loans in a sequence, or a combination of three covered short-term and longer-term balloon-payment loans in a sequence, there will now be a 30-day cooling-off period for all covered short-term and longer-term balloon-payment loans. Because the Bureau views covered short-term and longer-term balloon-payment loans as having similar risks, as noted above in the section-by-section analysis for § 1041.4, the Bureau’s analysis on why cooling-off periods are warranted for short-term loans made under § 1041.5 is applicable to longer-term balloon-payment loans made under § 1041.5. Three longer-term balloon-payment loans in a sequence, or a combination of three covered short-term or longerterm balloon-payment loans in a sequence, indicates both that the lender’s ability-to-repay determinations have not been reasonable, and that the consumer has begun a cycle of reborrowing. 827 Relatedly, the Bureau is not finalizing proposed comment 6(f)-1, which clarified that the cooling-off period did not limit a lender’s ability to make covered longer-term loans. That is still the case for most longer-term loans, because the cooling-off period only applies to loans made under §§ 1041.5 and 1041.6. However, as § 1041.5 now includes covered longer-term balloonpayment loans, the cooling-off period now prohibits that subset of longer-term loans. Again, as noted above, the Bureau is concerned that covered longer-term balloon-payment loans, where a large amount of funds are due at once and can potentially drive consumers to need to re-borrow, may be joined together, or joined with covered short-term loans to form a re-borrowing sequence. For this reason, the Bureau believes covered short-term loans and covered longerterm balloon-payment loans should be treated the same with regard to the cooling-off period. In crafting the preventive remedy to the unfair and abusive practice identified, the Bureau is attempting to maintain a significant amount of flexibility and not unduly restrain access to credit. And the Bureau recognizes that, as one commenter put it, “life happens.” There are likely to be a number of consumers who have an ability to repay when they take out the first loan, and who do repay the loan, but then encounter a new emergency expense or other independent borrowing need, and seek to take out a second loan to cover it (though as stated earlier, the Bureau continues to believe that most will in fact be re-borrowing even after the first loan due to the spillover effects of that loan). That this would happen again, two times in a row, is much less likely, but in the interest of maintaining access to credit and flexibility, the Bureau does not wish to categorically prevent such loans where there are likely to be at least some of these instances. There may even be a few instances where this would occur three times in a row, but the Bureau has made the judgment that at this point the likelihood that the consumer is 828 instead re-borrowing is overwhelmingly more likely. The Bureau believes that very few consumers who return for a fourth loan in row would have the ability to repay that loan. With regard to comments about the negative revenue impacts of the cooling-off period for lenders, the Bureau recognizes that this cooling-off period will reduce revenue for covered lenders. The Bureau has accounted for that revenue reduction in the costs, benefits, and impacts analysis below. As the Bureau has previously noted, the Bureau’s data suggest that many payday lenders rely on continuous re-borrowing for a substantial amount of their revenue. While a majority of consumers currently finish their payday loan sequences within the first three loans in a sequence, the majority of loans, and thus revenue, comes from loans made in sequences of 10 or more in a row.845 And as noted in the proposal, 21 percent of payday loans made to borrowers paid weekly, bi-weekly, or semi-monthly are in loan sequences of 20 loans or more. It is this very business model that is at the core of the unfair and abusive act or practice identified in § 1041.4, and thus, the Bureau cannot prevent the identified unfair and abusive practice without significantly impacting revenue made by lenders with this kind of business model. The Bureau is sensitive to the comments from many individual consumers who expressed concern and frustration over the proposed cooling-off period. The Bureau has carefully considered these comments, as well as related comments from consumers and other stakeholders about whether consumers affected by the cooling-off period will have available credit alternatives, and whether the rule will cause these consumers to seek out loans from more expensive or less reputable sources. And the Bureau recognizes that consumers who have obtained three covered short-term or longer-term balloon-payment loans in a sequence will be 845 CFPB Report on Supplemental Findings, at Chapter 1. 829 unable to obtain a fourth for 30 days, and that these consumers may be at risk of defaulting on their loans, or alternatively, defaulting on other expenses or obligations. However, the Bureau concludes that by requiring an ability-to-repay determination for each loan in a sequence, it is unlikely that many consumers will obtain a third loan in a sequence and not be able to repay that loan. Moreover, the cooling-off period will create an incentive that would not otherwise exist for lenders to offer no-cost payment plans to consumers who come to the end of a sequence and cannot afford to repay since otherwise the lender may face a default. In contrast, the Bureau believes that the risk of perpetuating cycles of unaffordable loans would be far greater without a cooling-off period. Further, the Bureau declines commenters’ suggestions to create an exception to the cooling-off period where a consumer can individually prove an independent borrowing need. As discussed in detail above in connection with § 1041.5(a)(5) and (b)(1), differentiating between re-borrowing that is prompted by a prior unaffordable loan and a new need can be complicated in practice, such that an exception would be very difficult to administer and would introduce significant risks of evasion. Where consumers are already three loans into a sequence, the Bureau believes for the reasons stated above that there is a substantial risk that they have become trapped in what would otherwise become a long-term cycle of debt. Further, such an approach would effectively turn the cooling-off period into a presumption, which the Bureau now disfavors for the reasons noted above. Some industry commenters believed that requiring lenders to offer an off-ramp option after a certain number of loans would be more advisable than a prohibition on new loans during a cooling-off period. As discussed in Market Concerns—Underwriting and in the introduction to 830 the section-by-section analysis for 1041.5, the Bureau is concerned, however, that if lenders remained free to continue loan sequences, they would find ways to do so and discourage consumers from using an off-ramp. Thus, the Bureau does not believe that an off ramp can substitute for a cooling-off period. The Bureau notes, however, that under the rule a lender may offer a no-cost off ramp after a consumer hits a cooling-off period and, indeed, may be required to do so under some State laws. These further protections are not prohibited by the rule, and the Bureau encourages lenders to find ways to work with their customers on repayment plans within the boundaries of the rule. Similarly, the Bureau does not agree with the comment by a group of State Attorneys General that the Bureau should allow the States to set their own re-borrowing restrictions to better reflect local conditions and that the Bureau should exempt from the requirements of this section any State that has extended repayment plans. As discussed in Market Concerns— Underwriting and in the introduction to the section-by-section analysis for 1041.5, the Bureau has considered various policy alternatives suggested by commenters as well as current State laws, both of which include extended repayment plans, but the Bureau has concluded that a Federal rule is necessary to protect consumers and that extended repayment plans imposed at the State level would not be adequate to prevent the unfair and abusive practice identified by the Bureau in this rulemaking, in part because evidence suggests low take rates for State mandated off-ramps or extended repayment plans. The Bureau does not believe that the suggestion by some commenters of a more flexible “sandbox” approach to the cooling-off periods, or safe harbors while industry participants experiment with different cooling-off periods, is warranted. The Bureau’s rulemaking process 831 has involved several years of analysis and experience and the Bureau does not believe that the potential benefits from a period of further experimentation warrant delaying the consumer protection that would be provided by this rule. The Bureau set the length of the cooling-off period for the reasons described herein and in the proposed rule. This final rule does, however, take a more flexible approach than the proposal in prescribing how lenders must make ability-torepay determinations, which the Bureau accomplished, in part, by not finalizing the proposed presumptions after each loan in a three-loan sequence as described above. Given that those presumptions are not being finalized, the Bureau believes that the remaining bright-line backstop of a strict cooling-off period is warranted. Length of Cooling-off Period. The Bureau concludes that, when a consumer has borrowed three covered short-term or longer-term balloon-payment loans in a sequence, the cooling-off period before the consumer can take out another such loan should be set at 30 days rather than some longer or shorter period of time. The Bureau believes that a 30-day cooling-off period strikes the appropriate balance and accordingly is finalizing that duration in § 1041.5(d). The Bureau’s rationale for doing so is largely the same as the reasons the Bureau chose a 30-day period to define the parameters of a loan sequence: namely, that major financial obligations generally are due on a monthly basis. During the SBREFA process, and in considering the comments on the proposal, including from the SBA Office of Advocacy, the Bureau heard examples of some consumers who paid for major financial obligation on a different cycle—like weekly rent. However, that does not change the fact that the traditional billing cycle in the United States is monthly. The Bureau has concluded that a consumer who returns to a lender to borrow again after paying a loan within a period consisting of a 30-day billing cycle is 832 very likely to have shifted money around to pay the loan instead of expenses. Again, the Bureau’s test for whether a consumer has the ability to repay is whether the consumer has the ability to repay the loan as well as major financial obligations and still meet basic living expenses. By contrast, if a consumer makes it through an entire billing cycle without needing to re-borrow, then it is more likely that she reached equilibrium and if the consumer then returns to borrow that may well reflect a new and independent borrowing need. As noted in the proposal, there is always some chance that a consumer will have a new need for a new loan within any reborrowing period, no matter what time period it is based on. There also is some chance that the spillover effects of repaying an unaffordable loan will be felt for a prolonged period of time after the payment. Nonetheless, the Bureau has concluded that a 30-day re-borrowing period is the appropriate threshold for the definition of a sequence—accounting for one billing cycle, but not extending so far as to capture a significant number of genuine new credit needs. Similarly, the Bureau believes that a 30-day cooling-off period is the appropriate length of time to ensure that a consumer who has just re-borrowed twice in a row is sufficiently free from the spillover effects of those unaffordable loans before she borrows additional covered short-term or longer-balloonpayment loans. The Bureau is also aligning the length of the cooling-off periods with the length of the reborrowing period for purposes of greater simplicity and practicality. Extending the cooling-off period to 60 or 90 days, as some commenters recommended, would reduce access to credit to a significant extent. The Bureau does not judge that approach to be warranted at this time. The Bureau notes that it has considered whether to impose a cooling-off period of a different length than the re-borrowing period, and also has considered whether to impose a graduated cooling-off 833 period, an alternative on which the Bureau sought comment (e.g., 30 days after the first full loan sequence, 60 days after the second, 90 days after the third). The Bureau has judged these alternatives to be too complex to administer. The Bureau again believes that the logic for setting the re-borrowing period at 30 days is applicable here as well, and that in addition setting the cooling-off period and re-borrowing period at the same length is the simplest and most intuitive approach. Treatment of Covered Longer-Term Balloon-Payment Loans. As noted above, the Bureau proposed to subject covered longer-term balloon-payment loans to the same presumptions that would have applied to covered short-term loans in situations in which the consumer’s re-borrowing or struggles to repay a current loan suggested that they may not have the ability to repay a new loan. The Bureau did not specifically propose to impose a 30-day cooling-off period after the third longer-term balloon-payment loan in a sequence, but did seek comment on whether particular patterns of re-borrowing within a particular timeframe warranted additional protections. Consumer groups responded with proposals to strengthen the presumptions for longer-term loans, or add to the number of facts that would trigger a presumption. After additional consideration, the Bureau has concluded that covered longer-term balloon-payment loans should be treated in the same manner as covered short-term loans where there is a sequence of three loans (i.e., where the loans are each taken out within 30 days of each other). In such circumstances, three prior ability-to-repay determinations will have proven inconsistent with the consumer’s actual experience. For consumers who reach that point, the Bureau believes that terminating a loan sequence may assist the consumer to escape from the 834 cycle of indebtedness. Particularly for loans with terms that slightly exceed the limits for a covered short-term loan and that have very large end payments—such as a 46-day lump-sum loan structure—the Bureau believes that the risks of consumers becoming stuck in a long cycle of borrowing absent a mandatory cooling-off period would be similar to those for covered shortterm loans. Borrowing history. As in the proposal, a lender will need to obtain a report from a registered information system to assess whether a consumer has or had loans from other lenders that would make a new loan violate either § 1041.5(d)(2) or (3). The Bureau received comments about what happens (or should happen) if no registered information system is available. Section 1041.5(d)(1) requires that a lender obtain a consumer report from a registered information system only if such a report is available. If no report is available, either because no entity has been registered as an information system for 180 days or more or because no registered information system is capable of producing a report at the time the lender is contemplating making a covered loan (for example, due to temporary system outage), a lender does not violate § 1041.5 if it makes a covered loan without obtaining a consumer report from a registered information system.846 Regarding the comment from consumer groups that the rule should provide for mandatory checking of State databases when no report from a registered information system is available, the Bureau declines to impose this requirement because it does not believe it would be 846 This is in contrast to loans under § 1041.6, which are not permitted if a consumer report from a registered information system is unavailable. 835 useful for compliance with this part.847 The Bureau also does not believe such a requirement is necessary; State laws already require such activity, and this rule would not preempt any such requirements. With regard to comments asking whether lenders must obtain a consumer report from another registered information system in the event the registered information system from which the lender regularly obtains reports is unavailable for some reason (e.g., a temporary system outage), the Bureau believes that it is reasonable and appropriate to impose such a requirement given the importance of the information contained in a registered information system report in assessing whether the lending limitations contained in § 1041.5(d) are triggered. The Bureau notes that lenders are required to furnish information to every registered information system and thus a lender should not experience difficulty in maintaining a backup purchasing relationship with a registered information system other than the one from which the lender regularly obtains reports. Annual loan limits. The Bureau addresses the comments it received regarding annual loan limits. At the outset, the Bureau finds it necessary to address a common misperception in the comments, including those submitted by many individual commenters and a trade group commenter described above. Some commenters perceived that the restrictions in proposed § 1041.7 (now § 1041.6 of the final rule) on the number of exempt covered short-term loans and the time of indebtedness on such loans within a 12-month period applied to all covered short847 The Bureau does not believe that such State databases provide information that lenders would need to comply with this part. For example, the Bureau understands that most if not all of such databases issue an eligibility determination under State law to lenders contemplating making loans, rather than information about outstanding and prior loans that lenders will need to comply with this part. Such databases typically simply indicate whether the contemplated loan may or may not be made under State law. Further, certain information required for compliance with this part is specific to this part and likely will not be required to be reported to State databases by lenders under State law. For example, § 1041.10(c)(1)(iii) requires lenders to furnish whether the loan is a covered short-term loan or a covered longer-term balloon-payment loan as those terms are defined in this part. 836 term loans. However, under the proposal, if consumers took out the maximum number of covered short-term loans under proposed § 1041.7 in a 12-month period and therefore could no longer obtain an exempt covered short-term loan under that provision, the proposal still would have permitted them to obtain a covered short-term loan within the 12-month period as long as they met the ability-to-repay requirements under proposed § 1041.5. The final rule contains a similar framework. Section 1041.6 permits a consumer to obtain loans under that provision so long as the consumer has not taken out six covered short-term loans or become indebted on covered short-term loans for 90 days within a 12-month period. After reaching either of those caps, a consumer could continue obtaining loans under § 1041.5, subject to the requirements of § 1041.5, including the ability-to-repay determination and the cooling-off period that applies after three loans in a sequence. The Bureau received many comments from stakeholders who were supportive of the proposal in general, including consumer advocates, elected officials, and others, but who urged the Bureau to impose a cap on covered short-term lending of six loans or 90 days of indebtedness in a 12-month period. The Bureau declines to impose such a limit. The Bureau has imposed such a cap on loans made under § 1041.6 because such loans can be made without assessing the consumer’s ability to repay. As explained in the discussion of that section, the Bureau is concerned about the risks of making such loans to consumers who have demonstrated a pattern of extensive borrowing. However, that same logic does not extend to § 1041.5 since loans made under that section do require an ability-to-repay determination. The Bureau is concerned that blanket caps limiting all consumers to no more than six covered short-term loans in a 12-month period and to 90 days of indebtedness within a 12-month 837 period would unduly restrict access to credit. A consumer may have several unusual and nonrecurring borrowing needs over the course of a 12-month period, with several months in between any loan sequence. A cap of this sort would deny access to credit to such consumers later in the year, regardless of their particular circumstances, even if they have the ability to repay. This restriction also would mean that a consumer who takes the maximum number of permitted exempt covered short-term loans under § 1041.6 could not take out another covered short-term loan during the 12-month period—even one for which they have the ability to repay. The Bureau is also mindful of the high number of individual consumers who commented on the concerns they had about potential restrictions on access to credit. The provisions in § 1041.5 of the final rule requiring ability-to-repay underwriting according to specific criteria directly address the risks and harms created by the identified unfair and abusive practice. That practice of making loans without reasonably determining the borrower’s ability to repay the loan according to its terms enables lenders to make unaffordable loans that mire many consumers in extended loan sequences through repeat re-borrowing—or else leads them to experience default, delinquency, or the collateral consequences of forgoing basic living expenses or major financial obligations to avoid defaulting on their unaffordable loans. Without moving to the stricter specification of an overall loan cap, the Bureau believes that the measures in § 1041.5 are sufficiently calibrated to prevent consumers from experiencing the risks and harms associated with the unfair and abusive practice. Furthermore, the Bureau has eliminated the specific regulatory requirements around noncovered bridge loans—in proposed §§ 1041.6(h) and 1041.10(f)—because it has determined that these requirements would be too complex to implement. At the same time, the Bureau 838 recognizes, as noted by consumer groups, that any kind of non-covered loan could be used as a means to bridge over a re-borrowing period or cooling-off period. Thus, the Bureau is addressing the concerns animating these proposed provisions by adding an example in comment 5(b)-2.iv.E, noting that frequent instances of using any kind of non-covered loans to bridge between loan sequences could indicate that the ability-to-repay determinations are not reasonable. In § 1041.5(d)(3), the Bureau has finalized the prohibition against making covered shortterm loans or longer-term balloon payment loans under § 1041.5 within 30 days of a loan made under § 1041.6 (as was proposed in proposed §§ 1041.6(g) and, to a certain extent, 1041.10(e)). These provisions were designed to ensure that protections in proposed § 1041.7 requiring a stepdown of the amount of principal over three loans in a sequence worked as intended, and is otherwise based on the same rationale as was in the proposal.848 5(e) Prohibition on evasion of § 1041.5 The Bureau is also adding a new § 1041.5(e), which states that a lender must not take any action with the intent of evading the requirements of § 1041.5 of the final rule. The Bureau had proposed a general anti-evasion provision in proposed § 1041.19, and is finalizing that more generalized anti-evasion provision at § 1041.13 of the final rule. Nonetheless, the Bureau has decided to add this more specific paragraph to § 1041.5 so that it can provide guidance on antievasion within the specific context of that section. Comment 5(e)-1 clarifies that the standard for what constitutes evasion is the same as that in the broader provision, § 1041.13 of the final rule, 848 As noted above, § 1041.6(d), which is also based on proposed § 1041.10(e), places a related limitation on lenders and their affiliates making loans within 30 days of a prior outstanding loan under § 1041.6 by the same lender or its affiliates. 839 which is applicable to part 1041 in its entirety. The Bureau addresses comments about that more general standard below in the section-by-section analysis of § 1041.13. For illustrative purposes, the Bureau provided one example at comment 5(e)-2, which is a particular fact pattern that may be considered an evasion of § 1041.5.849 Modified in response to comments received, the substance of the example in comment 5(e)-2 is based on the illustrative example that had been presented in proposed comment 19-2.ii. For ease of reference, it has been moved here. Consumer groups requested that the Bureau alter the example to clarify that late fees are considered rollovers or re-borrowing, and that the example was not viewed as exhaustive, meaning other scenarios could lack elements from this fact pattern and still constitute possible evasions. The Bureau does not believe these clarifications are necessary. The example is not exhaustive. All late fees would not be considered rollovers or re-borrowing, but as noted in the example, when combined with other features, may prove intent to evade the rule. The final comment 5(e)-2 consists, among other things, of a covered short-term or longer-term balloon-payment loan structure that requires a consumer to accrue a late fee for every two weeks of non-payment, in an amount that meets or exceeds the normal finance charge. The comment further explains that depending on the relevant facts and circumstances, including the lender’s prior practices, the lender may have taken these actions with the intent of evading its obligations in § 1041.5(b) (underwriting) and § 1041.5(d) (cooling-off period, if the late fees accrue beyond the time when the cooling-off period would begin if the late fees instead were new loans) and as a result the lender may have violated § 1041.5(e). The explanation of how the conduct may 849 Note that this example is similar to a real-life fact pattern. See Press Release, S.D., Dep’t of Labor and Regulation, “Statement from Division of Banking on Dollar Loan Center,” (Sept. 13, 2017), http://dlr.sd.gov/news/releases17/nr091317_dollar_loan_center.pdf. 840 violate § 1041.5(e) was not contained in the proposed comment, but was added to provide more clarity on specific actions that may indicate an intent to violate the provision and thereby support a possible violation of § 1041.5(e) of the final rule. Section 1041.6 Conditional Exemption for Certain Covered Short-Term Loans Proposed § 1041.7 would have exempted covered short-term loans that satisfy certain conditions from proposed §§ 1041.4, 1041.5, and 1041.6. The Bureau is finalizing the proposed conditional exemption for certain covered short-term loans, largely as proposed, but with several substantive adjustments and renumbered as § 1041.6 in light of other changes to the rule. This section first describes the Bureau’s general approach to the exemption in the proposed rule, the Bureau’s legal authority for the exemption, some comments received on the general approach to the exemption, and a high-level summary of the final rule. Then the Bureau will discuss each portion of § 1041.6, the comments received, and the final rule in turn. General Approach in the Proposed Rule The Bureau proposed to exempt covered short-term loans under proposed § 1041.7 from proposed §§ 1041.4, 1041.5, and 1041.6. Because loans made under proposed § 1041.7 would not have been subject to the underwriting criteria in proposed § 1041.5 and the additional borrowing limitations in proposed § 1041.6, proposed § 1041.7 would have included a number of screening and structural provisions to protect consumers in place of those other requirements. The Bureau believed that these protections would reduce the likelihood and magnitude of the kinds of risks and harms to consumers from unaffordable payments on covered short-term loans that were discussed in the section in the proposal on Market Concerns—Short-Term Loans, 841 including the harms that result to consumers from extensive re-borrowing in long sequences of short-term loans.850 In the proposal, the Bureau recognized, based on its own research and that of others, that even where lenders do not engage in any meaningful underwriting, some consumers are in fact able to repay a short-term loan when it comes due without further re-borrowing. These consumers thus avoid at least some, if not all, of the risks and harms with which the Bureau is concerned. For example, as described in the CFPB Report on Supplemental Findings, approximately 22 percent of new payday loan sequences do not result in any re-borrowing within the ensuing 30 days.851 While the Bureau believed that most of these consumers would be able to demonstrate their ability to repay and thus could continue to obtain loans under the proposal, the Bureau recognized there may be a subgroup of consumers for whom this is not true and who would be denied loans even though they could, in fact, afford to repay them. The proposal noted that some of these consumers may take out a payday or title loan, repay it on the contractual due date, and never again use such a loan. Others may return on another occasion, when a new need arises, likely for another single loan or a short sequence.852 850 The Bureau’s legal authority to grant conditional exemptions from its rules in certain circumstances is discussed below, as is its authority to prescribe rules for accurate and effective disclosures as well as the use of model forms. 851 See CFPB Report on Supplemental Findings, Chapter 5. The Bureau’s finding may overstate the extent to which payday borrowers are able to avoid re-borrowing, since the Bureau’s study looked at borrowing from a single lender. A study that tracks borrowers across five large lenders, who together make up 20 percent of the storefront payday market, found that 21 percent of borrowers switch lenders and of those borrowers roughly two-thirds did so within 14 days of paying off a prior loan. See Clarity Services, “Finding the Silver Lining in Regulatory Storm Clouds: Consumer Behavior and Borrowing Capacity in the New Payday Market,” at 4, 9 (2015) (hereinafter “Finding the Silver Lining in Regulatory Storm Clouds”), available at https://www.nonprime101.com/wpcontent/uploads/2015/10/FISCA-10-15.pdf. 852 The study described in the previous footnote, using data over a four-year period, found that 16 percent of borrowers took out one payday loan, repaid it on the contractual due date, and did not return again during the period reviewed; that the median such borrower had 2 sequences over four years; and that the average such borrower had 842 Further, even among those who do re-borrow, the Bureau’s research indicated that about 16 percent of payday sequences ended with final repayment within three loans, without either defaulting or re-borrowing within 30 days after the last payment has been made.853 In addition, the proposal noted that the Bureau’s research suggested that even consumers who re-borrow many times might have shorter loan sequences if they were offered the option of taking out smaller loans each time they returned to re-borrow—instead of being presented only with the binary option of either rolling over the loan without paying down any principal (in States where rollovers are permitted) or repaying the full amount of the loan plus the finance charge, which often leads the borrower to take out another loan in the same amount.854 Finally, the Bureau recognized that the verification and other underwriting criteria in proposed §§ 1041.5 and 1041.6 would have imposed compliance costs that some lenders, especially smaller lenders, may have found difficult to absorb for covered short-term loans, particularly for those loans that are relatively small in amount. In light of these considerations, the Bureau believed that it would further the purposes and objectives of the Dodd-Frank Act to provide a simpler alternative to the specific underwriting criteria in proposed §§ 1041.5 and 1041.6 for covered short-term loans, but with robust alternative protections against the harms that consumers experience from loans with unaffordable payments. Proposed § 1041.7 would have permitted lenders to extend to consumers a sequence of up to three loans, in which the principal is reduced by one-third at each 3.37 sequences. (This study defined sequence, as did the Bureau’s 2014 Data Point, by using a 14-day period.). See Finding the Silver Lining in Regulatory Storm Clouds, at 8, 14. 853 CFPB Report on Supplemental Findings, Chapter 6. 854 CFPB Report on Supplemental Findings, Chapter 6. 843 stage and certain other conditions are met, without following the underwriting criteria specified in proposed § 1041.5 and without satisfying the limitations of proposed § 1041.6. The Bureau’s approach to a conditional exemption for covered short-term loans garnered discussion from stakeholders even before the proposal was issued. During the SBREFA process and the Bureau’s outreach following its release of the Small Business Review Panel Report, many lenders and other industry stakeholders argued that the alternative requirements for covered short-term loans presented in the Report would not provide sufficient flexibility to sustain a lender’s profitability in making covered short-term loans.855 In contrast, during the Bureau’s outreach before and after the release of the Report, many consumer advocates argued that permitting covered short-term loans to be made without meeting specified underwriting criteria would weaken the overall framework of an ability-to-repay rule, and urged the Bureau not to adopt any alternatives that would sanction a series of repeat loans.856 The Bureau carefully considered this feedback in developing the proposed rule and in particular in developing proposed § 1041.7. With regard to the industry argument that the approach described in the Report would not allow lenders to remain profitable, the Bureau believed that reflected the heavy reliance of many lenders on revenue from borrowers who experience long sequences of covered short-term loans. Since the Bureau began studying the 855 See Small Business Review Panel Report, at 22. During and after the SBREFA process, the Bureau was considering two options, one of which would have allowed three-loan sequences with a subsequent off-ramp stage for consumers who had not been able to repay the principal, and one that would have required principal step-downs similar to the approach the Bureau ended up proposing. SERs and other industry stakeholders criticized both approaches because they would have limited lending to three-loan sequences and imposed limits on how many alternative loans could be taken out per year. 856 Letter from Americans for Financial Reform to the Hon. Richard Cordray, Director, Bureau of Consumer Fin. Prot., (Oct. 23, 2014) (regarding proposed payday loan rules), available at http://www.nclc.org/images/pdf/high_cost_small_loans/payday_loans/payday_letter_director_cordray_cfpb_10231 4.pdf. 844 market for payday, vehicle title, and similar loans several years ago, it has noted its significant concern with the amount of long-term re-borrowing observed in the market, and the apparent dependence of many lenders on such re-borrowing for a significant portion of their revenues.857 The Bureau was sensitive to the impact that the proposed rule would have had on small entities, but to the extent they are relying on repeated re-borrowing and long loan sequences for a substantial portion of their revenues, the Bureau had the same concerns here about significant harm to consumers that it found to exist more generally with this market. Proposed § 1041.7 would have permitted consumers with emergencies or occasional shortfalls to receive a limited number of covered short-term loans without having to meet the underwriting criteria in proposed §§ 1041.5 and 1041.6, but would have addressed the risks and harms to consumers from such loans by providing them with an alternative set of protective requirements. The Bureau acknowledged in the proposal that a substantial number of loans currently being made in the marketplace would not qualify for the exemption under proposed § 1041.7 because they are part of extended cycles of re-borrowing that are very harmful to many consumers. The Bureau noted that some lenders may be able to capture scale economies and build a business model that relies solely on making loans under proposed § 1041.7, with their approach to underwriting such loans likely having to be adjusted to take account of substantial declines in re-borrowing revenue. For other lenders, the Bureau expected that loans made under proposed § 1041.7 would become one element of a business model that would also incorporate 857 See Market Concerns—Underwriting. See also Richard Cordray, Director, Bureau of Consumer Fin. Prot., “Prepared Remarks of CFPB Director Richard Cordray at the Field Hearing on Payday Lending,” (Mar. 26, 2015), Richmond, Virginia), available at http://www.consumerfinance.gov/newsroom/prepared-remarks-of-cfpb-directorrichard-cordray-at-the-field-hearing-on-payday-lending/. 845 covered short-term and longer-term loans, loans that are not covered by this rule, and perhaps other financial products and services as well. As for the consumer advocates that disfavored any alternatives to requiring lenders to meet specified underwriting criteria for covered short-term loans, the Bureau issued its proposal because it did not believe that providing a carefully constructed alternative to the specific underwriting criteria proposed in §§ 1041.5 and 1041.6 would significantly undermine consumer protections. The Bureau noted that the proposed exemption would provide a simpler means of obtaining a covered short-term loan for consumers where the loan is likely to prove less harmful. That was so, the Bureau noted, because proposed § 1041.7 included a number of safeguards, including the principal step-down requirements and the fixed limit on the number of loans in a sequence of such loans, to ensure that consumers cannot become trapped in long-term debt on an ostensibly short-term loan. The Bureau believed that those safeguards also would reduce the risk of harms from default, delinquency, re-borrowing, and the collateral consequences of making unaffordable loan payments while forgoing basic living expenses or major financial obligations during a short sequence of these loans. The proposal reflected the Bureau’s view that the requirements in proposed § 1041.7 would appropriately balance the goal of providing strong consumer protections with the goal of permitting access to less risky credit on less prescriptive terms. The Bureau noted that by including an alternative set of requirements under proposed § 1041.7, the Bureau was not suggesting that regulation of covered short-term loans at the State, local, or Tribal level should encompass only the provisions of proposed § 1041.7. On the contrary, proposed § 1041.7(a) would not have provided an exemption from any other provision 846 of law. The Bureau noted that many States and other non-Federal jurisdictions have made and likely will continue to make legislative and regulatory judgments about how to treat such loans, including usury limits, prohibitions on making high-cost covered short-term loans, and other strong consumer protections under legal authorities that in some cases extend beyond those conferred on the Bureau. The proposed regulation would have coexisted with—rather than supplanted—State, local, and Tribal regulations that impose a stronger framework that is more protective of consumers, as discussed in part IV. In the same vein, the Bureau noted that proposed § 1041.7 also would not have permitted loans to servicemembers and their dependents that would violate the Military Lending Act and its implementing regulations. The Bureau requested comment generally on whether to provide an alternative to the requirement that lenders meet the specific underwriting criteria in proposed §§ 1041.5 and 1041.6 for covered short-term loans that satisfy certain requirements. The Bureau also sought comment on whether proposed § 1041.7 would appropriately balance the considerations regarding consumer protection and access to credit that presents a lower risk of harm to consumers. The Bureau sought further comment on whether covered short-term loans could be made in compliance with proposed § 1041.7 in States and other jurisdictions that permit covered short-term loans. In addition, the Bureau sought comment on the appropriateness of each of the proposed requirements in proposed § 1041.7, and more generally on the costs and other burdens that would be imposed on lenders, including small entities, by proposed § 1041.7. General Comments Received 847 The Bureau here is addressing the general comments that it received on the conditional exemption in proposed § 1041.7, and discusses the comments pertaining to its more specific components when addressing them below. A significant number of industry members and trade associations opposed the Bureau’s proposed conditional exemption. Several argued that the conditions in the proposed exemption are too restrictive and would severely reduce revenue, profits, and access to credit. A number of State Attorneys General similarly argued that the exemption in proposed § 1041.7 was not workable and would generate too little revenue to allow lenders to remain in business. Some industry commenters argued that the Bureau had not adequately justified the conditions of the proposed exemption, arguing that there was no data supporting the structural limitations of the exemption. One commenter, in connection with its argument that the Bureau had not shown that payday loans cause consumer harm, contended that the Bureau has provided no justification for providing the exemption in proposed § 1041.7. Several industry commenters opposed § 1041.7 as proposed because, they argued, the conditionally exempt loans would fail to meet the needs of borrowers, especially those who needed a loan for an emergency expense.858 Commenters argued that the requirements of proposed § 1041.7 would reduce the speed and convenience of the product, diminishing its value and therefore harming borrowers who are currently able to repay. Some commenters argued that 858 Hereinafter these loans made pursuant to § 1041.7 of the proposed rule or § 1041.6 of the final rule will be referred to as “conditionally exempt loans.” 848 the Bureau had underestimated how much its proposed approach would reduce lending volumes and thus the availability of credit, citing either their own studies or the studies of others.859 One industry commenter argued that the disclosures that would have been required by proposed § 1041.7(e) for loans made under § 1041.7 demonstrate that disclosures can be effective and maintained that the rule as a whole should focus on disclosures rather than on imposing more restrictive provisions such as ability-to-repay requirements. Another industry commenter argued that instead of offering an exemption under proposed § 1041.7, the rule should consider setting limits on the number of consecutive transactions a consumer may obtain under proposed § 1041.5 or requiring an “off-ramp” after a certain period of indebtedness. Some commenters argued that the exemption in proposed § 1041.7 was not broad enough and that it should exempt lenders from other requirements. For example, several commenters affiliated with banks or credit unions urged the Bureau to expand the exemption. Commenters asserted that even conditionally exempt loans would require banks or credit unions to comply with other portions of the rule, and this compliance would impose significant costs, causing them to leave the market. Some State officials took a different tack, urging the Bureau to further limit the extent of the exemption in proposed § 1041.7 and arguing that if the exemption existed at all, it should be limited to loans with APRs below 25 percent because loans with higher interest rates risk being unaffordable to consumers. Another commenter urged the Bureau to require lenders to refund finance charges if the borrower paid back a loan early. The commenter asserted that requiring a partial refund of fees when a borrower paid back a loan sequence early would encourage 859 Comments assessing the Bureau’s estimates of the impact of proposed § 1041.7 are discussed below in part VII. 849 borrowers to make earlier payments and would reduce the amount of money that borrowers ultimately paid over the course of the loan sequence. Consumer groups and many individual commenters urged the Bureau to eliminate the conditional exemption in proposed § 1041.7. They argued that ability-to-repay determinations are necessary to prevent the identified unfair or abusive practice, and thus there should be no exemptions from those portions of the rule. A coalition of consumer groups argued that the exemption would not prevent substantial payments from coming due in a short amount of time, which would not be affordable to borrowers. Another commenter argued that lenders making covered short-term loans will exploit any loophole, and thus lenders would exploit the exemption. Some commenters also argued that the exemption would allow for unaffordable loans and that unaffordable loans cause substantial harm. Others pointed to data suggesting that conditionally exempt loans would be unaffordable for borrowers. They argued that even small payments are often unaffordable and that even one unaffordable loan can cause substantial harm. Because the exemption would allow loans to be made without meeting specific underwriting criteria, they argued that it would increase the incidence of these harms. Consumer groups also urged the Bureau not to adopt the exemption in proposed § 1041.7 because they viewed it as inconsistent with the rest of the rule. They said the Bureau had persuasively demonstrated in proposed § 1041.4 that loans made without an ability-to-repay determination cause substantial harm. Because the exemption would allow loans that did not meet that standard, they argued that it was inconsistent with the rest of the rule. These commenters also suggested that the proposal’s reasoning about why conditionally exempt loans under proposed § 1041.7 should not be permitted to include a security interest in an auto title 850 applies to payday loans as well. And they stated that they were unaware of any precedent from other regulators for adopting a similar exemption. A non-profit group argued that the exemption was likely to be ineffective because lenders would make more money on longer-term loans and therefore would not offer conditionally exempt loans under proposed § 1041.7. It also argued that the exemption would not allow lowercost lenders to make loans. Several State Attorneys General argued that the rule should not include any exemption from the ability-to-repay requirements, though one stated that if the Bureau were to retain an exemption, it should be structured as in proposed § 1041.7. One attorney general urged the Bureau to monitor the effectiveness of the exemption periodically in order to ensure that it did not permit lenders to continue to make unaffordable loans on a regular basis. Some consumer groups criticized proposed § 1041.7 because it would not have required lenders to verify income for conditionally exempt loans, which they argue is necessary for all loans. Others also urged the Bureau not to adopt the proposed exemption because it could risk undermining State laws that restrict payday lending if lenders were to cite the exemption as evidence that payday loans are deemed to be safe. Both consumer group and industry commenters asked the Bureau to clarify how the requirements of the proposed rule would interact with existing State law. One commenter noted that some cities allow loans to roll over three times—for a total of four loans—while the proposed rule would only allow two rollovers. This commenter also urged the Bureau to promulgate a definition to clarify when the provisions of the rule would provide “greater consumer protection” than other measures, especially State laws for purposes of preemption 851 under the Dodd-Frank Act. Industry commenters similarly expressed concerns about interactions with State law, asserting that many States mandate extended payment plans, and arguing that the Bureau does not have the authority to displace those State laws. Final Rule The Bureau is finalizing proposed § 1041.7 as § 1041.6 of the final rule to provide for conditionally exempt loans, with several technical changes to accommodate other changes in the rule, and with one more substantive change that is summarized below and explained in more detail in the section-by-section analysis of § 1041.6(d). Proposed § 1041.7(d) would have required that, for the purpose of calculating the period for determining whether loans made under proposed § 1041.7 would be part of the same loan sequence, a lender or its affiliate must not count the time when it had a non-covered bridge loan (as defined in proposed § 1041.2(a)(13)) outstanding with the consumer. As discussed in more detail in the section-by-section analysis of § 1041.6(d), in the final rule, the Bureau has replaced the “tolling” provision in proposed § 1041.7(d) relating to non-covered bridge loans with § 1041.6(d), which prohibits a lender or its affiliate from making any covered or non-covered loans (other than a loan under § 1041.6) within 30 days of a loan made under § 1041.6 of the final rule. The Bureau is finalizing the exemption substantially as proposed based on the grounds set forth in the proposal and discussed above. As described and explained further in § 1041.6(c)(3) and (d) below, the exemption has been carefully designed to minimize the risk of borrowers becoming trapped in cycles of re-borrowing. In § 1041.4 of the final rule, the Bureau has identified the substantial risks and harms to consumers associated with lending without making reasonable determinations that borrowers have the ability to repay—default, 852 delinquency, re-borrowing, and other harms associated with avoiding default. Because loans made under § 1041.6 would not be required to meet the specific underwriting criteria in § 1041.5, the specific features of this conditional exemption are designed to mitigate those harms. Certain requirements for loans made under § 1041.6 (and described in more detail below), including the 3-loan cap, the cooling-off period, and the specific limitation on indebtedness in a 12-month period, are all intended to prevent extended re-borrowing. Other requirements for loans made under § 1041.6, including the principal-reduction requirements, the prohibition on security interests in vehicle titles, and the limits on loan amounts, are intended to prevent re-borrowing, and prevent or reduce the risks and harms associated with default, delinquency, and forgoing basic living expenses or major financial obligations to avoid default. The Bureau also has concluded that, compared to specific alternatives suggested by certain commenters, the exemption in § 1041.6 is likely to be more effective at balancing the need for consumer protections with preservation of access to credit. As noted above, an industry commenter argued that instead of offering an exemption under proposed § 1041.7, the rule should consider setting limits on the number of consecutive transactions a consumer may obtain under proposed § 1041.5 or requiring an “off-ramp” after a certain period of indebtedness. The Bureau agrees that prescribing certain limits on sequential borrowing would help limit the harms that result from repeated re-borrowing and has prescribed certain limits in § 1041.6(c)(2) for conditionally exempt loans made under § 1041.6, as well as in § 1041.5(d) for loans made under the ability-to-repay requirements in § 1041.5. However, as discussed in the section-by-section analysis for §§ 1041.5 and 1041.6, the Bureau has concluded that additional protections are necessary to protect consumers against the risks and harm from unaffordable loans. 853 The Bureau is not persuaded by the commenter’s argument that because the disclosures proposed for these conditionally exempt loans under § 1041.6 can be effective; it follows that the entire substance of this rule can therefore be replaced with a disclosure-only rule. The Bureau recognizes that disclosures like those finalized in § 1041.6(e) can be valuable and effective in educating consumers on how their choices may be affected by the restrictions prescribed in the final rule. Yet the Bureau does not believe that prescribing disclosures to explain the provisions of § 1041.6 is inconsistent with the conclusion that disclosures alone do not suffice to protect consumers against the harms targeted in this rulemaking. As discussed above in the section-bysection analysis for § 1041.5, the Bureau has concluded that disclosures alone are not enough to protect consumers against the risks and harms of unaffordable loans. With respect to the recommendation to require off-ramps instead of providing for a conditional exemption, the Bureau concludes that off-ramps alone would not provide sufficient protection to consumers. As discussed in the section-by-section analysis of § 1041.6(b) through (e), the Bureau believes those provisions offer important protections against harms from default, delinquency, re-borrowing, and forgoing basic living expenses or major financial obligations to avoid default. While off-ramps likely would help consumers who are struggling to repay their loans by giving them additional time and reducing their payments, they would not mitigate the potential harms as effectively as the suite of protections in § 1041.6. Moreover, as some commenters noted, lenders frequently have managed to find ways to discourage consumers from taking advantage of off-ramp options under existing State laws, and therefore the Bureau has determined that off-ramps would be less effective at improving the chances that consumers will be able to repay covered short-term loans without becoming mired in extended loan sequences. 854 As noted above, the Bureau has concluded that the structural requirements of the exemption are well-designed to prevent or mitigate the harm that results from unaffordable shortterm loans, but the Bureau also has concluded that making the requirements of the exemption more demanding would restrict its value to consumers and lenders. A range of commenters argued that the exemption should be limited to loans with certain APRs, that conditionally exempt loans should remain subject to income verification, or that lenders should be required to pay back finance charges if borrowers repay early. While the requirements in § 1041.5 of the final rule are designed to prevent the harms identified in § 1041.4, the Bureau has recognized that those requirements may be burdensome to some lenders and consumers, and thus finds it prudent to offer a less restrictive alternative to address the identified harms. As noted above, some industry commenters argued that the underwriting requirements in proposed §§ 1041.5 and 1041.6 would be unworkable and that the exemption in proposed § 1041.7 would not provide a feasible alternative. The Bureau has endeavored to substantially address the concerns raised about the complexity and burdens of the underwriting requirements, as adopted in § 1041.5, through revisions to those requirements as discussed above. Section 1041.6 was intended to reduce burden and allow for a more feasible alternative to loans made under § 1041.5. In particular, it does not require lenders to meet the specific underwriting criteria set out in § 1041.5. It does, however, still impose some restrictions, which in turn involve some burden. The Bureau acknowledges this, but considers each of the restrictions imposed in § 1041.6 necessary or appropriate to ensure that the exemption does not allow significant amounts of harms to continue under the exemption. 855 Having said that, the Bureau recognizes, as commenters noted, that allowing lenders to continue making covered short-term loans without requiring the loans to meet the underwriting criteria specified in § 1041.5 poses some risk, even with the protections that are built into the exemption. Those risks include the likelihood that at least some loans meeting the conditions under § 1041.6 may be unaffordable at least to some consumers. The Bureau acknowledges these concerns, and agrees that finalizing § 1041.5 without this exemption would create a more rigid framework that would more completely prevent the risks and harms identified in § 1041.4. But a significant animating influence in the Bureau’s decision to include this exemption was the aim of acting prudently in fashioning its first underwriting rule for this market, while recognizing as noted above that some borrowers that likely cannot satisfy the ability-to-repay test may still be able to repay their loans without re-borrowing.860 As some commenters suggested, the Bureau will monitor how lenders use conditionally exempt loans to see if the risks and harms identified in this rule are being perpetuated, and stands ready to take action if it sees this occurring. Of course, lenders will also need to comply with more restrictive State laws as applicable, which is consistent with the notion that this rule is a floor and not a ceiling on consumer protections, both in general and for purposes of preemption as discussed in part IV.861 Additionally, the Bureau judges it likely that lenders will find it in 860 It should be recognized that with the modifications made to § 1041.5, the Bureau has determined that the population of people who cannot establish the ability to repay, yet can actually repay, has reduced substantially. 861 In response to commenters’ that expressed concerns that this exemption may influence State law, or be used by others to influence State law, the Bureau has no comment on what State legislatures should do in the future, and trusts that they will advance their own policy goals while keeping in mind that, as a matter of preemption, this rule acts as a floor rather than a ceiling on consumer protections, and beyond that threshold the States are free to engage in further regulation of covered loans as they may determine to be appropriate, including by imposing usury caps as a number of States have chosen to do, whereas Congress prohibited the Bureau from imposing any usury limits. See 856 their self-interest to engage in additional underwriting before making conditionally exempt loans given that the re-borrowing restrictions with respect to such loans will mean that lenders cannot count on revenue from extended loan sequences to cover the costs of defaults. Put differently, the distinct conditions for these loans will likely lead to modifications in the lending practices of those lenders choosing to utilize the provisions of § 1041.6. Those conditions are likely to prompt more caution in making such loans, because the costs incurred by making unaffordable loans cannot be offset by heavy volumes of re-borrowing fees. The Bureau also disagrees with the claim made by some commenters that after having identified as an unfair and abusive practice the making of covered short-term loans without reasonably determining that the borrower has the ability to repay the loans according to their terms, the Bureau must prohibit all such loans in all circumstances. As explained further below, the Bureau has express legal authority to issue exemptions from its rules. The Bureau agrees that the measures intended to mitigate the harms caused by the practice identified as unfair and abusive in § 1041.4 may not entirely mitigate those harms when lenders make conditionally exempt loans without underwriting according to the criteria laid out in § 1041.5. At this time, however, the Bureau deems it prudent to accept that level of risk in light of the positive effects that § 1041.6 will have on reducing burden and providing access to credit while continuing to mitigate most of the harms caused by the practice identified in § 1041.4. Both consumer and industry commenters asked the Bureau to clarify how the requirements of § 1041.6 would interact with existing State law. The provisions to which the 12 U.S.C. 5517(o) (Bureau may not impose a “usury limit”); see also part II (discussing different State approaches to these issues); part IV (discussing legal authorities and preemption under section 1041 of the Dodd-Frank Act). 857 commenters objected are merely conditions for loans to satisfy the § 1041.6 exemption, not new requirements that the Bureau is imposing on all loans. If a lender cannot legally offer a loan meeting such conditions in the State or city where a conflicting requirement exists, then that lender simply cannot offer loans that qualify for the § 1041.6 exemption, though it always can underwrite loans under the provisions of § 1041.5 where State law permits such loans to be made. To be clear, however, nothing in this rule categorically prohibits extended repayment plans. To the extent that some jurisdictions presently allow loans to be rolled over three times, the cap of two partial rollovers (subject to the prescribed limits on the amounts that can be rolled over) in § 1041.6 nevertheless must be met for loans to qualify for the conditional exemption. Legal Authority Section 1041.6 establishes an alternative set of requirements for covered short-term loans that, if complied with by lenders, conditionally exempts them from § 1041.4 and the specific underwriting criteria in § 1041.5.862 The requirements of § 1041.6 have been developed pursuant to section 1022(b)(3)(A) of the Dodd-Frank Act, which authorizes the Bureau to grant conditional exemptions in certain circumstances from its rules. With respect to § 1041.6(e), the Bureau developed the proposed disclosures by relying on its authority under section 1032(a) of the Act, which allows it to prescribe rules to ensure that the features of a consumer financial product or service are fully, accurately, and effectively disclosed to consumers, and section 1032(b) of the Act, which provides for the use of model forms. These sources of legal authority for § 1041.6 of the final rule are explained more fully below. 862 The Bureau finalizes those provisions pursuant to its separate authority under section 1031(b) of the Dodd-Frank Act to “prescribe rules identifying as unlawful unfair, deceptive or abusive acts or practices” and to include in such rules “requirements for the purpose of preventing such acts or practices.” 12 U.S.C. 5531(b). 858 Section 1022(b)(3)(A) of the Dodd-Frank Act—Exemption Authority. Section 1022(b)(3)(A) of the Dodd-Frank Act authorizes the Bureau, by rule, to “conditionally or unconditionally exempt any class of . . . consumer financial products or services” from any provision of Title X of the Act or from any rule issued under Title X as the Bureau determines “necessary or appropriate to carry out the purposes and objectives” of Title X.863 The purposes of Title X are set forth in section 1021(a) of the Act, which provides that the Bureau shall implement and, where applicable, enforce Federal consumer financial law consistently “for the purpose of ensuring that all consumers have access to markets for consumer financial products and services and that [such markets] are fair, transparent and competitive.”864 The objectives of Title X are set forth in section 1021(b) of the Dodd-Frank Act.865 This section authorizes the Bureau to exercise its authorities under Federal consumer financial law for five specified purposes, two of which are relevant here. In particular, the Bureau may exercise its authorities under Federal consumer financial law for the purposes of ensuring that, with respect to consumer financial products and services: (1) consumers “are provided with timely and understandable information to make responsible decisions about financial transactions;”866 (2) consumers “are protected from unfair, deceptive, or abusive acts and practices and from discrimination;”867 (3) “outdated, unnecessary, or unduly burdensome regulations are regularly identified and addressed in order to reduce unwarranted regulatory burdens;”868 (4) “Federal consumer financial law is enforced consistently, without regard to the status of a person as a 863 12 U.S.C. 5512(b)(3)(A). 12 U.S.C. 5511(a). 865 12 U.S.C. 5511(b). 866 12 U.S.C. 5511(b)(1). 867 12 U.S.C. 5511(b)(2). 868 12 U.S.C. 5511(b)(3). 864 859 depository institution, in order to promote fair competition;”869 and (5) “markets for consumer financial products and services operate transparently and efficiently to facilitate access and innovation.”870 When issuing an exemption under section 1022(b)(3)(A) of the Dodd-Frank Act, the Bureau is required under section 1022(b)(3)(B) of the Act to take into consideration, as appropriate, three factors: (1) the total assets of the class of covered persons;871 (2) the volume of transactions involving consumer financial products or services in which the class of covered persons engages;872 and (3) existing provisions of law which are applicable to the consumer financial product or service and the extent to which such provisions provide consumers with adequate protections.873 The conditional exemption for covered short-term loans in § 1041.6 is appropriate to carry out the purposes and objectives of Title X of the Dodd-Frank Act, for three primary reasons. First, § 1041.6 is consistent with the Bureau’s statutory purposes and its statutory objective under section 1021(b)(5) of the Dodd-Frank Act: seeking to implement Federal consumer financial law consistently to ensure that consumers have access to fair, transparent, and competitive markets for consumer financial products and services; and ensuring that such markets operate transparently and efficiently to facilitate access to consumer financial products and services. Section 1041.6 will help preserve access to credit by providing lenders with an option for making covered short-term loans that is an alternative to—and a conditional 869 12 U.S.C. 5511(b)(4). 12 U.S.C. 5511(b)(5). 871 12 U.S.C. 5512(b)(3)(B)(i). 872 12 U.S.C. 5512(b)(3)(B)(ii). 873 12 U.S.C. 5512(b)(3)(B)(iii). 870 860 exemption from—the requirements of § 1041.5. Because lenders making these conditionally exempt loans under proposed § 1041.6 will be conditionally exempt from complying with the specific underwriting criteria under § 1041.5, making loans under § 1041.6 will reduce the compliance costs for lenders that make covered short-term loans relative to the costs of complying with the underwriting requirements under § 1041.5.874 This reduction in compliance costs will help facilitate access to credit. Second, the conditional exemption for covered short-term loans is consistent with the Bureau’s statutory objective under section 1021(b)(2) of the Dodd-Frank Act, which is to ensure that consumers are protected from unfair or abusive acts and practices. In § 1041.4, the Bureau has stated that it is an unfair and abusive practice for a lender to make covered short-term loans without making a reasonable determination that consumers have the ability to repay the loans according to their terms. In § 1041.5, the Bureau prevents this unfair and abusive practice by prescribing specific underwriting criteria for lenders making certain covered loans. Although lenders making conditionally exempt loans are not required to satisfy these same requirements, they will be required to satisfy the alternative requirements for the conditional exemption under § 1041.6. These alternative requirements are designed to protect consumers from the harms that result from lenders making covered short-term loans that are unaffordable—namely, default, delinquency, repeat borrowing, and collateral harms from making unaffordable loan payments. These are the same kinds of harms that the requirements in § 1041.5 were designed to address. 874 Note that the relative difference in compliance costs and access in the proposal would likely be reduced in the final rule because the Bureau made changes to proposed § 1041.5 intended to reduce complexity and burden and to maintain access to credit. For example, in the proposal, the Bureau stated that borrowers who are paid in cash would be able to obtain a loan under proposed § 1041.7, even though they would be unable to obtain a loan under proposed § 1041.5. Now borrowers who are paid in cash can get a loan under either § 1041.5 or § 1041.6 of the final rule. 861 Third, the conditional exemption in § 1041.6 is consistent with the Bureau’s statutory objective under section 1021(b)(1) of the Dodd-Frank Act to ensure that consumers are provided with timely and understandable information to make responsible decisions about financial transactions. Under § 1041.6(e), the Bureau is prescribing a series of disclosure requirements in connection with the making of these conditionally exempt loans. The disclosures notify the consumer about important aspects of how these transactions operate, and are designed to contribute significantly to consumers having timely and understandable information about taking out these conditionally exempt loans. The Bureau also considered the statutory factors listed in section 1022(b)(3)(B) of the Dodd-Frank Act, as appropriate. The first two factors are not materially relevant because they pertain to exempting a class of covered persons, whereas § 1041.6 conditionally exempts a class of transactions from certain requirements of the rule. Nor did the Bureau base the conditional exemption on the third factor. Certain requirements under § 1041.6 are similar to requirements under certain applicable State and local laws. However, the Bureau is not aware of any State or locality that has combined all the elements that the Bureau has concluded are necessary or appropriate to adequately protect consumers from the risks and harms associated with unaffordable loans when covered short-term loans are not underwritten under the terms of § 1041.5.875 875 See also discussion in Market Concerns—Underwriting about the prevalence of harms in the short-term loan market in spite of existing regulatory approaches. 862 The Bureau emphasizes that the conditional exemption in § 1041.6 is a partial exemption. That is, these conditionally exempt loans are still subject to all of the requirements of the Bureau’s proposed rule other than the specific underwriting criteria in § 1041.5. Sections 1032(a) and (b) of the Dodd-Frank Act—Disclosures. In § 1041.6(e), the Bureau is requiring disclosures related to covered short-term loans made under § 1041.6. The Bureau is doing so pursuant to its authority under section 1032(a) and (b) of the Dodd-Frank Act. Section 1032(a) of the Act provides that the Bureau “may prescribe rules to ensure that the features of any consumer financial product or service, both initially and over the term of the product or service, are fully, accurately, and effectively disclosed to consumers in a manner that permits consumers to understand the costs, benefits, and risks associated with the product or service, in light of the facts and circumstances.”876 This authority is broad, and it empowers the Bureau to prescribe rules on disclosures about the features of consumer financial products and services generally. Accordingly, the Bureau may prescribe disclosure requirements for particular features even if other Federal consumer financial laws do not specifically require such disclosures. Specifically, the Bureau is requiring lenders to provide notices before making the first and third loan in a sequence of conditionally exempt loans, which would inform consumers of the risk of obtaining such a loan and restrictions on taking out further conditionally exempt loans in a sequence. Under section 1032(b)(1) of the Dodd-Frank Act, “any final rule prescribed by the Bureau under [section 1032] requiring disclosures may include a model form that may be used at 876 12 U.S.C. 5532(a). 863 the option of the covered person for provision of the required disclosures.”877 Any model form must contain a clear and conspicuous disclosure which, at a minimum, must use plain language comprehensible to consumers, contain a clear format and design, and succinctly explain the information that must be communicated to the consumer. Section 1032(b)(3) of the Act provides that any model form the Bureau issues shall have been validated through consumer testing. Accordingly, in developing the model forms for the proposed notices, the Bureau conducted two rounds of qualitative consumer testing in September and October of 2015, contracting with Fors March Group (FMG) to conduct qualitative user testing of the forms, which presented its results in the FMG Report. Dodd-Frank Act section 1032(d) provides that, “Any covered person that uses a model form included with a rule issued under this section shall be deemed to be in compliance with the disclosure requirements of this section with respect to such model form.”878 6(a) Conditional Exemption for Certain Covered Short-Term Loans Proposed Rule In proposed § 1041.7(a), the Bureau proposed to establish a conditional exemption for certain covered short-term loans. Under proposed § 1041.7(a), a covered short-term loan that is made in compliance with the requirements set forth in proposed § 1041.7(b) to (e) would have been exempt from §§ 1041.4-1041.6. The Bureau also proposed in § 1041.7(a) to require the lender, in determining whether the proposed requirements in paragraphs (b), (c), and (d) are satisfied, to obtain information about the consumer’s borrowing history from the records of the 877 878 12 U.S.C. 5532(b)(1). 12 U.S.C. 5532(d). 864 lender or its affiliates, and a consumer report from an information system registered under proposed § 1041.17(c)(2) or (d)(2). Proposed comment 7(a)-1 explained that a lender could make a covered short-term loan without making the ability-to-repay determination under proposed § 1041.5, provided it complied with the requirements set forth in proposed § 1041.7(b) to (e). Proposed comment 7(a)-2 clarified that a lender cannot make a covered short-term loan under proposed § 1041.7 if no information system is both registered under proposed § 1041.17(c)(2) or (d)(2) and available when the lender seeks to make the loan. Proposed comment 7(a)-2 also clarified that a lender may be unable to obtain a report on the consumer’s borrowing history if, for example, information systems are not yet operational or are temporarily unavailable. Comments Received Commenters urged the Bureau not to adopt the prohibition on making these conditionally exempt loans if no registered information system is operational and available. They argued that this requirement would be unfair or irrational because, even if a lender complied with all of the regulatory requirements under the alternative approach, the lender would still have to rely on a third-party reporting agency’s compliance with the new and untested rules. One commenter observed that this was especially problematic given that most lenders will come to depend primarily on the approach to lending provided in the conditional exemption, and hence this restriction will reduce access to credit for consumers. Consumer groups supported the requirement that a lender check a registered information system before making a conditionally exempt loan. They asserted that restrictions based on borrower history are the primary limit on conditionally exempt loans and that without this 865 requirement the exemption would only work on a lender-by-lender basis. Because of the risk of multiple lenders making loans to the same borrower absent the requirement, the commenters argued that this requirement is appropriate. Several commenters requested a safe harbor from the requirements in the rule where the lender relies on information from a registered information system where the information turns out to be incorrect. For example, if a borrower were to have previously taken out three consecutive conditionally exempt loans under proposed § 1041.7 at a different lender, and applied for a fourth such loan within 30 days at a new lender, and those prior three loans did not appear on the report obtained from the registered information system, one commenter believed the new lender should not be held liable for failing to comply with the requirements in proposed § 1041.7 when it makes the loan in accordance with the erroneous information that the registered information system had provided to it. Final Rule The final rule adopts § 1041.7(a) as proposed, renumbered in this final rule as § 1041.6(a), with some technical edits and one addition—that the information system from which the lender obtains a consumer report must have been registered for 180 days or more pursuant to § 1041.11(c)(2) or registered pursuant to (d)(2). In addition, the final rule clarifies that the lender must use this borrowing history information to determine a potential loan’s compliance with the requirements in § 1041.6(b) and (c); the reference to § 1041.6(d) is removed. Lenders will not need to obtain a separate report from a registered information systems to comply with § 1041.6(d), which prohibits a lender from making a loan within 30 days of a conditionally exempt loan made by that lender itself (other than another conditionally 866 exempt loan following the conditions of § 1041.6).879 And § 1041.6(c), as well as § 1041.5(d), restrict covered short-term loans made by other lenders, as well as loans made by the same lender and its affiliates. The Bureau added the provision specifying that, when a lender is relying on a report from an information system registered pursuant to § 1041.11(c)(2) to satisfy § 1041.6, the registered information system must have been registered for 180 days or more. Under § 1041.10(b), a lender is not required to begin furnishing information to registered information systems registered pursuant to § 1041.11(c)(2) until 180 days after they are registered. A consumer report obtained from an information system registered for less than 180 days would not contain any information about borrowers’ use of covered short-term and longer-term balloon payment loans. In the final rule, the Bureau is retaining the proposed requirement that, prior to making a covered short-term loan under § 1041.6, a lender must review the consumer’s borrowing history in its own records, the records of the lender’s affiliates, and a consumer report from a registered information system. The Bureau concludes that lenders should not be permitted to make conditionally exempt loans under § 1041.6 if lenders do not obtain and review a report from a registered information system, even in instances where a report from a registered information system is unavailable. The Bureau maintains its view that reports from registered information systems are important for ensuring that the protections put in place by § 1041.6 are fully realized, and, based on outreach during the rulemaking process, the Bureau expects to register at least one 879 Lenders that make covered short-term loans under § 1041.6 will have to check their own records and records of affiliates before making loans to ensure that they are complying with § 1041.6(b) and (c). 867 information system sufficiently in advance of the effective date of §§ 1041.5 and 1041.6 that reports from a registered information system will be available and may be relied upon on such effective date. If no report from a registered information systems is available and a lender is therefore unable to obtain reliable information about a consumer’s borrowing history with other lenders, the Bureau is concerned that conditionally exempt lending could result in consumers continuing to experience extended cycles of re-borrowing. Consumers could refinance a loan under § 1041.6 from one lender with another lender, and repeat continuously, severely undermining many of the protections contained in § 1041.6. In the unlikely circumstance that no information system has been registered for at least 180 days as of the effective date of §§ 1041.5 and 1041.6, the Bureau will consider its options at that time, but does not at this time wish to leave open the possibility of § 1041.6 lending without lenders first obtaining borrower history from a registered information system. If lenders are unable to make loans under § 1041.6 absent a report from a registered information system, the Bureau has concluded that lenders will have an incentive to ensure that there is at least one registered information system that has been registered for at least 180 days as of the effective date of §§ 1041.5 and 1041.6. If the Bureau were to allow lenders to make § 1041.6 loans without obtaining a report from a registered information system, the opposite could be true—industry would have an incentive to impede or slow the development of registered information systems. The Bureau is finalizing comment 6(a)-1 as proposed, with the addition of citations of §§ 1041.8 and 1041.9 to clarify the meaning of “other applicable laws” (which in essence means that these conditionally exempt loans are still subject to the payment-related provisions of this 868 rule). The Bureau has adjusted comment 6(a)-2 to clarify the requirement that the registered information system from which the lender obtains a consumer report must have been registered under § 1041.11(c)(2) for 180 days or more or must be registered under § 1041.11(d)(2). The Bureau has added comment 6(a)-3 in response to commenters requesting a safe harbor when they rely on information obtained from a registered information system to make a loan determination and the information they are provided later turns out to have been erroneous. This comment clarifies that a lender is not responsible for inaccurate or incomplete information contained in a consumer report from a registered information system. If a lender relies on information obtained from a registered information system that is inaccurate, and based on that inaccurate information makes a loan that does not comply with the requirements of § 1041.6 because of inaccurate information in that report, the loan nonetheless qualifies for the exemption in § 1041.6. 6(b) Loan Term Requirements In proposed § 1041.7(b), the Bureau proposed to require a covered short-term loan that is made under proposed § 1041.7 to comply with certain requirements as to the loan terms and structure. The requirements under proposed § 1041.7(b), in conjunction with the other requirements set forth in proposed § 1041.7(c) to (e), were presented as an alternative to the underwriting criteria specified in § 1041.5, and were likewise intended to reduce the likelihood that consumers who take out these conditionally exempt loans would suffer the competing harms of default, delinquency, re-borrowing, or the collateral harms from making unaffordable loan payments to avoid default. These proposed requirements were also intended to limit the harm to consumers if they are unable to repay the loan as scheduled. 869 6(b)(1) Proposed Rule In proposed § 1041.7(b)(1), the Bureau proposed certain principal amount limitations for a conditionally exempt loan. Specifically, proposed § 1041.7(b)(1)(i) would have required that the first loan in a sequence of conditionally exempt loans have a principal amount that is no greater than $500. Proposed § 1041.7(b)(1)(ii) would have required that the second loan in a sequence of conditionally exempt loans have a principal amount that is no greater than two-thirds the principal amount of the first loan in the sequence. Proposed § 1041.7(b)(1)(iii) would have required that the third loan in a sequence of conditionally exempt loans have a principal amount that is no greater than one-third of the principal amount of the first loan in the sequence. Proposed comment 7(b)(1)-1 cross-referenced the definition and commentary for loan sequences. Proposed comment 7(b)(1)-2 clarified that the principal amount limitations apply regardless of whether the loans are made by the same lender, an affiliate, or unaffiliated lenders. Proposed comment 7(b)(1)-3 noted that the principal amount limitations under proposed § 1041.7 apply to both rollovers of an existing loan when they are permitted under State law and new loans that are counted as part of the same loan sequence. Proposed comment 7(b)(1)-4 gave an example of a loan sequence in which the principal amount is stepped down or amortized in increments of one-third. The Bureau believed that the principal cap and principal reduction requirements under proposed § 1041.7(b)(1) were critical to reducing both the risk of extended loan sequences and the risk that the loan payments over a limited, shorter loan sequence would prove unaffordable for consumers. Because proposed § 1041.7 would not require the borrower to meet the 870 underwriting criteria set forth in proposed § 1041.5 for a covered short-term loan, some consumers may not be able to repay these loans as scheduled. Absent further protections, these consumers would be in the position of choosing among the harms that borrowers confront when they have to make the payments on an unaffordable loan—default on the loan, or re-borrow, or fail to meet basic living expenses or other major financial obligations in an effort to avoid default as the loan comes due. As discussed in the proposal, the Bureau found that in this predicament, consumers in the market today generally re-borrow for the same amount as the prior loan, rather than pay off a portion of the principal and reduce their debt burden. As a result, consumers may face a similar situation when the next loan comes due and all succeeding loans after that, except that they have paid substantial fees for re-borrowing with every additional loan. The Bureau has found that this lack of principal reduction, or “self-amortization,” over the course of a loan sequence is correlated with higher rates of re-borrowing and default.880 Proposed § 1041.7(b)(1) was designed to work in tandem with proposed § 1041.7(c)(3), which proposed to limit a loan sequence of these conditionally exempt loans to no more than three loans. The proposed requirements together would ensure that a consumer may not receive more than three consecutive covered short-term loans under proposed § 1041.7 and that the principal would decrease from a maximum of $500 on the first loan over the course of a loan sequence. The proposed principal reduction feature was intended to steadily reduce consumers’ debt burden and permit them to pay off the original loan amount in its entirety in more manageable increments over the course of a loan sequence with three loans. 880 See CFPB Data Point: Payday Lending, at 16, 17 panel A & panel B. 871 The Bureau believed that the proposed $500 limit for the first loan was appropriate in light of current State regulatory limits and would reduce the risks that unaffordable payments would cause consumers to default, re-borrow, or fail to meet basic living expenses or other major financial obligations during a loan sequence. Many State statutes authorizing payday loans impose caps on the loan amount, with $500 being a common limit.881 In States that have lower limits on loan amounts, those lower limits would prevail. In addition, the Bureau’s empirical research found that average loan sizes are well under this threshold.882 Finally, without applying the underwriting criteria under proposed § 1041.5, the Bureau believed that loans with a principal amount larger than $500 would carry a significant risk of unaffordable payments. The Bureau also gave extensive consideration to proposing an “off-ramp” for consumers who are struggling to repay a covered short-term loan, in lieu of the principal reduction structure.883 Under this approach, lenders would be required to provide a no-cost extension of the third loan in a sequence (the off-ramp) if a consumer is unable to repay the loan according to its terms. The Bureau believed that the off-ramp approach would have three significant disadvantages relative to the principal reduction structure outlined above. First, an off-ramp, 881 E.g., Ala Code sec. 5–18A–12(a); Alaska Code sec. 06.50.010; Col. Code sec. 5–3.1–101; Fla. Code sec. 560.402; Iowa Code sec. 533D.10(1)(b); Kan. Code secs. 16a–2–404–05; Ken. Code sec. 286.9–010; Miss. Code sec.75–67–501; Mo. Code secs. 408.500–06; Neb. Code sec. 45–901; N.H. Code sec. 399A:1; Ohio Rev. Code sec.1321.35; Okla. Code sec. 59–3101; R.I. Code secs. 19–14.1–11; S.D. Code sec. 54–4–36; Tenn. Code sec. 45– 17–101; Va. Code sec. 6.2–1800. 882 The Bureau’s analysis of supervisory data indicated that the median loan amount for payday loans is around $350. See CFPB Payday Loans and Deposit Advance Products White Paper, at 15. Another study found that the average loan amount borrowed was $375. See Pew Charitable Trusts, “Payday Lending in America: Who Borrows, Where They Borrow, and Why,” at 9 (Report 1, 2012), available at http://www.pewtrusts.org/~/media/legacy/uploadedfiles/pcs_assets/2012/pewpaydaylendingreportpdf.pdf. 883 See Small Business Review Panel Report, at 8. 872 which began after a sequence of three loans, would delay the onset of the principal reduction and compel consumers to carry the burden of unaffordable payments for a longer time, increasing the likelihood of default and collateral harms from making unaffordable loan payments. Second, the Bureau believed that an off-ramp provision likely could not be designed in a way so as to ensure that consumers actually receive the off-ramp. The Bureau’s analysis of State regulatory reports indicated that even where off-ramps are made available under State law, actual consumer use of available off-ramps has been quite limited.884 Third, to make an off-ramp approach less susceptible to such defects, additional provisions would be necessary, including disclosures alerting consumers to their rights to take the off-ramp and prohibitions on false or misleading information regarding off-ramp usage and collections activity prior to completion of the full loan sequence. These measures would be of uncertain effectiveness and would increase complexity, burdens on lenders, and challenges for enforcement and supervision. Comments Received Several industry commenters urged the Bureau not to adopt the $500 cap in proposed § 1041.7(b)(1) because it is too low. These groups argued that the Bureau had not sufficiently demonstrated that $500 was a large enough amount of money to meet consumer demand and that consumers routinely needed more money, especially for potential emergencies. One commenter was concerned that the $500 cap was inconsistent with the definition of small-dollar loans in 884 The experience in Florida also suggests that off-ramps are not likely to be made available to all consumers who struggle to repay covered short-term loans. For borrowers who indicate that they are unable to repay the loan when due and agree to attend credit counseling, Florida law requires lenders to extend the loan term on the outstanding loan by 60 days at no additional cost. Although 84 percent of loans were made to borrowers with seven or more loans in 2014, fewer than 0.5 percent of all loans were granted a cost-free term extension. See Brandon Coleman & Delvin Davis, “Perfect Storm: Payday Lenders Harm Florida Consumers Despite State Law,” Ctr. for Responsible Lending,” at 4 & n.7 (2016), available at http://www.responsiblelending.org/sites/default/files/nodes/files/researchpublication/crl_perfect_storm_florida_mar2016_0.pdf. 873 some States and could lead to compliance problems and costs, causing lenders to leave the market and producing a reduction in available credit. In contrast, consumer groups urged the Bureau not to adopt the $500 cap in the proposed rule because it is too high. The group argued that the median loan amount for current borrowers is $350 to $375 and this smaller median loan amount did not support the $500 cap. Several commenters supported the principal reduction requirements in proposed § 1041.7(b)(1). An academic commenter suggested this feature would benefit borrowers by helping them make incremental progress on their loans, and argued that a 3-loan sequence would help provide borrowers with sufficient time to repay their loans. Several consumer groups urged the Bureau not to adopt the conditional exemption, yet supported the 3-loan framework with an amortizing structure if the exemption was part of the final rule. Some commenters argued that roughly two-thirds of borrowers are unable to pay off these kinds of loans in three payments or less, so the provision would likely be ineffective, but stated that it may be worth trying nonetheless. Several consumer groups and a legal services organization supported the Bureau’s choice to use principal reduction and amortization instead of using off-ramps. These commenters asserted that consumers often are not informed about or are discouraged from using off-ramps, which makes them ineffective. In contrast, some industry commenters wrote in support of adding an off-ramp option. One said it would be more in keeping with existing approaches by the States and would adequately address the Bureau’s concerns about the number of consecutive transactions in extended loan sequences. 874 Some industry commenters urged the Bureau not to adopt the proposed structure of three loans with amortization. They asserted that emergency expenses are not predictable, and so a rigid 3-loan schedule with amortization would not meet borrower needs. Several industry commenters urged the Bureau to allow more conditionally exempt loans in order to reduce the size of the step-down between each loan, and thus reduce the amount that the borrower would be unable to re-borrow after each loan, which would also reduce the burden and impact on lenders by allowing more re-borrowing. A number of State Attorneys General similarly noted that some States have implemented smaller principal-repayment requirements that permit more rollovers and more time for consumers to repay. One commenter suggested that five step-down loans was a better limit than three because it would allow for smaller and more affordable payments. Another recommended a 4-loan sequence with an indebtedness limit of 104 days during a 12-month period. In contrast, consumer groups urged the Bureau not to extend the number of loans. These commenters argued that increasing the number of loans from the proposed level of three loans even to four loans would result in more harm to borrowers because of the longer payment period. Final Rule The Bureau has considered the comments and is adopting proposed § 1041.7(b)(1), renumbered in this final rule as § 1041.6(b)(1), as proposed. The Bureau adopts proposed comments 7(b)(1)-1 through 7(b)(1)-4 as proposed, renumbered in this final rule as comments 6(b)(1)-1 through 6(b)(1)-4, with only technical modifications. The Bureau does not agree with the industry commenters that urged the Bureau not to adopt the $500 cap because it is too low to meet consumer demand, especially for potential 875 emergencies. The Bureau also does not agree with consumer groups that the Bureau should set the cap closer to the median loan amount for current borrowers of $350 to $375. For the reasons discussed in the proposed rule and noted above, the Bureau has determined that the $500 limit for the first loan is appropriate in light of current State regulatory limits and ordinances, and will reduce the risks that unaffordable payments will cause consumers to default, re-borrow, or seek to avoid default by failing to meet basic living expenses or other major financial obligations over the course of a loan sequence. The Bureau’s empirical research, confirmed by commenters, has also found that average loan sizes are well under this threshold.885 In addition, without applying the underwriting criteria set out in § 1041.5, the Bureau concludes that short-term loans with a principal amount larger than $500 would carry a significant risk of having unaffordable payments with the ensuing harms to consumers that are discussed more fully above in Market Concerns—Underwriting. Of course, lenders could always choose to proceed by underwriting loans according to the criteria set out in § 1041.5, or they could instead make other types of loans that are not covered by the rule, in amounts higher than $500 to the extent permitted by State law. Similarly, the Bureau is not persuaded by the concern that the $500 cap is inconsistent with the definition of small-dollar loans in some States, and could lead to compliance problems and costs that would cause lenders to leave the market and reduce the availability of credit. The Bureau determined that many State statutes authorizing payday loans already impose caps on the 885 The Bureau’s analysis of supervisory data indicated that the median loan amount for payday loans is around $350. See CFPB Payday Loans and Deposit Advance Products White Paper, at 15. Another study found that the average loan amount borrowed was $375. See Pew Charitable Trusts, “Payday Lending in America: Who Borrows, Where They Borrow, and Why,” at 9 (Report 1, 2012), available at http://www.pewtrusts.org/~/media/legacy/uploadedfiles/pcs_assets/2012/pewpaydaylendingreportpdf.pdf. 876 loan amount, with $500 as a common limit.886 In States with lower limits on loan amounts, those lower limits would prevail. In States with higher limits, lenders could still make underwritten loans under § 1041.5 at those higher amounts. The Bureau also concludes that the 3-loan step-down will provide borrowers with sufficient time to repay the loan and mitigate harm to borrowers. It adopted this framework for § 1041.6(b)(1) of the final rule in an attempt to balance the interests of limiting re-borrowing while also providing for a gradual step-down. For each additional loan, the step-down would be less steep (i.e., the amount that would not be refinanced and thus would need to be “repaid” would decrease), but the borrower would incur that much more re-borrowing. For example, if the Bureau adopted a 5-loan limit, the second loan would be 80 percent of the original, the third loan 60 percent, the fourth loan 40 percent, and the fifth loan 20 percent. That would allow for more affordable payments, but would also add two additional loans, with the attendant costs. Ultimately, the Bureau had to determine where to draw the line, which is often an unavoidable exercise in the rulemaking process, and it concluded that the combination of the $500 cap and the 3-loan step-down, resulting in fees from three loans and a maximum “repayment amount” of $166.66 (the amount not refinanced on each step) in principal for each loan, strikes a reasonable balance between these competing concerns. The Bureau recognizes that some borrowers may not be able to use loans under § 1041.6 to meet new credit needs because of the step-down in loan amounts for the second and third 886 See, e.g., Ala Code sec. 5–18A–12(a); Alaska Code sec. 06.50.010; Col. Code sec. 5–3.1–101; Fla. Code sec. 560.402; Iowa Code sec. 533D.10(1)(b); Kan. Code secs. 16a–2–404–05; Ken. Code sec. 286.9–010; Miss. Code sec.75–67–501; Mo. Code secs. 408.500–06; Neb. Code sec. 45–901; N.H. Code sec. 399A:1; Ohio Rev. Code sec.1321.35; Okla. Code sec. 59–3101; R.I. Code secs. 19–14.1–1; S.D. Code sec. 54–4–36; Tenn. Code sec. 45– 17–101; Va. Code sec. 6.2–1800. 877 conditionally exempt loan. For example, a borrower who takes out a first loan of $300 under § 1041.6, and then has a new need arise before 30 days has passed, would only be able to take out a further loan of $200 (which is the remaining amount under the principal cap), which may not be sufficient to cover the need. But, as stated above, and in the discussion for § 1041.6(c) and (d), borrowers who return for loans within a 30-day period are often re-borrowing because of difficulty in repaying their previous loan and meeting their obligations rather than taking out a new loan in response to a new need that is separate and independent from the original need. Further, those borrowers may be able to get other types of credit from other lenders to supplement the amount obtainable under § 1041.6, including a loan that would be underwritten in accordance with the provisions of § 1041.5. One further benefit from the limitations on re-borrowing imposed in the principal cap and the principal reduction feature in § 1041.6(b)(1), as mentioned earlier, is that they are likely to improve the care and consideration with which lenders make these conditionally exempt loans, even though they are not required to be underwritten in accordance with the criteria specified in § 1041.5. As noted above in Market Concerns—Underwriting, a major reason why lenders in this market are willing to lend to borrowers who are unable to repay their loans is that the costs of default are substantially offset by the revenues generated by high levels of re-borrowing; and indeed, many defaults may be deferred rather than immediate because the borrower can opt to reborrow some number of times—and often in extended loan sequences—before finally defaulting. By strictly limiting the amount of re-borrowing that can occur with loans made under § 1041.6, the Bureau’s conditional exemption thus is likely to lead to improved underwriting of these loans, even without imposing any mandatory underwriting criteria upon their origination. 878 6(b)(2) Proposed Rule Proposed § 1041.7(b)(2) would have imposed certain safeguards in the event that a lender chose to structure the loan with multiple payments, such as a 45-day loan with three required payments. Under the proposed requirement, the loan would have required payments that are substantially equal in amount, fall due in substantially equal intervals, and amortize completely during the term of the loan. Proposed comment 7(b)(2)-1 provided an example of a loan with an interest-only payment followed by a balloon payment, which would not satisfy the loan structure requirement under proposed § 1041.7(b)(2). The requirement under proposed § 1041.7(b)(2) was intended to address covered shortterm loans made under proposed § 1041.7 that are structured to have multiple payments. Absent the requirements in proposed § 1041.7(b)(2), the Bureau was concerned that lenders could structure loans to pair multiple interest-only payments with a significantly larger payment of the principal amount at the end of the loan term. The Bureau believed that consumers are better able to manage repayment obligations for payments that are due with reasonable frequency, in substantially equal amounts, and within substantially equal intervals. Comments Received One commenter urged the Bureau not to adopt the approach in proposed § 1041.7(b)(2) that requires a payment schedule based on applying a fixed rate of interest. It observed that the States generally regulate payday loan finance charges by limiting fees charged per amount lent instead of using an interest rate, and argued that requiring a payment schedule based on an 879 interest rate would force lenders to reprogram their systems on a scale that goes beyond the Bureau’s statutory mandate. On the other hand, several consumer groups supported the Bureau’s proposal to allow multi-payment loans under the exemption, assuming it remained a part of the rule. They asserted that the risk of including the multi-payment loans did not increase the inherent risk of the exemption. They also supported the position taken in the proposal that permitting balloon payments for multiple-payment loans under the conditional exemption would be antithetical to the purpose of the exemption. Final Rule The Bureau has considered the comments and is adopting proposed § 1041.7(b)(2), renumbered in this final rule as § 1041.6(b)(2), as proposed. The Bureau also adopts proposed comment 7(b)(2)-1 as proposed, renumbered in this final rule as comment 6(b)(2)-1, with only technical modifications. As discussed in more detail in the proposed rule and above, § 1041.6(b)(2) provides certain safeguards in the event that a lender chooses to structure a covered short-term loan with multiple payments. Absent the requirements in § 1041.6(b)(2), the Bureau is concerned that lenders could structure loans to pair multiple interest-only payments with a significantly larger payment of the principal amount at the end of the loan term. The Bureau has concluded that consumers are better able to manage repayment obligations for payments that are due with reasonable frequency, in substantially equal amounts, and within substantially equal intervals. The Bureau agrees with commenters that the principal reduction feature will help borrowers make incremental progress on loans. The Bureau also judges that the concern regarding 880 supposed inconsistency with State laws is overstated. Section 1041.6(b)(2) only applies in circumstances where one individual loan has multiple payments, and there is nothing in the text of § 1041.6(b)(2) that limits the imposition of fees, so long as the fees are repaid equally during every scheduled payment. 6(b)(3) Proposed Rule In proposed § 1041.7(b)(3), the Bureau proposed to prohibit a lender, as a condition of making a covered short-term loan under proposed § 1041.7, from obtaining vehicle security, as defined in proposed § 1041.3(d). A lender seeking to make a covered short-term loan with vehicle security would have had to make an ability-to-repay determination under proposed § 1041.5 instead. Proposed comment 7(b)(3)-1 clarified this prohibition on a lender obtaining vehicle security on a conditionally exempt loan. The Bureau proposed this requirement because it was concerned that some consumers obtaining a loan under proposed § 1041.7, without meeting the underwriting criteria in proposed § 1041.5, would not be able to afford the payments required to pay down the principal over a sequence of three loans. Allowing lenders to obtain vehicle security in connection with such loans could substantially increase the harm to consumers by putting their vehicle at risk. The Bureau believed the proposed requirement would protect consumers from the harms of default, re-borrowing, and making unaffordable loan payments to avoid defaulting on covered short-term vehicle title loans. First, the Bureau was particularly concerned about default that could result in the loss of the consumer’s vehicle, which could jeopardize their livelihood or their ability to carry out essential everyday affairs. The Bureau found that sequences of short-term vehicle title 881 loans are more likely to end in default than sequences of payday loans are,887 and that fully 20 percent of loan sequences of single-payment vehicle title loans result in repossession of the consumer’s vehicle.888 Second, due to the potentially serious consequences of defaulting on title loans, the Bureau was concerned that consumers may take extraordinary measures to repay such loans and, as a result, would suffer harm from failing to meet basic living expenses or other major financial obligations. Third, even with the other protections against re-borrowing in proposed § 1041.7, the Bureau was concerned that, due to the serious consequences of defaulting on vehicle title loans, consumers may feel pressure to re-borrow up to the maximum allowed on unaffordable vehicle title loans.889 Furthermore, the Bureau believed that proposed § 1041.7(b)(3) is necessary or appropriate to restrict lenders’ incentives to make these conditionally exempt loans with unaffordable payments. Because loan sequences would be limited to a maximum of three conditionally exempt loans under proposed § 1041.7(c)(3) and subject to principal reduction under § 1041.7(b)(1), the Bureau believed a lender that makes these conditionally exempt loans would have a strong incentive to underwrite effectively, even without having to comply with the specific underwriting criteria in proposed § 1041.5. However, with vehicle title loans, in which the lender obtains a security interest in an asset of significantly greater value than the principal amount on the loan,890 the Bureau was concerned that a lender would have much less incentive to 887 CFPB Single-Payment Vehicle Title Lending, at 11; CFPB Report on Supplemental Findings, at 120. CFPB Single-Payment Vehicle Title Lending, at 23. 889 A single-payment short-term vehicle title loan is less likely to be repaid after one loan than a payday loan. See CFPB Single-Payment Vehicle Title Lending, at 11; CFPB Report on Supplemental Findings, at 120. 890 For further discussion of how vehicle security affects the market for such loans, see CFPB Single-Payment Vehicle Title Lending, and see also part II above. 888 882 evaluate the consumer’s ability to repay, because the lender could always simply repossess the vehicle if the loan were not repaid in full, even after the first loan in the sequence. Comments Received Consumer groups supported the proposed prohibition on auto title lending under the conditional exemption in proposed § 1041.7. They asserted that the repossession of a borrower’s vehicle represented significant harm, especially given the high rate of repossessions. They argued that the harm from repossession is so severe that lenders should not be allowed to make vehicle title loans without assessing ability to repay. In contrast, commenters associated with the vehicle title lending industry wrote in opposition to the proposed prohibition on title lending under the conditional exemption. An industry trade association argued that requiring all short-term vehicle title loans to satisfy the proposed ability-to-repay standards would have a devastating impact on lenders and on the availability of such loans. They argued that the Bureau had not sufficiently demonstrated that vehicle title lending presents greater risks than other forms of short-term lending and had overstated the rate and impact of repossession, asserting that only about 8 percent of title loans result in repossession. The commenter further argued that the Bureau had exaggerated the effects of repossession, contending that many consumers own a second vehicle and that surveys indicate consumers would have alternative transportation options if their vehicle were repossessed. The industry trade association also argued that the prohibition was inconsistent with the Bureau’s mandate to regulate the market fairly and consistently, and that by prohibiting vehicle title lenders from using the conditional exemption the proposed rule would provide an unfair advantage for other types of lenders. 883 Final Rule The Bureau has considered the comments and, for the reasons noted in the proposal and above and for the additional reasons discussed below, is adopting proposed § 1041.7(b)(3), renumbered in this final rule as § 1041.6(b)(3), as proposed. The Bureau is also adopting comment 7(b)(3)-1 as proposed, renumbered as comment 6(b)(3)-1. The Bureau concludes, as the consumer groups argued, that the risk of severe consumer harm from repossession of the borrower’s vehicle makes it inappropriate to allow lenders to make covered short-term vehicle title loans without satisfying the underwriting requirements in § 1041.5. The Bureau does not agree with the argument of the title lending industry commenters that the Bureau had not sufficiently demonstrated that vehicle title lending presents greater risks than other forms of short-term lending. The structure of § 1041.6 is intended to reduce defaults and re-borrowing, and if lenders were permitted to make vehicle title loans under this structure, the protections in § 1041.6 might reduce defaults and repossessions to some degree. But the Bureau is concerned that the reduction in defaults may be less likely than for unsecured short-term loans, such as payday loans. As noted in the proposal, as a general matter in this market, sequences of short-term vehicle title loans are more likely to end in default than sequences of payday loans are.891 Although an industry commenter argued that the Bureau had overstated the rate of repossession, that commenter focused on the per-loan default rate. As discussed in Market Concerns— Underwriting, the Bureau has concluded that a per-sequence rather than per-loan default rate provides a better measure for short-term loans. One in five loan sequences of single-payment 891 CFPB Single-Payment Vehicle Title Lending, at 11; CFPB Report on Supplemental Findings, at 120. 884 vehicle title loans result in repossession of the consumer’s vehicle.892 Moreover, as noted above, once the revenues from repeated re-borrowing are constrained, as they are by the conditions imposed in § 1041.6, the incentive for lenders to make unsecured loans on which the borrower is likely to default are sharply diminished. But the change in incentives is far less pronounced for vehicle title loans, where even as re-borrowing revenues decrease, the lender still has the leverage of a fully securitized loan available to cope with any defaults. Therefore, even with the protections of § 1041.6, there would still be some borrowers who cannot afford to repay loans made under § 1041.6. And for the reasons just stated, there are likely to be more such borrowers of vehicle title loans than of other covered short-term loans. In addition, the harm produced by unaffordable title loans is greater than for other such loans. If lenders could take vehicle security for loans under § 1041.6, then consumers who could not afford to repay their loans would face the threat of having their vehicles repossessed, and, in many cases, would suffer the severe harms of repossession. The harms from repossession (and comments about those harms) are discussed above in Market Concerns—Underwriting and in the section-by-section discussion of § 1041.4, and, contrary to the assertions by industry commenters, the Bureau has concluded that such harms are often severe. First, consumers facing repossession would suffer the potential loss of transportation to work or school and for many other everyday activities, such as securing food and health care, with consequential losses that may greatly exceed the original cost of the loan.893 Second, due to the potentially serious 892 CFPB Single-Payment Vehicle Title Lending, at 23. Even for those consumers who may have access to some other mode of transportation, the Bureau notes that there are hardships and inconveniences associated with having to use other forms of transportation, especially in nonurban areas of the country. And for at least 15 percent of title loan borrowers, their personal vehicles are essential 893 885 consequences of defaulting on title loans, the Bureau is concerned that consumers may take extraordinary measures to repay such loans and, as a result, would suffer greater harm more frequently from failing to meet basic living expenses or other major financial obligations. Third, even with the other protections against re-borrowing in § 1041.6, the Bureau is concerned that, due to the serious consequences of defaulting on vehicle title loans, consumers may feel greater pressure to re-borrow up to the maximum allowed on unaffordable vehicle title loans, since a vehicle title loan is less likely to be repaid after one loan than are other types of covered shortterm loans.894 In addition, there are still other economic collateral harms of repossession, which is usually a self-help process performed by agents of the lender and which often results in significant consumer fees associated with the costs of the repossession and preparing a vehicle for auction.895 These processes can put the consumer at greater risk of harm, and often more severe harm, than when a consumer defaults on an unsecured loan. The Bureau has observed typical repossession fees charged to borrowers ranging from $100 to $400 or even higher, which could be larger than the small balance of the defaulted loan made under § 1041.6 (with a maximum of $500 on the first loan, $333.33 on the second loan, and $166.66 on the third loan). And there are additional harms often associated with repossessions, including the potential loss for numerous transportation needs. See See Pew Charitable Trusts, “Auto Title Loans: Market Practices and Borrowers’ Experiences,” at 14 (2015), available at http://www.pewtrusts.org/~/media/Assets/2015/03/AutoTitleLoansReport.pdf?la=en. 894 See CFPB Single-Payment Vehicle Title Lending, at 11; CFPB Report on Supplemental Findings, at 120. 895 Uniform Commercial Code section 9–615 provides that cash proceeds of the sale of collateral should be applied first to the “reasonable expenses of retaking, holding, preparing for disposition, processing, and disposing” incurred by the secured party. Under the U.C.C., these expenses are repaid to the lender and other third parties even before satisfying the outstanding balances of the secured loan. 886 of any property in the vehicle.896 These harms persist even in States that limit vehicle title lending to so-called non-recourse loans. For all of these reasons, vehicle title loans that are not subject to the specific underwriting criteria of § 1041.5 present significant additional risks as compared to unsecured loans that are not subject to § 1041.5. Moreover, the harms to consumers that flow from these risks are greater for vehicle title loans. Accordingly, the Bureau has concluded that it is appropriate in § 1041.6 to require lenders making such loans not to take a security interest in the consumer’s vehicle. The Bureau recognizes that, because lenders making short-term vehicle title loans are highly dependent on the revenue from re-borrowing, requiring short-term vehicle title loans to comply with the ability-to-repay requirements in § 1041.5 will have a significant impact on such lenders. Title lenders that are unable to adjust their business models or obtain a license to make unsecured small-dollar loans or installment title loans thus may face greater challenges than payday lenders because they would not be able to make loans under § 1041.6 that would be exempt from the ability-to-repay requirements of § 1041.5. (The Bureau notes that, by its own count, 18 of the 24 States that permit title lending allow title installment lending that would not be covered by § 1041.5.) Nonetheless, the Bureau concludes that, under § 1041.6, covered shortterm loans with vehicle security would present more risks and more severe harms than unsecured covered short-term loans. The Bureau therefore is requiring that if a lender takes a security interest in the consumer’s vehicle, then it must underwrite any covered short-term loans that it makes pursuant to § 1041.5. Finally, since the rule does not differentiate based on whether a 896 See, e.g., Bureau of Consumer Fin. Prot., “Supervisory Highlights,” at 5–6 (Fall 2016), available at https://www.consumerfinance.gov/data-research/research-reports/supervisory-highlights-issue-no-13-fall-2016/. 887 lender is a depository or non-depository lender, or based on any other characteristics of the lender, and instead makes differentiations based on the loan products themselves and the risks associated with them, the Bureau is not imposing inconsistent obligations here on lenders based on their status as depository or non-depository lender. 6(b)(4) Proposed § 1041.7(b)(4) would have required that, as a condition of making a covered short-term loan under proposed § 1041.7, the loan must not be structured as an open-end loan. Proposed comment 7(b)(4)-1 clarified this prohibition on a lender structuring a conditionally exempt loan as an open-end loan. The Bureau was concerned that permitting open-end loans under proposed § 1041.7 would present significant risks to consumers, as consumers could repeatedly draw down credit without the lender ever determining the consumer’s ability to repay. In practice, consumers could re-borrow serially on a single conditionally exempt loan that was structured as an open-end loan. The Bureau also believed that attempting to develop restrictions for open-end loans in proposed § 1041.7 would add undue complexity without providing appreciable benefit for consumers. The Bureau received very limited comments on this provision, with consumer groups supporting the Bureau’s proposed prohibition on using the conditional exemption to extend open-end credit and agreeing with its rationale. For the reasons stated, the Bureau is adopting the proposed prohibition against structuring loans as open-end loans under the conditional exemption, now renumbered as § 1041.6(b)(4). The Bureau is also adopting proposed comment 7(b)(4)-1, renumbered as comment 6(b)(4)-1. 888 6(c) Borrowing History Requirements The Bureau proposed to require lenders to determine that the borrowing history requirements under proposed § 1041.7(c), renumbered in this final rule as § 1041.6(c), are satisfied before making a conditionally exempt loan. The Bureau is finalizing this paragraph as proposed, with a few adjustments to reduce redundancy and to reflect the fact that the Bureau is not finalizing the rule as to covered longer-term loans at this time, yet is finalizing the underwriting requirements for covered short-term and longer-term balloon-payment loans in one section, § 1041.5 of the final rule. One adjustment that the Bureau is making, in particular, is not to finalize proposed § 1041.7(c)(1), which would have required a lender to determine, before making a conditionally exempt loan, that the consumer does not have a covered outstanding loan made under proposed §§ 1041.5, 1041.7, or 1041.9, not including a loan made by the same lender or its affiliate under proposed § 1041.7 that the lender is rolling over. As a result of this change, the Bureau also is not adopting proposed comments 7(c)(1)-1 and 7(c)(1)-2. For purposes of simplification and in light of other changes made to the rule, the Bureau has concluded that this proposed provision could be consolidated with § 1041.7(c)(2), which addresses restrictions on taking out conditionally exempt loans in light of prior loans in specified circumstances. As a result of eliminating § 1041.7(c)(1), the other proposed sections of § 1041.7(c) and the proposed comments are all renumbered in the final rule to conform to this change. 889 6(c)(1) Proposed Rule Proposed § 1041.7(c)(2) would have required that, prior to making a covered short-term loan under proposed § 1041.7, the lender must determine that the consumer has not had an outstanding loan in the past 30 days that was either a covered short-term loan made under proposed § 1041.5 or a covered longer-term balloon-payment loan made under proposed § 1041.9. The requirement under proposed § 1041.7(c)(2) would have prevented a consumer from obtaining a covered short-term loan under proposed § 1041.7 soon after repaying a covered short-term made under proposed § 1041.5 or a covered longer-term balloon-payment loan made under proposed § 1041.9. Proposed comment 7(c)(2)-1 explained that this requirement would apply regardless of whether the prior loan was made by the same lender, an affiliate of the lender, or an unaffiliated lender. The proposed comment also provided an illustrative example. Proposed § 1041.7(c)(2) would have protected consumers who lack the ability to repay a current or recent covered short-term or longer-term balloon-payment loan from the harms of a covered short-term loan made without meeting the specific underwriting criteria in proposed § 1041.5. As explained above, the Bureau observed that such re-borrowing frequently reflects the adverse budgetary effects of the prior loan and the unaffordability of the new loan. Moreover, the Bureau believed that permitting a consumer to transition from a covered short-term loan made under proposed § 1041.5 or a covered longer-term balloon-payment loan made under proposed § 1041.9 to a covered short-term loan made under proposed § 1041.7 would be inconsistent with the basic purpose of proposed § 1041.7. As previously noted, proposed § 1041.7 creates an alternative to the underwriting criteria specified in proposed 890 § 1041.5 and features carefully structured consumer protections. If lenders were permitted to make a conditionally exempt loan shortly after making a covered short-term loan under proposed § 1041.5 or a covered longer-term balloon-payment loan under proposed § 1041.9, it would be very difficult to apply all of the requirements under proposed § 1041.7 that are designed to protect consumers. If a consumer were permitted to transition from a proposed § 1041.5 loan to a covered short-term loan made under proposed § 1041.7, for example, the principal reduction requirements under proposed § 1041.7(b)(1) would be undermined. The Bureau also believed that providing separate paths for covered short-term loans that are made under the specific underwriting criteria in proposed § 1041.5 and under the framework in proposed § 1041.7 would make the rule’s application more consistent across provisions and also simpler for both consumers and lenders. The Bureau intended these two proposed lending frameworks to work in tandem, but not in harness, to ensure that lenders could not transition consumers back and forth between covered short-term loans made pursuant to the underwriting criteria specified in proposed § 1041.5 and those made without the same criteria but subject to other consumer protections under proposed § 1041.7. Furthermore, with these proposed provisions in place, and with the two lending frameworks largely kept separate from one another, consumers and lenders would have clear expectations of the types of covered short-term loans that they could and could not make if the consumer were to re-borrow. Comments Received Several commenters, including a coalition of consumer groups, two non-profit groups, three faith-based groups, and a State Attorney General urged the Bureau to increase the coolingoff periods in proposed § 1041.7(c), including the cooling-off period in proposed § 1041.7(c)(2) 891 so that, after making a covered short-term loan under § 1041.5, a lender would have to wait 60 days, rather than 30 days, before it could make a conditionally exempt loan under § 1041.6. They argued that a 60-day cooling-off period was more appropriate and more protective, and would do more to help ensure that loans were affordable. On the other hand, industry commenters generally opposed having a cooling-off period of any length, arguing that it would restrict access to credit for consumers with emergency or unexpected needs that may arise during the cooling-off period. Commenters argued that covered loans are often used for unexpected expenses, which can happen at any time, and that a coolingoff period would harm consumers by restricting their flexibility and reducing access to credit when borrowers needed it. A large number of individual commenters, including payday loan customers, also criticized the cooling-off periods, objecting to the prospect that they would be restricted from getting more credit after paying off a prior loan. Final Rule The Bureau is finalizing proposed § 1041.7(c)(2), renumbered as § 1041.6(c)(1), with a few adjustments. For purposes of simplification and in light of other changes made to the rule, the Bureau has concluded that proposed § 1041.7(c)(1) and (c)(2) can be consolidated together, with technical corrections to accommodate changes to other sections of the rule, including the fact that the underwriting requirements for covered longer-term loans (other than those with balloon payments) are not being finalized. Accordingly, § 1041.6(c)(1) provides that a condition of making a loan under § 1041.6 is that the consumer has not had in the past 30 days an outstanding covered short-term loan under § 1041.5 or a covered longer-term balloon-payment 892 loan under § 1041.5. The Bureau is also adopting proposed comment 7(c)(2)-1, renumbered as 6(c)(1)-1, with similar adjustments. In response to the commenters that had advocated extending the cooling-off period to 60 days, the Bureau continues to rely on the research and analysis that were used initially to set the 30-day re-borrowing period. In the proposal, the Bureau had chosen the cooling-off period to match the re-borrowing period because the primary objective served by cooling-off periods in this rule is to prevent re-borrowing. The main approach to preventing re-borrowing is to separate out any linkage between different types of loans or different permitted loan sequences by having sufficient time pass to diminish the plausibility that the prior loan was paid off only by taking out another loan that provided the money to do so. Under the Bureau’s definition, based on its analysis of the market, loans made after 30 days would not be considered re-borrowing. The Bureau’s research found that the number of loans in the average loan sequence increases when the re-borrowing window for identifying a sequence increases from 14 days to 30 days, suggesting that borrowers are returning to re-borrow within 30 days. The Bureau also concluded that a 30-day cooling-off period is a reasonable and sufficient representation of most consumers’ debt and payment cycles. Because payments for basic living expenses and most major financial obligations are due at least monthly, if not more frequently, the Bureau concludes that a consumer who goes more than 30 days between two short-term loans is more likely to be experiencing a new need, rather than continuing to service the need that gave rise to the prior loan, and thus extending the same cycle of indebtedness. The Bureau thus has concluded that setting a cooling-off period of 30 days between a § 1041.5 loan and a § 1041.6 loan is a reasonable exercise in line-drawing that is likely to prevent the perpetuation of hard-to893 escape cycles of indebtedness, while allowing consumers greater flexibility for borrowing to cover emergency or other unexpected expenses. While the Bureau acknowledges that a 60-day cooling-off period would do even more to prevent re-borrowing, as some consumers might be able to shuffle around certain expenses in order to reach day 31 in order to re-borrow, the Bureau concludes that the number of such loans is likely to be small given the data noted above, and that preventing relatively few additional consumers from remaining in a cycle of debt is not worth restricting credit to other consumers who may need it for genuine emergency expenses and new needs that may arise during that period (and subject to the protections conferred by this rule). As for the commenters who objected to cooling-off periods of any kind, including many individual commenters, the effect of this provision is that for 30 days after a § 1041.5 loan, a borrower would not be eligible for a § 1041.6 loan. The Bureau notes that where a lender has already made a § 1041.5 loan, the borrower has succeeded in demonstrating the ability to repay the loan in accordance with the underwriting criteria set forth in § 1041.5 and presumably is likely to continue to qualify for further loans by meeting that same standard. Therefore, if borrowers in that situation are now seeking a § 1041.6 loan instead, that may be because their circumstances have changed and they are now struggling to repay their loans and could no longer meet the underwriting criteria required by § 1041.5. To prevent lenders from using a mixture of § 1041.5 loans and § 1041.6 loans to create continuous cycles of debt where the borrower is confronting unaffordable loans, which would defeat a central purpose of § 1041.6, the Bureau has set this specific restriction. For the same reason of avoiding a mix of loans that could defeat the protections that the Bureau has intended to confer upon consumers under § 1041.6 (although 894 the circumstances are somewhat different), the Bureau has also specified that no lender can make a § 1041.5 loan within 30 days of a § 1041.6 loan.897 6(c)(2) Proposed Rule Proposed § 1041.7(c)(3) would have provided that a lender cannot make a covered shortterm loan under proposed § 1041.7 if the loan would result in the consumer having a loan sequence of more than three conditionally exempt loans made by any lender. Proposed comment 7(c)(3)-1 would have clarified that this requirement applies regardless of whether any or all of the loans in the loan sequence are made by the same lender, an affiliate, or unaffiliated lenders, explained that loans that roll over count toward the sequence as well, and included an example. The Bureau proposed § 1041.7(c)(3) for several reasons. First, the limitation on the length of loan sequences was aimed at preventing further harms from re-borrowing. Second, the Bureau believed that a 3-loan limit would be consistent with evidence presented in the Bureau’s Supplemental Findings on Payday Loans, Deposit Advance Products, and Vehicle Title Loans, that approximately 38 percent of new loan sequences end by the third loan without default.898 Third, a 3-loan limit would work in tandem with the main restrictions in proposed § 1041.7(b)(1) to allow consumers to repay a covered short-term loan in manageable one-third increments over a loan sequence. Fourth, the Bureau concluded that a 3-loan limit would provide lenders with a strong incentive to evaluate the consumer’s ability to repay before making conditionally exempt loans, albeit without complying with the specific underwriting criteria in proposed § 1041.5. 897 898 See § 1041.5(d)(3). See CFPB Report on Supplemental Findings, at 122. 895 Comments Received As noted above, a number of commenters urged the Bureau to increase the cooling-off periods in proposed § 1041.7(c) from 30 days to 60 days, including also the period after a borrower had received three loans under the conditional exemption in proposed § 1041.7. It should be noted that though proposed § 1041.7(c)(3) simply prohibited a lender from making a loan that would result in a consumer having a loan sequence of more than three loans under proposed § 1041.7, this provision in combination with the definition of loan sequence under proposed § 1041.2(a)(12) in effect created a 30-day cooling-off period after a three-loan sequence. Here too, consumer groups and others argued that a 60-day cooling-off period would be more protective of consumers and would help ensure that loans were more affordable. Industry commenters again were generally opposed to a cooling-off period after the loan sequence had ended, contending that it would restrict access to credit for consumers generally, including those with unexpected needs that could come up during a time when the borrower is not permitted to obtain another loan. Relatedly, and as discussed above, several industry commenters raised concerns about whether a three-loan sequence was the appropriate length for sequences of loans made under the conditional exemption, and suggested that the conditional exemption should permit longer loan sequences. As previously mentioned, large number of individual commenters, including payday loan customers, took issue with the cooling-off period and expressed concern that they might be blocked from getting a loan when they need it. 896 Final Rule The Bureau is finalizing proposed § 1041.7(c)(3), renumbered as § 1041.6(c)(2) with certain technical edits. The Bureau is also adopting proposed comment 7(c)(3)-1, renumbered as comment 6(c)(2)-1, with technical edits. Again, for much the same reasons as explained in the preceding discussion, the Bureau has relied on the same basic research and analysis to set the 30-day re-borrowing period and then has chosen this cooling-off period to match the re-borrowing period. Again, at the end of a 3loan sequence the purpose of the cooling-off period remains essentially the same, which is to prevent re-borrowing by preventing the borrower from linking different types of loans or different permitted sequences in such a manner as to continue taking out new loans or reborrowing as the means of paying off the prior loans. Again, loans made after 30 days would not be considered re-borrowing under the Bureau’s definition. As discussed above, the Bureau has determined that a 30-day period is a sound representation of most consumers’ debt and payment cycles. Because payments for basic living expenses and most major financial obligations are due at least monthly, if not more frequently, the Bureau concludes that a consumer who goes more than 30 days between loans is more likely to be experiencing a new need, rather than continuing to labor under pressure from the need that gave rise to the prior loan, and thus to be extending a cycle of indebtedness. The Bureau therefore determines that 30 days is a reasonable line to draw in setting a cooling-off period after completing a 3-loan sequence. Again, it helps prevents the perpetuation of hard-to-escape cycles of indebtedness, while allowing greater flexibility for further borrowing as needed to cover emergency or other unexpected expenses. While the Bureau acknowledges that a 60-day 897 cooling-off period would be even more protective of consumers, as some might be able to stretch certain expenses in order to exceed the 30-day cycle before having to re-borrow, the Bureau concludes that the number of such loans will be small and is outweighed by the benefits of having more credit available (with the other protections afforded by this rule) to consumers to meet any new needs that may arise during that period. As for the commenters that opposed a 30-day cooling-off period after three § 1041.6 loans, the Bureau acknowledges that some borrowers may experience a bona fide new need during that 30-day period and would be prevented from obtaining a new loan. As noted above when discussing the re-borrowing period, the Bureau concludes that borrowing within 30 days of a prior covered short-term loan will more typically reflect the continuing pressure that leads to re-borrowing rather than the emergence of a separate and independent need that prompts the borrower to take out a new loan. One of the primary purposes of this rule is to prevent consumers from falling into long-term re-borrowing cycles that result from loans with unaffordable payments. The Bureau concludes that the rule would fall far short of one of its chief purposes of preventing the risks and harms associated with unaffordable loans if § 1041.6 were to allow re-borrowing to create extended loan sequences in the period immediately after a 3-loan sequence has just been completed. Some built-in mechanism to disrupt a re-borrowing cycle is necessary or appropriate, and the Bureau has concluded that a cooling-off period of 30 days is the most effective way to accomplish that. Finally, the rationale for limiting loan sequences under § 1041.6(c)(3) to three loans is discussed above in the section-by-section analysis of § 1041.6(b)(1) and that discussion is incorporated by reference here. 898 6(c)(3) Proposed Rule Proposed § 1041.7(c)(4) would have required that a covered short-term loan made under proposed § 1041.7 cannot result in the consumer having more than six such loans outstanding during any consecutive 12-month period or having covered short-term loans outstanding for an aggregate period of more than 90 days during any consecutive 12-month period. The lender would have to determine whether any such loans were outstanding during the consecutive 12month period. If a consumer obtained a covered short-term loan prior to that period and was obligated on the loan during part of the period, this loan and the time it was outstanding during the consecutive 12-month period would count toward these overall limits. Under proposed § 1041.7(c)(4), the lender would have to count the proposed new loan toward the loan limit and count the anticipated contractual duration of the new loan toward the indebtedness limit. Under the proposal, because the new loan and its proposed contractual duration would count toward these limits, the lookback period would not start at the consummation date of the new loan. Instead, the lookback period would start at the proposed contractual due date of the final payment on the new loan and consider the full 12 months immediately preceding this date. Proposed comment 7(c)(4)-1 would have clarified that a consecutive 12-month period begins on the date that is 12 months prior to the proposed contractual due date of the new conditionally exempt loan and ends on the proposed contractual due date. Proposed comment 7(c)(4)-1 would have explained further that the lender would have to obtain information about 899 the consumer’s borrowing history on covered short-term loans for the 12 months preceding the proposed contractual due date on that loan, and it also provided an example. As a general matter, the Bureau was concerned about consumers’ frequent use of covered short-term loans made under proposed § 1041.7 for which lenders would not have been required to underwrite the loan in accordance with the criteria specified in proposed § 1041.5. The frequent use of covered short-term loans that do not require such an assessment may be a signal that consumers are struggling to repay such loans without re-borrowing. For purposes of determining whether the making of a loan would satisfy the loan and indebtedness limits in proposed § 1041.7(c)(4), the Bureau proposed to require the lender also to count covered shortterm loans made under both proposed § 1041.5 and proposed § 1041.7. Although loans made under proposed § 1041.5 would require the lender to make a reasonable determination of a consumer’s ability to repay, the Bureau believed that the consumer’s decision to seek a conditionally exempt loan, after previously obtaining a covered short-term loan based on the underwriting criteria in proposed § 1041.5, suggested that the consumer may now lack the ability to repay the loan and that the earlier loan approval may not have fully captured this particular consumer’s expenses or obligations. Under proposed § 1041.7(c)(4), consumers could receive up to six conditionally exempt loans and accrue up to 90 days of indebtedness on these loans, assuming the consumer did not also have any covered short-term loans made under proposed § 1041.5 during the same period. Because the duration of covered short-term loans is typically tied to how frequently a consumer receives income, the Bureau believed that the two overlapping proposed requirements were necessary to provide more complete protections for consumers. 900 Proposed § 1041.7(c)(4)(i) included the proposed requirement that a covered short-term loan made under proposed § 1041.7 cannot result in the consumer having more than six covered short-term loans outstanding during any consecutive 12-month period. Proposed comment 7(c)(4)(i)-1 explained certain aspects of proposed § 1041.7(c)(4)(i) relating to the proposed loan limit. Proposed comment 7(c)(4)(i)-1 clarified that, in addition to the new loan, all covered short-term loans made under either proposed § 1041.5 or proposed § 1041.7 that were outstanding during the consecutive 12-month period would count toward the proposed loan limit. Proposed comment 7(c)(4)(i)-1 also clarified that, under proposed § 1041.7(c)(4)(i), a lender may make a loan that when aggregated with prior covered short-term loans would satisfy the loan limit even if proposed § 1041.7(c)(4)(i) would prohibit the consumer from obtaining one or two subsequent loans in the sequence. Proposed comment 7(c)(4)(i)-2 provided examples. The Bureau believed that a consumer who seeks to take out a new covered short-term loan after having taken out six covered short-term loans during a consecutive 12-month period may very well be exhibiting an inability to repay such loans. The Bureau believed that if a consumer were seeking a seventh covered short-term loan under proposed § 1041.7 in a consecutive 12-month period, this would be an indicator that the consumer may, in fact, be using covered short-term loans to cope with regular expenses and compensate for chronic income shortfalls, rather than to cover an emergency or other non-recurring need.899 In these circumstances, the Bureau believed that the lender should make an ability-to-repay determination 899 See Market Concerns—Underwriting; Rob Levy & Joshua Sledge, “A Complex Portrait: An Examination of Small-Dollar Credit Consumers,” at 12 (Ctr. for Fin. Servs. Innovation, 2012), available at https://www.fdic.gov/news/conferences/consumersymposium/2012/A%20Complex%20Portrait.pdf. 901 in accordance with proposed § 1041.5 before making additional covered short-term loans and ensure that the payments on any subsequent loan are affordable for the consumer. Proposed § 1041.7(c)(4)(ii) included the proposed requirement that a covered short-term loan made under proposed § 1041.7 cannot result in the consumer having covered short-term loans outstanding for an aggregate period of more than 90 days during any consecutive 12-month period. Proposed comment 7(c)(4)(ii)-1 clarified certain aspects of the proposed rule as they relate to the proposed indebtedness limit. Proposed comment 7(c)(4)(ii)-1 explained that, in addition to the new loan, the period in which all covered short-term loans made under either proposed § 1041.5 or proposed § 1041.7 were outstanding during the consecutive 12-month period would count toward the indebtedness limit. The same proposed comment also clarified that, under proposed § 1041.7(c)(4)(ii), a lender may make a loan with a proposed contractual duration, which when aggregated with the time outstanding of prior covered short-term loans, would satisfy the indebtedness limit even if proposed § 1041.7(c)(4)(ii) would not prohibit the consumer from obtaining one or two subsequent loans in the sequence. Proposed comment 7(c)(4)(ii)-2 provided examples. The Bureau believed it was important to complement the proposed 6-loan limit with the proposed 90-day indebtedness limit in light of the fact that loan durations could vary under proposed § 1041.7. For the typical two-week payday loan, the two thresholds would have reached the same result, since a limit of six loans under proposed § 1041.7 means that the consumer could have been in debt on such loans for up to approximately 90 days per year or one quarter of the year. For 30- or 45-day loans, however, a 6-loan limit would have meant that the consumer could have been in debt for 180 or even 270 days out of a 12-month period. The 902 Bureau believed these kinds of results would be inconsistent with protecting consumers from the harms associated with long cycles of indebtedness. Given the income profile and borrowing patterns of consumers who borrow monthly, the Bureau believed that the proposed indebtedness limit is an important protection for these consumers. Consumers who receive 30-day payday loans are more likely to live on fixed incomes, and typically are recipients of Social Security.900 Fully 58 percent of monthly borrowers were identified as recipients of government benefits in the Bureau’s 2014 Data Point.901 These borrowers are particularly vulnerable to default and collateral harms from making unaffordable loan payments. The Bureau found that borrowers receiving public benefits are more highly concentrated toward the lower end of the income range. Nearly 90 percent of borrowers receiving public benefits reported annual incomes of less than $20,000, whereas less than 30 percent of employed borrowers reported annual incomes of less than $20,000.902 Furthermore, because public benefits are typically fixed and do not vary from month to month,903 in contrast to wage income that is often tied to the number of hours worked in a pay period, the Bureau believed that monthly borrowers are more likely than bi-weekly borrowers to use 900 Due dates on covered short-term loans generally align with how frequently a consumer receives income. Consumers typically receive public benefits, including Social security and unemployment, on a monthly basis. See CFPB Payday Loans and Deposit Advance Products White Paper, at 19. 901 See CFPB Data Point: Payday Lending, at 14. 902 The Bureau previously noted in the CFPB White Paper from April 2013 that a significant share of consumers (18 percent) reported a form of public assistance or other benefits as an income source (e.g., Social Security payments); these payments are usually of a fixed amount, typically occurring on a monthly basis; and that borrowers reporting public assistance or benefits as their income source are more highly concentrated toward the lower end of the income range for the payday borrowers in our sample. See CFPB Payday Loans and Deposit Advance Products White Paper, at 18–20. 903 CFPB Payday Loans and Deposit Advance Products White Paper, at 19. 903 covered short-term loans to compensate for a chronic income shortfall rather than to cover an emergency or other non-recurring need. The Bureau found that borrowers on fixed incomes are especially likely to struggle with repayments and face the burden of unaffordable loan payments for an extended period. As noted in the Supplemental Findings on Payday Loans, Deposit Advance Products, and Vehicle Title Loans, for loans taken out by consumers who are paid monthly, more than 40 percent of all loans to these borrowers were in sequences that, once begun, persisted for the rest of the year for which data were available.904 The Bureau also found that approximately 20 percent of borrowers905 who were paid monthly averaged at least one loan per pay period. In light of these considerations, the Bureau believed that a consumer who has been in debt for more than 90 days on covered short-term loans, made under either proposed § 1041.5 or proposed § 1041.7, during a consecutive 12-month period may very well be exhibiting an inability to repay such loans. If a consumer is seeking a covered short-term loan under proposed § 1041.7 that would result in a total period of indebtedness on covered short-term loans of greater than 90 days in a consecutive 12-month period, the Bureau believed that this consumer may, in fact, be using covered short-term loans to cover regular expenses and compensate for chronic income shortfalls, rather than to cover an emergency or other non-recurring need.906 Under these circumstances, the Bureau believed that the lender should make an ability-to-repay determination in accordance with the underwriting criteria proposed § 1041.5 before making 904 CFPB Report on Supplemental Findings, at 131. CFPB Report on Supplemental Findings, at 125. 906 See Market Concerns—Underwriting; Rob Levy & Joshua Sledge, “A Complex Portrait: An Examination of Small-Dollar Credit Consumers,” at 12 (Ctr. for Fin. Servs. Innovation, 2012), available at https://www.fdic.gov/news/conferences/consumersymposium/2012/A%20Complex%20Portrait.pdf. 905 904 additional covered short-term loans and ensure that the payments on any subsequent loan are affordable for the consumer. Comments Received Consumer groups wrote in support of the Bureau’s proposal to have both a 6-loan limit and a 90-day limit. Some asserted that having these overlapping limits was important because a limit that only covered the number of loans would not protect borrowers who took out somewhat longer 30-day or 45-day loans. A State Attorney General supported the 90-day limitation because it would limit many borrowers in that State to three loans a year, which would be significant. Two faith-based groups went further and urged the Bureau to further limit the number of short-term conditionally exempt loans. They argued that any re-borrowing is a sign of unaffordability, and suggested that the rule allow at most a single short-term conditionally exempt loan per year. Consumer groups and legal aid organizations further suggested that the 6-loan cap and the limitation of 90 days of indebtedness in a 12-month period should apply to all loans. They pointed to existing guidance from prudential regulators that provides no exceptions to the limit of six deposit advances in a year. A coalition of consumer groups also proposed that the Bureau adopt a further restriction on loans where the borrower would be unable to take out a full sequence of three conditionally exempt loans. The commenter noted that if a borrower took out a conditionally exempt loan but was close to either the 6-loan limit or the 90-day limit then the borrower would be unable to take advantage of the principal step-down requirements. The commenter asserted that this was inconsistent with the importance of the principal step-down 905 requirement and suggested either that no loan be permitted in these circumstances or that the loan be capped at a lower value based on the number of loans the borrower would still be permitted to take out. Some commenters urged the Bureau to expand some of the definitions relevant to the conditional exemption to capture more conduct. In response to the Bureau’s solicitation, commenters suggested that when computing the 90-day indebtedness limit it would be better to measure the days by the longer of contractual indebtedness or actual indebtedness because this measure is more relevant to whether borrowers are able to afford a loan. They also argued that loans which fall partially within the 12-month measuring period should be counted toward the 6loan limit. They further suggested that the look-back period for determining whether a borrower had six loans or 90 days of indebtedness should involve a two-step process: first the lender should look back 365 days from the first day of a new loan, then the lender should consider whether any days when the borrower has the loan would result in a violation of the 6-loan or 90day limit. Industry commenters urged the Bureau not to adopt the proposed 6-loan and 90-day limits. They asserted that rigid limits on re-borrowing were inappropriate because short-term loans are generally used to pay for emergency expenses and thus are not predictable, so the limits would be too inflexible to meet borrower needs. Industry commenters also argued that the restrictions would negatively affect borrowers who were paid monthly because they would only be able to take out three loans. Some commenters asserted that limits on days of indebtedness and numbers of loans would cause small lenders to go out of business, reducing the supply of credit. One industry commenter argued that the limit of six loans in a year was not supported by 906 the data and urged the Bureau to adopt a limit of eight loans per year instead, a comment also discussed in the section-by-section analysis for § 1041.5(d). Another industry commenter suggested that the Bureau consider engaging in more tests and experiments on loan limits. It argued that a limit on the number of loans may encourage borrowers to take out larger loans than they need because of uncertainty about their continuing ability to access credit. Another commenter opposed the proposed conditional exemption because of concerns about communications with borrowers and adverse action notices. This commenter observed that the rule might prohibit a conditionally exempt loan during some periods and not others, because of the restrictions, and that these variations would be difficult to explain adequately to consumers both more generally, and in adverse action notices. Final Rule For the reasons set forth in the proposal and discussed above and for the further reasons explained below, the Bureau is adopting proposed § 1041.7(c)(4), renumbered as § 1041.6(c)(3) of the final rule with certain technical edits. In addition, the Bureau is adopting proposed comment 7(c)(4)-1, renumbered as 6(c)(3)-1, with only technical edits. The Bureau adopts proposed comments 7(c)(4)(i)-1, 7(c)(4)(i)-2, 7(c)(4)(ii)-1, and 7(c)(4)(ii)-2, renumbered in this final rule as 6(c)(3)(i)-1, 6(c)(3)(i)-2, 6(c)(3)(ii)-1, and 6(c)(3)(ii)-2, with only technical adjustments. The Bureau modified the respective examples in comments 6(c)(3)(i)-2 and 6(c)(3)(ii)-2, however, in order to clarify that a lender could not make a conditionally exempt loan if either the 6-loan cap or the limit of 90 days of indebtedness was reached, even if that means a borrower had not yet reached the end of his 3-loan limit for a particular loan sequence. 907 The limits on making conditionally exempt loans pursuant to § 1041.6 during a 12-month period are intended to ensure that the conditional exemption does not become a mechanism that would allow for extensive repeat borrowing of potentially unaffordable covered short-term loans. The Bureau concludes that these limits on overall lending are not necessary for loans made under § 1041.5 because those loans must be underwritten according to criteria designed to prevent them from becoming unaffordable loans that pose special risks and harms to consumers as described above in Market Concerns—Underwriting. The Bureau has carefully considered the competing arguments that many commenters raised about the appropriate limits on lending under § 1041.6 of the final rule, with some suggesting that the Bureau should tighten the limits further from the proposed levels and others arguing that the limits as proposed should either be increased or eliminated entirely. The Bureau originally proposed the 6-loan limit based on considerable feedback as a reasonable limitation on the use of the conditionally exempt loans, which generally comprises two full loan sequences under § 1041.6. As noted above, the Bureau also proposed the overlapping 90-day limitation on indebtedness for such loans out of concerns specific to borrowers who are paid monthly and take out 30-day or 45-day loans, which, in the absence of a 90-day (or other durational) limit, could result in borrowers being indebted on covered short-term loans under § 1041.6 for half or even three-quarters of the year. If a borrower has the need to seek a loan more frequently than the exemption contemplates, the borrower can still receive an underwritten loan under § 1041.5 of the final rule or many other types of loans not covered by this rule. If in fact a borrower’s credit needs can only be met by arranging more extended credit than the limits under § 1041.6 would allow, the Bureau believes this may be a strong indicator that forms of underwritten longer-term 908 credit would be better suited to that consumer than the kinds of covered short-term loans under consideration here. In sum, the Bureau has considered the comments on both sides of this issue and declines to set higher limits. The limits set on loans made under § 1041.6 are the conditions that lenders must follow in order to be exempted from the underwriting criteria required in § 1041.5, which do not include any similar annual lending limitations. In setting these limitations, the Bureau has also relied in part on norms and precedents that have been set in this market by other Federal regulators, most notably the FDIC and the OCC, which both have issued guidance to the banks under their supervisory authority and have effectively limited borrowers of these kinds of loans to six loans in a 12-month period.907 As noted in the proposal, and in the Section 1022(b)(2) Analysis at part VII below, the Bureau recognizes that the broader combination of regulatory requirements in this rule, including the limitations on making conditionally exempt loans under § 1041.6 within a 12-month period, will have a significant economic impact on lenders that rely on extensive repeat re-borrowing for their operating revenue. The Bureau has also concluded, however, that the availability of loans under the exemption in § 1041.6, as well as underwritten loans made under § 1041.5 and other loans not covered by this rule, taken altogether, will still allow a large, albeit reduced, volume of 907 FDIC, “Financial Institution Letters: Guidelines for Payday Lending,” (Revised Nov. 2015), available at https://www.fdic.gov/news/news/financial/2005/fil1405a.html. (stating that institutions should ensure payday loans are not provided to customers who had payday loans for a total of 3 months during the previous 12-month period); Guidance on Supervisory Concerns and Expectations Regarding Deposit Advance Products, 78 FR 70624 (Nov. 26, 2013); Guidance on Supervisory Concerns and Expectations Regarding Deposit Advance Products, 78 FR 70552 (Nov. 26, 2013). 909 lending to continue in this market.908 As noted in the Section 1022(b)(2) Analysis below at part VII, even as some market consolidation occurs, consumers nevertheless are likely to retain convenient access to covered short-term loans. As clarified in Comments 6(c)(3)(i)-2 and 6(c)(3)(ii)-2, borrowers who reach one of the 12-month lending limitations in the midst of a loan sequence will not receive the full step-down for that sequence. In this particular situation, the Bureau has weighed the alternatives and concludes that the overall goal of limiting extensive repeat re-borrowing—a concern that is closely tied to the unfair and abusive practice identified in § 1041.4 and its harmful effects on consumers—takes precedence over the narrower goal of providing full amortization on each conditionally exempt loan that is made in this market. This decision involved a line-drawing exercise, and the Bureau has determined that this resolution steers a middle course between prohibiting the loan altogether in these circumstances, which seems too restrictive of access to credit, or allowing a full loan sequence to run its course, which would undermine the broader goal noted above of imposing the aggregate limits on re-borrowing over a 12-month period. The Bureau believes that were it to permit the full 3-loan sequence as long as the first loan would comply with the limitations on lending within a 12-month period, it could create incentives for lenders to structure their lending practices in order to ensure that a sixth loan is the beginning of a new sequence, and/or that a first loan in a sequence would end right at 90 days of indebtedness for that 12-month period, significantly undermining the effect of these limitations. Other ways to 908 The Bureau’s estimate in the Section 1022(b)(2) Analysis below is that the rule will reduce credit by approximately 6 percent, which would be much higher without the exemption in § 1041.6. 910 resolve this situation are also possible, but as this example demonstrates, as they become more complex, they would also become more difficult to administer. As for whether the 90-day limitation will negatively affect borrowers who are paid monthly because they would only be able to take out three conditionally exempt loans pursuant to § 1041.6 in a 12-month period, the Bureau notes that the situations of borrowers who are paid monthly were one of the reasons that the 90-day limitation was included in the rule. The Bureau is concerned that borrowers who take out 30-day or 45-day loans, without the 90-day limit, could find themselves indebted more often than not, which would be antithetical to the purpose of the conditional exemption to allow for credit for an emergency or other non-recurring need without having to comply with the full underwriting regime in § 1041.5. The Bureau recognizes that this framework will limit the ability of some borrowers to take out loans under the exemption, but reiterates that underwritten loans under § 1041.5 remain available, as do various loans not covered by this rule. 6(d) Restrictions on making other loans following a loan made under the conditional exemption Proposed Rule The proposed rule included a number of provisions designed to address the concern that lenders might seek to evade the protective features of proposed § 1041.7, such as the cooling-off period or principal step-down—and thereby keep consumers in long cycles of re-borrowing— through a combination of conditionally exempt loans and other loans. That proposed framework would have worked as follows. Under proposed § 1041.6(g), lenders would not have been allowed to make covered short-term loans pursuant to proposed § 1041.5 while a conditionally exempt loan is outstanding and for 30 days thereafter. That provision, modified to include 911 longer-term balloon-payment loans, is being finalized as § 1041.5(d)(3), as discussed above. Similarly, under proposed § 1041.10(e), lenders would not have been allowed to make covered longer-term loans under proposed § 1041.9 while a conditionally exempt loan made by the lender or its affiliate is outstanding and for 30 days thereafter. And under proposed § 1041.7(d), if the lender or its affiliate made a non-covered bridge loan (a certain type of non-recourse pawn loan) while a conditionally exempt loan made by the lender or its affiliate is outstanding and for 30 days thereafter, the days during which the non-covered bridge loan is outstanding would “toll” the running of the 30-day re-borrowing and cooling-off periods included in proposed § 1041.7. The latter two provisions are discussed immediately below because they are the basis of final § 1041.6(d). Proposed § 1041.10(e) provided that, during the time period in which a covered shortterm loan made by a lender or its affiliate under proposed § 1041.7 is outstanding and for 30 days thereafter, the lender or its affiliate must not make a covered longer-term loan under proposed § 1041.9 to a consumer. Proposed comment 10(e)-1 clarified that, during the time period in which a covered short-term loan made by a lender or its affiliate under proposed § 1041.7 is outstanding and for 30 days thereafter, a lender or its affiliate could make a covered longer-term loan under proposed § 1041.11 or proposed § 1041.12 to a consumer. In the proposal, the Bureau explained that although proposed § 1041.10(e) was functionally a component of the proposed conditional exemption in § 1041.7, it was being included in proposed proposed § 1041.10 for ease of reference for lenders so they could look to a single provision of the rule for a list of prohibitions and presumptions that affect the making of covered longer-term loans under proposed § 1041.9. More substantively, the Bureau explained 912 that it was proposing the prohibition contained in § 1041.10(e) to effectuate the principal reduction requirements under proposed § 1041.7(b)(1) and the three-loan limit on a sequence of loans under proposed § 1041.7(c)(3), which were designed to allow consumers to repay the principal gradually over a three-loan sequence. The Bureau noted that this proposed protection could be circumvented if, in lieu of making a loan subject to such principal reduction, a lender were free to make a high-cost covered longer-term loan under proposed § 1041.9 during the 30 days following repayment of the first loan—or second loan—in a sequence of covered short-term loans made under proposed § 1041.7 or while such first or second loan in the sequence was outstanding. Furthermore, the Bureau stated its belief that the prohibition in proposed § 1041.10(e) would prevent lenders from using a covered short-term loan made under proposed § 1041.7 to induce consumers into taking a covered longer-term loan made under proposed § 1041.9. The Bureau noted that, in the absence of the proposed requirement, as a covered short-term loan made under proposed § 1041.7 that was unaffordable comes due, the lender could leverage the consumer’s financial vulnerability and need for funds to make a covered longer-term loan that the consumer otherwise would not have taken. For a lender, this business model would generate more revenue than a business model in which the lender adhered to the proposed path for a sequence of loans made under proposed § 1041.7 and would also reduce the upfront costs of customer acquisition on covered longer-term loans. Lenders who desire to make covered longerterm loans under proposed § 1041.9 ordinarily would have to take steps and perhaps incur costs to acquire customers willing to take those loans and to disclose the terms of those loans upfront. 913 For the consumer, what is ostensibly a short-term loan may, contrary to the consumer’s original expectations, result in long-term debt. The Bureau sought comment, inter alia, on whether any alternative approaches exist that would address the Bureau’s concerns related to effectuating the conditional exemption in proposed § 1041.7 while preserving the ability of lenders to make covered longer-term loans under proposed § 1041.9 close in time to covered short-term loans under proposed § 1041.7. Turning to proposed § 1041.7(d), it provided that if a lender or an affiliate made a noncovered bridge loan during the time any covered short-term loan made by the same lender or an affiliate under proposed § 1041.7 is outstanding and for 30 days thereafter, the lender or affiliate would have had to modify its determination of loan sequence for the purpose of making a subsequent conditionally exempt loan. Specifically, the lender or an affiliate would not have been able to count the days during which the non-covered bridge loan is outstanding in determining whether a subsequent conditionally exempt loan made by the lender or an affiliate is part of the same loan sequence as the prior conditionally exempt loan. Non-covered bridge loan was defined in proposed § 1041.2(a)(13) as a non-recourse pawn loan made within 30 days of an outstanding covered short-term or longer-term balloon-payment loan that must be substantially repaid within 90 days. Proposed comment 7(d)-1 provided a cross-reference to proposed § 1041.2(a)(13) for the definition of non-covered bridge loan. Proposed comment 7(d)-2 clarified that proposed § 1041.7(d) would provide certain rules for determining whether a loan is part of a loan sequence when a lender or an affiliate makes both covered short-term loans under § 1041.7 and a noncovered bridge loan in close succession. Proposed comment 7(d)-3 provided an example. 914 The Bureau intended proposed § 1041.7(d) to maintain the integrity of a core protection in proposed § 1041.7(b). If a lender could make a non-covered bridge loan to keep a consumer in debt and reset a consumer’s loan sequence after 30 days, it could make a lengthy series of $500 loans and evade the principal step-down requirements in proposed § 1041.7(b)(1). In the absence of this proposed restriction, the Bureau believed that a consumer could experience an extended period of indebtedness after taking out a combination of covered short-term loans under § 1041.7 and non-covered bridge loans and not have the ability to gradually pay off the debt obligation and exit the loan sequence by means of the principal reduction requirement in proposed § 1041.7(b)(1). Proposed § 1041.7(d) paralleled the restriction in proposed § 1041.6(h) applicable to covered short-term loans made under proposed § 1041.5. The Bureau sought comment on whether this proposed restriction is appropriate, and also sought comment on whether lenders would anticipate making covered short-term loans under proposed § 1041.7 and non-covered bridge loans to consumers close in time to one another, if permitted to do so under a final rule. Comments Received The Bureau received a number of comments from consumer groups generally supporting both proposed § 1041.10(e) and proposed § 1041.7(d). Echoing the rationale provided by the Bureau for proposed § 1041.10(e), they asserted that, absent the prohibition, lenders would entrap consumers into an initial loan without assessing their ability-to-repay and then switch them to a longer-term installment loan. But they urged the Bureau to extend the 30-day period specified in proposed § 1041.10(e) to 60 days. As regards proposed § 1041.7(d), the consumer groups urged the Bureau to expand the definition of non-covered bridge loan to include any loan 915 from a lender or affiliate because the risks of evasion presented by non-covered bridge loans were equally present with other types of loans. In addition, they recommended that the proposed “tolling’ approach be replaced with a “reset” approach. That is, instead of tolling the running of the 30-day re-borrowing and cooling-off periods in proposed § 1041.7 during the pendency of a non-covered bridge loan, the period should reset to 30 days at the end of such a loan. Here as well, they urged the Bureau to extend the applicable periods from 30 to 60 days.909 Final Rule In the final rule, the Bureau has made a number of changes to the way it addresses the risk of lenders using other loans or a combination of loans to undercut the limitations in the conditional exemption as a way to evade the specific protections in § 1041.6 of the final rule and keep consumers in extended cycles of indebtedness. These changes have been made against the backdrop of the Bureau’s decisions not to finalize the underwriting requirements for covered longer-term loans (other than those with balloon payments) in proposed § 1041.9; to move the provisions on covered longer-term balloon-payment loans into § 1041.5; and not to finalize the presumptions in proposed § 1041.6 and proposed § 1041.10. As noted, one such change consists in modifying proposed 1041.6(g) to include covered longer-term balloon-payment loans as well as covered-short term loans, such that lenders would not be allowed to make either type of loan while a conditionally exempt loan is outstanding and for 30 days thereafter. The resulting provision is being finalized as § 1041.5(d)(3). 909 The Bureau’s response to the recommendations to extend the re-borrowing and cooling-off periods from 30 to 60 days are provided in the discussion of § 1041.6(c)(2) above. 916 In the same vein, § 1041.6(d) of the final rule is another example of these changes. It combines aspects of proposed § 1041.10(e) and proposed § 1041.7(d) into a single provision that applies to a broader range of loans. Under § 1041.6(d) of the final rule, a lender or its affiliate may not make any loan to a consumer, other than one governed by § 1041.6, for 30 days after making a prior § 1041.6 loan to that consumer. It thus applies to all loans other than § 1041.6 loans, not just covered longer-term loans (as in proposed § 1041.10(e)). Moreover, it prohibits all such loans being made during that 30-day period, rather than merely tolling the running of this period when the lender or its affiliate makes a non-covered bridge loan. With this restriction in place, a lender may or may not choose to opt in to the alternative lending framework created by the conditional exemption by making a loan to a consumer under § 1041.6 without meeting the specific underwriting criteria under § 1041.5. But if the lender does choose to make a loan to the borrower pursuant to § 1041.6, then it must make any further loans to that same consumer pursuant only to § 1041.6 until 30 days after any such conditionally exempt loans are no longer outstanding. As noted, under § 1041.5(d)(3), the lender also cannot make a conditionally exempt loan under § 1041.6 while a loan made under § 1041.5 is outstanding and for 30 days thereafter. The upshot is that if lenders want to make covered short-term loans without meeting the specified underwriting criteria under § 1041.5, one temporary condition they must accept is that the only loan they can make to the same borrower during the 30-day periods following the first and second loans in a sequence of loans made under § 1041.6 is another conditionally exempt loan, and that they cannot make any loans to the borrower during the 30-day cooling-off period following the third loan in such a sequence of loans 917 The Bureau has concluded that § 1041.6(d) of the final rule is necessary or appropriate for several reasons. As discussed, proposed § 1041.10(e) and proposed § 1041.7(d) had each been proposed to effectuate and prevent evasion of the protections provided by the principalreduction requirement and 30-day cooling-off period, as such evasion could result in long cycles of indebtedness. Proposed § 1041.7(d) was focused only on the limited bridging concern presented by making certain non-recourse pawn loans. In considering whether this restriction is appropriate—a point on which the Bureau explicitly sought comment—the Bureau came to view this treatment of the issues as much too narrow. The Bureau had been aware of some mergers and dual-channel operations that had created increased links between payday lending and pawn lending. But in thinking about the problems posed by any kind of loan that could be used by lenders to bridge between successive covered loans, the Bureau came to recognize that if the same lender could make other loans to the same borrower during the temporary period when the lender has opted into the alternative framework of the conditional exemption, then the lender could disrupt and potentially evade the alternative lending framework so carefully established in proposed § 1041.7. Instead of being restricted only to making amortizing loans in limited stepdown sequences that were established as a means of protecting consumers against the dangers of unaffordable loans that did not comply with the underwriting criteria specified in § 1041.5, it became clear that lenders could potentially move in and out of this framework and gain certain advantages by doing so. In considering the ways in which the proposed restriction might or might not be appropriate, the Bureau needed to confront two distinct issues: whether the tolling provision as proposed was properly calibrated and adequate to the task at hand, and which loans in addition to 918 certain non-recourse pawn loans should be identified as improper bridge loans when viewed from within the framework of the conditional exemption. As noted, consumer groups urged the Bureau to expand proposed § 1041.7(d) in two ways: (1) by including any type of loan made by the lender or its affiliate, not just non-covered bridge loans; and (2) by replacing the “tolling” approach with a “reset” approach. As regards the first comment, the Bureau agrees that there is no significant difference between non-covered bridge loans and all other loans when it comes to the potential to use the loan to bridge between conditionally exempt loans and loan sequences, and thus to potentially exacerbate their effects upon the borrower. Accordingly, the Bureau has designed final § 1041.6(d) of the final rule to apply to any loan made by the lender or its affiliate (other than a loan made under § 1041.6 itself, of course). Regarding longer-term loans, in particular, the Bureau has concluded that the prohibition in proposed § 1041.10(e) on lenders making such loans during the 30-day period following a conditionally exempt loan is needed for the reasons set forth in the proposal and reiterated above. Indeed, the fact that the Bureau has decided not to finalize the underwriting requirements on such loans in proposed § 1041.9, and the attendant presumptions in proposed § 1041.10, only heightens the need for this prohibition— which is now incorporated in § 1041.6(d) of the final rule. As regards the second comment, the Bureau generally agrees with the commenters’ concerns about the proposed tolling provision. The Bureau has concluded that merely tolling the cooling-off or re-borrowing periods is an inadequate measure to prevent lengthy debt cycles or bridging between conditionally exempt loans or sequences in an effort to evade the requirements of the rule. Merely tolling the running of the 30-day re-borrowing period or the 30-day coolingoff period for the duration of any loan—including those the proposed rule defined as non919 covered bridge loans—could negate the purpose of the period being tolled because the time periods are intended to run continuously. For example, a non-covered bridge loan made in the middle of the cooling-off period would mean that a consumer would not be in debt for only 15 days at a time, on either end of the non-covered bridge loan, which may be an inadequate period for the consumer’s finances to recover. Similarly, the justification for setting the re-borrowing period at 30 days is undermined where a borrower only has 15 days between a § 1041.6 loan and a bridge loan, on either end. The bridge loan would effectively be a re-borrowing of the prior loan, and the loan after the bridge loan would effectively be a re-borrowing of the bridge loan, if there was only 15 days in between each. Further, the principal step-down would not work as designed if a second or third conditionally exempt loan under § 1041.6 came after an intervening non-covered bridge loan in a higher amount than the prior loan. The Bureau recognizes that the reset approach suggested by consumer groups would be somewhat more protective than the tolling approach in certain respects. However, several of the weaknesses of the tolling approach detailed above likewise apply to the reset approach. In addition, the reset approach would not address the concern animating proposed § 1041.10(e)— which has been intensified by the Bureau’s decision not to finalize the underwriting requirements for covered longer-term loans—that a lender could leverage the consumer’s financial vulnerability and need for funds after having taken out an unaffordable conditionally exempt loan to make a longer-term loan that the consumer otherwise would not have taken, indeed one that would be unaffordable in its own right. Further, the tolling provision would have added considerable complexity to the rule, and for that reason may have been difficult to comply with and enforce. The same would be largely true of a revised provision using the reset approach. 920 For all of these reasons, the Bureau concludes that the most effective means of fully achieving the purposes of proposed § 1041.10(e) and proposed § 1041.7(d)—as well as the simplest means—is a straightforward limitation on any other lending occurring between the specific lender and borrower who had opted in to the § 1041.6 framework by choosing to consummate a conditionally exempt loan during the 30-day re-borrowing and cooling-off periods of § 1041.6. The Bureau also concludes, as discussed above in the discussion of § 1041.6(c), that by prohibiting loans within 30 days of a conditionally exempt loan, the finalized approach will protect the effectiveness of the principal reduction requirements of § 1041.6(b), and will also best serve the purposes of the 30-day re-borrowing and cooling-off periods. The Bureau therefore has reframed § 1041.6(d) to prohibit all loans that may be made within 30 days after a covered short-term loan is made under the exemption, rather than prohibiting covered loans and tolling or resetting time periods during non-covered bridge loans. The final rule provides that the only loan that a lender or its affiliate may make to a borrower, while a loan made under § 1041.6(d) from that lender is outstanding to the borrower or for 30 days thereafter, is a short-term loan that complies with the principal reduction and other provisions of § 1041.6. As was true of both proposed § 1041.10(e) and proposed § 1041.7(d), § 1041.6(d) of the final rule does not apply to all lenders, but only to the lender or affiliate that has made a § 1041.6 loan to the consumer, for essentially the same reasons provided in the proposal with respect to this aspect of proposed § 1041.10(e) and proposed § 1041.7(d). A lender in a non-covered market would not otherwise have a reason or a need to check a registered information system, and thus would be unaware of a prior § 1041.6 loan. This also reduces the impact that § 921 1041.6(d) will have on limiting access to credit that is not used for bridging, but nonetheless falls within the period of a conditionally exempt loan. If, for example, a borrower wants to take out a 5-year installment loan 15 days after he obtains a loan under § 1041.6, the borrower could do so, as long as he did so with a different lender. Moreover, the concerns that animated proposed § 1041.10(e) and are in part the basis for final § 1041.6(d)—that a lender could use an unaffordable loan it had made under § 1041.6 to induce a consumer to take out a different kind of loan—are not present or are present to a much lesser degree if a consumer is considering a loan from a different lender. Two new comments have been added to reflect the revisions to § 1041.6(d). Comment 6(d)-1 explains that while a covered short-term loan made under § 1041.6 is outstanding from a lender to a consumer, and for 30 days thereafter, that lender and its affiliates may only make a covered short-term loan to that borrower if it complies with § 1041.6. The comment also expressly clarifies that the lender and its affiliates may not make any other types of loans to the same borrower during that period. Comment 6(d)-2 includes an example involving a consumer who seeks a loan from a lender during the 30 days after repaying a prior conditionally exempt loan from that lender. The example explains that the rule does not prohibit the lender from making a covered short-term loan under § 1041.6, and clarifies that the consumer could obtain a non-covered installment loan from a lender that is unaffiliated with the original lender. The example also illustrates how the 30-day cooling-off period works by identifying the first date on which the lender or its affiliate could make a non-covered installment loan (or a covered loan under § 1041.5) to the consumer. 922 6(e) Disclosures Proposed Rule In proposed § 1041.7(e), renumbered in this final rule as § 1041.6(e), the Bureau proposed to require a lender to provide disclosures before making the first and third loan in a sequence of conditionally exempt loans under § 1041.6. Under the proposal, the notices in proposed § 1041.7(e)(2)(i) and (e)(2)(ii) would have had to be substantially similar to model forms provided in the proposal. Proposed § 1041.7(e) would have required a lender to provide the notices required under proposed § 1041.7(e)(2)(i) and (e)(2)(ii) before the consummation of a loan. Proposed comment 7(e)-1 would have clarified the proposed disclosure requirements. The proposed disclosures were designed to provide consumers with key information about how the principal amounts and the number of loans in a sequence would be limited for covered short-term loans made under proposed § 1041.7 before they take out their first and third loans in a sequence. The Bureau developed model forms for the proposed disclosures through consumer testing.910 The Bureau believed that the proposed disclosures would, consistent with section 1032(a) of the Dodd-Frank Act, ensure that these costs, benefits, and risks are fully, accurately, and effectively disclosed to consumers. In the absence of the proposed disclosures, the Bureau was concerned that consumers would be less likely to appreciate the risk of taking out a loan with mandated principal reductions or understand the proposed restrictions on conditionally exempt loans that were designed to protect consumers from the harms of unaffordable loan payments. 910 See generally FMG Report, “Qualitative Testing of Small Dollar Loan Disclosures, Prepared for the Consumer Financial Protection Bureau,” at 2–6 (Apr. 2016), available at http://files.consumerfinance.gov/f/documents/Disclosure_Testing_Report.pdf 923 The Bureau believed that it was important for consumers to receive the proposed notices before they would be contractually obligated on a conditionally exempt loan. By receiving the proposed notices before consummation, a consumer could make a more fully informed decision, with greater awareness of the features of such loans, including specifically the limits on taking out more conditionally exempt loans in the near future. Proposed § 1041.7(e)(1), renumbered in this final rule as § 1041.6(e)(1), provided the form of disclosures that would be utilized under proposed § 1041.7. The format requirements generally would have paralleled the format requirements for disclosures related to payment transfers under proposed § 1041.15 (now renumbered as § 1041.9 of the final rule). Proposed § 1041.7(e)(1)(i) would have required that the disclosures be clear and conspicuous. Proposed § 1041.7(e)(1)(ii) would have required that the disclosures be provided in writing or through electronic delivery. Proposed § 1041.7(e)(1)(iii) would have required the disclosures to be provided in retainable form. Proposed § 1041.7(e)(1)(iv) would have required the notices to be segregated from other items and to contain only the information in proposed § 1041.7(e)(2), other than information necessary for product identification, branding, and navigation. Proposed § 1041.7(e)(1)(v) would have required electronic notices to have machine readable text. Proposed § 1041.7(e)(1)(vi) would have required the disclosures to be substantially similar to the model forms for the notices set out under proposed § 1041.7(e)(2)(i) and (ii). Proposed § 1041.7(e)(1)(vii) would have allowed lenders to provide the disclosures that would have been required by proposed § 1041.7(e) in a foreign language, provided that the disclosures must be made available in English upon the consumer’s request. 924 Proposed comment 7(e)(1)(i)-1, renumbered in this final rule as 6(e)(1)(i)-1, clarified that disclosures are clear and conspicuous if they are readily understandable and their location and type size are readily noticeable to consumers. Proposed comment 7(e)(1)(ii)-2, renumbered in this final rule as 6(e)(1)(ii)-2, explained that the disclosures required by proposed § 1041.7(e)(2) may be provided electronically without regard to the Electronic Signatures in Global and National Commerce Act.911 Proposed comment 7(e)(1)(iii)-1, renumbered in this final rule as 6(e)(1)(iii)-1, explained that electronic disclosures are considered retainable if they are in a format that is capable of being printed, saved, or emailed by the consumer. Proposed comment 7(e)(1)(iv)-1, renumbered in this final rule as 6(e)(1)(iv)-1, explained how segregated additional content can be provided to a consumer. Proposed comment 7(e)(1)(vi)-1, renumbered in this final rule as 6(e)(1)(vi)-1, explained the safe harbor provided by the model forms, providing that although the use of the model forms and clauses is not required, lenders using them would be deemed to be in compliance with the disclosure requirement with respect to such model forms. In proposed § 1041.7(e)(2), renumbered in this final rule as § 1041.6(e)(2), the Bureau proposed to require a lender to provide notices to a consumer before making a first and third loan in a sequence of conditionally exempt loans. Proposed § 1041.7(e)(2)(i) would have required a lender before making the first loan in a sequence of conditionally exempt loans to provide a notice. Proposed § 1041.7(e)(2)(ii) would have required a lender before making the third loan in a sequence of conditionally exempt loans to provide another, different notice. More generally, these proposed notices were intended to help consumers understand the availability of conditionally exempt loans in the near future. 911 Also known as the E-Sign Act, 15 U.S.C. 7001 et seq. 925 In proposed § 1041.7(e)(2)(i) the Bureau proposed to require a lender before making the first loan in a sequence of conditionally exempt loans to provide a notice that warns the consumer of the risk of a conditionally exempt loan that is unaffordable and informs the consumer of the Federal restrictions governing subsequent conditionally exempt loans. Specifically, the proposed notice would have warned the consumer not to take the loan if the consumer is unsure whether the consumer can repay the loan amount, which would include the principal and the finance charge, by the contractual due date. In addition, the proposed notice would have informed the consumer, in text and tabular form, of the Federally-required restriction, as applicable, on the number of subsequent loans and their respective amounts in a sequence of conditionally exempt loans. The proposed notice would have been required to contain the identifying statement “Notice of restrictions on future loans,” using that phrase. The other language in the proposed notice would have had to be substantially similar to the language provided in proposed Model Form A-1 in appendix A. Proposed comment 7(e)(2)(i)-1, renumbered in this final rule as 6(e)(2)(i)-1, explained the “as applicable” standard for information and statements in the proposed notice. It stated that, under proposed § 1041.7(e)(2)(i), a lender would have to modify the notice when a consumer is not eligible for a sequence of three covered short-term loans under proposed § 1041.7. The Bureau believed the proposed notice would ensure that certain features of conditionally exempt loans are fully, accurately, and effectively disclosed to consumers in a manner that permits them to understand certain costs, benefits, and risks of such loans. Given that the restrictions on obtaining covered short-term loans under proposed § 1041.7 would be new and conceptually unfamiliar to many consumers, the Bureau believed that disclosing them 926 would be critical to ensuring that consumers understand the restriction on the number of and principal amount on subsequent loans in a sequence of conditionally exempt loans. The Bureau’s consumer testing of the notice under proposed § 1041.7(e)(2)(i) indicated that it aided consumer understanding of the proposed requirements on conditionally exempt loans.912 In contrast, the consumer testing of notices for covered short-term loans made under § 1041.5 indicated that these notices did not improve consumer understanding of the ability-to-repay requirements under proposed § 1041.5.913 Since the notice under proposed § 1041.7(e)(2)(i) would be provided in retainable form, the Bureau believed that the incremental informational value of providing the same or similar notice before the consummation of the second loan in a sequence of conditionally exempt loans would be limited. Proposed § 1041.7(e)(2)(ii), renumbered in this final rule as § 1041.6(e)(2)(ii), would have required a lender before making the third loan in a sequence of conditionally exempt loans to provide a notice that informs a consumer of the restrictions on the new and subsequent loans. Specifically, the Bureau’s proposed notice would state that the new conditionally exempt loan must be smaller than the consumer’s prior two loans and that the consumer cannot take another similar loan for at least another 30 days after repaying the new loan. Under the proposal, the 912 In Round 1 of consumer testing of the notice under proposed § 1041.7(e)(2)(i), “[n]early all participants who saw this notice understood that it was attempting to convey that each successive loan they took out after the first in this series had to be smaller than the last, and that after taking out three loans they would not be able to take out another for 30 days.” FMG Report, “Qualitative Testing of Small Dollar Loan Disclosures, Prepared for the Consumer Financial Protection Bureau,” at 11(Apr. 2016), available at http://files.consumerfinance.gov/f/documents/Disclosure_Testing_Report.pdf. In Round 2 of consumer testing of the notice under proposed § 1041.7(e)(2)(i), “participants . . . noticed and understood the schedule detailing maximum borrowable amounts, and the schedule appeared to influence their responses when asked about the form’s purpose.” Id. at 40. 913 See FMG Report, “Qualitative Testing of Small Dollar Loan Disclosures, Prepared for the Consumer Financial Protection Bureau,” at 9–11, 38–39 (Apr. 2016), available at http://files.consumerfinance.gov/f/documents/Disclosure_Testing_Report.pdf. 927 language in this proposed notice must be substantially similar to the language provided in proposed Model Form A-2 in appendix A. The proposed notice would have to contain the identifying statement “Notice of borrowing limits on this loan and future loans,” using that phrase. The other language in this proposed notice would have to be substantially similar to the language provided in proposed Model Form A-2 in appendix A. The Bureau believed the proposed notice would be necessary to ensure that the restrictions on taking conditionally exempt loans are fully, accurately, and effectively disclosed to consumers. Since several weeks or more may have elapsed since a consumer received the notice under proposed § 1041.7(e)(2)(i), this proposed notice would remind consumers of the prohibition on taking another similar loan for at least the next 30 days. Importantly, it would present this restriction more prominently than it is presented in the notice under proposed § 1041.7(e)(2)(i). The Bureau’s consumer testing of the notice under proposed § 1041.7(e)(2)(ii) indicated that it would aid consumer understanding of the prohibition on taking a subsequent conditionally exempt loan.914 Proposed § 1041.7(e)(3), renumbered in this final rule as § 1041.6(e)(3), proposed to require a lender to provide the notices required under proposed § 1041.7(e)(2)(i) and (ii) before the consummation of a loan. Proposed comment 7(e)(3)-1, renumbered in this final rule as 6(e)(3)-1, explained that a lender can provide the proposed notices after a consumer has 914 In Round 1 of consumer testing of the notice under proposed § 1041.7(e)(2)(ii), “[t]he majority of participants who viewed this notice understood it, acknowledging that it would not be possible to refinance or roll over the full amount of the third loan they had taken out, and that they would have to wait until 30 days after it was paid off to be considered for another similar loan.” FMG Report, “Qualitative Testing of Small Dollar Loan Disclosures, Prepared for the Consumer Financial Protection Bureau,” at 14–15 (Apr. 2016), available at http://files.consumerfinance.gov/f/documents/Disclosure_Testing_Report.pdf. The notice under proposed § 1041.7(e)(2)(ii) was not tested in Round 2. 928 completed a loan application but before the consumer has signed the loan agreement. It further clarified that a lender would not have to provide the notices to a consumer who merely makes an inquiry about a conditionally exempt loan but does not complete an application for this type of loan. Proposed comment 7(e)(3)-2, renumbered in this final rule as 6(e)(3)-2, stated that a lender must provide electronic notices, to the extent permitted by § 1041.7(e)(1)(ii), to the consumer before a conditionally exempt loan is consummated. It also offered an example of an electronic notice that would satisfy the timing requirement. The Bureau believed that it would be important for consumers to receive the proposed notices before they are contractually obligated on a conditionally exempt loan. By receiving the proposed notices before consummation, a consumer could make a more fully informed decision, with an awareness of the restrictions on the current loan and on additional conditionally exempt loans or similar loans in the near future. Comments Received A number of stakeholders commented on the Bureau’s consumer testing process for the model forms. Some commenters believed that the Bureau’s sample size of 28 consumers was too small, noting that the Bureau and other agencies had used larger sample sizes for the qualitative testing of other disclosures (such as the TILA-RESPA integrated disclosure),915 and supplemented them with quantitative testing. These commenters asked the Bureau to clarify that the notices do not need to be exactly the same as the model forms, so that lenders could conduct their own testing. Others claimed that the level of research rigor for the model disclosures was weak as compared to what would be considered a best practice in the industry. One commenter 915 78 FR 79730 (Dec. 31, 2013). 929 criticized both the sample size and the geographical representation of the sample, and recommended that the Bureau remove the model forms from the proposal. This commenter stated that it conducted its own user testing of the “Notice of Restrictions on Future Loans,” a notice that would have been required by § 1041.7(e), with 50 participants, and found that 18 percent understood the table accurately (with 54 percent having a limited understanding and 24 percent who did not understand) and 22 percent had a solid understanding of the purpose of the notice (with 48 percent noting limited knowledge and 30 percent having no knowledge or an inaccurate understanding). The commenter also argued that the Bureau’s use of qualitative testing on its own, without pairing it with quantitative testing, suggested that its findings may not be projectable to the broader population. However, other industry commenters supported the Bureau’s use of a model form. Several consumer groups commented that the proposed disclosures were well designed. But they doubted that disclosures would effectively prevent the harm they perceived as persisting under the exemption. They did support the Bureau’s proposed requirements that disclosures contain machine readable text, be clear and conspicuous, be retainable, be segregated, contain only the specified information, and be substantially similar to the model forms. Industry commenters generally supported the proposal’s approach to electronic disclosures, and urged the Bureau not to adopt a rule requiring email or paper disclosures. Commenters argued that if a borrower chooses to receive disclosures via text, including texts with click-through links, then the borrower should not need email or paper disclosures. The Bureau received a number of comments about the proposed approach to foreign language disclosures. Several commenters argued against requiring foreign language notices 930 (which the Bureau did not propose but did seek comment on) because doing so would impose substantial costs and could involve wide-ranging consequences that deserve thoughtful consideration in a separate rulemaking. Other commenters argued that lenders should offer the model form in the language of the consumer’s preference, or in the language that the lender uses to negotiate the transaction. A consumer group asked the Bureau to go further and prescribe specific contract language in addition to the specific language for disclosures. A legal aid group proposed that the Bureau add a provision that would make the failure to provide any required disclosure or provision of a dissimilar disclosure a deceptive act. A coalition of consumer groups wrote in support of more extensive requirements regarding disclosures, urging the Bureau to go further by: requiring a disclosure for the second loan in a sequence; requiring disclosures at application and just before consummation; requiring paper disclosures for in-person transactions (with electronic disclosures as a supplement); allowing text or mobile disclosures only as supplements to paper or email disclosures because of problems with retainability; imposing a requirement that a URL should be persistent for at least three years after the final payment; imposing a requirement that the full text of a disclosure be provided in an email without a click-through; imposing a requirement that a paper disclosure should be sent if an email is returned; and imposing a requirement that lenders follow E-SIGN requirements, specifically requiring confirmation that borrowers are able to receive and view electronic communications. Final Rule The Bureau is finalizing proposed § 1041.7(e) and all of its subsections as § 1041.6(e) of the final rule with identical subsections. The only differences between proposed § 1041.7(e) and 931 final § 1041.6(e) are numbering changes: the number of the section itself is updated to § 1041.6, and one internal reference to proposed § 1041.7 is replaced with an internal reference to § 1041.6 of the final rule. The Bureau is also finalizing all proposed commentary to proposed § 1041.7(e), again only making renumbering changes. The Bureau continues to believe that the disclosures will, consistent with section 1032(a) of the Dodd-Frank Act, ensure that costs, benefits, and risks associated with § 1041.6 loans are fully, accurately, and effectively disclosed to consumers. The Bureau concludes, based on its considerable experience with consumer testing, that the qualitative user testing process for the model forms and notices is sufficient for purposes of this rule. That is because, unlike the TILA-RESPA model disclosures, the model forms for this rule are relatively short and less complicated. The Bureau contracted with FMG to conduct qualitative user testing of the forms. While the sample size was relatively small—28 test subjects—each subject was given a one-on-one interview with FMG for about an hour. The interviews were conducted in two geographical locations—New Orleans and Kansas City. After the round of testing in New Orleans, Bureau staff used the feedback to improve the model forms before the second round of testing in Kansas City. The Bureau did not conduct quantitative testing, which could have provided some additional information, but the Bureau finds that the testing suffices to show that the disclosures use plain language that is comprehensible to consumers, contains a clear format and design, and succinctly explains the information that must be imparted to the consumer. The commenter that tested the notice of restrictions on future lending, which purportedly found that 18 percent understood the table accurately and 54 percent had a limited understanding, while 22 percent had a solid understanding of the purpose and 48 percent had a 932 limited knowledge of the form’s purpose, does not necessarily discount the efficacy of the model forms. The Bureau does not know whether participants were shown the letters in an appropriate environment and manner, and does not know whether the wording or substance of the questions asked could have contributed to the lower numbers. Participants who did not understand the content of the table may not have had enough of the context to understand the form being tested (in fact, the commenter suggested that the participants did not understand its purpose). In response to comments relating to text message disclosures, the Bureau notes that nothing in § 1041.6(e) prohibits transmission by text. Without being able to review a specific method of delivery, the Bureau cannot opine on whether any specific provision of disclosures via text with a click-through link satisfies the requirements for disclosures in § 1041.6(e)— particularly the requirement of retainability in § 1041.6(e)(iii)—but the Bureau acknowledges that such disclosures could, if correctly administered, satisfy the requirements of § 1041.6(e). In response to the commenter contending that the initial disclosure, if sent by email, could be prevented by a spam filter, the Bureau does not find this to be a valid ground for not finalizing the text of § 1041.6(e). While the Bureau understands that email disclosures may not be feasible for all lenders, it concludes that providing paper disclosures in those instances where companies cannot provide an adequate text or email message notification to all borrowers is necessary or appropriate to ensure that borrowers receive notice of their first scheduled payment—receipt of such notice is particularly important to both borrowers and lenders, as it will begin the repayment cycle. More broadly, the Bureau is not convinced that it is difficult for industry to provide a written or electronic disclosure to borrowers before the borrower enters a loan agreement. After all, the Bureau would expect that the lender would need to transmit or 933 provide a loan agreement and TILA disclosure to the borrower through some means; and the lender could use those means to provide the disclosure. As proposed, the Bureau is not requiring non-English disclosures; instead, it is finalizing the rule as proposed, which merely allows non-English disclosures. Certain of the Bureau’s rules, like its remittance rule,916 require disclosures in foreign languages in certain circumstances. The Bureau continues to view disclosures in languages other than English as a positive development in all markets for consumer financial products or services, where the customer base has become increasingly more diverse. The Bureau is not, however, prepared to make non-English disclosures mandatory at this time with respect to these forms. The Bureau so concludes for several reasons, including its recognition that the current final rule will involve a significant amount of implementation work, including the work needed to design and implement the disclosures in English. The Bureau is making the judgment not to add required foreign language notices at this time, but may consider supplemental rulemakings or model forms in the future when industry has fewer regulatory adjustments to manage and has developed more experience with the English-language forms. In response to commenters asking the Bureau to go further and prescribe specific contract language in addition to the specific language for disclosures, the Bureau concludes that a loan made pursuant to any contract which creates terms that are incompatible with the requirements of § 1041.6 would disqualify the loan from coverage under the § 1041.6 exemption. Accordingly, the Bureau believes there would be minimal benefit to prescribing specific contract language, 916 12 CFR 1005.31(g). 934 and that doing so would restrict the ability of individual lenders to comply with specific requirements of local contract law. In response to commenters proposing that the Bureau add a provision to the rule that would make failure to provide any required disclosure or provision of a dissimilar disclosure a deceptive act, the Bureau concludes that such a provision is unnecessary. A lender that fails to make required disclosures would already be in violation of the rule, and labeling that violation as deceptive would not add anything to the lender’s liability. The Bureau does not find that it needs to require a notice before the second loan. That would be inconsistent with the more general approach the Bureau is taking in finalizing this rule, which is to attempt to make the rule more streamlined and capable of being administered more easily and practically. The payment notices, for example, now only require a notice before the first withdrawal and any unusual withdrawals, under the theory that borrowers could refer back to the initial notice. Similarly, borrowers here could refer back to the notice sent before the first loan was made under § 1041.6 of the final rule. The Bureau also finds insufficient evidence to support the claim that additional prescriptive requirements are necessary to ensure that borrowers receive electronic or written notices in any particular manner. Unlike with the payment notices, the Bureau concludes that the risk associated with borrowers missing the notice is lower. The payment notices are intended to warn borrowers of an impending event—thus, borrowers are not engaged in a decision at the very moment when those notices are sent. For this reason, the Bureau has provided further requirements for those notices to ensure they are received. However, here, the Bureau expects that the notices associated with making loans under § 1041.6 would be provided as part of the 935 pre-loan package when the borrower is inquiring about the contours of the transaction. In order to take out the loan, the borrower already must engage with that pre-loan package, so the Bureau concludes that a more permissive approach to transmission is sufficient for these specific notices. Subpart C—Payment Practices Overview of the Proposal In the proposed rule, the Bureau proposed to identify it as an unfair and abusive act or practice for a lender to attempt to withdraw payment from a consumer’s account in connection with a covered loan after the lender’s second consecutive attempt to withdraw payment from the account has failed due to a lack of sufficient funds, unless the lender obtains the consumer’s new and specific authorization to make further withdrawals from the account. To avoid committing this unfair and abusive practice, a lender would have to cease attempting to withdraw payments from the consumer’s account or obtain a new and specific authorization to make further withdrawals. Using the Bureau’s authority in section 1031 of the Consumer Financial Protection Act, the proposed rule would have prevented the unlawful practice by prohibiting further payment withdrawal attempts after two unsuccessful attempts in succession, except when the lender has obtained a new and specific authorization for further withdrawals. It also included requirements for determining when the prohibition on further payment withdrawal attempts has been triggered and for obtaining a consumer’s new and specific authorization to make additional withdrawals from the consumer’s account. The predicate for the proposed identification of an unfair and abusive act or practice that the Bureau identified in the proposed rule—and thus for the prevention requirements—was a set 936 of preliminary findings with respect to certain payment practices for covered loans and the impact on consumers of those practices. Those preliminary findings, the comments received on them, and the Bureau’s responses to the comments are addressed below in Market Concerns— Payments. The proposed rule would have provided a different set of interventions based on the Bureau’s disclosure authority found in section 1032, which would have required lenders to provide a notice to a consumer prior to initiating a payment withdrawal from the consumer’s account. It also proposed to require lenders to provide a notice alerting consumers to the fact that two consecutive payment withdrawal attempts to their accounts have failed—thus triggering operation of the new authorization requirements—so that consumers can better understand their repayment options and obligations in light of the severely distressed condition of their accounts. Market Concerns—Payments As the Bureau laid out in the proposal, at the time of loan origination, it is a common practice among many lenders to obtain authorization to initiate payment withdrawal attempts from the consumer’s transaction account. Such authorization provides lenders with the ability to initiate withdrawals without further action from the consumer. Like other industries that commonly use such authorizations for future withdrawals, consumers and lenders have found that they can be a substantial convenience for both parties. However, they also expose the consumer to a range of potential harms. Indeed, Congress has recognized that such authorizations can give lenders a special kind of leverage over borrowers, for instance by 937 prohibiting in EFTA the conditioning of credit on the consumer granting authorizations for a series of recurring electronic transfers over time.917 This section reviews the available evidence on the outcomes that consumers experience when lenders obtain and use the ability to initiate withdrawals from consumers’ accounts to secure payments on covered loans, including the comments that were submitted on the proposed rule. As detailed below, the available evidence reinforces the Bureau’s conclusion that despite various regulatory requirements, lenders in this market are using their ability to initiate payment withdrawals in ways that harm consumers. Moreover, the Bureau finds that, as a practical matter, consumers have little ability to protect themselves from the injuries caused or likely caused by these practices, and that private network attempts to restrict these behaviors are limited in various ways. The Bureau’s research with respect to payment practices focused on online payday and payday installment loans, where payment attempts generally occur through the ACH network and thus can be readily tracked at the account and lender level. Other publicly available data and the Bureau’s enforcement experience indicate that returned payments likewise occur with great frequency in the storefront payday market; indeed, a comparison of this data with the Bureau’s findings suggests that the risks to consumers with respect to failed payments may be as significant or even greater in the storefront market than in the online market. The Bureau reviewed the available evidence, which can be summarized as follows: • Lenders in these markets often take broad, ambiguous payment authorizations from consumers and vary how they use these authorizations, thereby increasing the risk that 917 Electronic Fund Transfer Act, 15 U.S.C. 1693k(1); Regulation E, 12 CFR 1005.10(e). 938 consumers will be surprised by the amount, timing, or channel of a particular payment and will be charged overdraft or NSF fees as a result. Commenters took both sides on these factual points, with industry commenters arguing that the Bureau had overstated the extent of the problems and any lack of understanding on the part of consumers, and consumer groups arguing that problems exist and cause harm that often is not understood by consumers. • When a particular withdrawal attempt fails, lenders in these markets often make repeated attempts at re-presentment, thereby further multiplying the fees imposed on consumers. Some commenters said that the Bureau had overstated the occurrence of re-presentments, arguing that the Bureau’s reliance on data from 2012 was improper in light of recent developments that may have driven down re-presentment rates; others disagreed.918 • These cumulative practices contribute to return rates that vastly exceed those in other markets, substantially increasing consumers’ costs of borrowing, their overall financial difficulties, and the risk that they will lose their accounts. Here again, commenters offered perspectives on both sides of these factual issues, with critics disputing the fact and the evidence that return rates here are disproportionately higher than in other markets and taking issue with the extent of the effect on consumers having their accounts closed, and others providing additional evidence that return rates were in fact disproportionately high. 918 Note that in this rule preamble, the Bureau uses “presentment,” and “re-presentment” to refer to payment attempts and payment re-attempts. Technically, these terms are often reserved for ACH payment attempts only. However, in the context of this rule, which is applicable across all payment methods, the Bureau uses the terms interchangeably with other types of payment withdrawals. 939 • Consumers have little practicable ability to protect themselves from these practices. This point was sharply disputed by industry and trade association commenters, with others such as consumer groups and some research organizations offering support for this point. • Private network protections necessarily have limited reach and impact, and are subject to change. This point was also disputed by commenters who argued that the private networks do provide appropriate and sufficient protections, while others strongly disagreed and supported the preliminary views as stated by the Bureau. a. Multiple Presentments Varied by Timing, Frequency, and Amount of Payments As discussed in the proposal and in the Background section, obtaining authorization to initiate withdrawals from consumers’ transaction accounts is a standard practice among payday and payday installment lenders. Lenders often control the parameters of how these authorizations are used. Storefront payday lenders typically obtain a post-dated paper check signed by the consumer, which in fact can be deposited before the date listed and can be converted into an ACH withdrawal. Online lenders typically obtain bank account information and authorizations to initiate ACH withdrawals from the consumer’s account as part of the consumer’s agreement to receive the funds electronically.919 Many lenders obtain authorization for multiple payment methods, such as taking a post-dated check along with the consumer’s ACH authorization or debit card information. Banks and credit unions often have additional 919 Although, as noted above, the EFTA and Regulation E prohibit lenders from conditioning credit on a consumer “preauthorizing” recurring electronic fund transfers, in practice online payday and payday installment lenders are able to obtain such authorizations from consumers for almost all loans through various methods. Lenders are able to convince many consumers that advance authorizations will be more convenient, and some use direct incentives such as by making alternative methods of payment more burdensome, changing APRs, or providing slower means of access to loan proceeds for loans without preauthorized withdrawals. The Bureau is not addressing in this rulemaking the question of whether any of the practices described are consistent with the EFTA and Regulation E. 940 payment channel options, such as using internal transfers from a consumer’s deposit account to collect loan payments. One commenter provided additional information on internal bank transfers, explaining that, when initiating internal bank transfers, financial institutions do not necessarily coordinate internally so that the initiator knows the amount of funds in a consumers’ account. Generally, commenters did not take issue with this account of the types of payment methods obtained by lenders. Once lenders have obtained the authorizations, payday and payday installment lenders frequently execute the withdrawals in ways that consumers do not expect. In some cases, these actions may violate authorizations, contract documents, Federal and State laws, and/or private network rules, and in other cases they may exploit the flexibility provided by these sources, particularly when the underlying contract materials and authorizations are broadly or vaguely phrased. The unpredictability for consumers can be exacerbated by the fact that lenders often also obtain authorizations to withdraw varying amounts up to the full loan amount, in an apparent attempt to bypass EFTA notification requirements that would otherwise require notification of transfers of varying amounts.920 The Bureau’s study on online payday and payday installment loan payments shows how common multiple payment presentments are.921 In the study, the Bureau reviewed presentment activity relating to online payday and payday installment loans using checking account files from several large depository institutions. The data was from 2011-2012. The study showed that 920 See part II.D for a more detailed discussion of the flexibility provided under laws and private network rules and other lender practices with regard to obtaining initial authorizations. 921 CFPB, Online Payday Loan Payments (April 2016), available at http://files.consumerfinance.gov/f/201604_cfpb_online-payday-loan-payments.pdf. 941 lenders re-presented after one failed attempt 75 percent of the time, re-presented after the second failed attempt 66 percent of the time, re-presented after the third failed attempt 50 percent of the time, and re-presented after the fourth failed attempt 29 percent of the time.922 The data also showed that re-presentments tend to come much sooner than do withdrawal attempts that follow a successful payment.923 Industry commenters disputed the Bureau’s point that withdrawals are executed in ways that consumers do not expect, or at least asserted that the Bureau failed to present sufficient evidence to support this point. Part of this criticism took issue with the Bureau’s partial reliance on confidential supervisory data to support its position, which some commenters viewed as improper. This line of comments echoed a broader concern from several commenters, who argued that it was improper for the Bureau to rely on confidential data in the rulemaking. Some commenters argued that data from 2012 is no longer indicative of current practices, given several changes in the market since that time in light of enforcement actions and adjustments to the NACHA Rules. They also argued that the data may have been based on only a few lenders, or lenders that were no longer in the market. Commenters further argued that the Bureau did not establish that these negative payment practices extended to all lenders, and should not have lumped together online and storefront lenders, unlicensed and State-licensed lenders, and bank products with non-bank products. On the other side, consumer groups and some research organizations submitted comments and data in support of the Bureau’s points, providing consumer stories about payment experiences and citing several reports that are publicly available 922 923 Id. at 14. Id. at 16. 942 on overdraft and NSF fees caused by lender re-presentments and irregular debiting of consumer accounts. The Bureau also does not agree that it is improper to cite supervisory information in the rulemaking process; this is information the Bureau collects as part of its lawful and authorized activities, and it provides insight into the issues addressed here. Data from the Bureau’s published reports were collected through its supervision function, and the Bureau’s regulations protect confidential supervisory information from disclosure.924 Courts have held that an agency can rely on confidential information in its rulemaking so long as the agency discloses information to allow interested parties to comment on the methodology and general data.925 The Bureau disclosed how it obtained the data, the methodologies used to analyze the data, the number of accounts reviewed, characteristics about the accounts reviewed, and the results of the various studies.926 For example, in the Bureau’s payments report, most applicable to this section, the Bureau disclosed the number of accounts reviewed (19,685) and the methodology and results in a 25-page report.927 That was enough information to allow commenters to adequately comment on the proposed rule. The Bureau believes that more detail could have revealed the identity of depository institutions, running counter to the Bureau’s rules governing confidential supervisory information. 924 12 U.S.C. 5512(c)(6)(A); 12 CFR Part 1070. See NRDC v. Thomas, 805 F.2d 410, 418 n.13 (D.C. Cir. 1986); see also Riverkeeper Inc. v. EPA, 475 F.3d 83, 112 (2d Cir. 2007) (Sotomayor, J.); rev’d on other grounds, 556 U.S. 208 (2009). 926 For a summary of the Bureau’s reports in this market, see CFPB, Payday Loans, Auto Title Loans, and High-cost Installment Loans: Highlights from CFPB Research (June 2, 2016), available at http://files.consumerfinance.gov/f/documents/Payday_Loans_Highlights_From_CFPB_Research.pdf. 927 CFPB, Online Payday Loan Payments (April 2016), available at http://files.consumerfinance.gov/f/201604_cfpb_online-payday-loan-payments.pdf. 925 943 The Bureau continues to adhere to the view that its study based on 2012 data is relevant. Commenters were very concerned about impacts of the NACHA same-day ACH program, the impact of more recent enforcement actions, and more recent innovations like ApplePay, arguing that more recent market developments render the 2012 data stale. It is true that NACHA has revised some of its rules, and provided more explicit guidance on others. The NACHA Rule most relevant to lender payment presentments—the reinitiation limit of a total of three presentments per entry—was already in place during the sample period, though NACHA has since provided further guidance on that rule. Various enforcement actions relating to problematic use of payment authorizations (or lack thereof) by payday lenders—including various cases pursued by the FTC—had become public before the 2012.928 It is also true that various enforcement actions have come after,929 but it is the Bureau’s common experience that industry often does not react uniformly to the Bureau’s enforcement actions. Despite preexisting enforcement actions, the NACHA reinitiation cap, other NACHA Rules about authorizations, and Regulation E requirements, the Bureau observed a high amount of returned presentments that were causing harm to consumers. Even if industry has stopped or lessened the 928 See, e.g., Press Release, FTC (Aug. 1, 2011), FTC Charges Marketers with Tricking People Who Applied for Payday Loans; Used Bank Account Information to Charge Consumers for Unwanted Programs, available at https://www.ftc.gov/news-events/press-releases/2011/08/ftc-charges-marketers-tricking-people-who-appliedpayday-loans; Press Release, FTC, FTC Obtains Court Order Halting Internet Payday Lenders Who Failed to Disclose Key Loan Terms and Used Abusive and Deceptive Collection Tactics (Feb. 23, 2009), available at https://www.ftc.gov/news-events/press-releases/2009/02/ftc-obtains-court-order-halting-internet-payday-lenderswho. 929 See, e.g., Press Release, Bureau of Consumer Fin. Prot., CFPB Takes Action Against Moneytree for Deceptive Advertising and Collection Practices (Dec. 16, 2016), https://www.consumerfinance.gov/about-us/newsroom/cfpbtakes-action-against-moneytree-deceptive-advertising-and-collection-practices/; Press Release, Bureau of Consumer Fin. Prot., CFPB Orders EZCORP to Pay $10 Million for Illegal Debt Collection Tactics (Dec. 16, 2015), available at http://www.consumerfinance.gov/newsroom/cfpb-orders-ezcorp-to-pay-10-million-for-illegaldebt-collection-tactics/; Press Release, Bureau of Consumer Fin. Prot., CFPB Takes Action Against Online lender for Deceiving Borrowers (Nov. 18, 2015), available at https://www.consumerfinance.gov/about-us/newsroom/cfpbtakes-action-against-online-lender-for-deceiving-borrowers/. 944 prevalence of problematic payment practices since the report sample period—a claim that the Bureau did not receive any evidence on and is purely speculative—consumer harm from repeated re-presentments continues to be of concern to the Bureau. Furthermore, as some commenters acknowledged, recent changes in the market (such as the NACHA return rate inquiry threshold) do not apply to all payment channels and lenders may be continuing problematic practices through other payment channels, like remotely created checks. Moreover, the Bureau continues to receive complaints on payment practices. Some commenters raised that NACHA has passed a 15 percent return rate inquiry threshold, which allows NACHA to request information from merchants who have high return rates, and that NACHA issued guidance to reiterate the two re-presentment threshold. For reasons discussed below, the Bureau believes that there are still significant risks to consumers despite these rule changes and clarifications. Even if this inquiry threshold has affected ACH payment practices, NACHA Rules do not apply to other types of payments. As for the 2014 clarification regarding NACHA’s re-presentment cap, even assuming that clarification significantly impacted compliance rates for the pre-existing rule, there are a number of ways for lenders to avoid the cap, the cap allows more re-presentments than this rule, and again, it only applies to ACH and not other payment methods. NACHA itself raised concerns that lenders are shifting towards other payment methods when they tightened the restrictions—suggesting that the practices that the NACHA Rules were trying to address may have shifted off of the ACH network. As for the makeup of the participants included in the study, the participant with the largest amount of ACH transactions accounted for 14 percent of the transactions, while the next 945 largest accounted for six percent. Given the high number of transactions and that individual participants accounted for a relatively small share of the transactions, the Bureau believes that it is unlikely the overall results of its 2012 study would be primarily driven by potential departure of any one participant from the market. More generally, the commenters only questioned whether the data is still relevant as to the current prevalence of lenders making multiple repeated payment presentments. They did not suggest that the practice has ceased entirely or that the likelihood that a payment attempt would succeed has been impacted by new NACHA Rules or intervening enforcement actions. Thus the Bureau does not find any reason to conclude that the last few years have cast in doubt the relevance of those aspects of its study. The Bureau acknowledges that the payments report was based on online payday and payday installment loans only, and did not include loans by storefronts or depository institutions. The study, however, is informative of what occurs when a lender re-presents multiple times, and data from other sources—including public enforcement actions about depository institution practices, public filings for storefront lenders, and industry data about return rates—shows that these lenders have outlier payment practices. The Bureau believes that this information shows that lenders of loans covered by this rule are more likely to engage in harmful payment practices. The data and analysis that the Bureau presented in the proposal is further bolstered by the studies cited by other commenters such as consumer groups and other research organizations. One published study on checking account activity showed that one-third of payday borrowers experienced at least one incident in which their checking account was overdrawn on the same day that the payday lender withdrew a payment, triggering one or more fees, even where the 946 payment withdrawal itself succeeded.930 Nearly half of them incurred an overdraft or NSF fee in the two weeks after a payday loan transaction. A 2013 report found that 27 percent of payday borrowers said that a payday lender making a withdrawal from their bank account caused an overdraft.931 Among storefront borrowers, 23 percent had this experience while 46 percent of online borrowers reported that a payday lender’s withdrawal caused an overdraft.932 The same study went on to note that while these borrowers may choose payday loans in order to avoid overdrafts, a finding consistent with an earlier national survey which found that 90 percent of those who overdrew their account did so by mistake, many end up paying both payday loan and overdraft fees. Another national survey showed that 22 percent of borrowers reported closing their checking accounts or having them closed by the bank in connection with an online payday loan.933 Going back to the discussion in the proposal, these payment practices increase the risk that the payment attempt will be made in a way that triggers fees on a consumer’s account. Unsuccessful payment attempts typically trigger bank fees. According to deposit account agreements, banks charge an average NSF fee of approximately $34 for returned ACH and check payments.934 Some prepaid card providers charge fees for returned or declined payments.935 930 Center for Responsible Lending, Payday Mayday: Visible and Invisible Payday Lending Defaults (March 31, 2015), available at http://www.responsiblelending.org/research-publication/payday-mayday-visible-and. 931 The PEW Charitable Trusts, Payday Lending in America: Report 2, How Borrowers Choose and Repay Payday Loans, p. 35 (Feb. 2013), available at http://www.pewtrusts.org/~/media/assets/2013/02/20/pew_choosing_borrowing_payday_feb2013-(1).pdf. 932 Id. 933 The PEW Charitable Trusts, Payday Lending in America: Report 4, Harmful Practices in Internet Payday Lending, p. 16 (Oct. 2014). 934 CFPB Study of Overdraft Programs White Paper, at 52. 935 There does not appear to be a standard charge for returned and declined payments by prepaid card providers, though the fees currently appear to be lower than those on depository accounts. The Bureau has observed fees ranging from 45 cents to $5. 947 Even if the payment goes through, the payment may exceed the funds available in the consumer’s account, thereby triggering an overdraft fee, which also averages approximately $34, and in some cases “extended” overdraft fees ranging from $5 to $38.50, if the consumer is unable to clear the overdraft within a specified period of time.936 These failed payment fees charged to the consumer’s deposit account may be exacerbated by returned payment fees and late fees charged by lenders, since many lenders also charge a returned-item fee for any returned check or returned electronic payment.937 The Bureau noted in the proposal that some depository institutions have charged overdraft and NSF fees for payments made within the institutions’ internal systems, including a depository institution that charged overdraft and NSF fees on payments related to its own small-dollar loan product.938 The commenters generally did not dispute that attempted withdrawals generate these kinds of fees to consumers, though some said 936 CFPB Study of Overdraft Programs White Paper. Some extended overdraft fees are charged repeatedly if the overdraft is not cleared. 937 See, e.g., ACE Cash Express, Loan Fee Schedule—Texas, available at https://www.acecashexpress.com/~/media/Files/Products/Payday/Internet/Rates/TX_FeeSchedule.pdf (last visited May 18, 2016) (charging $30 “for any returned check, electronic payment, or other payment device”); Cash America, Rates and Fees—Texas, available at http://www.cashamerica.com/LoanOptions/CashAdvances/RatesandFees/Texas.aspx (last visited May 18, 2016) (“A $30 NSF charge will be applied for any returned payment.”); Advance America 2011 Annual Report (Form 10-K), at 8 (“Fees for returned checks or electronic debits that are declined for non-sufficient funds (‘NSF’) vary by State and range up to $30, and late fees vary by State and range up to $50. For each of the years ending December 31, 2011 and 2010, total NSF fees collected were approximately $2.9 million and total late fees collected were approximately $1 million and $0.9 million, respectively.”); Mypaydayloan.com, FAQs, https://www.mypaydayloan.com/faq#loancost (last visited May 17, 2016) (“If your payment is returned due to NSF (or Account Frozen or Account Closed), our collections department will contact you to arrange a second attempt to debit the payment. A return item fee of $25 and a late fee of $50 will also be collected with the next debit.”); Great Plains Finance, Installment Loan Rates, https://www.cashadvancenow.com/rates.aspx) (last visited May 16, 2016) (explaining returned payment fee of $25 and, for payments more than 15 days late, a $30 late fee). 938 See, e.g., CFPB Consent Order, Regions Bank, CFPB No. 2015-CFPB-0009 (Apr. 28, 2015), available at http://files.consumerfinance.gov/f/201504_cfpb_consent-order_regions-bank.pdf (finding that Regions charged overdraft and NSF fees with its deposit advance product, despite stating that it would not do so after a change in policy. Specifically, if the bank collected payment from the consumer’s checking account and the payment was higher than the amount available in the account, it would cause the consumer’s balance to drop below zero. When that happened, the bank would either cover the transaction and charge an overdraft fee, or reject its own transaction and charge an NSF fee.), available at http://files.consumerfinance.gov/f/201504_cfpb_consent-order_regionsbank.pdf. 948 that if the issue is the high fees that are charged, then the Bureau should pursue that problem separately rather than by adopting this rule. Despite these potential risks to consumers, many lenders vary the timing, frequency, and amount of payment attempts over the course of the lending relationship. For example, the Bureau has received a number of consumer complaints about lenders initiating payments before the due date, sometimes causing the borrower’s accounts to incur NSF or overdraft fees. The Bureau has received consumer complaints about bank fees triggered when lenders initiated payments for more than the scheduled payment amount. The Bureau is also aware of payday and payday installment lender policies that vary the days on which a payment is initiated based on prior payment history, payment method, and predictive products provided by third parties. Bureau analysis of online loan payments shows differences in how lenders space out payment attempts and vary the amounts sought in situations when a payment attempt has previously failed.939 Same-Day Attempts The Bureau also noted in the proposal that some lenders make multiple attempts to collect payment on the same day, contributing to the unpredictable nature of how payment attempts will be made and further exacerbating fees on consumer accounts. For example, the Bureau has observed storefront940 and online payday and payday installment lenders that, as a matter of course, break payment attempts down into multiple attempts on the same day after an initial attempt fails. This practice has the effect of increasing the number of NSF or overdraft 939 CFPB Online Payday Loan Payments, at 16-17 figs.2-3. See Consent Order, EZCORP, CFPB No. 2015–CFPB–0031 (Dec. 16, 2015), available at http://files.consumerfinance.gov/f/201512_cfpb_ezcorp-inc-consent-order.pdf. 940 949 fees for consumers because, in most cases when the account lacks sufficient funds to pay the balance due, attempts will trigger NSF or overdraft fees.941 In the Bureau’s analysis of ACH payments submitted by online payday lenders, approximately 35 percent 942 of the payments were attempted on the same day as another payment attempt. This includes situations in which a lender makes three attempts in one day (four percent of payments observed) and four or more attempts in one day (two percent of payments observed). The most extreme practice the Bureau has observed was a lender who attempted to collect payment from a single account 11 times in one day. The Bureau also has received consumer complaints about lenders making multiple attempts to collect in one day, including an instance of a lender reported to have made nine payment attempts in a single day. When multiple payment requests are submitted to a single account on the same day by an online payday lender, the payment attempts usually all succeed (76 percent) or all fail (21 percent), leaving only three percent of cases where one but not all attempts succeed.943 In other words, multiple presentments are seven times more likely to result in multiple NSF events for the consumer than they are to result in a partial collection by the lender. Re-presentment The Bureau also finds that when a lender’s presentment or multiple presentments on a single day fail, online payday lenders typically repeat the attempt to collect payment multiple 941 With the exception that overdraft fees cannot be charged on one-time debit card transactions when a borrower does not opt in. 12 CFR 1005.17. 942 CFPB Online Payday Loan Payments, at 20 tbl.3. 943 Id. at 21 tbl.4. 950 times on subsequent days.944 According to the Bureau’s analysis of ACH payments, 75 percent of ACH payments presented by online payday lenders that initially fail are re-presented by the lender.945 Because six percent of initial payments originally fail, the result is that four and half percent of all initial payments had an accompanying re-presentment. Of those re-presentments, 70 percent fail, and after the second failed attempt, 66 percent of failed payments are represented. That means a little over two percent of all initial payments involved three presentments (this rule would cut off the third presentment). Of these third re-presentments, 73 percent fail, and 50 percent are re-presented after three failures. Consumers have complained to the Bureau that lenders attempt to make several debits on their accounts within a short period of time, including one consumer who had taken out multiple loans from several online payday lenders and reported that the consumer’s bank account was subject to 59 payment attempts over a two-month period.946 Online payday lenders appear to make a second payment attempt more quickly after a failed payment than after a successful payment. According to Bureau analysis, 60 percent of payment attempts following a failed payment came within one to seven days of the initial failed attempt, compared with only three percent of payment attempts following a successful 944 See, e.g., First Cash Fin. Servs., 2014 Annual Report (Form 10-K), at 5 (Feb. 12, 2015), available at https://www.sec.gov/Archives/edgar/data/840489/000084048915000012/fcfs1231201410-k.htm (explaining that provider of online and storefront loans subsequently collects a large percentage of returned ACH and check payments by redepositing the customers’ checks, ACH collections, or receiving subsequent cash repayments by the customers); CashNet USA, FAQs, https://www.cashnetusa.com/faq.html (last visited Dec. 18, 2015) (“If the payment is returned for reason of insufficient funds, the lender can and will re-present the ACH Authorization to your bank”). 945 CFPB Online Payday Loan Payments, at 14. In the CFPB analysis, any payment attempt following a failed payment attempt is considered a “re-presentment.” Failed requests submitted on the same day are analyzed separately from re-presentments submitted over multiple days. 946 This consumer reported that their bank account was ultimately closed with charges of $1,390 in bank fees. 951 payment.947 The Bureau observed a lender that, after a returned payment, made a payment presentment every week for several weeks. In addition to deviations from the payment schedule, some lenders adopt other divergent practices to collect post-failure payments. For example, the Bureau preliminarily found in the proposal that after an initial failure, one storefront payday and payday installment lender had a practice of breaking an ACH payment into three smaller pieces on the consumer’s next payday: one for 50 percent of the amount due, one for 30 percent of the amount due, and one for 20 percent of the amount due.948 Approximately 80 percent of these smaller attempts resulted in all three presentments being returned for non-sufficient funds, thus triggering multiple NSF fees. Some commenters suggested that they believe the Bureau’s points about same-day attempts and re-presentment were overstated. For example, they cited the Bureau’s data showing a high level of storefront payment failures by ACH transfer failures and bounced checks, and suggested that these figures did not take sufficient account of other cash transactions that were completed successfully. It is true that many payday loan payments are made in cash, and so not implicated by this rule. The Bureau’ study also focused on only online payday and payday installment lenders, which do not take cash payments. Online payday and payday installment lenders continue to have high outlier return rates despite having all payments included in the denominator. The Bureau believes, however, that many cash transactions are likely to come from the population of consumers who would have funds in their accounts if instead the only method of payment were ACH (as in the studied online payday markets), and many would not 947 CFPB Online Payday Loan Payments, at 16. See Consent Order, EZCORP, CFPB No. 2015–CFPB–0031 (Dec. 16, 2015), available at http://files.consumerfinance.gov/f/201512_cfpb_ezcorp-inc-consent-order.pdf. 948 952 come out of the population for which a payment withdrawal fails (because we know those consumers do have the funds to cover a payment). The Bureau received a number of comments, including some from industry, asserting that lenders continue to engage in making repeat attempts to debit payments from consumer accounts. b. Cumulative Impacts These practices among payday and payday installment lenders have substantial cumulative impacts on consumers. Industry analyses, outreach, and Bureau research suggest that the industry is an extreme outlier with regard to the rate of returned items. As a result of payment practices in these industries, consumers suffer significant NSF, overdraft, and lender fees that substantially increase financial distress and the cumulative costs of their loans. Outlier Return Rates Financial institution analysis and Bureau outreach indicate that the payday and payday installment industry is an extreme outlier with regard to the high rate of returned items generated. These returns are most often for non-sufficient funds, but also include transactions that consumers have stopped payment on or reported as unauthorized. The high rate of returned payment attempts suggests that the industry is causing a disproportionate amount of harm relative to other markets.949 949 High return rates for non-sufficient funds may also be indicative of lenders’ problematic authorization practices. In developing its rules to monitor overall ACH return rates, NACHA explained: Moreover, while some level of Returns, including for funding-related issues such as insufficient funds or frozen accounts, may be unavoidable, excessive total Returns also can be indicative of problematic origination practices. For example, although some industries have higher average return rates because they deal with consumers with marginal financial capacity, even within such industries there are outlier originators whose confusing authorizations result in high levels of Returns for insufficient funds because the Receiver did not even understand that s/he was authorizing an ACH transaction. Although such an Entry may be better characterized as “unauthorized,” as a 953 A major financial institution has released analysis of its consumer depository account data to estimate ACH return rates for payday lenders, including both storefront and online companies.950 In a 2014 analysis of its consumer account data, the institution found that industry lenders had an overall return rate of 25 percent for ACH payments.951 The institution observed individual lender return rates ranging from five percent to almost 50 percent. In contrast, the average return rate for debit transactions in the ACH network across all industries was just 1.36 percent. Among individual industries, the industry with the next highest return rate was cable television at 2.9 percent, then mobile telephones at 1.7 percent, insurance at 1.2 percent, auto and practical matter it may be returned for insufficient funds before a determination regarding authorization can be made. NACHA, Request for Comment and Request for Information—ACH Network Risk and Enforcement Topics, Rule Proposal Description, at 3 (Nov. 11, 2013), available at https://www.shazam.net/pdf/ach_networkRisk_propRulesDesc_1113.pdf (last visited May 17, 2016). See also Federal Financial Institutions Examinations Council (“FFIEC”), Bank Secrecy Act/Anti-Money Laundering Exam Manual, at 237 (2014), available at https://www.ffiec.gov/bsa_aml_infobase/documents/BSA_AML_Man_2014_v2.pdf (“High levels of RCCs and/or ACH debits returned for insufficient funds or as unauthorized can be an indication of fraud or suspicious activity. Therefore, return rate monitoring should not be limited to only unauthorized transactions, but include returns for other reasons that may warrant further review, such as unusually high rates of return for insufficient funds or other administrative reasons.”); FDIC, Financial Institution Letter FIL-3-2012, Payment Processor Relationships, at 5 (rev’d July 2014), available at https://www.fdic.gov/news/news/financial/2012/fil12003.pdf (“Financial institutions that initiate transactions for payment processors should implement systems to monitor for higher rates of returns or charge backs and/or high levels of RCCs or ACH debits returned as unauthorized or due to insufficient funds, all of which often indicate fraudulent activity.”). 950 JP Morgan Chase is one of the largest banks in the country, with $2.4 trillion in assets and an average of $200 billion in consumer checking accounts. See JP Morgan Chase, About Us, https://www.jpmorganchase.com/corporate/About-JPMC/about-us.htm (last visited Mar. 17, 2015); JP Morgan Chase & Co., Annual Report 2014 (2015), available at http://files.shareholder.com/downloads/ONE/1717726663x0x820066/f831cad9-f0d8-4efc-9b68f18ea184a1e8/JPMC-2014-AnnualReport.pdf. 951 Monitoring for Abusive ACH Debit Practices, Presentation by Beth Anne Hastings of JP Morgan Chase at Spring 2014 NACHA Conference in Orlando, FL (Apr. 7, 2014). This RDFI analysis included returns due to non-sufficient funds, stop-payment orders, and unauthorized activity; administrative returns were not included. However, most of these returns were triggered by non-sufficient funds; lenders generally had an unauthorized return rate below 1 percent. See also First Cash Fin. Servs., 2014 Annual Report (Form 10-K), at 5 (“Banks return a significant number of ACH transactions and customer checks deposited into the Independent Lender’s account due to insufficient funds in the customers’ accounts.”) (discussion later in the document indicates that the CSO section covers both online and storefront loans). 954 mortgage at 0.8 percent, utilities at 0.4 percent, and credit cards at 0.4 percent.952 Clearly, the numbers for the kinds of loans covered under this rule are so high as to contrast dramatically with consumer’s experience with payment practices in the markets for all of these other types of consumer services, including consumer financial services. The Bureau also considers this evidence that the practices identified in § 1041.7 are more common or more likely to occur in the covered markets than in other markets. In addition to this combined financial institution analysis, Bureau research and outreach suggest extremely high rates of returned payments for both storefront and online lenders. As noted earlier, for example, storefront lenders report failure rates of approximately 60 to 80 percent when they deposit consumers’ post-dated checks or initiate ACH transfers from consumer accounts in situations where the consumer has not come into the store to repay in cash.953 Bureau research of ACH payments finds that online lenders experience failure rates upwards of 70 percent where they attempt to re-present an ACH withdrawal one or more times after an initial failure.954 Moreover, of the 30 percent of second attempts and 27 percent of third 952 NACHA Q4 2014. QC Holdings 2014 Annual Report (Form 10-K), at 7 (reporting a return rate of 78.5 percent); Advance America 2011 Annual Report (Form 10-K), at 27 (reporting return rates of 63 percent for checks and 64 percent for ACH attempts). 954 Bureau analysis of ACH payments by online lenders shows an initial ACH payment failure rate due to NSFs of six percent. However, among the “successful” payments, Bureau research indicates that approximately six percent are paid only by overdrawing the consumer’s account. CFPB Report: Online Payday Loan Payments, Table 1, at 13. The Bureau’s analysis includes payday lenders and payday installment lenders that only operate online; the dataset excludes lenders that provide any storefront loans. In comparison, the Chase dataset includes both storefront and online payday lenders. As discussed in the proposal, many payments to storefront lenders are provided in person at the store. The fact that the consumer has not shown up at the store is a sign that the consumer may be having trouble making the payment. In contrast, online lenders generally collect all payments electronically and succeed more often on the initial payment attempt. Given that storefront lenders have higher rates of return on the first payment attempt, this sample difference may explain the relatively lower failure rate for first-attempt online ACH payments observed by the Bureau. 953 955 attempts that succeed, Bureau research indicates that approximately a third of them only do so by creating overdrafts on the consumer’s account, which trigger further fees.955 It may be the case that, as commenters noted, high return payment rates are influenced significantly by the fact that lenders are making loans to borrowers who are less likely to have funds in their accounts, or that the one-time balloon payment structure of these loans are more prone to failed payment attempts. But that argument also implies that borrowers in this market are more vulnerable to harm from engaging in multiple presentments than consumers are in other markets. Account Fees The proposal cited the Bureau’s analysis, consumer complaints, and public litigation documents, which show that the damage done to consumers from these payment attempts can be substantial.956 Fifty percent of checking accounts of online borrowers in the Bureau’s analysis of online payday and payday installment loans incurred at least one overdraft or NSF return in connection with their loans, with average fees for these consumers at $185.957 Indeed, 10 percent of these accounts experienced at least 10 payment withdrawal attempts that resulted in an overdraft or NSF return over an 18-month period.958 A small but significant percentage of consumers suffer extreme incidences of overdraft and NSF fees on their accounts; for consumers with at least one online payday attempt that resulted in an overdraft or NSF return, 10 percent 955 CFPB Online Payday Loan Payments, at 13, tbl. 1. See, e.g., Complaint at 19, Baptiste v. JP Morgan Chase Bank, No. 1:12-CV-04889 (E.D.N.Y. Oct. 1, 2012) (alleging that during a two-month period, 6 payday lenders debited the plaintiff’s bank account 55 times, triggering a total of approximately $1,523 in NSF, overdraft, and service fees). 957 CFPB Online Payday Loan Payments, at 10-11. 958 Id. at 10. 956 956 were charged at least $432 in related account fees over the 18-month sample period.959 This recounting of the types and amounts of fees charged to consumers in these circumstances was generally accepted by commenters on both sides of the proposed rule, though one commenter took issue with the Bureau’s use of averages, noting that they can be skewed by outliers and that citing the median experience would be more reliable. While that may be so as a logical matter, the Bureau cited the average fees because it was interested in assessing the total harm of the conduct in question, and not just the harm incurred by the typical borrower. Account Closure Lender attempts to collect payments from an account may also contribute to account closure. The Bureau has observed that the accounts of borrowers who use loans from online payday lenders are more likely to be closed than accounts generally (17 percent versus three percent, respectively).960 In particular, 36 percent of borrowers had their account closed involuntarily following an unsuccessful attempt by an online payday lender to collect a payment from the account, a rate that is four times greater than the closure rate for accounts that only had NSFs from non-payday transactions. Additionally, the Bureau found that borrowers with two consecutive failures by the same lender are significantly more likely to experience an involuntary closure than accountholders generally (43 percent versus three percent, respectively).961 For accounts with failed online payday loan transactions, account closures typically occur within 90 days of the last observed online payday loan transaction; in fact, 74 percent of account closures in these situations occur within 90 days of the first NSF return triggered by an online payday or 959 Id. at 12. Id. at 24 tbl. 5. 961 CFPB Report on Supplemental Findings, at p. 151. 960 957 payday installment lender.962 This suggests that the online loan played a role in the closure of the account, or that payment attempts failed because the account was already headed toward closure, or both.963 Commenters provided further data suggesting a connection between payment presentment practices and account closures. For example, a Pew survey found that 22 percent of online payday borrowers claimed to have lost bank accounts because of online payday loans.964 Some commenters took issue with the Bureau’s reliance on its 2016 report on online payday loan payments to establish the link between payday payment practices and account closures. They asserted certain methodological limitations of the report and accused the Bureau of using the data to assert causation when all it showed was correlation. They noted that the report itself had recognized the possibility that other confounding factors might explain the correlation. But the Bureau did not fail to recognize these points; on the contrary, the Bureau had been careful to note the limitations of its study and to caution that correlation is not necessarily show causation. Similarly, commenters contended that the Bureau’s report did not sufficiently distinguish between truly voluntary and truly involuntary account closures. Yet the Bureau did distinguish between voluntary account closures by the consumer and involuntary account closures initiated by the bank. Practically, it would be quite difficult to parse individual circumstances any further. 962 Id. at 23. See also Complaint at 14, Baptiste, No. 1:12-CV-04889 (alleging plaintiff’s bank account was closed with a negative balance of $641.95, which consisted entirely of bank’s fees triggered by the payday lenders’ payment attempts); id. at 20-21 (alleging plaintiff’s bank account was closed with a negative balance of $1,784.50, which consisted entirely of banks fees triggered by the payday lender’s payment attempts and payments provided to the lenders through overdraft, and that plaintiff was subsequently turned down from opening a new checking account at another bank because of a negative ChexSystems report stemming from the account closure). 964 Pew Charitable Trusts, “Payday Lending in America: Report 4, Harmful Practices in Internet Payday Lending,” at 16 (Oct. 2014). 963 958 A consumer might have pulled all of his money out of an account, making the eventual bank closure seem more “voluntary,” but those kinds of individual circumstances are difficult to account for in a broader study. Due to variations in borrower circumstances, the Bureau agrees that the study does not necessarily show that the presentment practices described were the actual cause of every observed involuntary account closure. However, the Bureau believes the high correlation between account closure and problematic payment practices indicates that these consumers may be experiencing harms beyond the fees immediately triggered by the transactions. c. Limited Consumer Control Consumers’ ability to protect their accounts from these types of payment attempt problems is limited due to a combination of factors, including the nature of the lender practices themselves, lender revocation procedures (or lack thereof), costs imposed by depository institutions in connection with consumer efforts to stop-payment attempts, and the operational limits of individual payment methods. In some cases, revoking authorization and stopping payment may be infeasible, and at a minimum they are generally both difficult and costly. Consumers Have Difficulty Stopping Lenders’ Ability to Access Their Accounts In the proposal, the Bureau indicated its preliminary view that lenders and accountholding institutions may make it difficult for consumers to revoke account access or stop withdrawals. 965 One way that consumers could attempt to stop multiple attempts to collect from their accounts would be to direct their lender to stop initiating payments. To do so, however, the 965 The Bureau is not addressing in this rulemaking the question of whether any of the practices described are consistent with the EFTA and Regulation E. 959 consumer must be able to identify and contact the lender, which can be difficult or impossible for consumers who have borrowed from an online lender. Moreover, lenders that can be contacted often make it difficult to revoke access. For example, several lenders require consumers to provide another form of account access in order to effectively revoke authorization with respect to a specific payment method—some lenders require consumers to provide this back-up payment method as part of the origination agreement.966 Some lenders require consumers to mail a written revocation several days before the effective date of revocation.967 These same lenders automatically debit payments through another method, such as a remotely created check, if a consumer revokes the ACH authorization. Others explicitly do not allow revocation, even though ACH private network rules require stop-payment rights for both one-time and recurring ACH transactions.968 For example, one lender website states that ACH revocation is not allowed for its single-payment online loans.969 Other lenders may not have obtained proper authorization 966 See, e.g., Castle Payday Loan Agreement, Ex. A, Parm v. BMO Harris Bank, N.A., No. 13-03326 (N.D. Ga. Dec. 23, 2013), ECF No. 60-1 (“You may revoke this authorization by contacting us in writing at ach@castlepayday.com or by phone at 1-888-945-2727. You must contact us at least three (3) business days prior to when you wish the authorization to terminate. If you revoke your authorization, you authorize us to make your payments by remotelycreated checks as set forth below.”); Press Release, Bureau of Consumer Fin. Prot., CFPB Takes Action Against Online Lender for Deceiving Borrowers (Nov. 18, 2015), available at https://www.consumerfinance.gov/aboutus/newsroom/cfpb-takes-action-against-online-lender-for-deceiving-borrowers/. 967 See id. 968 See NACHA Rule 3.7.1.2, RDFI Obligation to Stop Payment of Single Entries (“An RDFI must honor a stoppayment order provided by a Receiver, either verbally or in writing, to the RDFI at such time and in such manner as to allow the RDFI a reasonable opportunity to act upon the order prior to acting on an ARC, BOC, POP, or RCK Entry, or a Single Entry IAT, PPD, TEL, or WEB Entry to a Consumer Account.”). 969 Advance America provides the following frequently asked question in regard to its online loan product: Can I revoke my ACH payment? No. The ACH Authorization can only be revoked AFTER we have received payment in full of the amount owed. Because our advances are single payment advances (that is, we advance a sum of money that is to be repaid in a lump sum), we are permitted to require ACH repayment in accordance with the Federal Electronic Funds Transfer Act (“EFTA”). See Advance America, Frequently Asked Questions, https://www.onlineapplyadvance.com/faq (last visited May 17, 2016). 960 in the first place970 or take broad authorizations to debit any account associated with the consumer.971 Consumer complaints sent to the Bureau also indicate that consumers struggle with anticipating and stopping payment attempts by lenders of covered loans. As of December 31, 2016, complaints where the consumer has identified the issues “can’t stop lender from charging my bank account” or “lender charged my bank account on wrong day or for wrong amount” account for nearly 10 percent of the more than 16,600 payday loan complaints the Bureau has handled since November 2013.972 In addition, the Bureau handled approximately 31,000 debt collection complaints relating to payday loans during this same period. More than 11 percent of debt collection complaints received by the Bureau stem from payday loans. The Bureau also handled more than 15,800 installment loan complaints. Review of those complaints suggests that there are consumers who labeled their complaints as falling under those categories who also experience difficulties anticipating and stopping payment attempts. 970 Hydra Group, a purported online payday lender against which the Bureau brought an enforcement action, allegedly used information bought from online lead generators to access consumers’ checking accounts to illegally deposit payday loans and withdraw fees without consent. The Bureau alleged that Hydra Group falsified loan documents to claim that the consumers had agreed to the phony online payday loans. The scam allegedly added up to more than $100 million worth of consumer harm. Hydra had been running its transactions through the ACH system. Complaint, CFPB v. Moseley, No. 4:14-CV-00789 (W.D. Mo. Sept. 8, 2014), ECF No. 3, available at http://files.consumerfinance.gov/f/201409_cfpb_complaint_hydra-group.pdf. See also Stipulated Order, FTC v. Michael Bruce Moneymaker, Civil Action No. 2:11-CV-00461 (D. Nev. Jan. 24, 2012), available at https://www.ftc.gov/sites/default/files/documents/cases/2012/02/120201moneymakerorder.pdf (purported lead generator defendants used information from consumer payday loan applications to create RCCs to charge consumer accounts without authorization). 971 See, e.g., Great Plains Lending d/b/a Cash Advance Now, Frequently Asked Questions (FAQs), https://www.cashadvancenow.com/FAQ.aspx (last visited May 16, 2016) (“If we extend credit to a consumer, we will consider the bank account information provided by the consumer as eligible for us to process payments against. In addition, as part of our information collection process, we may detect additional bank accounts under the ownership of the consumer. We will consider these additional accounts to be part of the application process.”). 972 This figure excludes debt collection payday loan complaints because consumers filing debt collection payday loan complaints have a different set of issues to choose from when completing the complaint form. 961 The other option for consumers is to direct their bank to stop payment, but this too can be challenging. Depository institutions typically charge a fee of approximately $32 for processing a stop-payment order, making this a costly option for consumers.973 In addition, some lenders charge returned-item fees if the stop-payment order successfully blocks an attempt.974 The Bureau has received complaints from consumers who were charged overdraft and NSF fees after merchants with outstanding stop-payment orders were able to withdraw funds despite the presence of the orders; in some instances, banks refused to refund these charges. The ease of successfully stopping a payment also varies by channel. To execute a stoppayment order on a check, banks usually use the check number provided by the consumer. As ACH payments do not have a number equivalent to a check number for the bank to identify them, ACH payments are more difficult to stop. To block the payment, banks may need to search the ACH transaction description for information that identifies the lender. Determining an effective search term is difficult, given that there is no standardization of how originators of a payment—in this case, lenders—identify themselves in the ACH network. Lenders may use a parent company name or an abbreviated name, or may vary names based on factors like branch location. Other lenders use the name of their third-party payment processor. During the Bureau’s outreach, some depository institutions indicated that certain payday lenders use 973 This is the median stop-payment fee for an individual stop-payment order charged by the 50 largest financial institutions in 2015. Informa Research Services, Inc. (Aug. 7, 2015), Calabasas, CA. www.informars.com. Although information has been obtained from the various financial institutions, the accuracy cannot be guaranteed. 974 See, e.g., Complaint at 19, Baptiste v. JP Morgan Chase Bank, No. 1:12-CV-04889 (E.D.N.Y. Oct. 1, 2012) (alleging that during a two-month period, six payday lenders debited the plaintiff’s bank account 55 times, triggering a total of approximately $1,523 in non-sufficient funds, overdraft, and service fees); CFPB Online Payday Loan Payments. 962 multiple merchant ID codes and different names on their ACH transactions in an apparent attempt to reduce the risk of triggering scrutiny for their ACH presentments. Moreover, remotely created checks (RCCs) and remotely created payment orders (RCPOs) are virtually impossible to stop because the consumer does not know the check number that the payee will generate, and the transaction information does not allow for payment identification in the same way that an ACH file does. RCCs and RCPOs have check numbers that are created by the lender or its payment processor, making it unlikely that consumers would have this information.975 Industry stakeholders, including members of the Bureau’s Credit Union Advisory Council, indicate that it is virtually impossible to stop payments on RCCs and RCPOs because the information needed to stop the payment—such as check number and payment amount—is generated by the lender or its payment processor. Consumers also may not realize that a payment will be processed as a RCC, so they may not even know to ask their bank to look for a payment processed as a check rather than as an ACH payment. Some financial institutions impose additional procedural hurdles, for instance by requiring consumers to provide an exact payment amount for a stop-payment order and allowing payments that vary by a small amount to go through.976 Others require consumers to provide the 975 See Letter to Ben Bernanke, Chairman, Board of Governors of the Federal Reserve System, from the National Consumer Law Center, Consumer Federation of America, Center for Responsible Lending, Consumer Action, Consumers Union, National Association of Consumer Advocates, National Consumers League, and U.S. PIRG, Comments on Improving the U.S. Payment System, at 8 (Dec. 13, 2013), available at https://fedpaymentsimprovement.org/wp-content/uploads/2013/12/Response-Natl_Consumer_Law_Center_et_al121313.pdf. 976 For example, Regions Bank instructs consumers that “If you are attempting to stop payment on an ACH draft, you must provide the exact amount of the draft or the stop payment cannot be placed.” See Regions Bank, Frequently Asked Questions, http://www.regions.com/FAQ/lost_stolen.rf (last visited May 17, 2016). 963 merchant identification code that the lender used in the ACH file.977 Because there is no standardization of merchant names or centralized database of merchant identification codes in the ACH system, however, the only way for consumers to know the exact merchant identification code is if they observed a previous debit by that lender. Even if a consumer located a lender’s identification code on a previous debit, which may or may not be practicable, lenders may vary this code when they are debiting the same consumer account again.978 As mentioned previously, during the Bureau’s outreach, some depository institutions indicated that payday lenders sometimes use multiple merchant ID codes and different names on their ACH transactions in an apparent attempt to reduce the risk of triggering scrutiny for their ACH presentments. Moreover, banks may require consumers to navigate fairly complex procedures in order to stop a payment, and these procedures may vary depending on whether the payment is presented through the ACH system or the check system. For example, one major depository institution allows consumers to use its online system to stop payment on a check, but requires notification over the phone to stop a payment on an ACH item.979 The Bureau also identified in the proposal some risk that bank personnel may misinform consumers about their rights. During outreach, the Bureau learned that the ACH operations 977 See Wells Fargo, Instructions for Stopping Payment, https://www.wellsfargo.com/help/faqs/order-checks/ (last visited May 17, 2016) (“ACH items—Please provide the Company Name, Account Number, ACH Merchant ID and/or Company ID (can be found by reviewing a previous transaction) and Amount of item.”). 978 Through market outreach, the Bureau has learned that the ACH channel used to be allowed only for recurring authorizations. Future transactions could be stopped relatively easily because the bank could use the merchant identification information (in this case, the name that the lender or its payment processor puts in the ACH file) that was on prior preauthorized debits. However, now that the ACH network can also be used to initiate one-time payments, a bank may not know which merchant identifier to use. In addition, some merchants (including lenders) seem to be gaming the system by changing the merchant identifiers to work around stop payments. 979 See Wells Fargo Instructions for Stopping Payment (“You can request a stop payment online (check only), by phone (check and ACH items) or by visiting your local store and speaking with a banker.”), https://www.wellsfargo.com/help/faqs/order-checks/ (last visited May 17, 2016). 964 personnel at some banks do not believe consumers have any right to stop payment or send back unauthorized transactions initiated by payday lenders. The Bureau has received consumer complaints to this effect.980 Recent Federal court cases and information from legal aid organizations981 also provide evidence that bank personnel may not correctly implement consumer payment rights in all cases.982 d. Private Network Protections Have Limited Impact Finally, while the presentment practices of the payday industry are so severe that they have prompted recent actions by the private rulemaking body that governs the ACH network, the Bureau stated in the proposal that these efforts likely would be insufficient to solve the problems discussed above. The private NACHA Rules do provide some protections in addition to those currently provided by law. Specifically, the NACHA Rules now limit the re-presentment of any one single failed payment to two additional attempts and provide that any lender with a total return level of 15 percent or above may be subject to an inquiry process by NACHA. They also impose a “company name rule” mandating that originators of ACH transactions use names that consumers would recognize, and impose a fee on payment originators when payments are returned. NACHA has also undertaken various efforts to improve the enforcement of their rules 980 The Bureau has received complaints from consumers alleging that banks told consumers that the bank could not do anything about unauthorized transactions from payday lenders and that the bank would not stop future debits. 981 See also, New Economy Project Letter to Federal Banking Regulators, at 1-2 (September 2014), available at http://www.neweconomynyc.org/wp-content/uploads/2014/11/letter.pdf (“People have often found that their financial institution fails to honor requests to stop payment of recurring payments; has inadequate systems for implementing stop payment orders and preventing evasions of those orders; charges inappropriate or multiple fees; and refuses to permit consumers to close their accounts.”). 982 See Jessica Silver-Greenberg, Major Banks Aid in Payday Loans Banned by States, NY Times (Feb. 23, 2013), available at http://www.nytimes.com/2013/02/24/business/major-banks-aid-in-payday-loans-banned-by-states.html (discussing allegations against JP Morgan Chase about consumer difficulties in revoking authorization and stopping payment on online payday loans); Complaint at 11, Baptiste, No. 1:12-CV-04889 (alleging that a bank employee told the plaintiff that the bank “could not stop the debits from payday lenders, and that she should instead contact the payday lenders to tell them to stop debiting her account”). 965 in recent years, and to encourage more developed self-monitoring across all industries. As NACHA set forth in its comment responding to this rulemaking, it has engaged in a number of reforms more recently, including several reforms in 2014. However, the narrower scope of these rules, the limited private network monitoring and enforcement capabilities over them, and their applicability to only one payment method, taken together, mean that private network protections are not well positioned to completely solve problematic practices in the payday and payday installment industries. Re-initiation Cap The Bureau observed in the proposal that the NACHA Rules have historically provided a re-initiation cap, which limits re-presentment of a failed payment to two additional attempts. Compliance with this requirement is difficult to monitor and enforce.983 Although ACH files are supposed to distinguish between collection of a new payment and the re-initiation of a prior one, some originators do not comply with this requirement to label re-initiated transactions.984 Because the ACH system does not record whether the payment is for a loan and accordingly cannot identify the terms of the loan, including whether it is a single-payment loan or an installment loan with a series of scheduled payments, there is limited ability to distinguish re983 See FFIEC, Bank Secrecy Act/Anti-Money Laundering Exam Manual, at 238 (“Transactions should be monitored for patterns that may be indicative of attempts to evade NACHA limitations on returned entries. For example, resubmitting a transaction under a different name or for slightly modified dollar amounts can be an attempt to circumvent these limitations and are violations of the NACHA Rules.”). 984 NACHA Request for Comment and Request for Information—ACH Network Risk and Enforcement Topics, Rule Proposal Description, at 6-7 (proposing amendments in response to lack of compliance with requirement to label reinitiated transactions) (“NACHA has reason to believe that some high-risk Originators may ignore or attempt to evade the requirements of the Reinitiation Rule, including by changing content in various fields to make an Entry appear to be a new Entry, rather than a reinitiation. . . . For additional clarity, NACHA proposes to include in the Reinitiation Rule common examples that would be considered reinitiating an Entry to avoid arguments, for example, that adding a fee to an Entry creates a new Entry or that attempting to resubmit for a lesser amount takes the Entry outside of these limitations.”). 966 initiations (and potential NACHA rule violations) from the next installment payment. Unless a lender explicitly labels the attempt as a re-initiation, the ACH system cannot otherwise distinguish between, for example, the second attempt to collect a payment for January 1 and the first attempt to collect the next payment that is due on February 1.985 Even if the rule were not subject to ready evasion by originating entities, the cap also does not apply to future payments in an installment payment schedule. Accordingly, if a failed payment on a previously scheduled payment is followed by a payment attempt on the next scheduled payment, that second attempt is not considered a re-initiation and does not count toward the cap. For example, for a loan payment that does not go through, NACHA Rules allow that payment to be presented a total of three times, thereby generating three fees to the consumer, and the following payment due can still proceed despite any prior failures. Commenters suggested that the Bureau should distinguish between re-presentments and new payments on the payment schedule, and suggested that the Bureau should not have counted payments 14 days out as “re-presentments” in its studies. The Bureau did include them because payments in short succession would look quite similar to re-presentments from the consumer’s perspective. And as the Bureau’s study showed, even when counting presentments 14 days apart as “representments,” the rates of rejection are quite high for second, third, fourth, and further presentments, especially when compared to the rejection rate for the first presentment. 985 NACHA explicitly excludes scheduled payments from its reinitiation rule. See id. at 7 (explaining that “the proposal would clarify that a debit Entry in a series of preauthorized recurring debit Entries will not be treated as a reinitiated Entry, even if the subsequent debit Entry follows a returned debit Entry, as long as the subsequent Entry is not contingent upon whether an earlier debit Entry in the series has been returned.”). 967 There were a number of comments stating that NACHA has recently clarified its reinitiation cap in 2014, and that the Bureau should wait to see if that effort fixed the problems the Bureau had identified. In a similar vein, commenters suggested that the Bureau’s study is stale because it was based on data from 2012, which was before these NACHA reforms were enacted. On this topic, NACHA wrote to the Bureau that its pre-existing re-initiation cap has acted to protect consumers against excessive debits to their accounts for many years, while providing a reasonable opportunity for duly authorized transactions to be paid when the account to be debited inadvertently has inadequate funds at the time of the original charge. NACHA also clarified that it took steps in 2014 to clarify the application of the re-initiation rule because of concerns regarding evasion and non-compliance with the rule. As discussed in the proposal, NACHA had concerns that originators were not labeling re-initiated transactions or were using scenarios where a payment had changed in some way—such as by adding a fee—to avoid considering it as a re-initiation under the cap. The Bureau notes that the cap is longstanding and existed during the 2011-2012 study period, and the Bureau’s data shows that the problems identified above remained. NACHA clarified the application of the rule thereafter, and the Bureau agrees that this effort to focus more industry attention on the cap is likely to be helpful in addressing the problem. However, as noted in the NPRM, NACHA has limited to ability to monitor and enforce its reinitiation cap. Although it is possible that fewer industry participants violated the cap after the 2014 clarification, the fact that industry was evading the rule pre-2014 suggests that they may be still evading it today. NACHA claims that the overall NSF return rates for all ACH debits fell by 21 percent since 2012, and by 31 percent for online payments. Those are marketwide numbers, and it is unclear whether the payday industry made similar improvements. But 968 even if it did, much of the problems that the Bureau’s study identified would remain, though they may have been depressed somewhat. Furthermore, NACHA stated in its comment that the new NACHA Rules have resulted in a shift to other riskier payment methods, such as remotely created checks and debit network transactions that are not governed by the NACHA Rules. The Bureau believes that this final rule will be a beneficial supplement to the NACHA Rules in that this rule will apply across multiple payment methods (including those riskier methods that the NACHA Rules cannot reach). Additionally, the NACHA Rules cap re-presentments at two of the original entry, which allows one more re-presentment than does this rule (and, as discussed above, allows the reinitiation clock to re-start with the next scheduled payment). A substantial amount of the consumer harm found in the Bureau’s study data occurred on the second representment, and since the NACHA Rules did not affect that, the Bureau concludes that its data is not stale as to that issue. Lastly, as stated earlier in this section, while the NACHA reforms may impact the prevalence of re-presentment practices to some degree, they would not alter the type and extent of consumer harm that re-presentments cause when they do occur. Total Return Rate Level According to a NACHA rule that went into effect in September 2015, originators986 with a total return rate of 15 percent or above are subject to an inquiry process by NACHA.987 This 986 The return rate level is calculated for individual entities like lenders and payment processors that direct an ODFI to debit a consumer’s account on the entities’ behalf. See NACHA Rule 2.17.2; NACHA Rule 8.6 (defining “originator”). 987 See NACHA Rule 2.17.2; NACHA, ACH Network Risk and Enforcement Topics, https://www.nacha.org/rules/ach-network-risk-and-enforcement-topics (last visited May 17, 2016) (“The Rule will establish an inquiry process that will provide NACHA with a preliminary evaluation point to research the facts behind an Originator’s ACH activity. Preliminary research, as part of the inquiry process, begins when any Originator exceeds the established administrative return rate or overall return rate level. The review process involves eight steps, and includes an opportunity for NACHA and an industry review panel to review an 969 return rate threshold includes returns for reasons such as non-sufficient funds, authorization that was revoked by the consumer, administrative issues (such as an invalid account number), and stop-payment orders. It does not include the returns of re-presented checks, which are ACH representments of payments that were first attempted through the check-clearing network. Exceeding this threshold does not necessarily violate NACHA Rules, but rather simply allows NACHA to demand additional information from the lender’s originating depository financial institution (ODFI) for the purpose of determining whether the ODFI should lose access to the ACH system.988 During this process, the ODFI may be able to justify a high return rate depending on the lender’s business model and other factors.989 NACHA set the threshold at 15 percent to allow flexibility for a variety of business models while identifying originators that were burdening the ACH system.990 However, in the proposal the Bureau stated its concern that lenders can adopt Originator’s ACH activity prior to any decision to require a reduction in a return rate. The inquiry process does not automatically trigger a Rules enforcement activity.”) (“The rule does not automatically require an ODFI to reduce an Originator’s return rate below 15 percent; as such, it is meant to be flexible in accounting for differing needs of a variety of businesses. The rule would require an ODFI to reduce an Originator’s return rate below 15 percent if directed to do so by the industry review panel.”). 988 See NACHA Rule 2.17.2. 989 See NACHA, ACH Network Risk and Enforcement Topics: FAQs, available at https://www.nacha.org/rules/achnetwork-risk-and-enforcement-topics (last visited May 16, 2016). The inquiry process is an opportunity for the ODFI to present, and for NACHA to consider, specific facts related to the Originator’s or Third-Party Sender’s ACH origination practices and activity. At the conclusion of the preliminary inquiry, NACHA may determine that no further action is required, or may recommend to an industry review panel that the ODFI be required to reduce the Originator’s or Third-Party Sender’s overall or administrative return rate below the Return Rate Level…. In reviewing the results of a preliminary inquiry, the industry review panel can consider a number of factors, such as: (1) The total volume of forward and returned debit Entries; (2) The return rate for unauthorized debit Entries; (3) Any evidence of Rules violations, including the rules on reinitiation; (4) Any legal investigations or regulatory actions; (5) The number and materiality of consumer complaints; (6) Any other relevant information submitted by the ODFI. 990 See NACHA, Request for Comment and Request for Information, at 5(“By setting the threshold at approximately 10 times the ACH Network average, NACHA believes that sufficient leeway will be permitted for businesses that attempt to service high risk communities without creating return rates that significantly increase costs on RDFIs and raise questions about the quality of the origination practices.”). 970 problematic payment practices and remain below this inquiry level. This concern is borne out by the data, as the Bureau in fact has observed an overall lender NSF return rate of 10.1 percent in its analysis of ACH payments attempts by online payday and payday installment lenders.991 Monitoring and Enforcement of the New Total Return Rate Level In the proposal, the Bureau preliminarily found that NACHA has a limited ability to monitor return rates. First, NACHA has no ability to monitor returns based on a particular lender. All of the return information it receives is sorted by the ODFIs that are processing the transactions, rather than at the level of the individual lenders that are accessing the ACH network. Because lenders sometimes use multiple ODFI relationships to process their payments,992 the returns used in the NACHA threshold do not provide a full picture of those lenders’ payment activity. In addition, NACHA has no ability to monitor or calculate return rates on an ongoing basis. Although it receives return volume reports from the ACH operators (the Federal Reserve and The Clearinghouse), these reports do not contain the successful payment volume information that is necessary to calculate a return rate. Rather, NACHA relies on financial institutions to bring suspect behavior to its attention, which even then only provides it with a basis to investigate further and request more detailed payment reports. The Bureau also emphasized in the proposal that lenders often obtain access to multiple payment methods, such as check, ACH, and debit card. As private payment networks do not 991 This return rate does not include same-day presentments; with same-day presentments included, the overall return rate is 14.4%. The NACHA reinitiation cap was in effect during the Bureau’s sample period of 2011-2012. The NACHA rule on overall return rate levels went into effect in September 2015. 992 In order to access the ACH network, lenders must use an ODFI. A lender may not have a direct ODFI relationship if it is sending payments through a third-party payment processor. In that case, the processor would have an ODFI relationship. A lender may have multiple ODFI and processor relationships, such as different relationships for different loan products or regions. 971 combine return activity, there is no monitoring of a lender’s overall returns across all payment types. Payments that begin as checks and then are re-presented as ACH payments, a practice that is not uncommon among storefront payday lenders, are excluded from the NACHA return rate threshold. The Bureau is also aware that lenders sometimes alternate between payment networks to avoid triggering scrutiny or violation of particular payment network rules. Processor marketing materials, Bureau staff conversations with industry, and documents made public through litigation indicate that the NACHA unauthorized return and total return rate thresholds have already prompted migration to remotely created checks and debit network transactions, none of which is covered by the NACHA Rules.993 In light of the available evidence, including the comments received on the points discussed in this section of the proposed rule, the Bureau concludes that substantial risk to consumers remains. Although private network rules may improve lender practices in some respects, they have many gaps, impose limited consequences, and do not eliminate all consumer harm. There is no systematic way to monitor lender payment practices in the current ACH system, or more broadly for practices across all payment channels, leaving only weak enforcement mechanisms in place for applying the NACHA Rules. In addition, because these rules are private, the public has no guarantee or assurance of any kind that they will exist in the same form or an improved form in the future. And perhaps most importantly, the NACHA Rules only apply to the ACH system, and not all payment methods. For all of these reasons, the Bureau concludes that the private ACH network rules do not provide an adequate solution to the 993 See, e.g., FTC Final Amendments to Telemarketing Sales Rule, 80 FR 77520, 77532 (Dec. 14, 2015) (discussing marketing by payment processors). 972 problematic payment practices in this market. The Bureau values NACHA’s continued efforts to improve payment practices, both for this lending market and across the entirety of the ACH networks, and will continue to consider NACHA as a partner while the Bureau proceeds with its own work to address the harms it identifies to consumers. Section 1041.7 Identification of Unfair and Abusive Practice—Payments The Bureau’s Approach in the Proposal In the proposal, the Bureau stated its belief that the act or practice of obtaining a consumer’s authorization in advance to initiate electronic fund transfers (EFTs) from the consumer’s bank account often can be beneficial for creditors and consumers alike by providing a relatively speedy, predictable, and low-cost means of repayment. Nonetheless, for all of the reasons discussed in Markets Concerns—Payments of the proposed rule, the Bureau also stated its belief that lenders in the markets for payday and payday installment loans often use such payment authorizations in ways that may cause substantial harms to consumers who are especially vulnerable, particularly when lenders continue making payment withdrawal attempts after one or more attempts have failed due to non-sufficient funds. Based on the available evidence and pursuant to its authority under section 1031 of the Dodd-Frank Act, the Bureau proposed in § 1041.13994 to identify it as both an unfair and an abusive act or practice for a lender to attempt to withdraw payment from a consumer’s account in connection with a covered loan after the lender’s second consecutive attempt has failed due to a 994 Because changes made to the proposal have led to omissions of certain sections, the sections on payment attempts, along with certain others, have been renumbered in the final rule. Thus, for example, § 1041.13 of the proposed rule has now become § 1041.7 of the final rule. The numbering of the sections in the final rule will be used here, unless specifically indicated otherwise. 973 lack of sufficient funds, unless the lender obtains the consumer’s new and specific authorization to make further withdrawals from the account. In this context, an “attempt to withdraw payment from a consumer’s account” was defined, in proposed § 1041.14, to mean a lender-initiated debit or withdrawal from the account for purposes of collecting any amount due or purported to be due in connection with a covered loan, regardless of the particular payment method used by the lender to initiate the debit or withdrawal. The proposed identification thus would apply to all common methods of withdrawing payment from consumers’ accounts, including but not limited to the following methods: EFTs (including preauthorized EFTs), without regard to the particular type of payment device or instrument used; signature checks; remotely created checks; remotely created payment orders; and an account-holding institution’s withdrawal of funds held at the same institution. The Bureau sought comment on the evidence it had presented on these issues, and on the preliminary findings and conclusions it had reached in the proposal. General Comments Received The Bureau received a number of general comments about the Bureau’s use of its authority to prohibit unfair, deceptive, or abusive acts or practices. The Bureau addresses those more general comments immediately below; the specific comments on the prongs of unfairness or abusiveness are addressed further below. Some commenters claimed the proposed intervention was not necessary because of the NACHA Rules described above or, alternatively, that the data the Bureau used was stale because of the new NACHA Rules. Other commenters suggested that the Bureau should simply enforce Regulation E, or use its UDAAP enforcement authority, to address the issue. Others argued that State law sufficiently addressed the issues identified by the Bureau, or that leveraged payment 974 mechanisms were required by State law, and that this meant the rule was in conflict with those requirements. Some commenters argued that it was improper or inappropriate to write a rule that only implicates a small subset of the total market’s transactions, and that these issues should be addressed instead by supervisory oversight or enforcement activity. Several commenters argued that the rule was overbroad, arguing that the Bureau’s primary source of data was from online payday lenders, and that the data were not applicable to depository institutions, traditional installment loans, or storefront lenders. Other commenters argued that the Bureau had not shown that there was any difference in payment presentment practices between covered industries and industries the rule would not cover—for example, longer-term installment lending with interest rates below 36 percent APR. Still others argued that the Bureau had not identified, as an unfair or abusive practice, the failure to provide the consumer notice before initiating a transfer, and thus did not properly identify any UDAAP predicate to support the notice interventions in the proposed rule (proposed § 1041.15, final § 1041.9). Lastly, commenters argued that this part of the rule was unnecessary because proposed §§ 1041.4 to 1041.6 would ensure that more borrowers have an ability to repay, and thus would be much more likely to have funds in their accounts when the first presentment is made (meaning there would be no need for multiple payment attempts). Final Rule The Bureau now concludes that the practice of making attempts to withdraw payment from consumers’ accounts in connection with a covered loan after the lender’s second 975 consecutive attempts to withdraw payments from the accounts from which the prior attempts were made have failed due to a lack of sufficient funds, unless the lender obtains the consumers’ new and specific authorization to make further withdrawals from the accounts, is unfair and abusive. The Bureau’s analysis of why this practice meets the elements of unfairness and the elements of abusiveness, as well as its responses to the comments received on those topics, are provided below. But first the Bureau responds to the broader comments concerning the Bureau’s general approach. The Bureau addressed the comments regarding whether the Bureau’s data are stale because of new NACHA Rules in the Market Concerns—Payments section above. This final rule would only allow one re-presentment, as opposed to the two re-presentments allowed by the NACHA Rules, and this marginal difference will have a significant impact on an identifiable set of consumers. Additionally, as noted above, this rule governs all payment methods, which is important because NACHA only addresses ACH payments and accordingly has seen many lenders shift towards other, non-ACH payment methods in response to NACHA’s efforts to address the payment practices at issue in this rule. Further, the final rule clarifies that, as further explicated in the section-by-section analysis for § 1041.8, the payment presentment cap applies across multiple loans, contrary to the NACHA Rules. The Bureau values NACHA’s efforts and looks forward to working in a partnership on these issues, but concludes that the provisions in the NACHA Rules do not eliminate the need for regulatory intervention here. In addition, the Bureau concludes that merely continuing to enforce Regulation E would not be enough to remedy the harms from the identified practice. Regulation E does not impose a limit on multiple failed presentments. It does give consumers certain rights to stop payments and 976 cancel authorizations, which may mitigate some of the harm caused by multiple failed presentments, if exercised successfully. However, as the Bureau highlighted in the Market Concerns—Payments section above, consumers often have difficulty exercising these rights, and many of the reasons for this difficulty result from conduct and other factors that may not violate Regulation E or even be subject to that regulation. Furthermore, even when entities are in compliance with Regulation E, consumers may not be aware of their rights under that regulation, and may not be able to exercise them quickly enough. Given these limitations, the Bureau believes that individual enforcement actions under Regulation E would not sufficiently address the problematic payment practices and resulting consumer harms in markets for payday and payday installment loans. As discussed below, the Bureau is now deciding to use its UDAAP authority to address these problems in a more fundamental and comprehensive manner, instead of on a case-by-case basis. To the extent there are State laws that could address the problems identified, the Bureau believes, based on the evidence of payments-related consumer harms in markets for payday and payday installment loans, that those laws have not succeeded in preventing the harms caused by the identified practice, and the Bureau has thus decided that a more fundamental and comprehensive approach is in order. The Bureau has authority to bring UDAAP enforcement actions without issuing a rule. It could do so on a case-by-case basis, focusing only on those actors that engage in the most egregious payment practices. And it has already been doing so.995 However, the Bureau believes that addressing only the most egregious payment practices on a case-by-case basis would not sufficiently address consum er harms that occur when lenders in markets for payday 995 See, e.g., Consent Order, In the Matter of EZCORP, Inc., CFPB No. 2015–CFPB–0031 (Dec. 16, 2015). 977 and payday installment loans make multiple failed attempts to withdraw payment from consumers’ accounts. Accordingly, the Bureau has decided to address those harms more holistically with a rule. Several industry commenters made the point that the Bureau was proposing to take action on the basis of a fairly small set of payment presentments, as compared to the total presentments in the industry (which are often successful on the first try). The Bureau acknowledges this point, but finds that it does not undermine the case for this portion of the rule. The Bureau finds that there is substantial injury to a significant population of consumers, even though those affected do not constitute a majority of all consumers. The Bureau finds that this practice meets the prongs of unfairness and abusiveness, as discussed below, and believes this finding suffices for a rule that is narrowly tailored to address the minority of transactions at issue. The Bureau’s primary study on this topic was a report based on online payday and high cost payday installment lenders only, which includes covered short-term loans and covered longer-term loans as defined in this rule. The report and other evidence showed, generally, what happens to consumers when lenders re-present after two previous and consecutive failed attempts. The Bureau’s decision to apply the rule specifically to covered loans (short-term loans, high-cost longer-term loans, and long-term balloon payment loans), but not other lending markets, was based on the fact that consumers in the markets for covered loans have similar characteristics—as discussed in the proposal, Market Concerns—Underwriting, and Market Concerns—Payments—which make them vulnerable to harms that occur from the identified unfair and abusive practice. The Bureau also has evidence suggesting that lenders making covered loans are more likely to engage in the practice. Based on the higher return rates 978 observed in the markets for covered loans, the payments report, the Bureau’s enforcement experience, and consumer complaints, the Bureau believes the practice of continuing to make attempts to withdraw payment from a consumer’s account after two consecutive attempts have failed is more likely to occur in the markets for covered loans, and that consumers of loans in those markets are therefore more likely to incur the observed harms that result from that practice. The Bureau has not observed similar evidence in other markets, and thus makes the reasonable determination to confine the rule to those markets where it has data, evidence, and experience. Additionally, the fact that leveraged payment mechanisms are generally a feature of loans covered by the rule suggests that these lenders are more likely to have the opportunity to engage in the practice than are lenders in credit markets that are not so dependent on leveraged payment mechanisms. Of course, if the Bureau were to receive evidence suggesting that participants in other markets are engaging in this practice in ways that similarly harm consumers, it would consider expanding the rule to those markets, or perhaps taking supervisory or enforcement action as appropriate. With respect to the Bureau’s determination to apply the final rule to covered longer-term loans with an APR of more than 36 percent but not to those with a lower APR, the Bureau has substantial evidence that the identified practice is occurring in the market for higher-cost installment loans, specifically as shown in the payments report and through enforcement actions.996 The Bureau does not have similar evidence as to installment loans of all kinds, 996 CFPB, Online Payday Loan Payments (April 2016), available at http://files.consumerfinance.gov/f/201604_cfpb_online-payday-loan-payments.pdf; Consent Order, EZCORP, CFPB No. 2015-CFPB-0031 (Dec. 16, 2015), available at http://files.consumerfinance.gov/f/201512_cfpb_ezcorp-incconsent-order.pdf. Both involved high-cost installment or longer-term payday loans. 979 including traditional lower-cost credit, which makes up a much broader and more varied portion of the credit market, and is therefore limiting application of the rule so as to not reach all credit markets. If the Bureau were to obtain evidence that lenders in other installment loan markets are engaged in the identified practice or similarly harmful payment practices, it could initiate supervisory or enforcement actions, or expand the coverage of the rule, depending on the circumstances.997 The Bureau chose the 36 percent threshold specifically because of the long history of States and Federal regulators that have exercised their judgment to rely on that particular rate as a point of distinction between high-cost loans and other loans, as described in more detail in the Background section. Commenters are correct in asserting that the Bureau did not identify an unfair or abusive practice that would warrant the notice requirements in § 1041.15 of the proposed rule (§ 1041.9 of the final rule). But the Bureau did not attempt to do so. Instead, as discussed in the sectionby-section analysis of § 1041.9, below, the notice requirements were proposed pursuant to the Bureau’s disclosure authority under section 1032 of the Dodd-Frank Act. Lastly, the Bureau acknowledges that covered lenders may have less opportunity to subject consumers to the practice identified in § 1041.7 after the underwriting provisions in §§ 1041.4 to 1041.6 are implemented. As a covered lender’s customer base for covered short997 In the Military Lending Act rule limiting the terms of consumer credit extended to servicemembers and their dependents, the Department of Defense noted its unwillingness to define the total cost of credit so as to exclude “certain fees, or all non-periodic fees, [which] could be exploited by a creditor who would be allowed to preserve a high-cost, open-end credit product by offering a relatively lower periodic rate coupled with an application fee, participation fee, or other fee.” 80 FR 43563. Under the cost of credit adopted here from Regulation Z to govern the applicability of subpart C to covered lenders, the Bureau would note that if a lender sought to structure its loans in such a manner as to shift the cost of credit from the periodic rate to unusual application fees, participation fees, or other fees that bore no relation to the actual cost of credit in order to avoid coverage under this rule, then supervisory or enforcement authority could be invoked and this structuring of the loans could be cited as evidence of attempted evasion of the rule. 980 term loans skews more towards borrowers with an ability to repay their loans, fewer initial payments will be returned, and thus lenders will have fewer opportunities to make multiple failed payment attempts. This will not be the case, however, for covered longer-term loans, which are not subject to § 1041.5. The Bureau also notes that covered short-term loans made under § 1041.6 will not be subject to rigorous underwriting requirements. Additionally, it is implausible that the underwriting requirements in §§ 1041.4 to 1041.6 will eliminate all failed payment attempts. No provisions in §§ 1041.4 to 1041.6 would stop a lender from engaging in the practice the Bureau identified in § 1041.7 if a borrower did not have enough funds in his account. In every credit market, even ones with substantial underwriting, consumers experience some rate of default. For these reasons, and those set forth below, the Bureau finalizes the language in § 1041.7, identifying the specified practice of payment attempts on covered loans as unfair and abusive, in the same form as it was proposed in the comparable section of the proposed rule, with two exceptions. The Bureau has added official commentary, at comment 7-1, which clarifies that a lender who complies with § 1041.8 with regard to a covered loan has not committed the unfair and abusive practice under § 1041.7. This comment is added to clarify that § 1041.8 is intended to prevent the practice in § 1041.7. Thus, if a lender complies with § 1041.8, then it will not be in violation of § 1041.7. Second, during inter-agency consultations, the Bureau received input from a Federal prudential regulator about the singular nature of the statement of the unfair and abusive act or practice. The regulator believed that supervisory or enforcement actions of this particular rule should be based on a pattern or practice of activity, rather than an isolated and inadvertent 981 instance, which the regulator believed could deter responsible lenders from making covered loans. In the interest of inter-agency cooperation, the Bureau is adopting the suggestion to pluralize the statement of the unfair and abusive practice. Relatedly, the Bureau does not intend to bring supervisory or enforcement actions against a lender for a single isolated violation of § 1041.8. a. Unfair Practice Under section 1031(c)(1) of the Dodd-Frank Act, the Bureau has no authority to declare an act or practice unfair, unless it has a reasonable basis to conclude that it “causes or is likely to cause substantial injury to consumers which is not reasonably avoidable by consumers,” and such substantial, not reasonably avoidable injury “is not outweighed by countervailing benefits to consumers or to competition.”998 In the proposal, the Bureau indicated that it could be an unfair act or practice to attempt to withdraw payment from a consumer’s account in connection with a covered loan after the second consecutive attempt has failed due to a lack of sufficient funds, unless the lender obtained the consumer’s new and specific authorization to make further withdrawals from the account. The Bureau received many comments from stakeholders on all sides of this issue, which are reviewed and addressed below. In sum, after having reviewed the comments, the Bureau concludes that the practice preliminarily identified in the proposal is unfair. It causes substantial injury to consumers because borrowers subjected to the practice incur repeated fees. Based on the Bureau’s study of online payday and payday installment lending, about two percent of borrowers in the market are subject to the practice, and of those subject to the practice, most previously incurred NSF or 998 12 U.S.C. 5531(c). 982 overdraft fees associated with the second failed attempt and more than 80 percent incurred additional NSF or overdraft fees as a result of the third, fourth, and further attempts, which are now prohibited. The practice is not reasonably avoidable because it is difficult to stop payments at the borrower’s account-holding institution, and difficult to revoke payment authorizations. The injury is not outweighed by countervailing benefits to consumers or competition. Third and subsequent re-presentments have low expected values because of how often they fail, and consumers otherwise see very little benefit when lenders are allowed to re-present after two failed attempts without a new borrower authorization. 1. Causes or Is Likely to Cause Substantial Injury Proposed Rule As noted in part IV, the Bureau’s interpretation of the various prongs of the unfairness test is informed by the FTC Act, the FTC Policy Statement on Unfairness, and FTC and other Federal agency rulemakings and related case law.999 Under these authorities, substantial injury 999 Over the past several decades, the FTC and Federal banking regulators have promulgated a number of rules addressing acts or practices involving financial products or services that the agencies found to be unfair under the FTC Act (the 1994 amendments to which codified the FTC Policy Statement on Unfairness). For example, in the Credit Practices Rule that the FTC promulgated in 1984, the FTC determined that certain remedies that creditors frequently included in credit contracts for use when consumers defaulted on the loans were unfair, including confessions of judgments, irrevocable wage assignments, security interests in household goods, waivers of exemption, pyramiding of late charges, and cosigner liability. 49 FR 7740 (Mar. 1, 1984) (codified at 16 CFR part 444). The D.C. Circuit upheld the FTC rule as a permissible exercise of unfairness authority. AFSA, 767 F.2d at 957 (1985). The Federal Reserve Board adopted a parallel rule applicable to banks in 1985. (The Federal Reserve Board’s parallel rule was codified in Regulation AA, 12 CFR part 227, subpart B. Regulation AA has been repealed as of March 21, 2016, following the Dodd-Frank Act’s elimination of the Federal Reserve Board’s rule-writing authority under the FTC Act. See 81 FR 8133 (Feb. 18, 2016)). In 2009, in the HPML Rule, the Federal Reserve Board found that disregarding a consumer’s repayment ability when extending a higher-priced mortgage loan or HOEPA loan, or failing to verify the consumer’s income, assets, and obligations used to determine repayment ability, is an unfair practice. See 73 FR 44522 (July 30, 2008). The Federal Reserve Board relied on a statutory basis for its exercise of unfairness authority pursuant to TILA section 129(l)(2), 15 U.S.C. 1639(l)(2) (renumbered to 15 U.S.C. 1639(p)(2), which incorporated the provisions of HOEPA. The Federal Reserve Board interpreted the HOEPA unfairness standard to be informed by the FTC Act unfairness standard. See 73 FR 44529 (July 30, 2008). That same year, the Federal Reserve Board, the OTS, and the NCUA issued the interagency Subprime Credit Card 983 may consist of a small amount of harm to a large number of individuals or a larger amount of harm to a smaller number of individuals. As the Bureau discussed in the proposal, the lender act or practice of attempting to withdraw payment from a consumer’s account in connection with a covered loan after the lender’s second consecutive attempt has failed due to a lack of sufficient funds, unless the lender obtains the consumer’s new and specific authorization to make further withdrawals from the account, appears to cause or to be likely to cause substantial injury to consumers. And each additional attempt by the lender is likely to trigger substantial additional fees for the consumer but is unlikely to result in successful collection for the lender. These additional attempts can cause serious injury to consumers who are already in substantial financial distress, including the cumulative fees that the consumers owe to both the lender and their account-holding institution. Specifically, the Bureau conducted an analysis of online lenders’ attempts to collect payments through the ACH system on loans with various payment structures, including payday loans with a single balloon payment and high-cost installment loans, typically with payments timed to coincide with the consumer’s payday. The Bureau’s analysis indicated that the failure rate after two consecutive unsuccessful attempts is 73 percent, even when re-presentments appear to be timed to coincide with the consumer’s next payday or the date of the next scheduled payment, and further worsens on subsequent attempts.1000 Return rates for resubmissions of Practices Rule, where the agencies concluded that creditors were engaging in certain unfair practices in connection with consumer credit card accounts. See 74 FR 5498 (Jan. 29, 2009). 1000 The analysis indicates that of the 20 percent of payment requests following a second failed payment request that occur between 14 and 15 days, 84 percent fail. CFPB Online Payday Loan Payments, at 16. In addition, the analysis indicates that while re-presentments at 30 days are rare, more than half of all that occur at 30 days fail. Id. at 17. In the Bureau’s analysis, these data show that even if the re-presentment is on the consumer’s next payday, 984 returned signature checks, RCCs, and RCPOs through the check system are not as readily observable. Nonetheless, it is reasonable to assume that lenders’ resubmissions of failed payment withdrawal attempts through the check-clearing system would yield high failure rates as well.1001 Similarly, when a lender that is also the consumer’s account-holding institution has already initiated two consecutive failed internal transfers to withdraw payment on a loan, despite having more information about the condition of the consumer’s account than other lenders generally have, there is no reason to assume that the lender’s next attempt to withdraw payment from the severely distressed account is any more likely to yield better results.1002 Consumers who are subject to the lender practice of attempting to withdraw payment from an account after two consecutive attempts have failed are likely to have incurred two NSF fees from their account-holding institution1003 and, where permitted, two returned-payment fees from the lender by the time the third attempt is made. Accordingly, these consumers already may have incurred more than $100 in fees in connection with the first two failed attempts. As a result of lenders’ attempts to withdraw payment from their accounts after the failure of a second which is likely to be the date of the consumer’s next scheduled payment on an installment loan, it is also likely to fail. Id. At 17 fig. 3. 1001 Indeed, as discussed in the proposal, information reported by storefront lenders suggests that when such lenders make payment withdrawal attempts using the consumer’s check—typically in cases where the consumer does not come into the store to repay—the failure rates for such attempts are as high as or higher than those for presentments through the ACH system. 1002 As discussed in the proposal, the Bureau is aware of some depository institutions that have charged overdraft and NSF fees for payments made within the institutions’ internal systems, including a depository institution that charged overdraft and NSF fees on payments related to its small-dollar loan product. The Bureau has decided to exempt depository institutions from this rule when the depository institution is also the account-holding institution and when that depository institution does not charge fees for failed attempts or allow an internal transfer to cause an overdraft or account closure. That decision was made not because these presentments are more likely to succeed, but because in those instances, no fees are charged (either by the lender or by the account-holding institution, which are one and the same), and thus no injury occurs. 1003 Although lenders do not directly charge these particular fees, their actions cause the fees to be charged by the account-holding institution. Furthermore, lenders know that consumers generally will incur fees from their accountholding institutions for failed payments. 985 consecutive attempt, most of these consumers will incur significant additional monetary and other harms. In the vast majority of cases, the third withdrawal attempt fails and thereby triggers additional NSF fees charged by the consumer’s account-holding institution and may trigger additional returned-item fees charged by the lender. Indeed, the Bureau’s evidence with respect to online payday and payday installment loans indicated that 73 percent of consumers who experience a third withdrawal attempt after two prior failures incur at least one additional NSF fee (bringing their total to three and total cost in NSF fees to over $100), 36 percent end up with at least two additional fees, and 10 percent end up with at least three additional fees (meaning in most cases they will have been charged approximately $175 in fees by their account-holding institution). When returned-item fees are added, that can double these costs. These lender fees may be imposed even for returned or declined payment withdrawal attempts for which the account-holding institution may not charge a fee, such as attempts made by debit cards and certain prepaid cards. Moreover, in the relatively small number of cases in which such a withdrawal attempt does succeed, Bureau research suggests that roughly one-third of the time, the consumer is likely to have been charged an overdraft fee of approximately $34.1004 In addition to incurring these types of fees, in the proposal, the Bureau preliminarily found that consumers who experience two or more consecutive failed lender payment attempts appear to be at greater risk of having their accounts closed by their account-holding institution. 1004 Thus, even when the consumer does not incur NSF fees from her account-holding institution as a result of a lender payment withdrawal attempt made in connection with a covered loan after two consecutive attempts have failed, the consumer still has a roughly one-in-three chance of incurring an overdraft fee as a result of the subsequent lender attempt. Moreover, at the time lenders choose to make further attempts to withdraw payment from the account, the lenders should be on notice that the account is severely distressed (as evidenced by the prior two consecutive returns) and that additional attempts thus are likely to cause further injury to the consumer, be it from NSF fees, lender-charged returned-item fees or, as the Bureau’s analysis indicates, overdraft fees charged by the consumer’s account-holding institution. 986 Specifically, the Bureau’s analysis of ACH payment withdrawal attempts made by online payday and payday installment lenders indicates that 43 percent of accounts with two consecutive failed lender payment withdrawal attempts were closed by the depository institution, as compared with only three percent of accounts generally.1005 Comments Received The primary thrust of the comments that claimed the Bureau had not satisfied this element was that the Bureau either had insufficient evidence or had evidence that was inapplicable to certain sub-categories of products—such as longer-term installment loans, bank loans, or loans made by Tribal entities or, relatedly, that the Bureau’s evidence was only applicable to online lending. There were also various other discrete comments. Some commenters suggested that identification of the third payment attempt as injurious as opposed to, for example, the fifth attempt, was arbitrary. Others suggested that even the second payment attempt is injurious and should be constrained under the terms of the rule. Commenters claimed that the Bureau had not shown why submitting payments more than two times is a unique characteristic of covered lenders, and had not shown why it was not similarly injurious when other industries did so. Several commenters identified that the third presentment after two consecutive failed presentments was a small portion of the total number of presentments initiated by lenders of covered loans, thereby suggesting that the injury was not substantial. Some commenters also noted that the Bureau had not provided evidence showing that covered lenders have knowledge of the fact that their actions will result in repeated fees at 1005 CFPB Report on Supplemental Findings, at Chapter 6. 987 consumers’ authorizing banks. Others claimed that the lenders covered by the proposed rule were not the cause of the injury, but rather it was the consumers’ banks that caused the injury. A number of commenters objected to the Bureau’s assertion that its evidence suggested that some account closures were caused by the identified practice. A few commenters argued that fees were not necessarily injury, and others suggested that some of the affected consumers were fraudsters or never intended to repay, and thus should not be considered injured parties. Final Rule After having reviewed the comments received, the Bureau concludes that the practice of attempting to withdraw payment from a consumer’s account in connection with a covered loan after the lender’s second consecutive attempt to withdraw has failed due to a lack of sufficient funds, unless the lender obtains the consumer’s new and specific authorization for the withdrawal, causes or is likely to cause substantial injury. It is true that the Bureau’s proposal relied significantly on a study of re-presentments and ACH withdrawal attempts in the online payday and payday installment lending market. But the Bureau relied on other data as well. For example, as stated above, one very large depository institution presented its own statistical analysis demonstrating that storefront and online lenders shared a 25% overall return rate, as compared to the 1.36% return rate industry-wide. And the Bureau reviewed the financial records of lenders that provide covered loans other than online loans, and preliminarily found disclosures of high return rates and/or a practice of engaging in representments.1006 1006 QC Holdings 2014 Annual Report (Form 10-K), at 7 (reporting a return rate of 78.5 percent); Advance America 2011 Annual Report (Form 10-K), at 27 (reporting return rates of 63 percent for checks and 64 percent for ACH 988 But more generally, the Bureau agrees with commenters that injury would result when any vendor initiates a third withdrawal attempt after two failed attempts (absent a new and specific authorization). The Bureau decided to take action as to lenders of the loans covered by this rule because the Bureau has reason to find, based on evidence and data available to it, that lenders in these markets are or were engaged in the identified practice, per the discussion in Market Concerns—Payments above. Were the Bureau presented with evidence that other markets are also engaged in the practice, it would consider expanding this rule. The Bureau does not agree that the evidence before it suggests that third and subsequent presentments (which, again, are second re-presentments) result in a small amount of injury. Of the borrowers who are subjected to a third presentment, the data showed that 73 percent incur an NSF fee and an additional 8 percent incur an overdraft fee. As the Bureau noted in the Market Concerns—Payments section, and as commenters correctly noted, the Bureau’s study showed that around two percent of all initial presentments were followed by two more attempts. The average overdraft and NSF fee was around $34, which means 1.6 percent of all initial payment attempts involved an estimated $34 in injury from a third payment attempt. Given the size of the market, the injury caused just by third presentments alone is substantial, amounting to millions of dollars. The Bureau also analyzed the harms of the practice in a different manner—by looking at the total percentage of payment requests that this rule would prevent, and the average overdraft attempts); First Cash Fin. Servs., 2014 Annual Report (Form 10-K), at 5 (Feb. 12, 2015) (explaining that provider of online and storefront loans subsequently collects a large percentage of returned ACH and check payments by redepositing the customers’ checks, ACH collections, or receiving subsequent cash repayments by the customers); CashNet USA, “Frequently Asked Questions,” https://www.cashnetusa.com/faq.html (last visited Dec. 18, 2015) (“If the payment is returned for reason of insufficient funds, the lender can and will re-present the ACH Authorization to your bank”). 989 and NSF fees that the rule will prevent from being charged per impacted borrower. Based on the Bureau’s study, around seven to ten percent of all presentments in the studied market consisted of a presentment after at least two consecutive failed attempts, while the average borrower subjected to the practice incurred an average of $64 to $87 in overdraft and NSF fees as a result of the practice.1007 Notably, these estimates do not take into consideration all the further risks and harms that occur to some consumers whose accounts are closed as a result of these situations. When adding to that the fee amounts charged cumulatively for further re-presentments, which occur in certain instances, plus the unquantifiable amounts for return fees charged by lenders themselves, the injury is even more substantial. Additionally, this injury would be incurred by borrowers who are more likely to be unable to absorb small to midsized financial burdens. The impact is likely to be significant given that impacted borrowers will have already incurred fees after the first two failed payment attempts. Also, as noted in Market Concerns—Underwriting, consumers of covered loans are typically in financial distress, which is often the reason for seeking covered loans in the first place. For a borrower that is in financial distress, incurring an average of $64-$87 in bank fees, plus any lender return fees and the risk of account closure, after having already incurred approximately $70 in bank fees and additional lender fees due to the first two failed payment attempts, would be quite substantial. As for the decision to finalize a limit of two representments, the Bureau recognizes that every re-presentment—whether the first, second, third, fourth, or any other ordinal—individually generates fees, and hence causes injury to consumers. 1007 CFPB Report on Supplemental Findings, at Chapter 6. 990 In fact, looking individually at each presentment, the fee injury is likely identical for each instance (one NSF fee, overdraft fee, and perhaps return fee). But the Bureau does not view the injury and benefits of each additional presentment individually. Instead, it takes into account the cumulative impact of the string of presentments. The Bureau did not decide on a limit of two representments because the first re-presentment does not cause injury. It did so because the injury after each failed attempt is cumulative, meaning the injury after two re-presentments is approximately double the injury after one, and the first re-presentment implicates certain additional countervailing benefits.1008 Lenders may have simply tried the first presentment at the wrong time, and consumers may find it convenient to not have to reauthorize after one failed attempt. The Bureau draws the line at two re-presentments in an abundance of caution, in an attempt to avoid regulating potentially more legitimate justifications for re-presentment. But this discussion should not be interpreted to minimize the harms that can occur even from a single representment. Indeed, depending on the facts and circumstances, even payment practices involving a single re-presentment may be unfair, deceptive, or abusive. The Bureau also notes that this rule does not provide a safe harbor against misconduct that is not explicitly addressed by the rule, and the Bureau can and will continue to monitor these practices under its supervisory and enforcement authorities, and will take appropriate action as warranted by the circumstances.1009 1008 Note that the Bureau’s study, CFPB Online Payday Loan Payments, found that the second payment request had a 70 percent failure rate, while the third had a 73 percent failure rate. CFPB Online Payday Loan Payments at 13. 1009 This discussion reflects the fact that rules identifying and preventing certain unfair or abusive practices as determined on a categorical basis—as is true, for example, of this rule—do not divest the Bureau of authority to 991 The Bureau disagrees with commenters’ assertions that the identified practice does not cause the injury, either because consumers’ banks were the primary cause or because the Bureau did not prove that the lender knew fees would result. One commenter argued more specifically that lenders are not responsible for overdraft fees because borrowers opt in to overdraft fees with their banks. Another argued that fees are not necessarily an injury. As an initial matter, actual knowledge of the harm is not a requirement for an unfairness finding.1010 Even if it were, the Bureau assumes that market participants understand the natural consequences of their actions. Additionally, the fact that consumers’ banks are the actors that actually charge the fees does not suggest that the identified practice does not cause the substantial injury. The “contribution of independent causal agents” does not erase the role lenders play in causing the harm.1011 The Bureau’s proposal provided ample evidence that lenders are aware of high rejection rates, and any industry participant should know that a natural consequence of rejected transfers is that the consumer will incur fees. The Bureau study analyzed overdraft fees charged in connection with ACH transactions. Fees on such transactions are not subject to an opt-in requirement like overdraft fees on debit card transactions, meaning that while it is true borrowers may have opted into overdraft fees for some instances, that is not true for many instances in which overdraft fees address other unfair, deceptive, or abusive acts or practices that are identified in the particular facts or circumstances of a specific examination or enforcement investigation. For example, the Bureau has taken enforcement action in cases that involved payment practices which do not specifically track the unfair and abusive practice that is identified in § 1041.7. See, e.g., Consent Order, In the Matter of EZCORP, Inc., No. 2015-CFPB-0031 (Dec. 16, 2015). 1010 FTC v. Neovi, 604 F.3d 1150, 1156 (9th Cir. 2010). 1011 Neovi, 604 F.3d at 1155 (9th Cir. 2010). 992 are incurred. Further, it is a settled matter that fees which borrowers cannot reasonably avoid should be considered injury.1012 It may be true that some of the affected consumers may be fraudsters, or never intended to repay their loans. To the extent a person had used another individual’s account number, any re-presentments would further victimize a victim of identity theft. But the Bureau agrees that there may be a small population of borrowers who took out a loan with no intention of trying to repay either the loan or any associated bank fees. This small population of borrowers does not change the Bureau’s overall assessment of whether there was substantial injury, or whether that injury was outweighed by countervailing benefits. Lastly, several commenters stated that the Bureau’s evidence on high account-closure rates did not prove that the identified practice caused all of the closures. The Bureau acknowledged in the proposal that some accounts could be closed for other reasons. To the extent depository institutions do involuntarily close accounts as a result of repeated failed presentments, that result is injury. And one commenter provided a study in which 22 percent of the surveyed payday consumers did self-report that their account was closed because of payday loans.1013 The Bureau does not know the full extent of how often borrowers’ accounts are closed due to multiple presentments, but it can point to evidence showing that payday borrowers’ accounts are closed involuntarily much more often than other consumers. It is reasonable to assume that some portion of the closures result from the practice and some are a result of other 1012 FTC Statement on Unfairness, Appended to International Harvester Co., 104 F.T.C. 949, 1070 (1984) (“In most cases a substantial injury involves monetary harm.”). 1013 Pew Charitable Trusts, “Payday Lending in America Fraud and Abuse Online: Harmful Practices in Internet Payday Lending, at 16 (Report 4, 2014), available at http://www.pewtrusts.org/~/media/Assets/2014/10/PaydayLending-Report/Fraud_and_Abuse_Online_Harmful_Practices_in_Internet_Payday_Lending.pdf. 993 circumstances. Either way, the Bureau neither thinks this injury is necessary to make the total injury “substantial,” nor that it tips the balance regarding whether the injury is outweighed by countervailing benefits. 2. Injury Not Reasonably Avoidable Proposed Rule As previously noted in part IV, under the FTC Act and Federal precedents that inform the Bureau’s interpretation and application of the unfairness test, an injury is not reasonably avoidable where “some form of seller behavior . . . unreasonably creates or takes advantage of an obstacle to the free exercise of consumer decision-making,” or unless consumers have reason to anticipate the injury and the means to avoid it. In the proposal, the Bureau observed that in a significant proportion of cases, unless the lender obtains the consumer’s new and specific authorization to make further payment withdrawals from the account, consumers may be unable to reasonably avoid the injuries that result from the lender practice of attempting to withdraw payment from a consumer’s account in connection with a covered loan after two consecutive payment withdrawal attempts by the lender have failed. The Bureau noted that consumers could avoid the above-described substantial injury by depositing into their accounts enough money to cover the lender’s third payment withdrawal attempt and every attempt that the lender may make after that, but that for many consumers this is not a reasonable or even an available way of avoiding the substantial injury discussed above. Even if a consumer had sufficient funds to do so and knew the amount and timing of the lender’s next attempt to withdraw payment, which are unlikely to be the case, any funds deposited into the consumer’s account likely would be claimed first by the consumer’s bank to repay the NSF 994 fees charged for the prior two failed attempts. Thus, even a consumer who had some available cash could have difficulties in avoiding the injury resulting from the lender’s third attempt to withdraw payment, as well as in avoiding the injury resulting from any attempts that the lender may make after the third one.1014 Moreover, as a practical matter, in the vast majority of cases in which two consecutive attempts to withdraw payment have failed, the consumer is in severe financial distress and thus does not have the money to cover the next payment withdrawal attempt.1015 Although the Bureau’s consumer testing indicates that consumers generally have a strong commitment to repaying their legal obligations,1016 a consumer who has already experienced two consecutive failed payment attempts and incurred well over $100 in related fees may at that point consider, as the only other options to avoid further fee-related injury, either closing the account or attempting to stop payment or revoking authorization. Given that consumers use their accounts to conduct most of their household financial transactions, the Bureau did not believe that voluntarily closing down the account was a reasonable means for consumers to avoid injury. Further, as discussed in the proposal, the option of attempting to stop payment or revoke authorization is not a reasonable means of avoiding the injuries either, for several reasons. First, 1014 In proposed § 1041.15, the Bureau proposed to require lenders to provide a notice to consumers in advance of each payment withdrawal attempt. The Bureau believed that the notices would help consumers make choices that may reduce potential harms from a payment withdrawal attempt—by reminding them, for example, to deposit money into their accounts prior to the attempt and thus avoid a late payment fee. The Bureau’s treatment of these issues is discussed further below in the section-by-section analysis of § 1041.9 of the final rule. 1015 The Bureau noted that even when consumers have agreed to make a series of payments on an installment loan, the substantial injuries discussed above are not reasonably avoidable, based on its analysis of ACH payment withdrawal attempts made by online payday and payday installment lenders, which indicates that after two failed presentments, even payment withdrawal attempts timed to the consumer’s next payday, which is likely to be the date of the next scheduled payment on an installment loan, are likely to fail. 1016 FMG Report, “Qualitative Testing of Small Dollar Disclosures, Prepared for the Consumer Financial Protection Bureau,” at 53 (Apr. 2016) available at http://files.consumerfinance.gov/documents/Disclosure_Testing_Report.pdf. 995 as listed in the Market Concerns—Payments section above, consumers often face considerable challenges in issuing stop-payment orders or revoking authorization as a means to prevent lenders from continuing to attempt to make payment withdrawals from their accounts. Complexities in payment processing systems and the internal procedures of consumers’ accountholding institutions, combined with lender practices, often make it difficult for consumers to stop payment or revoke authorization effectively. With respect to preauthorized EFTs authorized by the consumer, for example, even if the consumer successfully stops payment on one transfer, the consumer may experience difficulties in blocking all future transfers by the lender. In addition, payment withdrawal attempts made via RCC or RCPO can be especially challenging for the consumer’s account-holding institution to identify and be able to stop payment on them. Various lender practices exacerbate these challenges. Lenders often obtain several different types of authorizations from consumers—e.g., authorizations to withdraw payment via both ACH transfers and RCCs—such that if the consumer successfully revokes one type of authorization, the lender has the ability to continue making payment collection attempts using another type of authorization. The procedures of consumers’ account-holding institutions for stopping payment often vary depending on the type of authorization involved. Thus, when a lender has obtained two different types of authorizations from the consumer, the considerable challenges associated with stopping payment or revocation in connection with just one type of authorization are effectively doubled. Many consumers also may not understand that they must navigate two different sets of stop-payment or revocation procedures to prevent the lender from making additional withdrawal attempts. 996 In addition, the costs to the consumer for issuing a stop-payment order or revoking authorization are often as high as some of the fees that the consumer is trying to avoid, as depository institutions charge consumers a fee of approximately $32, on average, for placing a stop-payment order. The consumer incurs this fee regardless of whether the consumer is seeking to stop payment on a check, a single EFT, or all future EFTs authorized by the consumer. Moreover, issuing a stop-payment order at a cost of $32 does not guarantee success. Some depository institutions require the consumer to provide the exact payment amount or the lender’s merchant ID code, and thus fail to block payments when the payment amount varies or the lender varies the merchant code. In addition, some depository institutions require consumers to renew stop-payment orders after a certain period of time. In such cases, consumers may incur more than one stop-payment fee in order to continue blocking future payment withdrawal attempts by the lender. As a result of these stop-payment fees, the cost to the consumer of stopping payment with the consumer’s account-holding institution is comparable to the NSF or overdraft fee that the institution would charge the consumer if the payment withdrawal attempt that the consumer is seeking to stop were made. Thus, even if the consumer successfully stops payment, they would not avoid this particular fee-related injury, but rather would be exchanging the cost of one comparable fee for another. In addition, some consumers may be charged a stop-payment fee by their account-holding institution even when, despite the stop-payment order, the lender’s payment withdrawal attempt goes through. In such cases, the consumer may be charged both a fee for the stop-payment order and an NSF or overdraft fee triggered by the lender’s payment withdrawal attempt. 997 In addition to the challenges consumers face when trying to stop payment or revoke authorization with their account-holding institutions, consumers often face lender-created barriers that prevent them from pursuing this option as an effective means of avoiding injury. Lenders may discourage consumers from pursuing this course of action by including language in loan agreements purportedly prohibiting the consumer from stopping payment or revoking authorization. In some cases, lenders may charge consumers a substantial fee in the event that they successfully stop payment with their account-holding institution. Lenders’ procedures for revoking authorizations directly with the lender create additional barriers. As discussed in the proposal, lenders often require consumers to provide written revocation by mail several days in advance of the next scheduled payment withdrawal attempt. A consumer who took out the loan online, but now wishes to revoke authorization, may have difficulty even identifying the lender that holds the authorization, especially if the consumer was paired with the lender through a third-party lead generator. These lender-created barriers make it difficult for consumers to stop payment or revoke authorization. Comments Received Several industry commenters stated that the substantial injury identified by the Bureau could be reasonably avoided by consumers because consumers could choose not to borrow, and do not need to agree to a leveraged payment mechanism. Others claimed that borrowers have the ability to revoke authorizations and stop payments, and that these options make the injury reasonably avoidable. Some also claimed that the Bureau overestimated or had no evidence of the difficulty in obtaining a stop-payment order or revoking the authorization. 998 A number of industry commenters argued that borrowers should simply place sufficient funds in their account or pay the lender before the scheduled transfer date, and should generally be aware that fees would result from failed payment withdrawals. Still other commenters claimed that borrowers could avoid the injury by re-borrowing. Final Rule After reviewing the comments received, the Bureau concludes that the substantial injury identified above is not reasonably avoidable by consumers. As an initial matter, the Bureau disagrees with comments that claimed that the Bureau did not have any convincing evidence of the difficulty of obtaining a stop-payment order or revoking an authorization. The proposed rule and the Market Concerns—Payments sections refer to significant evidence on this point.1017 As described above, many lenders have obfuscated or interfered with consumers’ ability to revoke authorization, and stop-payment orders can involve their own fees and are not always comprehensive. In particular, they are quite difficult to process for RCCs and RCPOs. One lender noted that it cancels hundreds of payment authorizations each year, and argued that lenders cannot be held responsible if third-party financial institutions mishandle stoppayments or charge excessive fees. Again, lenders are causing harm that is not reasonably avoidable. That harm manifests itself, and is difficult to avoid, in part because of the actions of third-party financial institutions. Although it is fair to say that lenders do not necessarily bear all the responsibility for any problems that ensue, this does not change the fact that consumers are 1017 See specific Market Concerns—Payments sub-section entitled “Consumers Have Difficulty Stopping Lenders’ Ability to Access Their Accounts” for that evidence. 999 not able to withdraw their prior authorizations or stop payments in a reasonably effective manner. That one lender may process hundreds of canceled payment authorizations each year neither suggests that all of its borrowers who seek to cancel payment authorization are successful, nor suggests that many other lenders do the same thing. The Bureau does not agree that simply repaying is a viable way to avoid the harm. Many borrowers will not have the funds (again, only approximately 20 percent of third presentments succeed without an overdraft fee). But, additionally, as laid out in the Market Concerns— Payments section, subsequent presentments can occur very quickly, often on the same day, making it difficult to ensure funds are in the right account before the re-presentment hits.1018 As in the section-by-section analysis for § 1041.4, the Bureau finds that simply replacing the injury with re-borrowing is not a satisfactory mechanism for reasonably avoiding the harm because it simply substitutes one injury for another. The Bureau has discussed, at length, the harms incurred by repeated re-borrowing in the section-by-section analysis of part B. Moreover, under the traditional unfairness analysis established by prior precedents, the suggestion that a consumer can simply decide not to participate in the market is not considered to be a valid means of reasonably avoiding the injury.1019 The Bureau addressed a similar line of comments in subpart B, and noted that if this view were adopted, no market practice could ever be determined to be unfair. That response is applicable here as well. 1018 In one demonstrative enforcement case, the Bureau found a payday and installment lender that regularly made three debit attempts on the same day. Consent Order, In the Matter of EZCORP, Inc., No. 2015-CFPB-0031 (Dec. 16, 2015). 1019 See, e.g., 49 FR 7740 (Mar. 1, 1984). 1000 As stated in the proposal and above, lenders often take broad, ambiguous payment authorizations from consumers and vary how they use these authorizations, thereby increasing the risk that consumers will be surprised by the amount, timing, or channel of a particular payment. Borrowers do not have the ability to shop, at the time of origination, for covered loans without leveraged payment mechanisms, as that is a central feature of these loans. As some commenters noted, leveraged payment mechanisms are sometimes even required by State law. 3. Injury Not Outweighed by Countervailing Benefits to Consumers or Competition Proposed Rule As noted in part IV, the Bureau’s interpretation of the various prongs of the unfairness test is informed by the FTC Act, the FTC Policy Statement on Unfairness, and FTC and other Federal agency rulemakings and related case law. Under those authorities, the countervailing benefits prong of the unfairness standard makes it appropriate to consider both the costs of imposing a remedy and any benefits that consumers enjoy as a result of the practice; yet this determination does not require a precise quantitative analysis of benefits and costs. The Bureau preliminarily found that the lender practice of making additional payment withdrawal attempts from a consumer’s account in connection with a covered loan after two consecutive attempts have failed does not generate benefits to consumers or competition that outweigh the injuries caused by the practice. As discussed above, a substantial majority of additional attempts are likely to fail. Indeed, the Bureau’s analysis in the proposal of ACH payment withdrawal attempts made by online payday and payday installment lenders preliminarily found that the failure rate on the third attempt is 73 percent, and it increases to 83 percent on the fourth attempt, and to 85 percent on the fifth attempt. Furthermore, of those 1001 attempts that succeed, 33 percent or more succeed only by overdrawing the consumer’s account and generally incurring fees for the consumer. When a third or subsequent attempt to withdraw payment does succeed, the consumer making the payment may experience some benefit in the form of avoiding further collection activity and consumer reporting, to the extent the lender is reporting the delinquency. According to the Bureau’s study, it appears that third presentments succeed approximately 20 percent of the time without an overdraft fee, while an additional eight percent succeed with an overdraft fee. In any event, the Bureau preliminarily found that to the extent some consumers are able, after two consecutive failed attempts, to muster sufficient funds to make the next required payment or payments, these consumers would be able to arrange to make their payment or payments even if lenders were first required to get a new and specific authorization from the consumer before making additional payment attempts. Turning to the potential benefits of the practice to competition, the Bureau recognizes that to the extent payment withdrawal attempts succeed when made after two consecutive failed attempts, lenders may collect larger payments or may collect payments at a lower cost by seeking payment from the consumer’s account rather than being required to seek payment directly from the consumer. Given their high failure rates, however, these additional attempts generate relatively small amounts of revenue for lenders. For example, the Bureau’s analysis of ACH payment withdrawal attempts made by online payday and payday installment lenders indicates that whereas the expected value of a first payment request is $152, the expected value of a third 1002 successive payment attempt is only $46, and that the expected value drops to $32 for the fourth attempt and to $21 for the fifth attempt.1020 Furthermore, the Bureau indicated that lenders could obtain much of this revenue without making multiple attempts to withdraw payment from demonstrably distressed accounts. For instance, lenders could seek payments in cash or “push” payments from the consumer or, in the alternative, could seek a new and specific authorization from the consumer to make further payment withdrawal attempts. Indeed, coordinating with the consumer to seek a new authorization may be more likely to result in successful payment withdrawal attempts than does the practice of repeatedly attempting to withdraw or transfer funds from an account in distress. Finally, in view of the pricing structures observed in the markets for loans that would be covered under the proposed rule, the Bureau preliminarily found that any incremental revenue benefit to lenders from subsequent attempts, including revenue from the fees charged for failed attempts, does not translate into more competitive pricing. In other words, the Bureau preliminarily found that prohibiting such attempts would not adversely affect pricing. In sum, the Bureau preliminarily determined in the proposal that consumers incur substantial injuries as a result of the identified practice that are not outweighed by the minimal benefits that this practice generates for consumers or competition. Comments Received Several industry commenters stated that the cost of credit would increase as a result of the remedy proposed by the Bureau, which the commenters interpreted to include the burden of sending payment reminders and of tracking unsuccessful debit attempts and new payment 1020 Expected values are calculated by multiplying the average successful payment amount by the success rate. 1003 authorizations. Many commenters argued more generally that covered loans help borrowers, improve financial health, or are otherwise beneficial. Some commenters argued that recurring payment authorizations are a benefit to consumers because they are more convenient and enable consumers to designate their due date around the timing of when they will have available funds. Some commenters argued that consumers would feel frustrated and inconvenienced whenever a lender is required to request a new and specific authorization. Still others argued that barring withdrawals after the second attempt would limit payment options that are available to consumers. Finally, some argued that limiting payment attempts would harm consumers by causing them to default or slip further into delinquency. Final Rule After reviewing the comments received, the Bureau concludes that the substantial injury identified above is not outweighed by countervailing benefits to consumers or competition. A number of industry commenters presented arguments that would be inappropriate to consider in the weighing of countervailing benefits against consumer injury. First, several commenters argued that the costs of complying with the notices and disclosures that would be provided in proposed § 1041.15 constitute compliance costs that should be considered as the Bureau weighs countervailing benefits. Because that remedy is a result of exercising the Bureau’s authority under section 1032 of the Dodd-Frank Act, and does not result from this finding of unfairness, the Bureau does not consider that remedy as part of its countervailing benefits analysis. Instead, it considers only the cost of those remedies that are being required to remediate the injury from the identified practice. It also did not identify the notices contained in proposed § 1041.15 as a remedy for the identified practice. 1004 Second, commenters’ claims that covered loans are generally beneficial, and that this should be accounted for in the weighing of benefits, cast too wide a net. The Bureau is not identifying the unfair practice as making covered loans, or even making covered loans with leveraged payment mechanisms. The Bureau is taking a much narrower approach here, by identifying the unfair practice as being limited to making a third payment request after two failed attempts, without first obtaining a new and specific payment authorization. The general benefits these commenters posit from the making of covered loans are not a result of that practice, and the Bureau has no reason to believe lenders will not make covered loans because they are unable to re-present after two attempts without obtaining a new authorization. Third, because the Bureau is not prohibiting leveraged payment mechanisms, it does not consider the convenience of recurring payment authorizations, or scheduled payments, to be a benefit for purposes of this analysis. Lenders can still provide the benefits to consumers of convenience and scheduling after this rule is finalized. In other words, those benefits are not a result of the identified practice, which is the initiation of additional payment requests after two failed attempts, absent a new and specific authorization. Commenters have correctly identified the cost of tracking unsuccessful debits and of either securing new payment authorizations or obtaining payment through other means if two consecutive presentments fail as a cost of compliance applicable to this analysis. The effect that this cost will have on pricing is mitigated by other market forces including the fact that, as noted in the proposal, many loans in this market are priced at the maximum possible price permitted under State law. Nonetheless, these are costs the market must bear and some of those costs may be passed to consumers. Our analysis suggests that those costs likely will not be overly 1005 substantial because lenders already have processes in place to track payment attempts, and thus will only need to augment them slightly to accommodate the particular details for this rule (see Section 1022(b)(2) Analysis in part VII for more on this point). These costs are not sufficient to change the Bureau’s overall conclusion that the substantial injury to consumers outweighs the countervailing benefits. The Bureau does not agree that the consumer frustration caused by requests for new and specific payment authorizations would be significant. These requests would provide consumers with a choice about whether the lender can debit the consumer’s bank account. Especially after two failed attempts, and the likely resulting fees, the Bureau judges that it is very likely that consumers will benefit from the opportunity to decide whether another attempt should occur. The Bureau’s conclusion on this point is consistent with its statutory objective to ensure that “consumers are provided with timely and understandable information to make responsible decisions about financial transactions.”1021 Commenters argued that some borrowers could default or slip further into delinquency if the payment would have succeeded, but had not gone through because of the limitations created by the rule. As the Bureau stated in the proposal, however, borrowers will retain the ability to choose to pay their loans as they wish, including by reauthorizing automatic debits. Although there may be some borrowers for whom a third or subsequent presentment would succeed but who would not manage to repay the loan absent such presentments, the Bureau believes that this population is too small to affect the countervailing benefits analysis. 1021 12 U.S.C. 5511(b)(1). 1006 Lastly, the Bureau addressed the fact that the rule will limit consumers’ payment options in the proposal. The rule covers all payment methods, and thus affects them evenly. To the extent that it limits payment options after two attempts, it limits them to any optional payment method at the specific initiation of the borrower. As consumers will have the choice of whether to re-authorize a payment authorization after two consecutive failed attempts—and they can always use any specifically initiated method for payment—the Bureau determines that the costs associated with limiting payment options (and thus the countervailing benefits of no limits) are quite minimal. 4. Consideration of Public Policy Proposed Rule Section 1031(c)(2) of the Dodd-Frank Act allows the Bureau to “consider established public policies as evidence to be considered with all other evidence” in determining whether a practice is unfair, as long as the public policy considerations are not the primary basis of the determination. This is an optional basis for justifying the rule, and in the proposal the Bureau did not make a preliminary determination to cite public policy as evidence to be considered in deciding that the identified payment practices are unfair. Yet some of the comments received invite further scrutiny of whether public policy should be viewed as a basis for either supporting or undermining the proposed rule. For that reason, the issue will be considered further here. Comments Received Some industry and other commenters suggested that the Bureau’s purported role here is superfluous, since State law governs consumer credit. They argued that some States already cap presentments. They also suggested that the proposed rule may obstruct State efforts to craft 1007 regulatory approaches that appropriately protect consumers, because the Bureau’s proposed intervention would interfere with policy experimentation by the States, and would shift the balance between consumer protection and access to credit in ways not intended by different State regulatory regimes. Rather than develop new provisions in a Federal rule to address these issues, these commenters argued that the Bureau instead should support changes in State law to address concerns about the misuse of payment instruments; or that it should increase its enforcement of existing Federal laws like the EFTA, Regulation E, and the Bureau’s authority to enforce against unfair, deceptive, or abusive acts or practices. Final Rule The Bureau does not find that the public policy considerations raised by some of the commenters militate against the adoption of this final rule. Federal law has governed consumer credit, and specifically electronic payments, for 50 years, dating as far back as the Truth in Lending Act (TILA). The EFTA is the most applicable example, and a Federal rule in this area would be consistent with that history. Ultimately, the issue here is simply whether the Bureau has the legal authority to adopt rules to address the identified practice of making repeated withdrawal attempts after two consecutive failures by first determining that the identified practice is unfair and abusive. Under the Dodd-Frank Act, the Bureau is authorized to do so. That authority is not affected by other provisions of Federal and State law, most notably because those provisions preceded this authorization by Congress. Thus, the more recent statute opened the door to policy changes that would affect the application of those pre-existing legal requirements. Moreover, Congress placed it within the Bureau’s discretion whether to address unfair, deceptive, or abusive acts or practices through enforcement, supervision, regulation, or 1008 some combination of these authorities.1022 By expressly permitting the Bureau to adopt UDAAP rules, as it is doing here, Congress authorized this very endeavor as fully consistent with current notions of sound public policy and the established framework of Federal and State law. b. Abusive Practice Under section 1031(d)(2)(A) and (B) of the Dodd-Frank Act, the Bureau may declare an act or practice abusive if it takes unreasonable advantage of “a lack of understanding on the part of the consumer of the material risks, costs, or conditions of the product or service,” or of “the inability of the consumer to protect the interests of the consumer in selecting or using a consumer financial product or service.”1023 In the proposal, the Bureau preliminarily found that, with respect to covered loans, it is an abusive act or practice for a lender to attempt to withdraw payment from a consumer’s account in connection with a covered loan after two consecutive failed attempts, unless the lender obtains the consumer’s new and specific authorization to make further withdrawals from the account. After reviewing the comments received, as described and responded to below, the Bureau now concludes that the practice identified in the proposal is abusive. Borrowers do not understand the material risks, costs, or conditions that are posed by lenders engaging in repeated re-presentments. Similarly, borrowers are unable to protect their interests in using the product by revoking authorizations or enacting stop payments. Lenders take advantage of these conditions by re-presenting, and those re-presentments are unreasonable. 1022 1023 See 12 U.S.C. 5531(c). 12 U.S.C. 5531(d). 1009 Before delving into the statutory prongs of abusiveness on which the Bureau relies for these conclusions, two broader comments can be addressed here. First, some commenters argued that the Bureau only has the authority to identify a practice as abusive if it “materially interferes with the ability of a consumer to understand a term or condition of a consumer financial product or service.” This suggestion, that section 1031(d)(1) must be satisfied in order to make a finding of abusiveness, is a misreading of the statute. Section 1031(d) articulates four disjunctive categories of abusive practices—this one set forth in section 1031(d)(1), and three others that are set forth in section 1031(d)(2). Congress defined a practice to be “abusive” if it satisfies any of these four independent criteria. Congress clearly indicated as much with its use of the conjunction “or” throughout the text of section 1031(d). Other commenters argued that Congress only intended abusiveness to cover conduct beyond what is prohibited as unfair or deceptive. The Bureau agrees that the abusiveness standard can reach practices that are not covered by the unfairness or deception standards if the prongs of abusiveness are met, but it does not agree that it can only reach practices that are not covered by the unfairness or deception standards. The Bureau is guided and limited by the definitional prongs of unfairness and abusiveness that are expressly articulated in the statute. A practice might meet these standards either alone or in combination (and, of course, lawful practices will meet none of the standards). There is little practical effect of any such overlap, as a practice is just as illegal if it violates one, two, or three of the standards. But as a matter of statutory interpretation, the Bureau has no textual basis to conclude that a practice meeting the statutory prongs of abusiveness cannot be considered abusive because it also meets the prongs of one of the other two standards. 1010 1. Consumers Lack Understanding of Material Risks and Costs Proposed Rule In the proposal, the Bureau stated that when consumers grant lenders an authorization to withdraw payment from their account, they understand as a general matter that they may incur an NSF fee from their account-holding institution as well as a returned-item fee charged by the lender. However, the Bureau preliminarily found that such a generalized understanding does not suffice to establish that consumers understand the material costs and risks of a product or service. Rather, the Bureau determined that it is reasonable to interpret “lack of understanding” in this context to mean more than mere awareness that it is within the realm of possibility that a particular negative consequence may follow or a particular cost may be incurred as a result of using the product. For example, consumers may not understand that such a risk is very likely to happen or that—though relatively rare—the impact of a particular risk would be severe. In this instance, precisely because the practice of taking advance authorizations to withdraw payment is so widespread across markets for other credit products and non-credit products and services, the Bureau preliminarily concluded that consumers lack understanding of the risk they are exposing themselves to by granting authorizations to lenders that make covered loans. Rather, consumers are likely to expect these payment withdrawals to operate in a convenient and predictable manner, similar to the way such authorizations operate when they are granted to other types of lenders and in a wide variety of other markets. Consumers’ general understanding that granting authorization can sometimes lead to fees does not prepare them for the substantial likelihood that, in the event their account becomes severely distressed, the lender will continue making payment withdrawal attempts even after the lender should be on notice (from two 1011 consecutive failed attempts) of the account’s distressed condition. Nor does it prepare them for the result that thereby they will be exposed to substantially higher overall loan costs in the form of cumulative NSF or overdraft fees from their account-holding institution and returned-item fees from their lender, as well as the increased risk of account closure. Moreover, this general understanding does not prepare consumers for the array of significant challenges they will encounter if, upon discovering that their lender is still attempting to withdraw payment after their account has become severely distressed, they take steps to try to stop the lender from using their authorizations to make any additional attempts. Comments Received Industry commenters argued that the Bureau’s findings on abusiveness rested on the unsubstantiated assumption that consumers did not understand the risks of covered loans, or the effects of leveraged payment mechanisms. These commenters questioned the Bureau’s purported reliance on “optimism bias.” Others commented that consumers generally did understand the risks and benefits of covered loans before taking them out. They advanced that awareness of due dates and the fact that payment requests will be initiated, often provided by lenders in conjunction with TILA disclosures, suggest that borrowers understand the material costs and risks of covered loans. Some commenters provided data on borrower expectations about default and re-borrowing, but not about practices around how a lender would use a leveraged payment mechanism to initiate multiple payment requests. Consumer group commenters suggested that the industry acknowledges that covered borrowers do not understand the risks, costs, and conditions of these loans. To support this assertion, one commenter cited a 2016 law review article written by Jim Hawkins, stating that consumers “are overly optimistic.” 1012 One industry commenter stated that “understanding” did not mean anything more than a general sense that a negative consequence would follow. It asserted that consumers did not need to understand both the probability and depth of potential adverse consequences, and cited as support a dictionary definition of “understanding,” which is “to know how (something) works or happens.” It further argued that the level of understanding the Bureau required under the proposed rule was equivalent to expecting a borrower to become an expert on the lending industry. Other commenters said that the Director of the Bureau had once publicly stated that whether a borrower has a lack of understanding is “unavoidably situational” and that abusiveness claims “can differ from circumstance to circumstance.” These commenters claimed that the statements confirmed that the Bureau could not address abusiveness in the market with a general rule, and must exercise its abusiveness authority on a case-by-case basis instead. Final Rule The Bureau now concludes that consumers lack understanding of material risks, costs, or conditions of the product or service, specifically the practice of repeated re-presentments. Evidence suggests that lenders in many non-covered markets take advanced authorizations to initiate electronic payments, yet do not appear to engage in the practice with any particular frequency. This means borrowers do not have experience with the practice, and thus, likely do not understand the specific risks at issue. The contrast in these markets again was shown by the analysis performed by a major financial institution of its consumer depository account data, which estimates ACH return rates for payday lenders, including both storefront and online companies, at 25 percent, with individual lender return rates ranging from five percent to 1013 almost 50 percent,1024 whereas the average return rate for debit transactions in the ACH network across all industries was just 1.36 percent (with the next highest return rate of any other industry being cable television at 2.9 percent, auto and mortgage at 0.8 percent, utilities at 0.4 percent, and credit cards at 0.4 percent).1025 It is reasonable to assume that many of that 25 percent consisted of rejected re-presentments, given that the Bureau’s own data showed a failure rate for first presentments of only six percent for transactions initiated by online payday and payday installment lenders.1026 Six percent is very close to the rejection rates of payday lenders with rejection rates at the low end in the financial institution’s analysis (five percent), suggesting that lenders at the low end may not have been re-presenting. Lenders at the high end, with 50 percent total rejection rates, were likely re-presenting, bringing up the average. The failure rates for representments in the Bureau’s study (70 to 85 percent) were much higher than those for initial presentments.1027 The comparatively much lower return rates in other markets do not similarly suggest high rates of re-presentment, and are more likely to simply constitute the typical rejection rate for initial presentments. This evidence suggests that the covered markets have much higher rates of re-presentment than consumers experience in other markets. Additionally, the Bureau concludes that the complexity of payment presentment practices and their effects makes it likely that a significant number of borrowers lack a sufficient understanding of those practices and their effects. These presentment practices are material 1024 Beth Anne Hastings, “Monitoring for Abusive ACH Debit Practices,” (Presentation by JP Morgan Chase at Spring 2014 NACHA Conference in Orlando, FL, Apr. 7, 2014). See also First Cash Fin. Servs., 2014 Annual Report (Form 10-K), at 5 (“Banks return a significant number of ACH transactions and customer checks deposited into the Independent Lender’s account due to insufficient funds in the customers’ accounts.”) (discussion later in the document indicates that the CSO section covers both online and storefront loans). 1025 NACHA Q4 2014. 1026 CFPB Online Payday Loan Payments, at 13. 1027 CFPB Online Payday Loan Payments, at 13. 1014 because they could result in significant risks and costs to the borrower, including NSF fees, overdraft fees, returned payment fees, and potentially account closures. The Bureau does not rest its legal conclusion on the premise that borrowers are unaware that when they take out covered loans with leveraged payment mechanisms, a payment will be deducted on the due date. Nor does it rest on the premise that borrowers are unaware that when a payment is deducted, and the account lacks the funds to cover the payment, they are likely to incur a fee. Rather, the Bureau concludes that consumers are unaware of the severity of the risk they are exposing themselves to in the circumstances of the identified practice. In other words, the Bureau’s analysis rests on the fact that borrowers are not aware of the risks and harms associated with engaging in the identified practice of multiple re-presentments. The risks, costs, or conditions of covered loans that borrowers do not understand are based on the fact that lenders will re-present repeatedly when borrowers default. Those risks, costs, or conditions are material because—as stated in the unfairness analysis above—borrowers incur substantial injury in the form of fees that are charged and other consequences of the identified practice when lenders repeatedly re-present payments. Data provided by commenters on borrower expectations about default and re-borrowing did not pertain to how lenders use leveraged payment mechanisms to initiate multiple payment requests and thus were not germane to the identified practice here. Many of the commenters’ arguments around whether consumers understand the risks, costs, or conditions of the covered loans focused on the fact that consumers knew a payment would be requested once, knew there would be fees, or knew about the likelihood of default. But those are not the risks, costs, or conditions at issue here, which, again, stem from multiple representments. Similarly, commenters’ assertions about the Bureau’s reliance on “optimism bias” 1015 —which rests on the assumption that borrowers are overly optimistic that they will be able to repay their loans—are misplaced here. The Bureau is not relying on the premise that borrowers underestimate the likelihood of default or re-borrowing for this part of the rule. Instead, the Bureau is merely concluding that borrowers underestimate the extent of fees resulting from default, because most of them have no basis to recognize that a lender will present multiple times in quick succession after the first payment request fails. The Bureau also disagrees with the complaint that the proposal sets too high a standard for what borrowers are able to understand. The statute merely states that when risks, costs, or conditions are material and consumers lack understanding of them, lenders cannot take unreasonable advantage of that fact. The Bureau agrees with the industry commenters that it is unreasonable to expect borrowers to understand the lending, banking, and payments system well enough to fully understand all the details of how lenders will initiate repeated re-presentments if the borrower defaults. But if the identified practice constitutes a material risk of the product, as the Bureau concludes here, then lenders are not at liberty to take unreasonable advantage of their consumers’ lack of understanding. The Bureau also disagrees with the claim that it is using a definition of “understanding” that differs from “to know how (something) works or happens.” This suggestion is flawed because it obfuscates the material risks, costs, or conditions to which that definition should be applied. The Bureau has found that most consumers do not realize that the identified practice involving multiple failed re-presentments happens. This conclusion is consistent with the accepted dictionary definition of “understanding.” 1016 Lastly, the Bureau rejects the claim that it cannot base any rule on the abusiveness authority defined in the statute, and instead can only enforce against abusive practices on a caseby-case basis, even where the Bureau has evidence and data that would justify a more general rule. Congress granted the Bureau explicit authority under section 1031(b) of the Dodd-Frank Act to issue rules grounded on its abusiveness authority. The Bureau believes that by giving the Bureau rulemaking authority using its abusiveness authority, Congress expressed its clear intent to give the Bureau authority to make more general assessments where it has evidence and data regarding an identified practice that meets the statutory prongs for abusiveness. Based on the facts and evidence described in the proposed rule, this section, and Market Concerns—Payments, the Bureau is concluding that consumers generally lack an understanding of the material costs, risks, or conditions of lenders’ repeated re-presentment practices, especially the extent of the risks and the severity of the costs. Accordingly, the Bureau is authorized to exercise its rulemaking authority in this area. 2. Consumers Are Unable to Protect their Interests Proposed Rule The Bureau proposed that when a lender attempts to withdraw payment from a consumer’s account in connection with a covered loan after the lender’s second consecutive failed attempt, unless the lender obtains the consumer’s new and specific authorization to make further withdrawals from the account, consumers are unable to protect their interests. By the time consumers discover that lenders are using their authorizations in this manner, it is often too late for them to take effective action. Although consumers could try to protect themselves from the harms of additional payment withdrawal attempts by closing down their accounts entirely, 1017 the Bureau did not interpret taking this action as being a practicable means for consumers to protect their interests, given that consumers use their accounts to conduct most of their household financial transactions. As discussed in the proposal, often the only option for most consumers to protect themselves (and their accounts) from the harms of lender attempts to withdraw payment after two consecutive attempts have failed is to stop payment or revoke authorization.1028 However, as also explained in the proposal, consumers often face considerable challenges and barriers when trying to stop payment or revoke authorization, both with their lenders and with their account-holding institutions. These challenges and barriers thus also make this option an impracticable means for consumers to protect themselves from the harms of further payment withdrawal attempts. As discussed in the proposal, lenders sometimes discourage consumers from stopping payment or revoking authorization by including language in loan agreements purporting to prohibit revocation. For instance, some lenders may charge consumers a substantial fee for stopping payment with their account-holding institutions. Others may have in place procedures for revoking authorizations directly with the lender that create additional barriers to stopping payment or revoking authorization effectively. For example, as discussed above, lenders often require consumers to provide written revocation by mail several days in advance of the next scheduled payment withdrawal attempt, among other requirements. Some consumers may even have difficulty identifying the lender that holds the authorization, particularly if the consumer 1028 As discussed in the proposal, even if consumers have enough money to deposit into their accounts prior to the next payment withdrawal attempt, those funds likely would be claimed first by the consumer’s account-holding institution to repay the NSF fees charged for the prior two failed attempts. Thus, there is still a risk of additional consumer harm from a third attempt in such situations, as well as from any attempts the lender may make after the third one, unless the consumer carefully coordinates the timing and amounts of the attempts with the lender, which is generally not possible. 1018 took out the loan online and was paired with the lender through a third-party lead generator. These and similar lender-created barriers—while challenging for consumers in all cases—can make it particularly difficult for consumers to revoke authorizations for repayment by recurring transfers, given that a consumer’s account-holding institution is permitted under Regulation E to confirm the consumer has informed the lender of the revocation (e.g., by requiring a copy of the consumer’s revocation as written confirmation to be provided within 14 days of an oral notification). Thus, if the institution does not receive the required written confirmation within this time frame, then it may continue to honor subsequent debits to the account. In the proposal, the Bureau explained that consumers encounter additional challenges when trying to stop payment with their account-holding institutions. For example, due to complexities in payment processing systems and the internal procedures of consumers’ accountholding institutions, consumers may be unable to stop payment on the next payment withdrawal attempt in a timely and effective manner. Even if the consumer successfully stops payment with her account-holding institution on the lender’s next payment attempt, the consumer may experience difficulties blocking all future attempts by the lender, particularly when the consumer has authorized the lender to make withdrawals from her account via recurring EFTs. Some depository institutions require the consumer to provide the exact payment amount or the lender’s merchant ID code, and thus fail to block payments when the payment amount varies or the lender varies the merchant code. Consumers are likely to experience even greater challenges in stopping payment on lender attempts made via RCCs or RCPOs, given the difficulty that account-holding institutions have identifying such payment attempts. Further, if the lender has obtained multiple types of authorizations from the consumer—such as authorizations to 1019 withdraw payment via both ACH transfers and RCCs—the consumer likely will have to navigate different sets of complicated stop-payment procedures for each type of authorization held by the lender, thereby making it even more challenging to stop the payment effectively. As further laid out in the proposal, the fees charged by consumers’ account-holding institutions for stopping a payment are often comparable to the NSF fees or overdraft fees from which the consumers are trying to protect themselves. Depending on the policies of their account-holding institutions, some consumers may be charged a second fee to renew a stoppayment order after a period of time. As a result of these costs, even if the consumer successfully stops payment on the next payment withdrawal attempt, the consumer will not have effectively protected herself from the fee-related injury that otherwise would have resulted from the attempt, but rather will have just exchanged the cost of one fee for another. Additionally, in some cases, consumers may be charged a stop-payment fee by their account-holding institution even when the stop-payment order fails to stop the lender’s payment withdrawal attempt from occurring. As a result, such consumers may incur both a fee for the stop-payment order and an NSF or overdraft fee for the lender’s withdrawal attempt.1029 Comments Received One commenter suggested that the statutory phrase “inability of the consumer to protect the interests of the consumer in selecting or using a consumer financial product or service” is similar to section 4(c)(1) of the Uniform Consumer Sales Practices Act. That provision bans 1029 Even when consumers’ account-holding institutions may not charge a fee for returned or declined payment withdrawal attempts made using a particular payment method, such as attempts made by debit cards and certain prepaid cards, consumers still incur lender-charged fees from which they cannot protect themselves. In addition, consumers sometimes incur lender-charged fees for successfully stopping payment or revoking authorization. 1020 unconscionable contracts that take “advantage of the inability of the consumer reasonably to protect his interests because of his physical infirmity, ignorance, illiteracy, [or] inability to understand the language of an agreement.” This commenter suggested that the Bureau should thus deem this prong met only if the consumers in question are physically infirm, ignorant, illiterate, or unable to understand. Several commenters suggested again that borrowers typically are able to appreciate the general consequences of failing to pay, or contended that this prong of the definition of abusiveness is only met where it is literally impossible for consumers to protect their interests in selecting or using the product. Many other comments pointed to the mechanisms that the Bureau identified in the proposal—authorization revocations, account closures, and stop payments—stating that these prove borrowers do have the ability to protect their interests. Some commenters argued more simply that borrowers can protect their interests by just making a payment when it is due, or by not taking out loans in the first place. Consumer groups, by contrast, argued that it is difficult, if not impossible, for consumers to revoke account access or stop payment withdrawals when lenders initiate multiple attempts. Final Rule The Bureau now concludes, as discussed below, that consumers are unable to protect their interests—specifically the interest of preventing the harms identified—in selecting or using a consumer financial product or service. The Bureau does not agree that the language in the Dodd-Frank Act should be interpreted as synonymous with the passage cited from the Uniform Consumer Sales Practices Act. In fact, there is no basis whatsoever for this suggestion. The statutory definition of abusiveness does not 1021 limit instances where a company can take advantage of an inability to protect one’s own interests to a narrow set of instances where that inability is caused by infirmity, ignorance, illiteracy, or inability to understand the language of an agreement. The Bureau also rejects the interpretation, presented by commenters, that the prong of “inability of the consumer to protect the interests of the consumer in selecting or using a consumer financial product or service” can be met only when it is literally impossible for consumers to take action to protect their interests.1030 One dictionary defines “inability” to mean a “lack of sufficient power, strength, resources, or capacity,”1031 and the Bureau believes the clause “inability of the consumer to protect” is similarly reasonably interpreted to mean that consumers are unable to protect their interests when it is impracticable for them to do so in light of the circumstances. As for comments that mechanisms are available to avoid undesirable outcomes, or that borrowers can protect their interests by just making a payment when it is due or by not taking out loans in the first place, these are arguments the Bureau already addressed in the “reasonable avoidability” part of the unfairness section above, and its responses to those points apply here. As stated in the proposal and discussed further above in Market Concerns—Payments, evidence in the record supports the conclusion that consumers are, in fact, unable to protect their own interests in relation to payment re-presentments by initiating stop payments or revoking 1030 At least one court has rejected a similar interpretation. See Consumer Financial Protection Bureau v. ITT Educational Services, Inc., 219 F. Supp. 3d 878, 919 (S.D. Ind. 2015). 1031 “Webster’s Third New International Dictionary,” (Merriam Webster Inc., 2002). 1022 authorizations.1032 Commenters’ assertions that borrowers have a literal ability to protect their interests in some conceivable but impractical circumstances rest on a misunderstanding of the statutory test and the actual facts of these types of situations. On the basis of the evidence presented, the Bureau thus concludes that consumers are generally and practicably unable to use these methods to protect their interests. 3. Practice Takes Unreasonable Advantage of Consumer Vulnerabilities Proposed Rule Under section 1031 of the Dodd-Frank Act, an act or practice is abusive when it takes “unreasonable advantage” of consumers’ lack of understanding of the material risks, costs, or conditions of selecting or using a consumer financial product or service or of their inability to protect their interests in selecting or using such a product or service. The Bureau proposed that, with respect to covered loans, the lender act or practice of attempting to withdraw payment from a consumer’s account after two consecutive attempts have failed, unless the lender obtains the consumer’s new and specific authorization to make further withdrawals, may take unreasonable advantage of consumers’ lack of understanding and inability to protect their interests and is therefore abusive. In making this proposal, the Bureau was informed by the evidence discussed in the proposal and above in Markets Concerns—Payments. In the proposal, the Bureau recognized that in any transaction involving a consumer financial product or service, there is likely to be some information asymmetry between the consumer and the financial institution. Often, the financial institution will have superior 1032 See specific Market Concerns—Payments subsection entitled “Consumers Have Difficulty Stopping Lenders’ Ability to Access Their Accounts” for that evidence. 1023 bargaining power as well. Section 1031(d) of the Dodd-Frank Act does not prohibit financial institutions from taking advantage of their superior knowledge or bargaining power to maximize their profit. Indeed, in a market economy, market participants with such advantages generally pursue their self-interests. However, section 1031 of the Dodd-Frank Act makes plain that at some point, a financial institution’s conduct in leveraging consumers’ lack of understanding or inability to protect their interests becomes unreasonable advantage-taking that is abusive.1033 The Dodd-Frank Act delegates to the Bureau the responsibility for determining when that line has been crossed. In the proposal, the Bureau stated that such determinations are best made with respect to any particular practice by taking into account all of the facts and circumstances that are relevant to assessing whether the practice takes unreasonable advantage of consumers’ lack of understanding or inability to protect their interests. The Bureau recognized that taking a consumer’s authorization to withdraw funds from her account without further action by the consumer is a common practice that frequently serves the interest of both lenders and consumers, and does not believe that this practice, standing alone, takes unreasonable advantage of consumers. However, at least with respect to covered loans, the Bureau proposed to conclude, based on the evidence discussed in the proposal and above in Markets Concerns—Payments, that when lenders use such authorizations to make another payment withdrawal attempt after two consecutive attempts have failed, lenders take unreasonable advantage of consumers’ lack of understanding and inability to protect their interests, absent the consumer’s new and specific authorization. 1033 A covered person also may take unreasonable advantage of one or more of the three consumer vulnerabilities identified in section 1031(d) of the Dodd-Frank Act in circumstances in which the covered person lacks such superior knowledge or bargaining power. 1024 As discussed above, with respect to covered loans, the lender practice of continuing to make payment withdrawal attempts after a second consecutive failure generates relatively small amounts of revenues for lenders, particularly as compared with the significant harms that consumers incur as a result of the practice. Moreover, the cost to the lender of re-presenting a failed payment withdrawal attempt is nominal; for this reason, lenders often repeatedly represent at little cost to themselves, and with little to no regard for the harms that consumers incur as a result of the re-presentments. Specifically, the Bureau’s analysis of ACH payment withdrawal attempts made by online payday and payday installment lenders, laid out in greater detail in the proposal, indicates that the expected value of a third successive payment withdrawal attempt is only $46 (as compared with $152 for a first attempt), and that the expected value drops to $32 for the fourth attempt and to $21 for the fifth attempt. And yet, despite these increasingly poor odds of succeeding, many lenders continue to re-present. This further suggests that at this stage, the consumers’ payment authorizations have ceased to serve their primary purpose of convenience, but instead have become a means for the lenders to seek to extract small amounts of revenues from consumers any way they can. In addition, lenders often charge consumers a returned-item fee for each failed attempt.1034 This provides lenders with an additional financial incentive to continue attempting to withdraw payment from consumers’ accounts even after two consecutive attempts have failed. Although lenders may not be able to collect such fees immediately, the fees are 1034 In addition, as discussed in the proposal, the Bureau is aware of some depository institutions that have charged NSF and overdraft fees for payment attempts made within the institutions’ internal systems, including a depository institution that charged such fees in connection with collecting payments on its own small-dollar loan product. 1025 added to the consumer’s overall debt and thus can be pursued and perhaps collected later through the debt collection process. The Bureau preliminarily concluded that lenders could obtain much of this revenue without engaging in the practice of trying to withdraw payment from consumers’ accounts after the accounts have exhibited clear signs of being in severe distress. For example, lenders could seek further payments in cash or ACH “push” payments from the consumer or, in the alternative, could seek a new and specific authorization from consumers to make further payment withdrawal attempts. Indeed, the Bureau determined that coordinating with the consumer to seek a new authorization may be more likely to result in successful payment withdrawal attempts than does the practice of repeatedly attempting to withdraw payments from an account that is known to be in distress. Comments Received Most of the comments relevant to this prong were already addressed in the two sections above. The Bureau also received comments suggesting that it provided no evidence that the practice takes unreasonable advantage of consumers. Commenters also argued that the Bureau should focus on how certain roadblocks imposed by financial institutions relating to stoppayment orders take unreasonable advantage of consumers rather than on the identified practice engaged in by lenders. Final Rule As described more fully above in Market Concerns—Payments, the Bureau does have ample evidence that the identified practice takes unreasonable advantage of consumers. Lenders take advantage by imposing financial harm on consumers when they make repeated efforts to extract funds from consumer accounts, and those actions are unreasonable in light of the low 1026 expected value of those re-presentments. Indeed, lenders should be well aware that borrowers will likely not have funds in their distressed accounts, as shown by the two prior failed presentments and the lenders’ general experience of the low expected value of multiple representments. They also should be well aware of the kinds of harms that consumers are likely to experience in these situations; nonetheless, they routinely make a conscious choice to engage in the identified practice by proceeding with their re-presentments. It may be the case that financial institutions engage in practices that hinder borrowers’ ability to stop payments. Whether this takes unreasonable advantage of consumers has no bearing on whether lenders also take unreasonable advantage of consumers by engaging in the identified practice. The Bureau finalizes its conclusion that the practice of attempting to withdraw payment from a consumer’s account in connection with a covered loan after the lender’s second consecutive failed attempt to withdraw payment from the account, unless the lender obtains the consumer’s new and specific authorization to make further withdrawals from the account, takes unreasonable advantage of consumers’ lack of understanding of the material risks, costs, or conditions of the product or service, as well as their inability to protect their interests in selecting or using a consumer financial product or service. Section 1041.8 Prohibited Payment Transfer Attempts For the reasons discussed in the section-by-section analysis of § 1041.7, the Bureau has concluded that it is an unfair and abusive practice for a lender to attempt to withdraw payment from a consumer’s account in connection with a covered loan after the lender’s second consecutive attempt to withdraw payment from the account has failed due to a lack of sufficient 1027 funds, unless the lender obtains the consumer’s new and specific authorization to make further withdrawals from the account. Thus, after a lender’s second consecutive attempt to withdraw payment from a consumer’s account has failed, the lender could avoid engaging in the unfair or abusive practice either by not making any further payment withdrawals or by obtaining from the consumer a new and specific authorization and making further payment withdrawals pursuant to that authorization. Section 1031(b) of the Dodd-Frank Act provides that the Bureau may prescribe rules “identifying as unlawful unfair, deceptive, or abusive acts or practices” and may include requirements in such rules for the purpose of preventing unfair, deceptive, or abusive acts or practices. The Bureau is preventing the unfair and abusive practice described above by including in § 1041.8 specific requirements for determining when making a further payment withdrawal attempt constitutes an unfair or abusive act and for obtaining a consumer’s new and specific authorization to make further payment withdrawals from the consumer’s account. In addition to its authority under section 1031(b), the Bureau is issuing two other provisions— § 1041.8(c)(3)(ii) and (iii)(C)—pursuant to its authority under section 1032(a) of the Dodd-Frank Act. Section 1032(a) authorizes the Bureau to prescribe rules to ensure that the features of consumer financial products and services, “both initially and over the term of the product or service,” are disclosed “fully, accurately, and effectively . . . in a manner that permits consumers to understand the costs, benefits, and risks associated with the product or service, in light of the facts and circumstances.”1035 Both of the proposed provisions relate to the requirements for 1035 12 U.S.C. 5532(a). 1028 obtaining the consumer’s new and specific authorization after the prohibition on making further payment withdrawals has been triggered. In addition to the provisions in § 1041.8, the Bureau is finalizing a complementary set of provisions in § 1041.9, pursuant to its authority under section 1032 of the Dodd-Frank Act, to require lenders to provide notice to a consumer prior to initiating a payment withdrawal from the consumer’s account. These disclosures inform consumers in advance of the timing, amount, and channel of upcoming initial and unusual withdrawal attempts, in order to help consumers detect errors or problems with upcoming payments and contact their lenders or account-holding institutions to resolve them in a timely manner. The disclosures will also help consumers take steps to ensure that their accounts contain enough money to cover the payments, when taking such steps is feasible for consumers. In § 1041.9, the rule also provides for a notice that lenders are required to provide to consumers, alerting them to the fact that two consecutive payment withdrawal attempts to their accounts have failed—thus triggering operation of the requirements in § 1041.8(b)—so that consumers can better understand their repayment options and obligations in light of their accounts’ severely distressed conditions. The two payments-related sections in the proposed rule thus complement and reinforce each other. As described earlier, because the Bureau is not finalizing at this time the provisions relating to the underwriting of covered longer-term loans by assessing the borrower’s ability to repay (other than for covered longer-term balloon-payment loans), various sections of the final rule have been renumbered differently than in the proposed rule. In particular, § 1041.14 of the proposed rule on prohibited payment transfer attempts, and § 1041.15 of the proposed rule on 1029 disclosure of payment transfer attempts, have now been renumbered, respectively, as §§ 1041.8 and 1041.9 of the final rule. 8(a) Definitions Proposed § 1041.14(a) defined key terms to be used throughout proposed §§ 1041.14 and 1041.15. The central defined term in both proposed sections was “payment transfer,” which would apply broadly to any lender-initiated attempt to collect payment from a consumer’s account, regardless of the type of authorization or instrument used. The Bureau also proposed to define “single immediate payment transfer at the consumer’s request,” which is described below. 8(a)(1) Payment transfer Proposed Rule Proposed § 1041.14(a)(1) defined a payment transfer as any lender-initiated debit or withdrawal of funds from a consumer’s account for the purpose of collecting any amount due or purported to be due in connection with a covered loan. It also provided a non-exhaustive list of specific means of debiting or withdrawing funds from a consumer’s account that would constitute payment transfers if the general definition’s conditions are met. They included a debit or withdrawal initiated through: (1) an EFT, including a preauthorized EFT as defined in Regulation E, 12 CFR 1005.2(k); (2) a signature check, regardless of whether the transaction is processed through the check network or another network, such as the ACH network; (3) a remotely created check as defined in Regulation CC, 12 CFR 229.2(fff); (4) a remotely created payment order as defined in 16 CFR 310.2(cc); and (5) an account-holding institution’s transfer of funds from a consumer’s account that is held at the same institution. 1030 The Bureau proposed a broad definition focused on the collection purpose of the debit or withdrawal rather than on the particular method by which the debit or withdrawal is made, to help ensure uniform application of the proposed rule’s payment-related consumer protections. In the proposal the Bureau stated that in markets for loans that would be covered under the proposed rule, lenders use a variety of methods to collect payment from consumers’ accounts. Some lenders take more than one form of payment authorization from consumers in connection with a single loan. Even lenders that take only a signature check often process the checks through the ACH system, particularly for purposes of re-submitting a returned check that was originally processed through the check system. At the proposal stage the Bureau believed that, for a rule designed to apply across multiple payment methods and channels, a single defined term was necessary to avoid the considerable complexity that would result if the rule merely adopted existing terminology that may be unique to every specific method and channel. The Bureau believed that defining payment transfer in this way would enable the rule to provide for the required payment notices to be given to consumers regardless of the payment method or channel used to make a debit or withdrawal. Similarly, the Bureau believed that the proposed definition would ensure that the prohibition in proposed § 1041.14(b) on additional failed payment transfers would apply regardless of the payment method or channel used to make the triggering failed attempts and regardless of whether a lender moves back and forth between different payment methods or channels when attempting to withdraw payment from a consumer’s account. Proposed comment 14(a)(1)-1 explained that a transfer of funds meeting the general definition would be a payment transfer regardless of whether it is initiated by an instrument, 1031 order, or other means not specified in § 1041.14(a)(1). Proposed comment 14(a)(1)-2 explained that a lender-initiated debit or withdrawal includes a debit or withdrawal initiated by the lender’s agent, such as a payment processor. Proposed comment 14(a)(1)-3 provided examples to illustrate how the proposed definition would apply to a debit or withdrawal for any amount due in connection with a covered loan. Specifically, proposed comments 14(a)(1)-3.i through (a)(1)3.iv explained, respectively, that the definition would apply to a payment transfer for the amount of a scheduled payment, a transfer for an amount smaller than the amount of a scheduled payment, a transfer for the amount of the entire unpaid loan balance collected pursuant to an acceleration clause in a loan agreement for a covered loan, and a transfer for the amount of a late fee or other penalty assessed pursuant to a loan agreement for a covered loan. Proposed comment 14(a)(1)-4 clarified that the proposed definition would apply even when the transfer is for an amount that the consumer disputes or does not legally owe. Proposed comment 14(a)(1)-5 provided three examples of covered loan payments that, while made with funds transferred or withdrawn from a consumer’s account, would not be covered by the proposed definition of a payment transfer. The first two examples, provided in proposed comments 14(a)(1)-5.i and (a)(1)-5.ii, were of transfers or withdrawals that are initiated by the consumer—specifically, when a consumer makes a payment in cash withdrawn by the consumer from the consumer’s account and when a consumer makes a payment via an online or mobile bill payment service offered by the consumer’s account-holding institution. The third example, provided in proposed comment 14(a)(1)-5.iii, clarified that the definition would not apply when a lender seeks repayment of a covered loan pursuant to a valid court order authorizing the lender to garnish a consumer’s account. 1032 Additionally, proposed comments relating to § 1041.14(a)(1)(i), (ii), and (v) clarified how the proposed payment transfer definition applies to particular payment methods. Specifically, proposed comment 14(a)(1)(i)-1 explained that the general definition of a payment transfer would apply to any EFT, including but not limited to an EFT initiated by a debit card or a prepaid card. Proposed comment 14(a)(1)(ii)-1 provided an illustration of how the definition of payment transfer would apply to a debit or withdrawal made by signature check, regardless of the payment network through which the transaction is processed. Lastly, proposed comment 14(a)(1)(v)-1 clarified, by providing an example, that an account-holding institution initiates a payment transfer when it initiates an internal transfer of funds from a consumer’s account to collect payment on a deposit advance product. Comments Received NACHA agreed with the Bureau’s decision to cover all payment methods with the rule, noting that their presentment cap is only applicable to payments processed on the ACH system and that since they clarified the cap on ACH presentments, they have seen vendors shift towards using other payment methods. The Bureau received a number of comments arguing that the compliance burden of, among other things, tracking payment presentments across multiple payment methods would be significant. Other commenters argued that payment withdrawal rules should be relaxed in cases where a depository institution is both the lender and the deposit account holder, provided that the depository institution does not charge a fee after attempting and failing to collect from the account. Similarly, a group representing community banks argued that the Bureau should not 1033 prohibit community banks from accessing consumer accounts held by the bank to pay for a loan made by the bank. This commenter claimed that the disclosures provided to borrowers before the authorization should suffice. More generally, commenters asked for further clarity on the rule’s treatment of internal transfers at account-holding institutions. Consumer group commenters were generally supportive of the proposed definition but argued that the Bureau should amend it in two ways. First, they argued that it should include both transactions initiated by the lender and transactions initiated by the lender’s agent in the definition of payment transfer. Second, the commenters argued that the definition should not be tied to the term “account” because a nonbank might be able to evade this requirement by pulling funds from a source of funds other than an “account.” Commenters suggested that the Bureau use the term “installment” instead of “payment” in the definition so as to clarify that the rule covers each payment on an installment contract, which the commenters believed would expand the rule and be more consistent with State and local laws. Several commenters, including State Attorneys General, argued that payments made using debit cards should be exempt because they generally do not engender NSF fees, and thus, the harm justifying the identified unfair and abusive act or practice is diminished for debit card payments. Final Rule The Bureau is generally finalizing the rule as proposed, with some technical changes, and the addition of an exclusion for lenders that are also acting as the borrower’s account-holding institution when certain conditions are met. The Bureau concludes, in particular, that it is 1034 essential for the rule to cover all payment methods in order to prevent harm to consumers from the practice identified as unfair and abusive. Additionally, the Bureau maintains its view that a single definition is a simpler approach that is more administrable as a practical matter than using separate terminology for each type of payment method. In adding the exclusion, the Bureau is reorganizing the numbering of § 1041.8(a)(1). The Bureau is also converting proposed comment 14(a)(1)-1 into the text of the regulation at § 1041.8(a)(1)(i). The initial examples of covered payment methods are now all listed there. The Bureau had proposed, as an example of a payment method included in the definition, “[a]n account-holding institution’s transfer of funds from a consumer’s account that is held at the same institution.” In light of the added conditional exclusion relating to account-holding institutions, the Bureau is adding at the end of that sentence “other than such a transfer meeting the description in paragraph (a)(1)(ii) of this section.” In response to the sound suggestion received from several commenters, the Bureau is adding to § 1041.8 paragraph (a)(1)(ii), which is a conditional exclusion for certain lenders that are also the borrower’s account-holding institution. That exclusion only applies to instances where the lender has set forth in the original loan agreement or account agreement that it will not charge the consumer a fee for payment attempts when the account lacks sufficient funds to cover the payment, and that it will not close the account in response to a negative balance that results from a transfer of funds initiated in connection with the covered loan. If lenders do not charge NSF, overdraft, return payment fees, or similar fees, and do not close accounts because of failed payment attempts, the harms underpinning the unfair and abusive practice identified in § 1041.7 would not occur, and thus the Bureau concludes that the rule does not need to cover those 1035 instances. The Bureau did not exclude transfers made by lenders that are also the borrower’s account-holding institution where the harms would continue (i.e., fees are charged or accounts are closed) because that would be inconsistent with the Bureau’s efforts in the rule to prevent the harms associated with the unfair and abusive practice. Paragraph (a)(1)(ii) would allow late fees because the Bureau considers those charges to be distinct from, and not caused by, the practice identified in § 1041.7. It bears emphasis that, under the terms of the rule, the borrower’s account or loan agreement must state, at the time the consumer takes out the first covered loan, that the account-holding institution does not charge such fees in connection with a failed payment attempt on a loan made by the institution or close the account in response to a negative balance resulting from the lender’s collection of a payment on the covered loan. This is meant to prevent lenders from avoiding the presentment cap for failed payments involving fees by simply switching back and forth between charging fees and not charging fees, as well as to ensure that both conditions apply for the duration of the covered loan. The Bureau has not finalized a similar exclusion for non-account-holding lenders where the account-holding institution otherwise does not charge fees or close accounts, because those lenders do not have control over whether those events occur, as do the lenders excluded by paragraph (a)(1)(ii). In light of changes made to the text of the rule and the incorporation of proposed comment 14(a)(1)-1 into the text, the commentary to the rule has been renumbered accordingly. In addition, the Bureau has amended proposed comment 14(a)(1)(v)-1, now comment 8(a)(1)(i)(E)-1 of the final rule, to reflect the changes made to accommodate the conditional exclusion. In response to requests from commenters, the Bureau also has added comment 1036 8(a)(1)(i)(E)-2, which to further clarifies the application of the payment transfer definition to internal transfers of funds within an account-holding institution. The Bureau notes that under the final rule, the payment transfer definition—and thus the cap on failed payment transfers—still applies to such lenders when the conditions for the exclusion from the definition are not met. The additional examples include: (1) initiating an internal transfer from a consumer’s account to collect a scheduled payment on a covered loan; (2) sweeping the consumer’s account in response to a delinquency on a covered loan; and (3) exercising a right of offset to collect against an outstanding balance on a covered loan. The Bureau also added some comments on the conditional exclusion. Comment 8(a)(1)(ii)(A)-1 clarifies that the loan or account agreement must contain a term to restrict the charging of fees that is in effect at the time the covered loan is made, which must remain in effect for the duration of the loan. Again, this comment is intended to ensure that lenders that are account-holding institutions do not avoid the rule’s cap on failed payment attempts by switching back and forth between charging fees and not charging fees for failed attempts. Comment 8(a)(2)(ii)(A)-2 provides examples of the types of fees that must be restricted in order to qualify for the conditional exclusion. It clarifies that those fees include NSF fees, overdraft fees, and returned-item fees. It also explains that a lender may charge late fees if such fees are permitted under the terms of the loan agreement, and still qualify for the conditional exclusion if the conditions in § 1041.8(a)(1)(ii) are met. Comment 8(a)(1)(ii)(B)-1 clarifies that in order to be eligible for the exclusion in § 1041.8(a)(1)(ii), the lender cannot close the borrower’s account in response to a negative balance that results from a lender-initiated transfer of funds in connection with the covered loan, 1037 but that the lender is not restricted from closing the account in response to another event. Specifically, the comment provides that a lender is not restricted from closing the consumer’s account in response to another event, even if the event occurs after a lender-initiated transfer of funds has brought the account to a negative balance. Further, the comment provides, as examples, that a lender may close the account at the consumer’s request, for purposes of complying with other regulatory requirements, or to protect the account from suspected fraudulent use or unauthorized access, and still meet the condition in § 1041.8(a)(1)(ii)(B). The Bureau believes it is important to clarify that lenders collecting payments pursuant to the conditional exclusion in § 1041.8(a)(1) are not restricted from closing a consumer’s account when circumstances unrelated to the covered loan payments dictate that they do so. Finally, comment 8(a)(1)(ii)(B)-2 clarifies that the loan or account agreement must contain a term providing that the lender will not close the consumer’s account in the circumstances specified in the rule at the time the covered loan is made, and that the term must remain in effect for the duration of the loan. The Bureau recognizes the industry commenters’ concern that lenders will incur compliance burdens associated with keeping track of payment presentments across different payment methods. However, as stated in the proposal, the Bureau continues to maintain ongoing compliance costs associated with tracking presentments will likely be minimal following the initial investment. There may be additional compliance burdens associated with tracking presentments across payment methods, but the alternative of only tracking presentments on certain payment methods would undermine the purposes of the rule, and would not fully prevent the full scope of consumer harm identified above in Market Concerns—Payments, and further 1038 discussed in the section-by-section analysis of § 1041.7. The Bureau also does not find it helpful to use the term “installment” to make clear that the rule applies to multiple payments initiated under an installment agreement. The definition of “payment transfer” is meant to cover any kind of payment attempt, including multiple attempts made to cover a single installment under a loan agreement. Replacing the term “payment” with “installment” may confuse that point. In addition, the Bureau does not see the need for further clarification with regard to how the rule covers agents of lenders that initiate payment presentments on the lender’s behalf. A lender’s use of third-party processors or servicers does not provide a basis to circumvent the payment presentment cap. In fact, a lender using a third-party service provider is still liable under the rule, as the service provider also may be, depending on the facts and circumstances. Lastly, the Bureau is not aware of any methods by which a non-bank lender could circumvent the rule based on the definition of the term “account.” The definition is the same as in 12 CFR 1005.2, and therefore includes normal deposit accounts at financial institutions, payroll card accounts, and (by the time this rule is effective) prepaid accounts. To the extent a lender is debiting something other than an “account,” that event may not involve the same kinds of fees associated with the identified practice. To provide greater clarity to industry, the Bureau finds it appropriate at this time to use a pre-existing definition. If in the future a lender or lenders cause repeated fees to consumers by attempting to take funds from something other than an “account” after multiple failed attempts, the Bureau would consider exercising its supervision, enforcement, or rulemaking authority to address the problem, as appropriate. Lastly, the Bureau has decided not to exempt payments made using debit cards from the 1039 rule. First, while failed debt card transactions may not trigger NSF fees, some of them do trigger overdraft fees, even after two failed attempts, as our study showed. Second, lenders may still charge return fees for each presentment. And third, the Bureau does not believe an exclusion based on payment type would work to alleviate much compliance burden associated with § 1041.8 because the lender would need to develop processes and procedures for those payment types that are covered regardless. In fact, juggling multiple, disparate processes and procedures depending on payment type would involve its own compliance burdens. 8(a)(2) Single immediate payment transfer at the consumer’s request Proposed Rule Proposed § 1041.14(a)(2) would have defined a single immediate payment transfer at the consumer’s request as, generally, a payment transfer that is initiated by a one-time EFT or by processing a consumer’s signature check within one business day after the lender obtains the consumer’s authorization or check. Such payment transfers would be exempted from certain requirements in the proposed rule. The principal characteristic of a single immediate payment transfer at the consumer’s request is that it is initiated at or near the time the consumer chooses to authorize it. During the SBREFA process, and in outreach with industry in developing the proposal, the Bureau received feedback that consumers often authorize or request lenders to make an immediate debit or withdrawal from their accounts for various reasons including, for example, to avoid a late payment fee. As discussed in the proposed rule, stakeholders expressed concerns primarily about the potential impracticability and undue burden of providing a notice of an upcoming withdrawal in advance of executing the consumer’s payment instructions in these circumstances. More generally, the SERs and industry stakeholders suggested that a transfer 1040 made at the consumer’s immediate request presents fewer consumer protection concerns than a debit or withdrawal authorized by the consumer several days or more in advance, presuming that the consumer makes the immediate request based on current and first-hand knowledge of their account balance. In the proposal, the Bureau stated that applying fewer requirements to payment transfers initiated immediately after consumers request the debit or withdrawal was both warranted and consistent with the important policy goal of providing consumers with greater control over their payments on covered loans. Accordingly, the proposed definition would be used to apply certain exceptions to the proposed rule’s payments-related requirements in two instances. First, a lender would not be required to provide the payment notice in proposed § 1041.15(b) when initiating a single immediate payment transfer at the consumer’s request. Second, a lender would be permitted under proposed § 1041.14(d) to initiate a single immediate payment transfer at the consumer’s request after the prohibition in proposed § 1041.14(b) on initiating further payment transfers has been triggered, subject to certain requirements and conditions. Proposed § 1041.14(a)(2) provided that a payment transfer is a single immediate payment transfer at the consumer’s request when it meets either one of two sets of conditions. The first of these prongs applied specifically to payment transfers initiated via a one-time EFT. Proposed § 1041.14(a)(2)(i) generally defined the term as a one-time EFT initiated within one business day after the consumer authorizes the transfer. The Bureau believed that a one-business-day time frame would allow lenders sufficient time to initiate the transfer, while providing assurance that the account would be debited in accordance with the consumer’s timing expectations. Proposed comment 14(a)(2)(i)-1 explained that for purposes of the definition’s timing condition, a one1041 time EFT is initiated at the time that the transfer is sent out of the lender’s control and that the EFT thus is initiated at the time the lender or its agent sends the payment to be processed by a third party, such as the lender’s bank. The proposed comment further provided an illustrative example of this concept. The second prong of the definition, in proposed § 1041.14(a)(2)(ii), applied specifically to payment transfers initiated by processing a consumer’s signature check. Under this prong, the term would apply when a consumer’s signature check is processed through either the check system or the ACH system within one business day after the consumer provides the check to the lender. Proposed comments 14(a)(2)(ii)-1 and -2 explained how the definition’s timing condition in proposed § 1041.14(a)(2)(ii) applies to the processing of a signature check. Similar to the concept explained in proposed comment 14(a)(2)(i)-1, proposed comment 14(a)(2)(ii)-1 explained that a signature check is sent out of the lender’s control and that the check thus is processed at the time that the lender or its agent sends the check to be processed by a third party, such as the lender’s bank. The proposed comment further cross-referenced proposed comment 14(a)(2)(i)-1 for an illustrative example of how this concept applies in the context of initiating a one-time EFT. Regarding the timing condition in proposed § 1041.14(a)(2)(ii), proposed comment 14(a)(2)(ii)-2 clarified that when a consumer mails a check to the lender, the check is deemed to be provided to the lender on the date it is received. As with the similar timing condition for a one-time EFT in proposed § 1041.14(a)(2)(i), the Bureau believed that these timing conditions would help to ensure that the consumer has the ability to control the terms of the transfer and that the conditions would be practicable for lenders to meet. In addition, the Bureau noted that the timing conditions would effectively exclude from 1042 the definition the use of a consumer’s post-dated check, and instead would limit the definition to situations in which a consumer provides a check with the intent to execute an immediate payment. The Bureau believed that this condition was necessary to ensure that the exceptions concerning single immediate payment transfers at the consumer’s request apply only when it is clear that the consumer is affirmatively initiating the payment by dictating its timing and amount. Under the proposal, these criteria would not be met when the lender already holds the consumer’s post-dated check. Comments Received The Bureau received some comments pertaining to the definition of a single immediate payment transfer at the consumer’s request. Because the definition is closely related to the exception in § 1041.8(d), the Bureau addresses those comments below in the discussion of final § 1041.8(d). Final Rule The Bureau is finalizing this definition as proposed, except for renumbering proposed § 1041.14(a) as § 1041.8(a). 8(b) Prohibition on initiating payment transfers from a consumer’s account after two consecutive failed payment transfers Proposed Rule Proposed § 1041.14(b) stated that a lender cannot attempt to withdraw payment from a consumer’s account in connection with a covered loan when two consecutive attempts have been returned due to a lack of sufficient funds. This proposal was made pursuant to section 1031(b) of the Dodd-Frank Act, which provides that the Bureau may prescribe rules for the purpose of 1043 preventing unlawful unfair, deceptive, or abusive acts or practices.1036 As discussed in the section-by-section analysis of proposed § 1041.13, it appeared that, in connection with a covered loan, it was an unfair and abusive practice for a lender to attempt to withdraw payment from a consumer’s account after the lender’s second consecutive attempt to withdraw payment from the account fails due to a lack of sufficient funds, unless the lender obtains the consumer’s new and specific authorization to make further payment withdrawals. This proposed finding would have applied to any lender-initiated debit or withdrawal from a consumer’s account for purposes of collecting any amount due or purported to be due in connection with a covered loan, regardless of the particular payment method or channel used. In accordance with this proposed finding, a lender would be generally prohibited under proposed § 1041.14(b) from making further attempts to withdraw payment from a consumer’s account upon the second consecutive return for nonsufficient funds, unless and until the lender obtains the consumer’s authorization for additional transfers under proposed § 1041.14(c), or obtains the consumer’s authorization for a single immediate payment transfer in accordance with proposed § 1041.14(d). The prohibition under proposed § 1041.14(b) would apply to, and be triggered by, any lender-initiated attempts to withdraw payment from a consumer’s checking, savings, or prepaid account. In addition, the prohibition under proposed § 1041.14(b) would apply to, and be triggered by, all lender-initiated withdrawal attempts regardless of the payment method used including, but not limited to, signature check, remotely created check, remotely created payment orders, authorizations for one-time or recurring EFTs, and an account-holding institution’s withdrawal of funds from a consumer’s account that is held at the same institution. 1036 12 U.S.C. 5531(b). 1044 In developing the proposed approach to restricting lenders from making repeated failed attempts to debit or withdraw funds from consumers’ accounts, the Bureau had considered a number of potential interventions. As detailed in Market Concerns—Payments of the proposal and final rule, for example, the Bureau is aware that some lenders split the amount of a payment into two or more separate transfers and then present all of the transfers through the ACH system on the same day. Some lenders make multiple attempts to debit accounts over the course of several days or a few weeks. Also, lenders that collect payment by signature check often alternate submissions between the check system and ACH system to maximize the number of times they can attempt to withdraw payment from a consumer’s account using a single check. These and similarly aggressive payment practices potentially cause harms to consumers and may each constitute more specific unfair, deceptive, or abusive acts or practices, as well as fitting within the broader unfair and abusive practice identified in the proposal. However, the Bureau believed that tailoring requirements in this rulemaking for each discrete payment practice would add considerable complexity to the proposed rule and yet still could leave consumers vulnerable to harms from aggressive practices that may emerge in markets for covered loans in the future. Accordingly, while the Bureau stated that it would continue to use its supervisory and enforcement authorities to address such aggressive payment practices in particular circumstances as appropriate, it proposed to address categorically the broader practice of making repeated failed attempts to collect payment on covered loans, which it preliminarily believed to be unfair and abusive. In addition, the Bureau proposed requirements to prevent that practice which would help protect consumers from a range of harmful payment practices in a considerably less complex fashion. For example, as applied to the practice of splitting payments into multiple 1045 same-day presentments, the proposed approach would effectively curtail a lender’s access to the consumer’s account when any two such presentments fail. As applied to checks, the proposed approach would permit a lender to resubmit a returned check no more than once, regardless of the channel used, before triggering the prohibition if the resubmission failed. The Bureau framed the proposed prohibition broadly so that it would apply to depository lenders that hold the consumer’s asset account, such as providers of deposit advance products or other types of proposed covered loans that may be offered by such depository lenders. Because depository lenders that hold consumers’ accounts have greater information about the status of those accounts than do third-party lenders, the Bureau believed that depository lenders should have little difficulty in avoiding failed attempts that would trigger the prohibition. Nevertheless, if such lenders elect to initiate payment transfers from consumers’ accounts when—as the lenders know or should know—the accounts lack sufficient funds to cover the amount of the payment transfers, they could assess the consumers substantial fees permitted under the asset account agreement (including NSF and overdraft fees), as well as any late fees or similar penalty fees permitted under the loan agreement for the covered loan. Accordingly, the Bureau believed that applying the prohibition in this manner would help to protect consumers from harmful practices in which such depository lenders may sometimes engage. As discussed above in Market Concerns—Payments, for example, the Bureau notably found that a depository institution that offered loan products to consumers with accounts at the institution charged some of those consumers NSF fees and overdraft fees for payment withdrawals initiated within the institution’s internal systems. 1046 Proposed comment 14(b)-1 explained the general scope of the prohibition. Specifically, it provided that the prohibition would restrict a lender from initiating any further payment transfers from the consumer’s account in connection with the covered loan, unless the requirements and conditions in either proposed § 1041.14(c) or (d) were satisfied. To clarify the ongoing application of the prohibition, proposed comment 14(b)-1 provided an example to show that a lender would be restricted from initiating transfers to collect payments that later fall due or to collect late fees or returned-item fees. The Bureau believed it was important to make clear that the proposed restriction on further transfers—in contrast to restrictions in existing laws and rules like the NACHA cap on re-presentments—would not merely limit the number of times a lender could attempt to collect a single failed payment. Lastly, proposed comment 14(b)-1 explained that the prohibition would apply regardless of whether the lender held an authorization or instrument from the consumer that was otherwise valid under applicable law, such as an authorization to collect payments via preauthorized EFTs under Regulation E or a post-dated check. Proposed comment 14(b)-2 clarified that when the prohibition is triggered, the lender is not prohibited under the rule from initiating a payment transfer in connection with a bona fide, subsequent covered loan made to the consumer, provided that the lender had not attempted to initiate two consecutive failed payment transfers in connection with the bona fide subsequent covered loan. The Bureau believed that limiting the restriction in this manner was appropriate to ensure that a consumer who had benefitted from the restriction at one time would not be effectively foreclosed from borrowing a covered loan from the lender after their financial situation had improved. 1047 Proposed 14(b)(1) General Proposed § 1041.14(b)(1) provided specifically that a lender must not initiate a payment transfer from a consumer’s account in connection with a covered loan after the lender has attempted to initiate two consecutive failed payment transfers from the consumer’s account in connection with that covered loan. It further proposed that a payment transfer would be deemed to have failed when it resulted in a return indicating that the account lacks sufficient funds or, for a lender that was the consumer’s account-holding institution, if it resulted in the collection of less than the amount for which the payment transfer was initiated because the account lacked sufficient funds. The specific provision for an account-holding institution thus would apply when such a lender elected to initiate a payment transfer resulting in the collection of either no funds or a partial payment. Proposed comments 14(b)(1)-1 to 14(b)(1)-4 provided clarification on when a payment transfer would be deemed to have failed. Specifically, proposed comment 14(b)(1)-1 explained that for purposes of the prohibition, a failed payment transfer included but was not limited to a debit or withdrawal that was returned unpaid or is declined due to nonsufficient funds in the consumer's account. This proposed comment clarified, among other things, that the prohibition applied to debit card transactions that were declined. Proposed comment 14(b)(1)-2 stated that the prohibition would apply as of the date on which the lender or its agent, such as a payment processor, received the return of the second consecutive failed transfer or, if the lender was the consumer’s account-holding institution, the date on which the transfer was initiated. The Bureau believed that, in contrast to other lenders, a consumer’s account-holding institution would or should have the ability to know that an account lacked sufficient funds before initiating a transfer 1048 (or immediately thereafter, at the latest). Proposed comment 14(b)(1)-3 clarified that a transfer that would result in a return for a reason other than a lack of sufficient funds was not a failed transfer for purposes of the prohibition, citing as an example a transfer that returned due to an incorrectly entered account number. Lastly, proposed comment 14(b)(1)-4 explained how the concept of a failed payment transfer would apply to a transfer initiated by a lender that was the consumer’s account-holding institution. Specifically, the proposed comment provided that if the consumer’s account-holding institution had initiated a payment transfer that resulted in the collection of less than the amount for which the payment transfer was initiated, because the account lacked sufficient funds, then the payment transfer would be a failed payment transfer for purposes of the prohibition. This would be the case regardless of whether the result was classified or coded as a return for nonsufficient funds in the lender’s internal procedures, processes, or systems. The Bureau believed that, unlike other lenders, such a lender would or should have the ability to know the result of a payment transfer and the reason for that result, without having to rely on a “return” as classified in its internal procedures, processes, or systems, or on a commonly understood reason code. Proposed comment 14(b)(1)-4 further stated that a consumer’s account-holding institution would not be deemed to initiated a failed payment transfer if the lender had merely deferred or forgone the debit or withdrawal of a payment from a consumer account, based on having observed a lack of sufficient funds. For such lenders, the Bureau believed it was important to clarify that the concept of a failed payment transfer incorporates the central concept of the proposed definition of payment transfer that the lender must engage in the affirmative act of initiating a debit or withdrawal from the consumer’s account in order for the term to apply. 1049 During the SBREFA process and in outreach with industry in developing the proposal, some lenders recommended that the Bureau take a narrower approach in connection with payment attempts by debit cards. One such recommendation suggested that the prohibition against additional withdrawal attempts should not apply when neither the lender nor the consumer’s account-holding institution charges an NSF fee in connection with a second failed payment attempt involving a debit card transaction that is declined. As explained in the proposal, the Bureau understood that depository institutions generally do not charge consumers NSF fees or declined authorization fees for such transactions, although it was aware that such fees are charged by some issuers of prepaid cards. It thus recognized that debit card transactions present somewhat less risk of harm to consumers. For a number of reasons, however, the Bureau did not believe that this potential effect was sufficient to propose excluding such transactions from the rule. First, the recommended approach would not protect consumers from the risk of incurring an overdraft fee in connection with the lender’s third withdrawal attempt. As discussed in Market Concerns—Payments, the Bureau’s research focusing on online lenders’ attempts to collect covered loan payments through the ACH system indicates that, in the small fraction of cases in which a lender’s third attempt succeeds—i.e., after the lender has sufficient information indicating that the account is severely distressed—up to one-third of the successful attempts are paid out of overdraft coverage. Second, the Bureau believed that the recommended approach would be impracticable to comply with and enforce, as the lender initiating a payment transfer would not necessarily know the receiving account-holding institution’s practice with respect to charging fees on declined or returned transactions. Additionally, the Bureau was concerned that lenders might respond to 1050 such an approach by seeking to evade the rule by re-characterizing their fees in some other manner. It thus believed that it was not appropriate to propose that payment withdrawal attempts by debit cards or prepaid cards be carved out of the rule, in light of the narrow range of those situations, the administrative challenges, and the residual risk to consumers. During the SBREFA process that preceded its issuance of the proposal, the Bureau received two other recommendations regarding the proposed restrictions on payment withdrawal attempts. One SER suggested that the Bureau delay imposing any restrictions until the full effects of NACHA’s recent 15 percent return rate threshold rule could be observed. As discussed in Markets Background—Payments, the NACHA rule that went into effect in 2015 can trigger inquiry and review by NACHA if a merchant’s overall return rate for debits made through the ACH network exceeds 15 percent. The Bureau considered the suggestion carefully but did not believe that a delay would be warranted. As noted, the NACHA rule applies only to returned debits through the ACH network. Thus, it places no restrictions on lenders’ attempts to withdraw payment through other channels. In fact, as discussed in the proposal (and confirmed by NACHA’s comment to the proposed rule), anecdotal evidence suggests that lenders are already shifting to use other channels to evade the NACHA rule. Further, exceeding the threshold merely triggers closer scrutiny by NACHA. To the extent that lenders making covered loans were to become subject to the review process, the Bureau believed that they might be able to justify their higher return rates by arguing that those higher rates are consistent with the rates for their market as a whole. Another SER recommended before the proposal was issued that lenders should be permitted to make up to four payment collection attempts per month when a loan is in default. 1051 The Bureau’s evidence indicates that for the covered loans studied, after a second consecutive attempt to collect payment fails, the third and subsequent attempts are also very likely to fail. The Bureau therefore believed that two consecutive failed payment attempts, rather than four presentment attempts per month, was the appropriate point at which to trigger the rule’s payment protections. In addition, the Bureau believed that in many cases where the proposed prohibition would apply, the consumer could technically be in default on the loan, considering that the lender’s payment attempts would have been unsuccessful. Thus, the suggestion to permit a large number of payment withdrawal attempts when a loan is in default could have effectively circumvented the proposed rule. Proposed 14(b)(2) Consecutive failed payment transfers Proposed § 1041.14(b)(2) would have defined a first failed payment transfer and a second consecutive failed payment transfer for purposes of determining when the prohibition in proposed § 1041.14(b) applies; the proposed commentary to this provision presented illustrative examples to explain and clarify the application of these terms. Proposed § 1041.14(b)(2)(i) provided that a failed transfer would be the first failed transfer if it met any of three conditions. First, proposed § 1041.14(b)(2)(i)(A) stated that a transfer would be the first failed payment transfer if the lender had initiated no other transfer from the consumer’s account in connection with the covered loan. This would apply to the scenario in which a lender’s very first attempt to collect payment on a covered loan had failed. Second, proposed § 1041.14(b)(2)(i)(B) provided that, generally, a failed payment transfer would be a first failed payment transfer if the immediately preceding payment transfer had been successful, regardless of whether the lender had previously initiated a first failed payment transfer. This proposed provision set forth the 1052 general principle that any failed payment transfer that followed a successful payment transfer would be the first failed payment transfer for the purposes of the prohibition in proposed § 1041.14(b). Lastly, proposed § 1041.14(b)(2)(i)(C) provided that a payment transfer would be a first failed payment transfer if it was the first failed attempt after the lender obtained the consumer’s authorization for additional payment transfers pursuant to proposed § 1041.14(c). Proposed comment 14(b)(2)(i)-1 provided two illustrative examples of a first failed payment transfer. Proposed § 1041.14(b)(2)(ii) provided that a failed payment transfer would be the second consecutive failed payment transfer if the previous payment transfer was a first failed transfer, and defined the concept of a previous payment transfer to include a payment transfer initiated at the same time or on the same day as the failed payment transfer. Proposed comment 14(b)(2)(ii)-1 provided an illustrative example of the general concept of a second consecutive failed payment transfer, while proposed comment 14(b)(2)(ii)-2 provided an illustrative example of a previous payment transfer initiated at the same time and on the same day. Given the high failure rates for same-day presentments, the Bureau believed it was important to clarify that the prohibition would be triggered when two payment transfers initiated on the same day fail, including instances where they had been initiated concurrently. Proposed comment 14(b)(2)(ii)3 clarified that if a lender initiated a single immediate payment transfer at the consumer’s request pursuant to the exception in § 1041.14(d), then the failed transfer count would remain at two, regardless of whether the transfer succeeded or failed. Thus, as the proposed comment further provided, the exception would be limited to the single transfer authorized by the consumer. Accordingly, if a payment transfer initiated pursuant to the exception failed, then the lender 1053 would not be permitted to reinitiate the transfer—e.g., by re-presenting it through the ACH system—unless the lender had first obtained a new authorization from the consumer, pursuant to § 1041.14(c) or (d). The Bureau believed this limitation was necessary, as the authorization for an immediate transfer would be based on the consumer’s understanding of their account’s condition only at that specific moment in time, as opposed to its possible condition in the future. Proposed § 1041.14(b)(2)(iii) would have provided the principle that alternating between payment channels does not reset the failed payment transfer count. Specifically, it proposed that a failed payment transfer meeting the conditions in proposed § 1041.14(b)(2)(ii) is the second consecutive failed transfer, regardless of whether the first failed transfer was initiated through a different payment channel. Proposed comment 14(b)(2)(iii)-1 provided an illustrative example of this concept. Comments Received Several industry representatives and lender commenters generally opposed the Bureau’s proposal. These commenters stated that new industry guidelines issued by NACHA were sufficient to address the harms identified by the Bureau. Specifically, those new rules set return thresholds, including a 15 percent rate of total returns, a three percent rate of administrative returns, and a 0.5 percent rate of unauthorized transaction returns, and clarified the limits on payment splitting and re-presentments, as noted above. Conversely, other commenters argued against delaying or forgoing the proposed approach because, as the Bureau noted in the proposal, NACHA’s new guidelines do not impact payment transfers initiated outside the ACH system. Various stakeholders commented on the number of failed payment transfers that the proposed rule allowed. Some noted that NACHA operating rules and general industry standards 1054 allow three attempts to collect a single payment. Others expressed concerns that the proposed rule would in effect reduce the allowance to two attempts, which would require NACHA to amend its operating rules, and depository institutions and lenders to adjust their systems. Yet others argued that the Bureau should not measure all presentments against the presentment cap, but should instead measure presentments of the same payment, consistent with NACHA’s approach. A few commenters objected to counting payment attempts towards the cap crosspayment method, and expressed concerns about the compliance costs associated with tracking payments across channels. However, some industry participants agreed with the proposed two-attempt limit proposed, which they claimed to already have adopted. Other stakeholders argued that the rule should prohibit payment transfer attempts after one failed attempt. One such commenter claimed that gaining the ability to debit a borrower’s account would reduce the lender’s incentive to determine whether the borrower would have the ability to repay the loan and cover other obligations. It also argued that even one overdraft or NSF fee could generate additional debt and fees that would quickly snowball. Some commenters argued that the Bureau should only declare the initiation of repeated presentments as unfair or abusive. In other words, this commenter believed that just finalizing this section, and not any of the ability-to-repay requirements, would suffice to address the identified harms without imposing significant industry costs. One commenter also was concerned that, as written, the proposal could be interpreted to require depository institutions to: (1) monitor lenders’ use of the payment system; (2) determine when a lender may be in violation of proposed §§ 1041.14 and 1041.15; and (3) act as an enforcer of the regulation even where the 1055 consumer authorized the transaction. This commenter asked the Bureau to clarify that the responsibility of ensuring compliance with these provisions would be exclusively an obligation of the lender, and not an obligation of the lender’s or the consumer’s depository institution. Other commenters stated that instead of prohibiting additional payment transfers after a number of previous failed attempts, the Bureau should require lenders to provide payment notices that include reminders that consumers have the ability to stop payments or revoke existing payment authorizations. These commenters shared the sentiment of commenters, discussed in the section-by-section analysis of § 1041.7 above, that borrowers should be able to avoid the harm by initiating stop payments or revoking payment authorizations with lenders, and argued that disclosure would help improve the efficacy of those mechanisms to a point where the harms would largely be eliminated. One commenter asked the Bureau to additionally require reauthorization from the consumer after three failed attempts in a 12-month period, even when those attempts are not consecutive. A number of comments from State Attorneys General and consumer groups also touted the benefits of the approach described in the proposed rule. These commenters noted that the limit on payment transfer attempts was essential because it would reduce fees and bolster the ability-to-repay determination. Final Rule The Bureau is finalizing the cap on payment presentments in § 1041.8(b), consistent with the conclusions reached above in the section-by-section analysis of § 1041.7 of the final rule. The Bureau is, however, making some changes to the proposed rule. 1056 First, to clarify that the presentment cap will apply across all loans with the lender, the Bureau is replacing, in two places in § 1041.8(b)(1), the phrase “in connection with a covered loan” with “in connection with any covered loan that the consumer has with the lender.” Similarly, the Bureau is adding “or any other covered loan that the consumer has with the lender” at the end of § 1014.8(b)(2)(i)(A). A lender will need to seek a new authorization, or cease payment attempts, after two failed attempts on any loan the borrower has with the lender. Accordingly, if a borrower has two outstanding covered loans and a lender makes a failed payment attempt for each such loan in succession, then the cap is met. The proposed rule could have been interpreted to apply only to two failed attempts on one loan, and then two failed attempts on a different loan, and so forth. Yet the Bureau has adopted this change in order to ensure that the rule fully prevents the scope of harms intended to be covered under the rule in light of its understanding and description of the practice that it has identified as unfair and abusive. Regardless of whether the multiple presentments are for one loan, or spread across multiple loans, the borrower harm and expected value would be the same.1037 To the extent lenders are not currently tracking payments across multiple loans, there may be some additional costs associated with this adjustment. However, the Bureau does not expect, once systems are updated, any additional compliance costs. Comment 8(b)-1 is amended to incorporate this point, and a new comment 8(b)-3 is added for further clarity and to add an example as well. In addition, the comments related to § 1014.8(b) have been revised to clarify the prohibition’s application to situations in which a 1037 The Bureau’s Online Payday Loans Payments report on online payday and payday installment lending did not distinguish between multiple payments for individual loans and multiple payments for multiple loans. CFPB Online Payday Loan Payments. 1057 consumer has more than one covered loan with a lender. The Bureau is also adding an example of a consumer with two covered loans who has a second failed payment transfer, in comment 8(b)(2)(ii)-1.ii. The second modification of this provision is intended to clarify, in § 1041.8(b)(1) and elsewhere in the final rule, that the presentment cap applies on a per-consumer-account basis. That means if a lender attempts to withdraw payments from multiple accounts, the lender is limited to two consecutive failed attempts each. The Bureau makes this clarification because the presumption that funds are unlikely to be available for a third presentment does not follow when the presentment is made from a different account. Two consecutive failed attempts from one account tell the lender nothing about the condition of another account. However, the prohibition applies to the other account if the lender then initiates two consecutive failed payment transfers from that account. The Bureau is adding a new comment 8(b)-2 to clarify this point. Third, the Bureau is making technical edits to the description, in § 1041.8(b)(1), of what constitutes a failed payment transfer when the lender is also the consumer’s account-holding institution. That description, both in the proposal and in the final rule, provides that for such lenders, presentments resulting in non-sufficient funds, partial payments, or full payments paid out of overdraft all count toward the cap. The Bureau is making these edits for consistency with the new conditional exclusion in § 1041.8(a)(1). The Bureau also is making similar conforming edits to comment 8(b)(1)-4. Lastly, the Bureau has made some other technical edits to § 1041.8(b)(2)(ii) for consistency with § 1041.8(b)(2)(i). 1058 In Market Concerns—Payments and the section-by-section analysis of § 1041.7, the Bureau has already addressed the comments about whether this rule is necessary in light of NACHA’s new guidelines. But to summarize again briefly, the Bureau believes that NACHA guidelines do not suffice to prevent all of the harms associated with the practice identified in § 1041.7. In particular, they would not prevent the second presentment or the third payment attempt. Commenters noted this difference and asserted that complying with the rule as proposed would require companies to change their systems. As explained in the section-bysection analysis of § 1041.7, the Bureau finds that there is a significant amount of injury in that third presentment: the Bureau’s study showed that approximately 80 percent of such presentments caused an overdraft fee or failed (and likely caused an NSF fee and/or returneditem fee). Importantly, not only do the NACHA Rules apply only to payments made through the ACH network, but NACHA’s own comment noted that it had already seen vendors shift to using other payment methods, likely in an effort to evade the NACHA Rules. The Bureau has chosen to use a two-presentment cap to prevent consumer harms from the practice that it has identified as unfair and abusive. It did so not because the first re-presentment causes no injury, but rather because the injury after each failed attempt is cumulative and thus the injury becomes more significant over time. In addition, the first re-presentment implicates certain additional countervailing benefits, as lenders may have simply tried the first presentment at the wrong time, and consumers may find it more convenient not to have to reauthorize after just one failed attempt. Additionally, if lenders only have one try, it may cause them to be overly circumspect about when to use it, which could undermine the benefits of ease and convenience for consumers. The Bureau therefore is drawing the line at two re-presentments in an abundance 1059 of caution, in an attempt to avoid regulating potentially more legitimate justifications for representment. Nonetheless, the Bureau is aware of the harms that can occur even from a single re-presentment, and that the manner in which a lender engages in re-presentment activities more generally could be unfair, deceptive, or abusive. The rule does not provide a safe harbor against misconduct that it does not explicitly address, and the Bureau could in appropriate circumstances address problems through its supervisory and enforcement authority.1038 For purposes of determining whether the cap has been met, the Bureau has decided not to distinguish between re-presentments of the same payment and new presentments to cover new loan installments, as NACHA does. As the Bureau stated in the proposal, and now affirms, the tailoring of individualized requirements for each discrete payment practice would add considerable complexity to the rule and yet still could leave consumers vulnerable to harms from aggressive and evasive practices that may emerge in markets for covered loans in the future. Accordingly, the Bureau is addressing a somewhat broader practice that it has determined to be unfair and abusive by providing significant consumer protections from a range of harms in a considerably less complex fashion. Notably, the Bureau’s study that showed very high rates of rejection and overdraft fees for third presentments did not distinguish between re-presentments of the same payment and new presentments for new installments. And the Bureau believes that after two failed attempts to the same account, even if two weeks or a month has passed, there is reason to believe a third would fail, and that obtaining a new authorization would be appropriate. 1038 See, e.g., Press Release, Bureau of Consumer Fin. Prot., “CFPB Orders EZCORP to Pay $10 Million for Illegal Debt Collection Tactics,” (Dec. 16, 2015), available at http://www.consumerfinance.gov/newsroom/cfpb-ordersezcorp-to-pay-10-million-for-illegal-debt-collection-tactics/; Press Release, Bureau of Consumer Fin. Prot., “CFPB Takes Action Against Online lender for Deceiving Borrowers,” (Nov. 18, 2015), available at https://www.consumerfinance.gov/about-us/newsroom/cfpb-takes-action-against-online-lender-for-deceivingborrowers/. 1060 The Bureau thus concludes that considerable injury is likely occurring from such new payment attempts and thus inclusion of those payments towards the cap is warranted. As noted above, one commenter suggested finalizing this portion of the rule as a standalone, without the underwriting provisions requiring lenders to make a reasonable, abilityto-repay determination. The Bureau declines to follow this approach, as it continues to believe that § 1041.8 alone could not prevent all of the harms that flow from the practice identified in § 1041.7, including those stemming from the practice identified in § 1041.4. If lenders continue to make covered loans without assessing borrowers’ ability to repay, consumers would still confront the harms associated with unaffordable loans—default, delinquency, re-borrowing, or other collateral injuries as described above in Market Concerns—Underwriting. The payment provisions of this rule address one of the potential collateral injuries from an unaffordable loan— which is itself an important source of harm—but they do not address the whole scope of harm that the Bureau seeks to address in part 1041. Therefore, the Bureau concludes that it would be quite insufficient to finalize subpart C of this rule by itself. Furthermore, the Bureau concludes that disclosures alone would not suffice to prevent all of the harms caused by the unfair and abusive practice identified in § 1041.7 of the final rule. As explained above in Market Concerns—Payments and the section-by-section analysis of § 1041.7, the Bureau has observed significant difficulty when borrowers seek to stop payments or revoke authorizations. Disclosures may be effective in helping consumers know their rights, and understand what is occurring, but they would not help consumers stop the multiple attempts. Furthermore, while the Bureau believes its model disclosures will be effective in informing some consumers, the Bureau knows there are many others they will not reach or for whom they will 1061 not be as effective. As discussed below, one commenter described that it had tested the Bureau’s “notice of restrictions on future loans,” which does not pertain to this particular part of the rule. The Bureau believes the methodology of that testing may have been flawed as noted in the section-by-section analysis of § 1041.6, but as we noted above, it is a reminder of the fact that disclosures in complicated areas, such as the payment attempt practices at issue here, are unlikely to be as effective as a substantive intervention shaped to respond more directly to the harms caused by the practice identified as unfair and abusive. That conclusion here is also consistent with the Bureau’s conclusion about the effectiveness of disclosures as a possible alternative to the ability-to-repay requirements laid out above in Market Concerns—Underwriting and the section-by-section analysis of § 1041.4. The principal obligation to comply with §§ 1041.8 and 1041.9 rests on the lender. Of course, if the lender uses a service provider to manage its payment withdrawals, that service provider may also be liable for any violation of the rule, as provided in the Dodd-Frank Act.1039 The Bureau does not intend for this rule to have the effect of changing the obligations of nonlender depository institutions. The Bureau also has decided not to require reauthorization after three failed attempts in a 12-month period. The effect of this change would be to establish a one-attempt cap where the lender had previously reached the two-attempt cap in the same 12-month period, or trigger the cap where, for example, every other payment fails. The Bureau has set the two-attempt cap to track the practice identified as unfair and abusive, and to avoid being overly restrictive by allowing the lender to make one more payment attempt after the first failed attempt following an 1039 12 U.S.C. 5531; 12 U.S.C. 5536(a). 1062 authorization. The Bureau concludes that adding this requirement about the number of attempts in a 12-month period would add further complexity to the rule and would increase the burdens associated with tracking payment attempts. 8(c) Exception for additional payment transfers authorized by the consumer Proposed Rule Whereas proposed § 1041.14(b) would have established the prohibition on further payment withdrawals, proposed § 1041.14(c) and (d) would have established requirements for obtaining the consumer’s new and specific authorization to make further payment withdrawals. Proposed § 1041.14(c) was framed as an exception to the prohibition, even though payment withdrawals made pursuant to its requirements would not fall within the scope of the unfair and abusive practice preliminarily identified in proposed § 1041.13 (now § 1041.7 of the final rule). Under the proposal, a new authorization obtained pursuant to proposed § 1041.14(c) would reset to zero the failed payment transfer count under proposed § 1041.14(b), whereas an authorization obtained pursuant to proposed § 1041.14(d) would not. Accordingly, a lender would be permitted under proposed § 1041.14(c) to initiate one or more additional payment transfers that are authorized by the consumer in accordance with certain requirements and conditions, and subject to the general prohibition on initiating a payment transfer after two consecutive failed attempts. The proposed authorization requirements and conditions in proposed § 1041.14(c) were designed to assure that, before a lender initiated another payment transfer (if any) after triggering the prohibition, the consumer did in fact want the lender to resume making payment transfers and that the consumer understands and had agreed to the specific date, amount, and payment channel for those succeeding payment transfers. The Bureau 1063 stated that requiring the key terms of each transfer to be clearly communicated to the consumer before the consumer decides whether to grant authorization would help assure that the consumer’s decision is an informed one and that the consumer understands the consequences that may flow from granting a new authorization and help the consumer avoid future failed payment transfers. The Bureau believed that, when this assurance was provided, it no longer would be unfair or abusive for a lender to initiate payment transfers that accord with the new authorization, at least until such point that the lender initiated two consecutive failed payment transfers pursuant to the new authorization. The Bureau recognized that, in some cases, lenders and consumers might want to use an authorization under this exception to resume payment withdrawals according to the same terms and schedule that the consumer had authorized prior to the two consecutive failed attempts. In other cases, lenders and consumers might want to establish a new authorization to accommodate a change in the payment schedule—as might be the case, for example, when the consumer entered into a workout agreement with the lender. Accordingly, the proposed exception was designed to be sufficiently flexible to accommodate both circumstances. In either circumstance, however, the lender would be permitted to initiate only those transfers authorized by the consumer under proposed § 1041.14(c). Proposed § 1041.14(c)(1) would establish the general exception to the prohibition on additional payment transfer attempts under § 1041.14(b), while the remaining subparagraphs would specify particular requirements and conditions. First, proposed § 1041.14(c)(2) would establish the general requirement that for the exception to apply to an additional payment transfer, the transfer’s specific date, amount, and payment channel must be authorized by the 1064 consumer. In addition, proposed § 1041.14(c)(2) would address the application of the specific date requirement to re-initiating a returned payment transfer and also address authorization of transfers to collect a late fee or returned item fee, if such fees are incurred in the future. Second, proposed § 1041.14(c)(3) would establish procedural and other requirements and conditions for requesting and obtaining the consumer’s authorization. Lastly, proposed § 1041.14(c)(4) would address circumstances in which the new authorization becomes null and void. Each of these sets of requirements and conditions is discussed in detail below. Proposed comment 14(c)-1 summarized the exception’s main provisions, and noted the availability of the exception in proposed § 1041.14(d). Proposed § 1041.14(c)(1) provided that, notwithstanding the prohibition in proposed § 1041.14(b), a lender would be permitted to initiate additional payment transfers from a consumer’s account after two consecutive transfers by the lender had failed if the transfers had been authorized by the consumer as required by proposed § 1041.14(c), or if the lender had executed a single immediate payment transfer at the consumer’s request under proposed § 1041.14(d). Proposed comment 14(c)(1)-1 explained that the consumer’s authorization required by proposed § 1041.14(c) would be in addition to, and not in lieu of, any underlying payment authorization or instrument required to be obtained from the consumer under applicable laws. The Bureau noted, for example, that an authorization obtained pursuant to proposed § 1041.14(c) would not take replace an authorization that a lender would be required to obtain under applicable laws to collect payments via RCCs, if the lender and consumer wished to resume payment transfers using that method. However, in cases where lenders and consumers wished to resume payment transfers via preauthorized EFTs, as that term is defined in 1065 Regulation E, the Bureau believed that—given the high degree of specificity required by proposed § 1041.14(c) —lenders could comply with the authorization requirements in Regulation E, 12 CFR 1005.10(b) and the requirements in proposed § 1041.14(c) within a single authorization process. Proposed § 1041.14(c)(2)(i) would establish the general requirement that for the exception in proposed § 1041.14(c) to apply to an additional payment transfer, the transfer’s specific date, amount, and payment channel must be authorized by the consumer. The Bureau believed that requiring lenders to explain these key terms of each transfer to consumers when seeking authorization would help ensure that consumers could make an informed decision between granting authorization for additional payment transfers, and other convenient repayment options—e.g., payments by cash or money order, “push” bill payment services, and single immediate payment transfers authorized pursuant to proposed § 1041.14(d) —which would help them avoid future failed payment transfers. With respect to lenders that wished to obtain permission to initiate ongoing payment transfers from a consumer whose account has already been subject to two consecutive failed attempts, the Bureau believed it was important to require such lenders to obtain the consumer’s agreement to the specific terms of each future transfer from the outset, rather than to provide for less specificity upfront and rely instead on the fact that under proposed § 1041.15(b), every consumer with a covered loan will receive notice containing the terms of each upcoming payment transfer. As discussed above, the Bureau believed that, in general, the proposed required notice for all payment transfers would help to reduce harms that may occur from payment transfers by alerting the consumers to the upcoming attempt in sufficient time for them to arrange to make a required payment when they could afford it, and to make choices that might 1066 minimize the attempt’s impact on their accounts when the timing of a payment is not aligned with their finances. However, the Bureau believed that consumers whose accounts have already experienced two failed payment withdrawal attempts in succession would have demonstrated a degree of financial distress that would make it unlikely that a notice of another payment attempt would enable them to avoid further harm. Proposed comment 14(c)(2)(i)-1 explained the general requirement that the terms of each additional payment transfer must be authorized by the consumer in order to qualify for the exception. It further clarified that for the exception to apply to an additional payment transfer, these required terms had to be included in the signed authorization that the lender would be required to obtain from the consumer. Proposed comment 14(c)(2)(i)-2 clarified that the requirement that the specific date of each additional transfer be expressly authorized would be satisfied if the consumer authorizes the month, day, and year of the transfer. Proposed comment 14(c)(2)(i)-3 clarified that the exception would not apply if the lender initiated an additional payment transfer for an amount larger than the amount authorized by the consumer, unless it satisfied the requirements and conditions in proposed § 1041.14(c)(2)(iii)(B) for adding the amount of a late fee or returned item fee to an amount authorized by the consumer. Proposed comment 14(c)(2)(i)-4 clarified that a payment transfer initiated pursuant to § 1041.14(c) would be initiated for the specific amount authorized by the consumer if its amount was equal to or smaller than the authorized amount. The Bureau recognized that in certain circumstances it might be necessary for the lender to initiate transfers for a smaller amount than 1067 specifically authorized including, for example, when the lender needed to exclude from the transfer the amount of a partial prepayment. In addition, the Bureau believed that this provision would provide useful flexibility in instances where the prohibition on further payment transfers is triggered at a time when the consumer has not yet fully drawn down on a line of credit. In such instances, lenders and consumers might want to structure the new authorization to accommodate payments on future draws by the consumer. With this provision for smaller amounts, the lender could seek authorization for additional payment transfers for the payment amount that would be due if the consumer had drawn the full amount of remaining credit, and then would be permitted under the exception to initiate the transfers for amounts smaller than the specific amount, if necessary. Proposed § 1041.14(c)(2)(ii) would establish a narrow exception to the general requirement that an additional payment transfer be initiated on the date authorized by the consumer. Specifically, it would provide that when a payment transfer authorized by the consumer pursuant to the exception is returned for nonsufficient funds, the lender would be permitted to re-present the transfer on or after the date authorized by the consumer, provided that the returned transfer had not triggered the prohibition on further payment transfers in proposed § 1041.14(b). The Bureau believed that this narrow exception would accommodate practical considerations in payment processing and noted that the prohibition in proposed § 1041.14(b) would protect the consumer if the re-initiation had failed. Proposed § 1041.14(c)(2)(iii) contained two separate provisions that would permit a lender to obtain the consumer’s authorization for, and to initiate, additional payment transfers to collect a late fee or returned-item fee. Both of these provisions were intended to permit lenders 1068 to use a payment authorization obtained pursuant to proposed § 1041.14(c)(2)(iii) to collect a fee that was not anticipated when the authorization was obtained, without having to go through a second authorization process under proposed § 1041.14(c). First, proposed § 1041.14(c)(2)(iii)(A) would permit a lender to initiate an additional payment transfer solely to collect a late fee or returned-item fee without obtaining a new consumer authorization for the specific date and amount of the transfer only if the lender, in the course of obtaining the consumer’s authorization for additional payment transfers, had informed the consumer of the fact that individual payment transfers to collect a late fee or returned-item fee might be initiated, and had obtained the consumer’s general authorization for such transfers in advance. Specifically, the lender could initiate such transfers only if the consumer’s authorization obtained pursuant to proposed § 1041.14(c) included a statement, in terms that were clear and readily understandable to the consumer, that the lender might initiate a payment transfer solely to collect a late fee or returned-item fee. In addition, the lender would be required to specify in the statement the highest amount for such fees that may be charged, as well as the payment channel to be used. The Bureau believed this required statement might be appropriate to help ensure that the consumer is aware of key information about such transfers—particularly the highest possible amount—when the consumer would be deciding whether to grant an authorization. Proposed comment 14(c)(2)(iii)(A)-1 clarified that the consumer’s authorization for an additional payment transfer solely to collect a late fee or returned item fee needed not satisfy the general requirement that the consumer must authorize the specific date and amount of each additional payment transfer. Proposed comment 14(c)(2)(iii)(A)-2 provided, as an example, that 1069 the requirement to specify the highest possible amount that might be charged for a fee would be satisfied if the required statement specified the maximum amount permissible under the loan agreement. Proposed comment 14(c)(2)(iii)(A)-3 provided that if a fee might vary due to remaining loan balance or other factors, then the lender had to assume the factors that would result in the highest possible amount in calculating the specified amount. The second provision, proposed § 1041.14(c)(2)(iii)(B), would have permitted a lender to add the amount of one late fee or one returned-item fee to the specific amounts authorized by the consumer as provided under proposed § 1041.14(c)(2) only if the lender had informed the consumer of the fact that such transfers for combined amounts might be initiated, and had obtained the consumer’s general authorization for such transfers in advance. Specifically, under the proposal, the lender could initiate transfers for such combined amounts only if the consumer’s authorization included a statement, in terms that were clear and readily understandable to the consumer, that the amount of one late fee or one returned-item fee might be added to any payment transfer authorized by the consumer. In addition, the lender would be required to specify in the statement the highest amount for such fees that may be charged, as well as the payment channel to be used. Proposed comment 14(c)(2)(iii)(B)-1 provided further clarification on that provision. Proposed § 1041.14(c)(3) provided a three-step process for obtaining a consumer’s authorization for additional payment transfers. First, proposed § 1041.14(c)(3)(ii) would contain provisions for requesting the consumer’s authorization. The permissible methods for requesting authorization would allow lenders considerable flexibility. For example, lenders would be permitted to provide the transfer terms to the consumer in writing or (subject to certain 1070 requirements and conditions) electronically without regard to the consumer consent and other provisions of the E-Sign Act. In addition, lenders would be permitted to request authorization orally by telephone, subject to certain requirements and conditions. In the second step, proposed § 1041.14(c)(3)(iii) provided that, for an authorization to be valid under the exception, the lender had to obtain an authorization that is signed or otherwise agreed to by the consumer and that includes the required terms for each additional payment transfer. The lender would be permitted to obtain the consumer’s signature in writing or electronically, provided the E-Sign Act requirements for electronic records and signatures were met. This was intended to facilitate requesting and obtaining the consumer’s signed authorization in the same communication. In the third and final step, proposed § 1041.14(c)(3)(iii) also would require the lender to provide to the consumer memorialization of the authorization no later than the date on which the first transfer authorized by the consumer is initiated. The comments to proposed § 1041.14(c)(3) specified and explained these points in greater detail. Under the proposal, the lender would be permitted to provide the memorialization in writing or electronically, without regard to the consumer consent and other provisions of the E-Sign Act, provided that it was in a retainable form. In developing this three-step approach, the Bureau endeavored to ensure that the precise terms of the additional transfers for which a lender sought authorization were effectively communicated to the consumer during each step of the process, and that the consumer had the ability to decline authorizing any payment transfers with terms that the consumer believed would likely cause challenges in managing her account. In addition, the Bureau designed the approach to be compatible with lenders’ existing systems and procedures for obtaining other types of payment authorizations, particularly authorizations for preauthorized, or “recurring,” EFTs under 1071 Regulation E. Accordingly, the proposed procedures generally were designed to mirror existing requirements in Regulation E, 12 CFR 1005.10(b). Regulation E requires that preauthorized EFTs from a consumer’s account be authorized “only by a writing signed or similarly authenticated by the consumer.”1040 Under EFTA and Regulation E, companies can obtain the required consumer authorizations for preauthorized EFTs in several ways. Consumer authorizations can be provided in paper form or electronically. The commentary to Regulation E explains that the rule “permits signed, written authorizations to be provided electronically,” and specifies that the “writing and signature requirements . . . are satisfied by complying with the [ESign Act] which defines electronic records and electronic signatures.”1041 Regulation E does not prohibit companies from obtaining signed, written authorizations from consumers over the phone if the E-Sign Act requirements for electronic records and signatures are met.1042 In addition, Regulation E requires persons that obtain authorizations for preauthorized EFTs to provide a copy of the terms of the authorization to the consumer.1043 The copy of the terms of the authorization must be provided in paper form or electronically.1044 The Bureau understands that this requirement in Regulation E, 12 CFR 1005.10(b), is not satisfied by providing the consumer with a recording of a telephone call. 1040 See 12 CFR 1005.10(b). 12 CFR part 1005, Supp. I, comment 10(b)-5. The E-Sign Act establishes that electronic signatures and electronic records are valid and enforceable if they meet certain criteria. See 15 U.S.C. 7001(a)(1). An electronic signature is “an electronic sound, symbol, or process, attached to or logically associated with a contract or other record and executed or adopted by a person with the intent to sign the record.” 15 U.S.C. 7006(5). An electronic record is “a contract or other record created, generated, sent, communicated, received, or stored by electronic means.” 15 U.S.C. 7006(4). 1042 In 2006, the Board explained that if certain types of tape-recorded authorizations constituted a written and signed (or similarly authenticated) authorization under the E-Sign Act, then the authorization would satisfy Regulation E requirements as well. 71 FR 1638, 1650 (Jan. 10, 2006). 1043 See 12 CFR 1005.10(b). 1044 See 12 CFR part 1005, Supp. I, comment 10(b)-5. 1041 1072 During the SBREFA process, an SER recommended that the procedures for obtaining consumers’ re-authorization after lenders trigger the proposed cap on failed presentments should be similar to existing procedures for obtaining consumers’ authorizations to collect payment by preauthorized EFTs under Regulation E. The Bureau believed that harmonizing the two procedures would reduce costs and burdens on lenders by permitting them to incorporate the proposed procedures for obtaining authorizations into existing systems. Accordingly, as discussed above, the proposed approach was designed to achieve this goal. Lastly, proposed § 1041.14(c)(4) would specify the circumstances in which an authorization for additional payment transfers obtained pursuant to proposed § 1041.14(c) expires or becomes inoperative. First, proposed § 1041.14(c)(4)(i) provided that a consumer’s authorization would become null and void for purposes of the exception if the lender obtained a subsequent new authorization from the consumer pursuant to the exception. This provision was intended to ensure that, when necessary, lenders could obtain a consumer’s new authorization to initiate transfers for different terms, or to continue collecting payments on the loan, and that such new authorization would supersede the prior authorization. Second, proposed § 1041.14(c)(4)(ii) provided that a consumer’s authorization would become null and void for purposes of the exception if two consecutive payment transfers initiated pursuant to the consumer’s authorization had failed, as specified in proposed § 1041.14(b). The Bureau proposed this provision for clarification purposes. Comments Received A number of commenters objected to the proposal that companies would have to obtain new authorizations after two failed attempts. More specifically, many of the commenters 1073 focused on how the rule would impact recurring debits or preauthorized EFTs. Under the proposal, if two recurring debits or EFTs failed, then the lender would have to receive a new authorization from the borrower under proposed § 1041.8(c) or (d) to continue processing payment transfers. Commenters argued that this could harm consumers because they might default or become delinquent on the loan if they believed the recurring transfers would continue, but the lender could not initiate further transfers because two previous transfers had been rejected. Commenters stated that a required notice informing borrowers of their right to revoke an authorization under Regulation E would be more appropriate for circumstances involving preauthorized EFTs. Commenters also argued that the rule would deter lenders from using recurring transfers, a convenience to borrowers, if it meant that the loan would then be considered a covered longerterm loan subject to the requirements of the rule. As stated previously, the Bureau also received a number of comments describing purported inconsistencies with the NACHA Rules. Specific to the proposed exception in § 1041.14(c), commenters noted that the NACHA Rules currently do not allow companies to add fees to an authorized amount, and instead only permit companies to initiate separate transfers for fees if the company had obtained the consumer’s authorization for such transfers. A consumer group asked the Bureau to clarify that the proposed “failed payment clock” would start again after reauthorization, meaning that if a lender reached the payment transfer limit, and then obtained reauthorization under proposed 1041.14(c), then the borrower would need to get another new authorization if the lender again reaches the payment transfer limit. 1074 Finally, the Bureau received comments generally supportive of the proposition that a lender should be required to, and allowed to, obtain a new authorization after two consecutive attempts have failed. Final Rule The Bureau is now finalizing § 1041.8(c)—which is renumbered from § 1041.14(c) of the proposed rule—with a few revisions to the content of the regulation and corresponding commentary. Most notably, the Bureau is modifying proposed § 1041.8(c)(2)(iii), which permits lenders to collect late fees and returned-item fees pursuant to the exception in § 1041.8(c). Specifically, in light of comments noting inconsistencies with NACHA Rules, the Bureau is deleting proposed paragraph (c)(2)(iii)(B), which would have permitted lenders to add the amount of such a fee to the amount of any payment transfer initiated pursuant to the exception, provided that the consumer authorized the addition of the fee amount. Accordingly, the Bureau is finalizing the provisions in § 1041.8(c)(2)(iii) to permit lenders to initiate a payment transfer to collect a late fee or returned-time fee under the exception in § 1041.8(c) only as a stand-alone transfer for the amount of the fee itself, and only if authorized by the consumer in accordance with the rule’s requirements. The Bureau notes that limiting such transfers in this way is consistent with existing practices of lenders that comply with NACHA Rules. Because the Bureau has deleted paragraph (c)(2)(iii)(B), paragraph (c)(2)(iii)(A) has been renumbered as (c)(2)(iii). The Bureau has also deleted the corresponding comment, and renumbered the remaining comments to reflect the change. The Bureau clarified the remaining paragraph (c)(2)(iii) as well. As discussed immediately above, that section allows lenders to initiate payment transfers for the collection of 1075 fees when a consumer has authorized such transfers. The Bureau replaced the word “authorized” with the phrase “has authorized the lender to initiate such payment transfers in advance of the withdrawal attempt” to indicate that the authorization cannot be obtained after-the-fact. The Bureau is making no other substantive changes to paragraph (c) or its corresponding comments, and finalizes the section as otherwise proposed. A number of the comment topics related to the prohibition on repeated failed payment attempts were already addressed above in Market Concerns—Payments or in the section-bysection analysis of § 1041.7, which identified this unfair and abusive practice. The Bureau recognizes that with recurring debits or preauthorized EFTs involving installment loans, if two scheduled payments fail, the recurring transfers would need to cease until after the lender has obtained a new authorization. It also recognizes that this could be an inconvenience, but nonetheless believes the interest of ceasing payment attempts when the consumer’s account has demonstrated that it lacks the funds to cover ongoing payment attempts warrants the inclusion of preauthorized EFTs. As stated in § 1041.8(c), borrowers who wish to continue making payments out of that account can simply reauthorize, including by setting up a new authorization for preauthorized EFTs. They can also request a single immediate payment transfer under § 1041.8(d) at any time. Concerns that the rule might deter lenders from offering recurring transfers on high-cost longer-term installment loans, because it would bring the loan under the requirements of the rule as proposed, are mitigated by the fact that the Bureau currently is not finalizing the ability-torepay underwriting criteria as to high-cost longer-term installment loans. As a result, the only provisions of the rule that could be triggered by a leveraged payment mechanism are the 1076 requirements relating to payment attempts. It is, however, still possible that a lender that is making high-cost longer-term installment loans might choose not to take a leveraged payment mechanism, including by not offering preauthorized EFTs. Borrowers in these circumstances could set up recurring “push” payments with their account-holding institution, instead of giving lenders authorization to initiate a “pull,” thereby still obtaining the convenience of recurring automatic transfers. The Bureau notes that these borrowers would also avoid all of the harms identified in § 1041.7 because the lender would not be authorized to initiate payment requests themselves. The Bureau does not find it necessary, contrary to some received comments, to clarify further that the “failed payment clock” under § 1041.8(b) restarts after a borrower provides a new authorization under § 1041.8(c). Section 1041.8(b)(2)(i)(C) makes clear that the clock does restart after a borrower reauthorizes under § 1041.8(c). 8(d) Exception for initiating a single immediate payment transfer at the consumer’s request Proposed Rule Proposed § 1041.14(d) set forth a second exception to the prohibition on initiating further payment transfers from a consumer’s account in proposed § 1041.14(b). In contrast to the exception available under proposed § 1041.14(c), which would allow lenders to initiate multiple recurring payment transfers authorized by the consumer in a single authorization, this exception would permit lenders to initiate a payment transfer only on a one-time basis immediately upon receipt of the consumer’s authorization, while leaving the overall prohibition in place. This limited approach was designed to facilitate the collection of payments that would be proffered by the consumer for immediate processing, without requiring compliance with the multi-stage 1077 process in proposed § 1041.14(c), and to ensure that consumers would have the option to continue making payments—one payment at a time—after the prohibition in proposed § 1041.14(b) had been triggered, without having to provide lenders with broader ongoing access to their accounts. In particular, subject to certain timing requirements, proposed § 1041.14(d) would permit lenders to initiate a payment transfer from a consumer’s account after the prohibition had been triggered, without obtaining the consumer’s authorization for additional payment transfers in accordance with proposed § 1041.14(c), if the consumer had authorized a one-time EFT or proffered a signature check for immediate processing. Under proposed § 1041.14(d)(1), a payment transfer initiated by either of these two payment methods would be required to meet the definition of a “single immediate payment transfer at the consumer’s request” in proposed § 1041.14(a)(2). Thus, for the exception to apply, the lender must initiate the EFT or deposit the check within one business day after receipt. Proposed § 1041.14(d)(2) provided that, for the exception to apply, the consumer had to authorize the underlying one-time EFT or provide the underlying signature check to the lender, as applicable, no earlier than the date on which the lender had provided to the consumer the consumer rights notice required by proposed § 1041.15(d) or on the date that the consumer affirmatively had contacted the lender to discuss repayment options, whichever date was earlier. The Bureau believed that many consumers who would elect to authorize only a single transfer under this exception would do so in part because they had already received the notice, had been informed of their rights, and had chosen to explore their options with the lender. The Bureau also believed that in some cases, consumers might contact the lender after discovering that the 1078 lender had made two failed payment attempts (such as by reviewing their online bank statements) before the lender had provided the notice. Moreover, by definition, this exception would not require the consumer to decide whether to provide the lender an authorization to resume initiating payment transfer from their account on an ongoing basis. Accordingly, the Bureau believed it was unnecessary to propose requirements similar to those proposed for the broader exception in proposed § 1041.14(c) to ensure that consumers had received the notice informing them of their rights at the time of authorization. Proposed comment 14(d)-1 cross-referenced proposed § 1041.14(b)(a)(2) and accompanying commentary for guidance on payment transfers that would meet the definition of a single immediate payment transfer at the consumer’s request. Proposed comment 14(d)-2 clarified how the prohibition on further payment transfers in proposed § 1041.14(b) continued to apply when a lender initiates a payment transfer pursuant to the exception in proposed § 1041.14(d). Specifically, the proposed comment clarified that a lender would be permitted under the exception to initiate the single payment transfer requested by the consumer only once, and thus would be prohibited under proposed § 1041.14(b) from re-initiating the payment transfer if it failed, unless the lender subsequently obtained the consumer’s authorization to reinitiate the payment transfer under proposed § 1041.14(c) or (d). The proposed comment further clarified that a lender would be permitted to initiate any number of payment transfers from a consumer’s account pursuant to the exception in proposed § 1041.14(d), provided that the requirements and conditions were satisfied for each such transfer. Accordingly, the exception would be available as a payment option on a continuing basis after the prohibition in proposed § 1041.14(b) had been triggered, as long as each payment transfer was authorized and initiated in 1079 accordance with the proposed exception’s timing and other requirements. In addition, the proposed comment cross-referenced proposed comment 14(b)(2)(ii)-3 for further guidance on how the prohibition in proposed § 1041.14(b) would apply to the exception in proposed § 1041.14(d). Proposed comment 14(d)-3 explained, by providing an example, that a consumer affirmatively had contacted the lender when the consumer called the lender after noticing on their bank statement that the lender’s last two payment withdrawal attempts had been returned for nonsufficient funds. The Bureau believed that the requirements and conditions in proposed § 1041.14(d) would prevent the harms that otherwise would occur if the lender—absent obtaining the consumer’s authorization for additional payment transfers under proposed § 1041.14(c)—were to initiate further transfers after two consecutive failed attempts. The Bureau believed that consumers who would authorize such transfers would do so based on their first-hand knowledge of their account balance at the time that the transfer, by definition, must be initiated. As a result of these two factors, the Bureau believed there was a significantly reduced risk that the transfer would fail. Comments Received Commenters argued that the proposed provisions in § 1041.14(d) that would not allow lenders to initiate single immediate payment transfers at the consumer’s request unless the borrower had received the consumer rights notice or the borrower affirmatively contacted the lender were detrimental to consumers. For borrowers who did not consent to electronic communications, commenters argued that it would take days to mail the notices, meaning 1080 borrowers might remain in delinquency for longer than they otherwise would if a collector could simply call and ask for a single immediate payment transfer. Commenters also argued that the proposed rule would result in situations where a collector would call the consumer, ask if they wanted to reauthorize payments, and then ask the consumer to call back to “affirmatively contact the lender,” which the Bureau agrees would be an unfortunate unintended consequence. One commenter argued that paragraph (d) would deter companies from reaching out to the consumer after a payment was rejected the first time to ask whether the consumer wanted to cover a required payment with a single immediate payment. It provided an example of a consumer authorizing a recurring ACH. If that recurring ACH was rejected, the commenter’s current practice was to call the borrower to ask if they wanted to cover the payment over the phone using a different method (under an independent authorization). The commenter stated that if the consumer authorized a different payment that was then also rejected, then the notice-andconsent requirements would be triggered. This commenter argued that as it would be hard to track payments across all non-cash methods, the proposed rule might deter companies from reaching out to the consumer after the first ACH was rejected. Final Rule The Bureau is finalizing paragraph (d) as proposed, with only technical edits to reflect the renumbering of this section to § 1041.8. The Bureau has decided not to eliminate the requirement that single immediate payment transfers only be processed after the consumer rights notice required under § 1041.9(c) is provided unless a borrower affirmatively reaches out to the lender to initiate the payment transfer. Commenters correctly noted that when combining the requirements in paragraphs (a), 1081 (b), (c), and (d), a lender will not be able to initiate any payment transfers after two failed payment transfers until after it they provide the notice under § 1041.9(c), unless the borrower affirmatively contact it to reauthorize. This means that for borrowers who do not accept electronic communications, there may be a period of several days before the notice under § 1041.9(c) is received, during which lenders cannot process payments unless the borrower affirmatively reaches out to the lender. Loans may continue to be delinquent during that period. And because lenders will be unable to process payments during this period on an outgoing collection call, they may be deterred from making collections calls during this brief window. For a number of reasons, the Bureau believes that this scenario does not present significant concerns. First, the Bureau’s study observed that only about 20 percent of third representments succeed without an overdraft fee, suggesting that a minority of borrowers will wish to re-initiate payments so quickly after the second failed payment attempt. Second, while the time necessary to process a mail notice, and delivery times, may add a few days of delinquency, often a few days of delinquency will not be likely to cause a significant amount of harm if the borrower is able to cure the delinquency soon after the notice is received, and a collection call can be made. Third, borrowers retain the option to affirmatively initiate payments through the lender, or avail themselves of a variety of payment options involving “pushes” from their account-holding institution, meaning that borrowers can still initiate payments, just not after being reminded to do so over an outgoing collection call. The Bureau does not believe the small fraction of consumers who may be harmed by this confluence of events is significant enough to outweigh the reasons for the restriction. Consumers would fall into this category only if they: (1) have experienced a second payment attempt failure; (2) nonetheless immediately have funds 1082 available for a third payment; (3) are unaware that the second payment did not go through (and thus do not have the information necessary to choose whether to make a payment through an affirmative contact); (4) have not consented to electronic notifications; and (5) are in the rare circumstances in which a few additional days of delinquency would have a negative impact. In this situation, these consumers will benefit from knowing their rights and understanding what occurred with the prior failed payment attempts before reinitiating payments. The Bureau similarly is not concerned about payments made at the borrower’s own affirmative initiation because, as stated in the proposal, such payments are more likely to be successful when the borrower knows what funds are available to process the payments. As for suggestions that the rule will result in lenders calling consumers and telling them to return the call in order to initiate a single immediate payment transfer after an affirmative consumer contact, the Bureau believes that this scenario may violate the prohibition against evasion set forth in paragraph (e), depending on the underlying facts and circumstances. The Bureau notes that if a lender reaches out after the first attempt fails in order to process a second attempt using a different payment method, then that second attempt would not be governed by paragraph (d) because it does not follow a second consecutive failed payment transfer. Instead, it simply would be an attempt to procure a payment after a first failed payment transfer. In other words, regardless of whether a lender reaches out to the borrower to arrange a new payment method after the first failed payment transfer, or simply re-presents under the original authorization, the cap and applicable notices would only trigger after the second failure. The Bureau expects that this may actually encourage lenders to reach out after the first failed payment transfer because a lender may be able to avoid the consequences of a second 1083 consecutive failed payment transfer by speaking with the consumer about the timing and amount of the transfer before initiating it. Finally, the Bureau concludes that after an initial investment, lenders should be able to track the number of failed payment attempts on a borrower level (and not a loan or payment method level) with relatively low burden. The Bureau thus is not persuaded that lenders will be reluctant to call consumers to procure payment after the first failed attempt because they are unaware of whether the cap has yet been initiated. 8(e) Prohibition against evasion. The Bureau is finalizing § 1041.8 with a new paragraph (e). Paragraph (e) states that a lender must not take any action with the intent of evading the requirements of this section (referring to § 1041.8). Proposed § 1041.14 did not include its own statement on evasion. Rather, the proposal included a general statement on evasion in proposed § 1041.19, which provided that a lender must not take any action with the intent of evading the requirements of part 1041. To clarify and reinforce this point, the Bureau is adding anti-evasion paragraphs to certain individual sections of the rule for ease of reference, and to allow it to provide specific examples relating to each section in the commentary. To that end, the Bureau is adding comment 8(e)-1 to clarify that the standard in § 1041.8(e) is same as that in § 1041.13. It also is finalizing an illustrative example in comment 8(e)-2, which formerly was an example for proposed § 1041.19, to clarify that, depending on the facts and circumstances, lenders might violate the prohibition against evasion if they process very small payments with the intent of evading the prohibition against three consecutive failed payment attempts without obtaining a new consumer authorization. 1084 Some commenters noted that the better way to address this issue would be to prohibit the initiation of additional transfers after any failed attempt. The Bureau addresses the feedback regarding whether the Bureau should impose a one re-presentment cap above. More general comments on the Bureau’s evasion authority also are found in the section-by-section analysis of § 1041.13. Section 1041.9 Disclosure of Payment Transfer Attempts Overview of the Proposed Rule As discussed in the proposal, consumers who use online payday and payday installment loans tend to be in economically precarious positions. They have low to moderate incomes, live paycheck to paycheck, and generally have no savings to fall back on. They are particularly susceptible to having cash shortfalls when payments are due and can ill afford additional fees on top of the high cost of these loans. At the same time, as discussed above in Market Concerns— Payments, many lenders in these markets may often obtain multiple authorizations to withdraw account funds through different channels, exercise those authorizations in ways that consumers do not expect, and repeatedly re-present returned payments in ways that can substantially increase costs to consumers and endanger their accounts. In addition to proposing in § 1041.14 (now § 1041.8 of the final rule) to prohibit lenders from attempting to withdraw payment from a consumer’s account after two consecutive payment attempts have failed, unless the lender obtains the consumer’s new and specific authorization to make further withdrawals, the Bureau proposed in § 1041.15 (which is now being finalized as § 1041.9) to use its authority under section 1032(a) of the Dodd-Frank Act to require two new disclosures to help consumers better understand and mitigate the costs and risks relating to 1085 payment attempt practices in connection with covered loans. While the interventions in proposed § 1041.14 were designed to protect consumers already experiencing severe financial distress in connection with their loans and depository accounts, the primary intervention in proposed § 1041.15 was designed to give all borrowers of covered loans who grant authorizations for payment withdrawals the information they need to prepare for upcoming payments and to take proactive steps to manage any errors or disputes before funds are deducted from their accounts. Specifically, proposed § 1041.15(b) would have required lenders to provide consumers with a payment notice before initiating each payment transfer on a covered loan. This notice was designed to alert consumers to the timing, amount, and channel of the forthcoming payment transfer and to provide consumers with certain other basic information about the payment transfer. The notice would specifically alert the consumer if the payment transfer would be for a different amount, at a different time, through a different payment channel than the consumer might have expected based upon past practice, or for the purpose of re-initiating a returned transfer. Where a lender had obtained consumer consent to deliver the payment notice through electronic means, proposed § 1041.15(c) would provide content requirements for an electronic short notice, which would be a truncated version of the payment notice formatted for electronic delivery through email, text message, or mobile application with a requirement to include in the short notice a hyperlink that would enable the consumer to access an electronic version of the full notice. In addition, proposed § 1041.15(d) would complement the intervention in proposed § 1041.14 by requiring lenders to provide a consumer rights notice after a lender triggered the 1086 limitations in that section. This consumer rights notice would inform consumers that a lender has triggered the provisions in proposed § 1041.14 and is no longer permitted to initiate payment from the consumer’s account unless the consumer chooses to provide a new authorization. The Bureau believed informing consumers of the past failed payments and the lender’s inability to initiate further withdrawals would help prevent consumer confusion or misinformation, and help consumers make an informed decision going forward on whether and how to grant a new authorization to permit further withdrawal attempts. For lenders to deliver the consumer rights notice required under proposed § 1041.15(d) through an electronic delivery method, proposed § 1041.15(e) would require the lenders to provide an electronic short notice that contains a link to the full consumer rights notice. Under the proposal, lenders would be able to provide these notices by mail, in person or, with consumer consent, through electronic delivery methods such as e-mail, text message, or mobile application. The Bureau sought to facilitate electronic delivery of the notices wherever practicable because it believed that such methods would make the disclosures timelier, more effective, and less expensive for all parties. Given that electronic delivery may be the most timely and convenient method of delivery for many consumers, the Bureau determined that facilitating electronic delivery was consistent with its authority under section 1032(a) of the Dodd-Frank Act to ensure that the features of any consumer financial product are “fully, accurately, and effectively disclosed” to consumers.1045 The Bureau proposed model clauses and forms in proposed § 1041.15(a)(7), which could be used at the option of covered persons for the provision of the notices that would be required 1045 12 U.S.C. 5532(a). 1087 under proposed § 1041.15. The proposed model clauses and forms were located in appendix A. Other than removing a line of APR information in one of the forms, the Bureau is finalizing them as proposed. These proposed model clauses and forms were validated through two rounds of consumer testing in the fall of 2015. The consumer testing results are provided in the FMG Report.1046 Legal Authority The payment notice, consumer rights notice, and short electronic notices in § 1041.9 of the final rule were proposed and are finalized under section 1032(a) of the Dodd-Frank Act, which authorizes the Bureau to prescribe rules to ensure that the features of consumer financial products and services “both initially and over the term of the product or service,” are disclosed “fully, accurately, and effectively” in a way that “permits consumers to understand the costs, benefits, and risks associated with the product or service, in light of the facts and circumstances.”1047 The authority granted to the Bureau in section 1032(a) is broad, and empowers the Bureau to prescribe rules regarding the disclosure of the “features” of consumer financial products and services generally. Accordingly, the Bureau may prescribe rules containing disclosure requirements even if other Federal consumer financial laws do not specifically require disclosure of such features. Dodd-Frank Act section 1032(c) provides that, in prescribing rules pursuant to section 1032, the Bureau “shall consider available evidence about consumer awareness, understanding 1046 FMG Report, “Qualitative Testing of Small Dollar Loan Disclosures, Prepared for the Consumer Financial Protection Bureau,” (Apr. 2016) available at http://files.consumerfinance.gov/f/documents/Disclosure_Testing_Report.pdf. 1047 12 U.S.C. 5532(a). 1088 of, and responses to disclosures or communications about the risks, costs, and benefits of consumer financial products or services.”1048 Accordingly, in developing the rule under DoddFrank Act section 1032(a), the Bureau considered consumer complaints, industry disclosure practices, and other evidence about consumer awareness, understanding of, and responses to disclosures or communications about the risks, costs, and benefits of consumer financial products or services. This included the evidence developed through the Bureau’s own consumer testing as discussed in the proposal, as well as in Market Concerns—Payments and the FMG Report. Section 1032(b)(1) also provides that “any final rule prescribed by the Bureau under [section 1032] requiring disclosures may include a model form that may be used at the option of the covered person for provision of the required disclosures.” Any model form issued pursuant to this authority shall contain a clear and conspicuous disclosure that, at a minimum, uses plain language that is comprehensible to consumers; contains a clear format and design such as an easily readable type font; and succinctly explains the information that must be communicated to the consumer.1049 Section 1032(b)(2) provides that any model form that the Bureau issues pursuant to section 1032(b) shall be validated through consumer testing. The Bureau conducted two rounds of qualitative consumer testing in September and October of 2015. The testing results are provided in the FMG Report. Section 1032(d) provides that “any covered person that uses a model form included with a rule issued under this [section 1032] shall be deemed to be in compliance with the disclosure requirements of this section with respect to such model form.” 1048 1049 12 U.S.C. 5532(c). Dodd-Frank Act section 1032(b)(2); 12 U.S.C. 5532(b)(2). 1089 The Bureau received a number of comments arguing that there was no UDAAP basis for the notices in proposed § 1041.15, or that the remedy the Bureau proposed for the identified unfair and abusive practice in proposed § 1041.13 (finalized as § 1041.7) was overbroad by requiring disclosures in addition to a prohibition on the identified practice. These commenters are correct in asserting that the Bureau did not identify an unfair or abusive practice that would warrant the notice requirements in proposed § 1041.15, but only because it did not attempt to do so. Instead, as described here, the Bureau proposed the section on notice requirements pursuant to its disclosure authority under section 1032 of the Dodd-Frank Act. Thus, the remedy in final § 1041.8 that is needed in order to prevent the practice identified in final § 1041.7 is not overbroad based on the existence of final § 1041.9, because § 1041.9 is intended for separate and additional reasons and finalized under separate authority. 9(a) General form of disclosures. Proposed Rule Proposed § 1041.15(a), finalized as § 1041.9(a), set basic rules regarding the format and delivery for all notices required under proposed § 1041.15 and set requirements for a two-step process for the delivery of electronic disclosures as further required under proposed § 1041.15(c) and (e). The format requirements generally paralleled the format requirements for other disclosures related to certain covered short-term loans as provided in proposed § 1041.7 (now final § 1041.6), but would also permit certain electronic disclosures by text message or mobile application. As proposed, a two-step electronic delivery process would involve delivery of short-form disclosures to consumers by text message, mobile application, or e-mail that would 1090 contain a unique website address for the consumer to access the full notices proposed under § 1041.15 Because the disclosures in proposed § 1041.15 involved the initiation of one or more payment transfers in connection with existing loans, the Bureau believed that electronic disclosures generally would be more timely, more effective, and less expensive for consumers and lenders than paper notices, as discussed below. At the same time, it recognized that there were some technical and practical challenges with regard to electronic channels. The two-stage process was designed to balance such considerations, for instance by adapting the notices in light of format and length limitations on text message and by accommodating the preferences of consumers who are using mobile devices in the course of daily activities and would rather wait to access the full contents until a time and place of their choosing. Proposed 15(a)(1) Clear and conspicuous Proposed § 1041.15(a)(1) provided that the disclosures required by proposed § 1041.15 must be clear and conspicuous, and could use commonly accepted or readily understandable abbreviations. Proposed comment 15(a)(1)-1 clarified that disclosures would be clear and conspicuous if they were readily understandable, and their location and type size were readily noticeable to consumers. This clear and conspicuous standard was based on the standard used in other Federal consumer financial laws and their implementing regulations, including Regulation E subpart B § 1005.31(a)(1). The Bureau believed that requiring the disclosures to be provided in a clear and conspicuous manner would help consumers understand the information in the disclosure about the costs, benefits, and risks of the transfer, consistent with the Bureau’s authority under section 1032(a) of the Dodd-Frank Act. 1091 Proposed 15(a)(2) In writing or electronic delivery. Proposed § 1041.15(a)(2) required disclosures mandated by proposed § 1041.15 to be provided in writing or through electronic delivery. The disclosures could be provided through electronic delivery as long as the requirements of proposed paragraph 15(a)(4) were satisfied. The disclosures would have to be provided in a form that can be viewed on paper or a screen, as applicable. The requirement in proposed § 1041.15(a)(2) would not be satisfied orally or through a recorded message. Proposed comment 15(a)(2) explained that the disclosures that would be required by proposed § 1041.15 may be provided electronically as long as the requirements of proposed § 1041.15(a)(4) were satisfied, without regard to the E-Sign Act.1050 The Bureau proposed to allow electronic delivery because electronic communications are more convenient than paper communications for some lenders and consumers. Given that some requirements of the E-Sign Act might not be necessary in this context, but other features like a revocation regime might be useful given the ongoing nature of these disclosures, the Bureau proposed a tailored regime that it believed would encourage lenders and consumers to identify an appropriate method of electronic delivery where consumers have electronic access. The Bureau understood that some lenders already contact their borrowers through electronic means such as text message and email.1051 Lenders that currently provide electronic notices had informed the Bureau that they provide both email and text message as communication options to consumers. A major trade association for online lenders reported that 1050 15 U.S.C. 7001 et seq. During the SBREFA process, several SERs explained that they currently provide consumers with text message reminders of upcoming payments. Other public information indicates that lenders contact consumers through many of these methods. See, e.g., ENOVA Int’l, Inc., 2014 Annual Report (Form 10-K), at 9 (“Call center employees contact customers following the first missed payment and periodically thereafter. Our primary methods of contacting past due customers are through phone calls, letters and emails.”). 1051 1092 many of its members automatically enroll consumers in an email notification system as part of the origination process but allow consumers to opt-in to receive text message notifications of upcoming payments. One member of this association asserted that approximately 95 percent of consumers opt in to text message notifications, so email effectively functions as a back-up delivery method. Similarly, during the Bureau’s SBREFA process, a SER from an online-only lender reported that 80 percent of its customers opt in to text message notifications. According to a major payday, payday installment, and vehicle title lender that offers loans through storefronts and the Internet, 95 percent of its customers have access to the Internet and 70 percent have a home computer.1052 Lenders may prefer contacting consumers through these methods given that they are typically less costly than mailing a paper notice. Given the convenience and timeliness of electronic notices, the Bureau believed the disclosure information would provide the most utility to consumers when it is provided through electronic methods. The Bureau believed that providing consumers with disclosures that they can view and retain would allow them to more easily understand the information, detect errors, and determine whether the payment is consistent with their expectations. In light of the detailed nature of the information provided in the disclosures required by proposed § 1041.15, including payment amount, loan balance, failed payment amounts, consumer rights, and various dates, the Bureau also believed that oral disclosures would not provide consumers with a sufficient opportunity to understand and use the disclosure information. 1052 Community Choice Fin. Inc., 2014 Annual Report (Form 10–K), at 4 (Mar. 30, 2015). At the time of the filing, most (about half) of Community Choice’s revenue was from short-term loans. Id. at 6. Both short-term loans and long-term installment loans were being offered online. Id. at 6–7. 1093 Proposed 15(a)(3) Retainable. Proposed § 1041.15(a)(3) would require disclosures mandated by proposed § 1041.15 to be provided in a retainable form, except for the electronic short notices delivered through mobile application or text message. Electronic short notices provided by email would still be subject to the retain-ability requirement. Proposed comment 15(a)(3) explained that electronic notices would be considered retainable if they were in a format that is capable of being printed, saved, or emailed by the consumer. The Bureau believed that having the disclosures in a retainable format would enable consumers to refer to the disclosure at a later point in time, such as after a payment has posted to their account or if they contact the lender with a question, allowing the disclosures to more effectively disclose the features of the product to consumers. The Bureau did not propose to require that text messages and messages within mobile applications be permanently retainable because of concerns that technical limitations beyond the lender’s control might make retention difficult. However, the Bureau anticipated that such messages would often be kept on a consumer’s device for a considerable period of time and could therefore be accessed again. In addition, proposed § 1041.15 would require that such messages contain a link to a website containing a full notice that would be subject to the general rule under proposed § 1041.15(a)(3) regarding retain-ability. A lender would also be required to maintain policies, procedures, and records to ensure compliance with the notice requirement under proposed § 1041.18 (now final § 1041.12). Proposed 15(a)(4) Electronic delivery. Proposed § 1041.15(a)(4) laid out various requirements designed to facilitate delivery of the notices required under proposed § 1041.15 through electronic channels. The proposal would 1094 allow disclosures to be provided through electronic delivery if the consumer affirmatively consents in writing or electronically to the particular electronic delivery method. Lenders would be able to obtain this consent in writing or electronically. The proposed rule would require that lenders provide e-mail as an electronic delivery option if they also offered options to deliver notices through text message or mobile application. Proposed § 1041.15(a)(4) would also set forth rules to govern situations where the consumer revokes consent for delivery through a particular electronic channel or is otherwise unable to receive notices through that channel. The consumer consent requirements for provision of the disclosures through electronic delivery were specified in the proposal. Proposed § 1041.15(a)(4)(i)(A) would require lenders to obtain a consumer’s affirmative consent to receive the disclosures through a particular method of electronic delivery. These methods might include e-mail, text message, or mobile application. The Bureau believed it was important for consumers to be able to choose a method of delivery to which they had access and that would best facilitate their use of the disclosures, and that viewable documentation would facilitate both informed consumer choice and supervision of lender compliance. The Bureau was concerned that consumers could receive disclosures through a method that they would not prefer or that would not be useful to them if they were automatically defaulted into an electronic delivery method. Similarly, the Bureau was concerned that a consumer might receive disclosures through a method that they would not expect if they had been provided with a broad electronic delivery option rather than an option specifying the method of electronic delivery. Proposed § 1041.15(4)(i)(B) stated that when obtaining consumer consent to electronic delivery, a lender had to provide the consumer with the option to select email as the method of 1095 electronic delivery, separate and apart from any other electronic delivery methods such as mobile application or text message. Proposed comment 15(a)(4)(i)(B) explained that the lender could choose to offer email as the only method of electronic delivery. The Bureau believed that such an approach would facilitate consumers’ choice of the electronic delivery channel that would be most beneficial to them, in light of differences in access, use, and cost structures between channels. For many consumers, delivery via text message or mobile application might be the most convenient and timely option. However, there would be some potential tradeoffs. For example, consumers might incur costs when receiving text messages and could have privacy concerns about finance-related text messages appearing on their mobile phones. During consumer testing, some of the participants had a negative reaction to receiving notices by text message, including privacy concerns about someone being able to see that they were receiving a notice related to a financial matter. The Bureau believed that mobile application messages might create similar privacy concerns, as such messages may generate alerts or banners on a consumer’s mobile device. Nonetheless, the Bureau believed that receiving notices by text message might be useful to some consumers. In general, most consumers have access to a mobile phone. According to a recent Federal Reserve study on mobile banking and financial services, approximately 90 percent of “underbanked” consumers—consumers who have bank accounts but use non-bank products like payday loans—have access to a mobile phone.1053 Fewer underbanked consumer have a 1053 Bd. of Governors of the Federal Reserve System, “Consumers and Mobile Financial Services 2015,” at 5 (Mar. 2015), available at http://www.federalreserve.gov/econresdata/consumers-and-mobile-financial-services-report201503.pdf. 1096 phone with Internet access, although the coverage is still significant at 73 percent. A few participants in the Bureau’s consumer testing indicated a preference for receiving notices by text message. The Bureau believed that text message delivery should be allowed as long as consumers had the option to choose email delivery, which for some consumers might be a strongly preferred method of disclosure delivery. The Bureau also maintained that requiring an email option might help ensure that the disclosure information is effectively disclosed to consumers, consistent with the Bureau’s authority under section 1032 of the Dodd-Frank Act. Proposed § 1041.15(a)(4)(ii) would have prohibited a lender from providing the notices through a particular electronic delivery method if there was a subsequent loss of consent as provided in proposed § 1041.15(a)(4)(ii), either because the consumer had revoked consent pursuant to proposed § 1041.15(a)(4)(ii)(A), or the lender had received notification that the consumer was unable to receive disclosures through a particular method, as described in proposed § 1041.15(a)(4)(ii)(B). Proposed comment 15(a)(4)(ii)(B)-1 explained that the prohibition applied to each particular electronic delivery method. It further provided that a lender that had lost a consumer’s consent to receive disclosures via text message but, for example, not the consent to receive disclosures via email, could continue to provide disclosures via email so long as all of the requirements in proposed § 1041.15(a)(4) were satisfied. Proposed comment 15(a)(4)(ii)(B)-2 clarified that the loss of consent would apply to all notices required under proposed § 1041.15. For example, if a consumer revoked consent in response to the electronic short notice text message delivered along with the payment notice under proposed § 1041.15(c), then that revocation also would apply to text message delivery of the electronic 1097 short notice that would be delivered with the consumer rights notice under proposed § 1041.15(e), or to delivery of the notice under proposed § 1041.15(d) if there were two consecutive failed withdrawal attempts that would trigger the protections of § 1041.14. Proposed § 1041.15(a)(4)(ii)(A) would prohibit a lender from providing the notices through a particular electronic delivery method if the consumer had revoked consent to receive electronic disclosures through that method. Proposed comment 15(a)(4)(ii)(A)-1 clarified that a consumer could revoke consent for any reason and by any reasonable means of communication. The comment provided that examples of a reasonable means of communication included calling the lender and revoking consent orally, mailing a revocation to an address provided by the lender on its consumer correspondence, sending an email response or clicking on a revocation link provided in an email from the lender, and responding to a text message sent by the lender. The Bureau was aware that burdensome revocation requirements could make it difficult for the consumer to revoke consent to receive electronic disclosures through a particular electronic delivery method. Accordingly, the Bureau believed it was appropriate to provide a simple revocation regime and require that lenders cannot provide the notices through a particular electronic delivery method if the consumer revokes consent through that method. Proposed § 1041.15(a)(4)(ii)(B) would prohibit a lender from providing the notices through a particular electronic delivery method if the lender had received notice that the consumer was unable to receive disclosures through that method. Such notice would be treated in the same manner as if the consumer had affirmatively notified the lender that the consumer was revoking authorization to provide notices through that means of delivery. Proposed comment 15(a)(4)(ii)(B)-1 provided 1098 examples of notice, including a returned email, returned text message, and statement from the consumer. The Bureau believed this was an important safeguard to ensure that consumers have ongoing access to the notices required under proposed § 1041.15. It also believed this requirement to change delivery methods after consent has been lost would ensure that the disclosure information had been fully and effectively disclosed to consumers, consistent with the Bureau’s authority under section 1032. Proposed 15(a)(5) Segregation requirements for notices. All required notices under proposed § 1041.15 would have to be segregated from all other written materials and contain only the information required by the proposed rule, other than information necessary for product identification, branding, and navigation. Under the proposal, segregated additional content that was required by proposed § 1041.15 could not be displayed above, below, or around the required content. Proposed comment 15(a)(5)-1 clarified that additional, non-required content could be delivered through a separate form, such as a separate piece of paper or web page. To increase the likelihood that consumers would notice and read the written and electronic disclosures required by proposed § 1041.15, the proposed notices had to be provided in a stand-alone format that is segregated from other lender communications. This requirement was intended to ensure that the disclosure contents would be effectively disclosed to consumers, consistent with the Bureau’s authority under section 1032 of the Dodd-Frank Act. Lenders would not be allowed to add additional substantive content to the disclosure. 1099 Proposed 15(a)(6) Machine readable text in notices provided through electronic delivery. Under the proposal, a payment notice and consumer rights notice provided through electronic delivery also had to use machine readable text that is accessible via both Web browsers and screen readers. As the Bureau stated in the proposal, graphical representations of textual content cannot be accessed by assistive technology used by the blind and visually impaired. Providing the electronically-delivered disclosures with machine readable text rather than as a graphic image file, thus would allow consumers with a variety of electronic devices and consumers that utilize screen readers, such as consumers with disabilities, to access the disclosure information. Proposed 15(a)(7) Model Forms. Proposed § 1041.15(a)(7) required all notices in proposed § 1041.15 to be substantially similar to the model forms and clauses proposed by the Bureau. Specifically, proposed § 1041.15(a)(7)(i) required the content, order, and format of the payment notice to be substantially similar to the Models Forms A-3 through A-5 in appendix A. Proposed § 1041.15(a)(7)(ii) required the consumer rights notice to be substantially similar to Model Form A-5 in appendix A. And similarly, proposed § 1041.15(a)(7)(iii) mandated the electronic short notices required under proposed § 1041.15(c) and (e) to be substantially similar to the Model Clauses A-6 through A-8 provided in appendix A. To explain the safe harbor provided by these model forms, proposed comment 15(a)(7)-1 provided that although the use of the actual model forms and clauses was not required, lenders using such model forms would be deemed to be in compliance with the disclosure requirement. 1100 As stated in the proposal, the model forms developed through consumer testing might make the notice information comprehensible to consumers while minimizing the burden on lenders who otherwise would need to develop their own disclosures. Consistent with the Bureau’s authority under section 1032(b)(1), the Bureau believed that its proposed model forms used plain language comprehensible to consumers, contained a clear format and design, such as an easily readable type font, and succinctly explained the information that must be communicated to the consumer. As described in the FMG Report and as discussed above, it further considered evidence developed through its testing of model forms pursuant to section 1032(b)(3). It also believed that providing these model forms would help ensure that the disclosures were effectively provided to consumers, while also allowing lenders to adapt the disclosures to their loan products and preferences. Proposed 15(a)(8) Foreign language disclosures. The proposal also would allow lenders to provide the required disclosures in a language other than English, provided that the disclosures were made available in English upon the consumer’s request. Comments Received Some industry commenters, many consumer groups, and many State Attorneys General supported the notice intervention. Several commenters raised concerns that consumers should have notice of upcoming transfers in order to minimize unexpected bank fees. A number of lenders stated that they already provide upcoming payment notices to their customers. One explained that it does not anticipate much additional compliance burden from the notices because 1101 it already provides payment reminders and does not use the payment practices described in the proposal, like re-presentments. However, many industry commenters raised concerns about the burden of the intervention. One supported the intervention overall but raised burden concerns about the frequency and delivery of the notice. Some disputed the need for the intervention, arguing that the proposed notices were too burdensome and complex, that consumers knew when an ACH will be pulled, that the practices the notices sought to prevent violated existing laws that needed to be enforced, and that it would be burdensome to create a payment notice for past due consumers because lender wanted to debit when funds come in. A number of stakeholders commented on the Bureau’s consumer testing process for the model forms. Some commenters believed that the Bureau’s 28 consumer sample size was too small, noting that the Bureau and other agencies had used larger sample sizes for the qualitative testing of other disclosures (such as the TILA-RESPA integrated disclosure),1054 and supplemented with quantitative testing. These commenters asked the Bureau to clarify that the notices do not need to conform to the model forms, such that lenders could conduct their own testing. Commenters claimed that the level of research rigor for the model disclosures was weak as compared to what would be considered a best practice in the industry. Another criticized both the sample size and the number of geographies represented, and recommended that the Bureau remove the model forms from the proposal. It also suggested that the Bureau’s use of just qualitative testing without quantitative testing meant that the findings might not be projectable to the broader population. However, others supported the Bureau’s use of a model form. 1054 See 78 FR 79730 (Dec. 31, 2013). 1102 Stakeholders also commented on the consent requirements around receiving notices electronically. Commenters argued that the consent scheme imposed by the E-Sign Act should suffice, and that the Bureau had not explained why the E-Sign Act requirements were not sufficient in this context. In particular, one commenter argued that the prohibition against providing electronic notices that would apply after the lender receives notification that the consumer is unable to receive notices through a given electronic medium would create uncertainty around when a consumer will be deemed to have “received notification.” It noted that this requirement was more onerous than the E-Sign Act, which allows the lender to give electronic disclosures to consumers who have affirmatively consented, and have not withdrawn such consent. Others similarly suggested that allowing borrowers to consent to electronic delivery over the phone, something E-Sign allows, would be beneficial. These commenters said the Bureau should instead follow the E-Sign Act’s requirements relating to consent. More generally, the Bureau heard from a variety of industry participants about the compliance burden of the notice requirements. Although each had somewhat different perspective on the compliance costs, many considered them to be too high and argued that they could lead to higher prices for loan products. One commenter argued that the proposed notice requirements would pose a significant cost when borrowers do not opt in to electronic notifications, because mailings would pose significant costs. It provided the example of a borrower who takes out a $1,000 loan payable over 12 months, in semi-monthly installments. It estimated that the payment notices would cost about $0.40 per notice at high scale, and $1 at low scale. In the commenter’s view, this meant that the notice requirement could cost more than two percent of the principal balance. In light of this significant cost, it asked that the Bureau allow 1103 borrowers to opt out of the notice requirement, or that it allow lenders to provide the notices through other methods, including pre-recorded phone calls. Other commenters asked the Bureau to similarly allow oral notices. Alternatively, a consumer group argued that lenders should be required to verify consent with a digital or print signature. Another industry participant argued that the allowance for electronic notifications would not alleviate the costs associated with mailed notices because the costs of tracking consent and withdrawals across channel are too complex operationally and technologically, and thus too costly. This commenter argued that the Bureau should abandon the notice requirements because the costs would result in higher pricing. Another entity commented that the proposal would impose high costs because a lender would have to invest in a system capable of recognizing that the consumer’s inability to receive notices through certain methods or at a certain address. Another commenter claimed that community banks would likely not attempt electronic notices, and thus would be left with the cost of providing paper notices. However, a different industry participant stated that electronic notices, for which consent is taken over the phone, are in their experience 80 times cheaper than mail notices. The Bureau received several comments about methods of consenting to electronic delivery of the notices. One commenter argued that email notifications should only be allowed if the consumer explicitly consented to such notices, and that print text via mobile phone should be prohibited. Some commenters urged the Bureau to allow consent to electronic delivery to be received orally over the phone. One lender stated that 90 percent of customers had consented to receive electronic disclosures via verbal consent that would be either captured by a retail agent or by a call center 1104 agent on a recorded line (they appeared to be obtaining the consent while also closing the loan over the phone). A number of commenters also addressed the foreign language disclosures in proposed § 1041.15(a)(8). Several argued that the final rule should not require foreign language notices (which it did not propose but did seek comment on) because this would impose substantial costs and could involve wide-ranging consequences that deserve thoughtful consideration in a separate rulemaking. Other commenters argued that lenders should offer the model form in the language they use to communicate with consumers, in the language of the consumer’s preference, or in the language that the lender uses to negotiate the transaction. One industry commenter suggested that the Bureau convene a Federal interagency and industry working group and address foreign language disclosures in a separate proceeding. Final Rule The Bureau is finalizing proposed § 1041.15(a) with no substantive changes except to renumber it as § 1041.9(a). It also made cosmetic or technical changes to § 1041.9(a)(2) and the commentary pertinent to § 1041.9(a) including, primarily, changes to section numbers in light of the reorganization of the rest of the regulatory text. Based on its considerable experience with consumer testing, the Bureau has made the judgment that the qualitative user testing process for the model forms and notices is sufficient for purposes of this rule, especially because unlike the TILA-RESPA model disclosures, the model forms for this rule are relatively short and uncomplicated. Lenders remain free to conduct their own user-testing, including quantitative testing, and to improve upon the Bureau’s model forms if their user-testing suggests further improvements are possible (and encourages lenders to share the results of that testing, and any specific improvements to the forms, which the Bureau may 1105 incorporate into the forms at a future date). The Bureau contracted with Fors March Group (FMG) to conduct qualitative user testing of the forms. While the sample size was indeed small—28 test subjects—each subject was given a one-on-one interview with an FMG staff member for about an hour. The interviews were conducted in two geographical locations, New Orleans and Kansas City. In addition, CFPB staff used the feedback after the round of testing in New Orleans to improve the model forms before the second round of testing in Kansas City. The Bureau did not conduct quantitative testing, though the Bureau agrees that quantitative testing could be advantageous. Regardless, it believes the testing it did suffices to show that the disclosures use plain language that is comprehensible to consumers, contains a clear format and design, and succinctly explains the information that must be communicated to the consumer. There are a few differences between the regime for obtaining consent set forth in the proposal, and now the final rule, in comparison to the regime set forth in the E-Sign Act. That statute does not set forth the only electronic disclosure and consent requirements that an agency can prescribe, but rather presents general rules of the road where requirements are not otherwise specifically prescribed. It was not designed for this specific disclosure requirement, but rather, set forth default rules where others are not enacted specifically. Under the E-Sign Act, companies can only obtain consent after providing certain disclosures set forth in 15 U.S.C. 7001(c)(1)(B) and (c)(1)(C)(i). This rule does not require those disclosures—which would add marginal burden to the regime in this final rule—though companies may provide them if they wish. These disclosures require consumers to confirm through the particular electronic method that they can receive notices through that particular electronic method. Given the steps and potential delay that this requirement could impose on the origination process, the Bureau 1106 believes that the consumer consent regime being finalized will make it easier for consumers to provide (and lenders to obtain) consent to electronic delivery at origination. The E-Sign Act also requires certain actions when a company changes hardware or software requirements, which are not found in the rule (companies may provide these as well).1055 The rule requires that the lender, when obtaining consent, must offer consumers the option to consent to the specific electronic method used (and not just general consent to electronic disclosures), and specifically requires that one method be provided—email. As the Bureau stated in the proposal, and now finds, consumers will benefit from being able to consent to specified electronic delivery methods—for example, a borrower may wish to consent to email but not mobile text messages (largely unavailable when the E-Sign Act was enacted).1056 In certain circumstances, consent can also be provided by phone under E-Sign, which this rule would not allow. As stated in the proposal, the Bureau continues to believe that consumers would benefit from being able to see the specific delivery location—for example, the email address or phone number for text messaging. Of course, none of this means the lender must provide electronic notices; it is just an option. The rule requires that lenders cease using an electronic method when a lender receives notification that the consumer is unable to receive disclosures through that method. Here, the Bureau contemplated a rejected email, text message, or other electronic communication, like an automated notification that a disclosure email or text was undeliverable. It does not agree with the commenters that this provision adds any particular level of uncertainty—when a lender 1055 15 U.S.C. 7001(c)(1)(D). The Bureau notes that lenders communicating by electronic means may be subject to additional requirements under the Telecommunications Consumer Protection Act (47 U.S.C. 227) or other authorities. 1056 1107 receives any notice that the delivery method is no longer available, the lender cannot continue using that method. To the extent it is more burdensome than the E-Sign Act, it is for good reason—the Bureau does not wish to permit a lender to continue sending disclosures to an inactive email account or phone number, especially with regard to the unusual withdrawal notice where the disclosure is intended to warn consumers about an impending event. The Bureau is not adding an option to allow oral consent to electronic delivery. It maintains that it would be helpful for consumers to see, and be able to retain, the type of delivery they are consenting to and which email address or phone number they are providing for this purpose. This requirement seems workable given lender practices. In the storefront, lenders could incorporate consent to electronic delivery into its in-person processes, and could have the consumer consent on paper or a computer screen. Online lenders could adjust their application process to have consumers consent to electronic delivery as part of the application process, even if they close the loan over the phone. They could even show the consent form electronically during application process or email it separately. The bulk of the comments the Bureau received on § 1041.9(a) and (b) pertained to the burdens associated with the notice requirements. The Bureau has made changes to § 1041.9(b) that will substantially reduce the total aggregate burden of the disclosures, most notably that the notices no longer have to be sent before every payment attempt. Under the final rule, a payment notice must be sent before the first payment withdrawal (and can be provided during the origination process) and thereafter, notices only will have to be sent when there is an unusual withdrawal (defined as a payment that varies from a regular payment or minimum payment in the case of open-end credit, occurs on a date other than the regularly scheduled payment date, is 1108 processed through a different payment channel from the previous channel used, or is a representment) or the payment attempt cap is met. Thus, taking the commenter’s example of the borrower with a $1,000 loan payable over 12 months in semi-monthly installments, instead of providing 24 notices, the lender would only have to provide one (assuming there were no unusual payments, and the borrower never hit the payment attempt cap). Using the commenter’s estimates, instead of costing more than two percent of the principal balance, it would cost 0.05 to 0.10 percent of principal. The lender would also able to provide that first and only payment notice during origination, thereby saving on postage as well. Given the changes discussed above, lenders may be able to avoid the need to send such paper notices at all if they avoid unusual withdrawals and hitting the cap, which should generally be rare events. To the extent the costs of tracking consent to receive electronic notifications or to detect whether electronic communications are being rejected is too burdensome, lenders can always provide paper notices. But in the Bureau’s experience, the technology to track borrower consent and detect rejected communications is readily available on the market today, and could be developed for this specific market, such that even small to mid-sized lenders would be able to procure that functionality from a vendor. The Bureau concludes that providing notices through a pre-recorded call or a robo-call, or orally over the phone or in person, would not suffice to meet the purposes of the rule. The Bureau has determined that it is important for the notices to be retainable, such that a borrower can refer back to it at a later time—for example, to check that the right amount was debited. This is especially important now that lenders will not be providing notices before every payment 1109 withdrawal. Also, the burden of providing the notices is lower now that they are not required before every payment and, after origination, should only be necessary in rare circumstances. The Bureau does not agree with consumer group commenters suggesting that it should not allow print text via mobile phones. In light of the constantly updating technology of the modern world—where some consumers may move frequently and may be more reliably communicated with through their phones—the Bureau believes this rule should allow communications to be made through the common communications means of the day. This means that for now, the Bureau will allow disclosures through mobile application or text message (provided that there is a link or PDF to the full disclosure); and that disclosures may be transmittable through other electronic means as they become available. As proposed, the Bureau is not requiring foreign language disclosures, and is instead finalizing the rule as proposed, which merely allows foreign language disclosures. Some of the Bureau’s rules, like 12 CFR 1005.31(g), require disclosures in foreign languages in certain circumstances. The Bureau continues to believe that disclosures in languages other than English are a positive development in all markets for consumer financial products or services, where the customer base has become increasingly more diverse. It is not, however, prepared to make foreign language disclosures mandatory at this time with respect to these forms, largely because it recognizes that the current final rule will require lenders to engage in a significant amount of implementation work in order to begin complying with the rule, including the work to design and implement disclosures in English. In finalizing this rule, the Bureau is attempting to minimize compliance burden to the extent possible while maintaining the core protections of the rule. Although it has decided to allow but not mandate foreign language notices at this time, it may consider supplemental 1110 rulemakings or model forms in the future, when industry has fewer regulatory adjustments to manage and has developed more experience with the English-language forms. 9(b) Payment notice. Proposed Rule Proposed § 1041.15(b) required lenders to provide to consumers a payment notice before initiating a payment transfer from a consumer’s account with respect to a covered loan. The Bureau notes here that under the final rule, this requirement has been scaled back to be required only in more limited payment transfer circumstances. As defined in proposed § 1041.14(a), a payment transfer would be any transfer of funds from a consumer’s account that was initiated by a lender for the purpose of collecting any amount due or purported to be due in connection with a covered loan. The proposed notice contained timing requirements that would vary depending on the method of delivery, along with additional required information if the payment transfer was unusual in that it involved changes in amount, timing, or payment channel from what the consumer would otherwise be expecting. As discussed in the proposal and above in Market Concerns—Payments, when a lender initiates a payment transfer for which the consumer’s account lacks sufficient funds, the consumer can suffer a number of adverse consequences. The consumer’s bank will likely charge an overdraft or NSF fee. If the payment is returned, the lender may also charge a returned-item fee and/or a late fee. These fees can materially increase the overall amount that the consumer is required to pay. Moreover, the incidence of returneditem fees and other payments of these kinds appear to increase the likelihood that the consumer’s account will be closed. 1111 The Bureau believed that the payment notice could help consumers mitigate these various harms by providing a timely reminder that a payment transfer will occur, the amount and expected allocation of the payment as between principal and other costs, and other information that consumers may need to follow up with lenders or their depository institutions if they anticipate a problem with the upcoming withdrawal or in covering the payment transfer. The Bureau believed that the notice could have value as a general financial management tool, but would be particularly valuable to consumers in situations in which lenders intend to initiate a withdrawal in a way that deviates from the loan agreement or prior course of conduct between the parties. As detailed above, the Bureau was aware that some lenders making covered loans sometimes initiate payments in an unpredictable manner, which may increase the likelihood that consumers will experience adverse consequences. Consumers have limited ability to control when or how lenders will initiate payment. Although paper checks specify a date and amount for payment, UCC § 4-401(c) allows merchants to present checks for payment on a date earlier than the date on the check. Lenders sometimes attempt to collect payment on a different day from the one stated on a payment schedule. The Bureau had received complaints from consumers who had incurred bank account fees after online payday and payday installment lenders attempted to collect payment on a different date from what was scheduled. It was also aware that lenders sometimes split payments into multiple pieces, make multiple attempts to collect in one day, add fees and charges to the payment amount, and change the payment method used to collect. The Bureau was aware that these notices would impose some cost on lenders, particularly the payment notice under proposed § 1041.15(c), which would be sent before each payment 1112 transfer. It considered requiring the payment notice only when a payment transfer qualified as unusual, such as when there is a change in the amount, date, or payment channel. However, at the time of the proposal the Bureau believed that once lenders had built the infrastructure to send the unusual payment notices, the marginal costs of sending notices for all upcoming payments would likely to be relatively minimal. The Bureau noted that a number of lenders already had a similar infrastructure for sending payment reminders (e.g., monthly bills). Indeed, a trade association representing online payday and payday installment lenders had expressed support for upcoming payment reminders.1057 These lenders currently may choose to send out payment reminders before all payments initiated from a consumer’s account. Others may be sending out notices for preauthorized EFTs that vary in amount in accordance with Regulation E § 1005.10(d), which requires payees to send a notice of date and amount ten days before a transfer that varies in amount from the previous transfer under the same authorization or from the preauthorized amount. The Bureau describes each subsection of proposed § 1041.15(b) and (c) below, discusses the comments received on § 1041.15(b) and (c) together thereafter, and discusses the changes made to final § 1041.9(b). 1057 “Bank account overdrafts are a lose-lose for online lenders and their customers. It is in the customers best interests as well as the lenders best interest for customers to not incur overdrafts. This is why we support payment reminders so that customers do not overdraft their accounts.” Lisa McGreevy, “OLA Releases Statement in Response to CFPB Online Loan Payment Study,” Online Lenders Alliance (Apr. 20, 2016), available at http://onlinelendersalliance.org/ola-releases-statement-in-response-to-cfpb-online-loan-payment-study/. 1113 Proposed 15(b)(1) General. The proposal would have specifically required lenders to send a payment notice to a consumer prior to initiating a payment transfer from the consumer’s account, subject to limited exceptions as specifically listed in proposed § 1041.15(b)(2) and the comments thereto. Proposed 15(b)(2) Exceptions. Proposed § 1041.15(b)(2)(i) would except covered loans made pursuant to proposed § 1041.11 or proposed § 1041.12 from the payment notice requirement. The Bureau had limited evidence that lenders making payday alternative loans like those covered by proposed § 1041.11 take part in questionable payment practices. Given the cost restrictions placed by the NCUA on payday alternative loans and on the loans conditionally exempt under proposed § 1041.12, the Bureau believed it might have been particularly difficult to build the cost of providing the payment disclosure into the cost of the loan. It was concerned that lenders might be unable to continue offering payday alternative loans or the loans encompassed by proposed § 1041.12 if the disclosure requirement is applied. Proposed § 1041.15(b) also provided a limited exception to the notice requirement for the first transfer from a consumer’s account after the lender obtains the consumer’s consent pursuant to proposed § 1041.14(c) (now final § 1041.8(c)), regardless of whether any of the conditions in proposed § 1041.15(b) apply. As discussed above, proposed § 1041.14 would have generally required a lender to obtain a consumer’s consent before initiating another payment attempt on the consumer’s account after two consecutive attempts have failed. Proposed § 1041.15(b) would allow lenders to forgo the payment notice for the first payment attempt made under the consumer’s affirmative consent as the consent itself will function like a payment notice. 1114 Proposed comment 15(b)(2)(ii)-1 clarified that this exception would apply even if the transfer otherwise triggered the additional disclosure requirements for unusual attempts under proposed § 1041.15(b)(5). Proposed comment 15(b)(2)(ii)-2 explained that this exception would apply only to the first transfer when a consumer had affirmatively consented to multiple transfers in advance. Proposed § 1041.15(b)(2) also provided an exception for an immediate single payment transfer initiated at the consumer’s request as defined in proposed § 1041.14(a)(5). This exception would carve out situations where a lender is initiating a transfer within one business day of receiving the consumer’s authorization. During the SBREFA process and other external outreach, lenders raised concerns about how the Bureau’s potential proposal would apply to one-time, immediate electronic payments made at the consumer’s request. Industry commenters stated that, unless these payments were excepted from the requirement, lenders could be prohibited from deducting payments from consumers’ accounts for several days in situations in which consumers had specifically directed the lender to deduct an extra payment or given approval to pay off their loans early. Similarly, if an advance notice were required before a one-time payment, consumers attempting to make a last-minute payment might incur additional late fees due to the waiting period required after the disclosure. The Bureau believed that these were valid policy concerns and accordingly proposed to except an immediate single payment transfer made at the consumer’s request. It also believed that because this category of payments involved situations in which the consumer’s affirmative request to initiate a transfer is processed within a business day of receiving the request, the consumer was unlikely to be surprised or unprepared for the subsequent withdrawal. 1115 Proposed 15(b)(3) Timing. Proposed § 1041.15(b)(3) set forth timing requirements applicable to each of the three methods through which the payment notice can be delivered, which were mail, electronic, and in-person delivery. The minimum time to deliver the notice would range from six to three business days before the transfer, depending on the channel, as specified in the proposal. In proposing the timing requirements, the Bureau was attempting to balance several competing considerations about how timing may impact consumers and lenders. First, it believed that the payment notice information is more likely to be useful, actionable, and effective for consumers if it is provided shortly before the payment will be initiated. Consumers could use this information to assess whether there were sufficient funds in their account to cover the payment and whether they need to make arrangements for another bill or obligation that is due around the same time. However, consumers also might need some time to arrange their finances, to discuss alternative arrangements with the lender, or to resolve any errors. For example, if the payment were not authorized and the consumer wanted to provide a notice to stop payment to their account provider in a timely fashion under Regulation E § 1005.10(c)(1), the regulation would require the consumer to take action three business days before the scheduled date of the transfer. The Bureau was also aware that the delay between sending and receiving the notice complicates timing considerations. For example, paper delivery via mail involves a lag time of a few days and is difficult to estimate precisely. Finally, as discussed above, the Bureau believed that electronic delivery might be the least costly and most reliable method of delivery for many consumers and lenders. However, some consumers would not have access to an electronic means of receiving notices, in which case a paper option would be their only option to receive 1116 the notices required under proposed § 1041.15(b). In light of these considerations, the Bureau believed that these timing requirements, which incorporate the delays inherent in various methods of delivery and the utility of the disclosure information for consumers, would help ensure that the content of the payment notice is effectively disclosed to consumers, consistent with the Bureau’s authority under section 1032 of the Dodd-Frank Act. Specifically, proposed § 1041.15(b)(3) would require the lender to mail the notice no earlier than 10 business days and no later than six business days prior to initiating the transfer. Proposed comment 15(b)(3)(i)-1 clarified that the six business days would begin when the lender placed the notice in the mail, rather than when the consumer received the notice. For a payment notice sent by mail, there might be a gap of a few days between when the lender sent the notice and when the consumer received it. The Bureau expected that in most cases this would result in the consumer receiving the notice between three and seven business days prior to the date on which the lender intended to initiate the transfer. This expectation was consistent with certain provisions of Regulation Z,1058 which consider consumers to have received disclosures delivered by mail three business days after they are placed in the mail. For a payment notice sent through electronic delivery along with the electronic short notice in proposed § 1041.15(c), consumers would be able to receive a notice immediately after it is sent and without the lag inherent in paper mail. Proposed § 1041.15(b)(3)(ii)(A) would therefore adjust the time frames and require the lender to send the notice no earlier than seven business days and no later than three business days prior to initiating the transfer. Proposed comment 15(b)(3)(ii)(A)-1 clarified that the three business days would begin when the lender 1058 12 CFR part 1026. 1117 sends the notice, rather than when the consumer received or was deemed to have received the notice. Proposed § 1041.15(b)(3) would require that if, after providing the payment notice through electronic delivery pursuant to the timing requirements in proposed § 1041.15(b)(3), the lender lost a consumer’s consent to receive notices through a particular electronic delivery method, then the lender would have to provide the notice for any future payment attempt, if applicable, through alternate means. Proposed comment 15(b)(3)(ii)(B)-1 clarified that in circumstances when the lender received the consumer’s loss of consent for a particular electronic delivery method after the notice has already been provided, the lender could initiate the payment transfer as scheduled. If the lender was scheduled to make any payment attempts following the one that was disclosed in the previously provided notice, then the lender would have to provide notice for that future payday attempt through alternate means, in accordance with the applicable timing requirements in proposed § 1041.15(b)(3). Proposed comment 15(b)(3)(ii)(B)-2 explained that alternate means could include a different electronic delivery method that the consumer has consented to in person or by mail. Proposed comment 15(b)(3)(ii)(B)-3 provided examples of actions that would satisfy the requirements in proposed § 1041.15(b)(3). The Bureau was concerned that requiring lenders to delay the payment transfer past its scheduled date could cause consumers to incur late fees and finance charges. For example, if the lender attempts to deliver a notice through text message three days before the transfer date and the lender received a response indicating that the consumer’s phone number was out of service, then the lender would not have sufficient time before the scheduled payment transfer date to deliver to payment notice by mail according to the timing requirements in proposed 1118 § 1041.15(b)(3). Although it would be preferable that consumers received the notice before any transfer in all circumstances, on balance the Bureau believed that the potential harms of causing payment delays outweighed the benefits of requiring delivery of the notice through another method. It was concerned that even if lenders were required to deliver the notice through another means, such as mail, alternative means also might not successfully deliver the notice to the consumer. Under the proposal, if a lender provided the payment notice in person, then there would be no lag between providing the notice and the consumer’s receipt. Similar to the timing provisions provided for the electronic short notice, proposed § 1041.15(b)(3) would provide that if the lender provided the notice in person, then the lender would have to provide the notice no earlier than seven business days and no later than three business days prior to initiating the transfer. Proposed 15(b)(4) Content requirements Proposed § 1041.15(b)(4) specified the required contents of the payment notice, including an identifying statement, date and amount of the transfer, truncated information to identify the consumer account from which the withdrawal will be taken, loan number, payment channel, check number (if applicable), the annual percentage rate of the loan, a breakdown of how the payment is applied to principal and fees, and lender contact information. The proposed rule and comments thereto added more detail about these items. When the payment transfer had changed in a manner that makes the attempt unusual, the disclosure title would have to reflect that the attempt is unusual. The Bureau believed that this content would enable consumers to understand the costs and risks associated with each loan payment, consistent with its authority 1119 under section 1032 of the Dodd-Frank Act. The Bureau was aware that providing too much or overly complicated information on the notice may prevent consumers from reading and understanding it. To maximize the likelihood that consumers would read the notice and retain the most importance pieces of information about an upcoming payment, it believed that the content requirements should be minimal. In particular, the Bureau considered adding information about other consumer rights, such as stop-payment rights for checks and EFTs, but had concerns that this information may be complicated and distracting. Consumer rights regarding payments are particularly complicated because they vary across payment methods, loan contracts, and whether the authorization is for a one-time or recurring payment. As discussed in Market Concerns—Payments, these rights are often burdensome and costly for consumers to utilize. On the requirement to disclose APR, which is the one content requirement the Bureau is not finalizing as discussed below, it believed that providing information about the cost of the loan in the disclosure would remind consumers of the cost of the product over its term and assist consumers in their financial management, for instance in choosing how to allocate available funds among multiple credit obligations or in deciding whether to prepay an obligation. The Bureau recognized that consumers generally do not have a clear understanding of APR, as confirmed by the consumer testing of these model forms. It also stated at the proposal stage that APR nonetheless may have some value to consumers as a comparison tool across loan obligations even by consumers who are not deeply familiar with the underlying calculation. 1120 Proposed 15(b)(5) Additional content requirements for unusual attempts. Under the proposal, if a payment transfer was unusual according to the circumstances described in the proposal, then the payment notice would have to include both the content provided in proposed § 1041.15(b)(4) (other than disclosure of the APR) and the content required by § 1041.15(b)(5), which would mandate the notice to state if the amount or the date or the payment channel differs from the amount of the regularly scheduled payment, and that the transfer would be for a larger or smaller amount than the regularly scheduled payment, as applicable. Proposed § 1041.15(b)(5) would require the notice to state, if the payment transfer date is not a date on which a regularly scheduled payment is due under the loan agreement, that the transfer will be initiated on a date other than the date of a regularly scheduled payment. For payment attempts using a payment channel different from the channel used for the previous transfer, proposed § 1041.15(b)(5) would require a statement to specify that the transfer would be initiated through a different payment channel, as well as the channel that the lender had used for the previous payment attempt. If the transfer was for the purpose of re-initiating a returned transfer, then proposed § 1041.15(b)(5) would require the notice to state that it was a reinitiation, along with a statement of the date and amount of the returned transfer and a statement of the reason for the return. Proposed comment 15(b)(5)-1 explained if the payment transfer was unusual according to the circumstances described in proposed § 1041.15(b)(5), then the payment notice had to contain contents required by proposed § 1041.15(b)(4) (except for APR) and (b)(5). Proposed comment 15(b)(5)(i)-1 explained that the content requirement for varying amount applies when a transfer was for the purpose of collecting a payment that was not specified by amount on the payment schedule, or when the transfer was for the purpose of collecting a 1121 regularly scheduled payment for an amount different from the regularly scheduled payment amount according to the payment schedule. Proposed comment 15(b)(5)(ii)-1 explained that the content requirement for the date other than due date would apply when a transfer was for the purpose of collecting a payment that was not specified by date on the payment schedule, or when the transfer was for the purpose of collecting a regularly scheduled payment on a date that differed from regularly scheduled payment date according to the payment schedule. The Bureau believed that all four of these circumstances—varying amount, date, payment channel and re-initiating a returned transfer—might be important to highlight for the consumer, so that the status of their loan is fully disclosed to them pursuant to section 1032(a) of the DoddFrank Act. If a lender initiated a payment that differed from the regularly scheduled payment amount authorized by the consumer, the payment was more likely to vary from consumer expectations and pose greater risk of triggering overdraft or NSF fees. The Bureau thus believed that these changes should be highlighted for consumers to understand the risks, attempt to plan for changed payments, and determine whether their authorization is being used appropriately. It also believed that changes in the date and channel of the payment could be important information for the consumer to prepare for the withdrawal and take steps as necessary. To effectively and fully understand their current loan status and alert consumers to a series of repeat attempts over a short period, the Bureau further found it important for the consumer to know if the past payment attempt failed and the lender is attempting to re-initiate a returned transfer. Proposed 15(c)(1) General. The Bureau is combining the content from § 1041.15(c) into final § 1041.9(b) as well, and thus addresses these provisions here. Proposed § 1041.15(c) provided content requirements 1122 for an electronic short notice, essentially a truncated version of the payment notice formatted for electronic delivery through e-mail, text message, or mobile application. This notice would be provided when the lender has obtained the consumer consent for an electronic delivery method and is proceeding to provide notice through such a delivery method. As described above, this electronic short notice would provide a web link to the complete payment notice that would be required by the proposed rule. The Bureau believed it was appropriate to tailor the notices in light of format limitations for electronic delivery channels that may be beyond the lender’s control; as well as considerations about the ways consumers may access e-mail, text messages, and mobile applications; privacy considerations; preferences for particular usage settings; and other issues. For all of these reasons, it found it appropriate for the electronic short notice to contain less information than the full payment notice, given that it links to the full notice. It was also persuaded that providing access to the full notice via the website link would appropriately balance related concerns to ensure that consumers could access the full set of notice information in a more secure, usable, and retainable manner. However, the Bureau asked for comment on this two-step structure in the proposal and, as discussed below, is finalizing additional ways to deliver the notices electronically, such as by providing the full text of the notice in the email and providing a PDF attachment of the full notice rather than a web link. 15(c)(2) Content. The proposed electronic short notice contained an abbreviated version of the proposed payment notice content, and would be an initial notice provided through a method of electronic delivery that the consumer has consented to, such as a text message or email, that would provide a link to a unique URL containing the full payment notice. It would include an identifying 1123 statement that describes the purpose of the notice and the sender of the notice; the date of the transfer, amount of the transfer, and consumer account information; and a unique website URL that the consumer may use to access the full payment notice. 15(c)(3) Additional Content Requirements. Under the proposal, if the electronic short notice was being provided under an unusual attempt scenario, then the notice would have to state what makes the payment attempt unusual by providing information about whether the amount, date, or payment channel has changed. Comments Received The Bureau received a number of comments about the payment notice requirements proposed in the rule. Some commenters noted that the notices were beneficial because they would provide information to consumers that might allow them to avoid unexpected bank fees. On the other hand, a commenter argued that the timing requirements of the payment notices could pose safety-and-soundness risks by creating a “loophole” for those seeking to avoid payment, and create barriers to borrowers repaying their contractual obligations. It appears this commenter suggested that because borrowers would be made aware of a pending payment, they might choose to stop that payment, which concerned the commenter because it would make it harder to collect. Many industry commenters raised burden concerns about providing the notice. Several raised concerns about providing the paper notices through the mail. For example, one lender explained that compliance costs for mailed notices are between $10 and $24 for a $1,000 12month loan and another stated that mailed written notices would be 80 times more expensive than electronic notices. 1124 Additionally, as noted above when discussing § 1041.9(a) of the final rule, several commenters asserted that the payment notice requirements create compliance complexity. One commenter argued that because these notice requirements may preempt some and overlay other State law requirements, the requirement could cause both regulatory and consumer confusion. For example, the commenter claimed that if finalized, the rule could potentially require lenders to provide multiple notices with the same information in different formats (one required by this rule and the other required by State law). The commenter also suggested that lenders would incur substantial costs to try to navigate this dynamic. Another commenter argued that a similar overlap dynamic could exist with TILA and Regulation Z, which imposes disclosure requirements for creditors at loan origination. The commenter claimed that companies which are lenders under this rule and “creditors” under TILA and Regulation Z would have potentially duplicative disclosure requirements that would be burdensome and perhaps confusing to consumers, thus recommending that the Bureau issue a revised proposal to better align with the requirements in TILA and Regulation Z. Several stakeholders commented on the proposed content of the payment notices, arguing that they merely would disclose information pertaining to an agreement into which the borrower had already entered, and thus would be unnecessary, or could frustrate or confuse consumers. A number of commenters asked the Bureau to provide a means for consumers to opt out of the notices, explaining that some consumers may not want to receive a stream of notices for normal payment activity. One commenter claimed that consumers might be disconcerted by receiving a comprehensive disclosure, and that it would be atypical to receive a disclosure that explains something to which a consumer already had agreed. This commenter claimed that consumers 1125 might not want the notices, or be frustrated by receiving them, and that their frustration would likely be aimed at the lenders. Many of these commenters focused their concerns on instances where a borrower agreed to regular automatic payments to make payments on installments. One consumer advocate suggested using the term “balance” instead of “principal.” Others suggested providing all of the notice information in the body of the email, given concerns that a link may be at times difficult for consumers to access. The Bureau did not receive any comments about privacy concerns from including the full notice in the body of the email or from a web link notice. Several commenters argued that instead of requiring lenders to obtain new payment authorizations after two failed attempts, the Bureau should include in these notices a disclosure requirement about consumers’ rights to revoke existing authorizations. Other commenters had specific comments about the content of the notices. Some generally agreed with the prohibition against providing the full account number, agreeing with the Bureau that a full account number could leave consumers vulnerable to fraud. One commenter argued that the Bureau should require that the name of the Originating Depository Financial Institution (ODFI) be included in the notices. Another argued that the Bureau should not require inclusion of a check number, which they claim may interfere with lenders’ ability to use remotely created checks and payment orders. A number of commenters expressed agreement with the requirement to include APR in the notices, including a suggestion to disclose an APR that includes credit insurance premiums. Others cited the Bureau’s findings in the mortgage context that borrowers find APR confusing or unhelpful, arguing that it should not be included in the payment notices. 1126 One commenter argued that credit union lenders, unlike other lenders, already provide most of the information in the proposed disclosures in monthly billing statements. Credit union commenters expressed concern that they would have to comply with the payment provisions, including by providing payment notices, when making loans under the NCUA’s PAL program. These commenters argued that credit unions that already provide the information via billing statement should be exempted from having to provide this information again in a separate disclosure. Finally, one commenter argued that depository institutions acting as service providers to lenders would have no way to know, under current technological means, whether transactions were related to covered loans, and would have no way to tell whether lenders had complied with notice requirements. For this reason, the commenter asked the Bureau to clarify under the final rule that the depository institutions holding the lender’s or borrower’s deposit account would not be held responsible for compliance with notice requirements. Final Rule The Bureau is now finalizing proposed § 1041.15(b) and (c), renumbered as § 1041.9(b), with significant deviations from the requirements proposed. In response to many comments about the burden of the notice, along with other concerns such as how consumers may be overwhelmed and desensitized by notices that are provided before every payment withdrawal, the Bureau is finalizing a scaled back payment notice requirement. Under the final rule, the notice will be required before (i) the first time a lender initiates a withdrawal and (ii) any unusual payment notices thereafter. There are also additional exceptions for open-end credit products, which already have periodic statement requirements under Regulation Z. 1127 In particular, in deciding to modify the proposal in this manner, the Bureau found compelling the comments it received about over-disclosure and burdens associated with notices before every automatic payment withdrawal on installment loans. The upcoming payment notices may not be necessary for long term loans that are not experiencing unusual payment activity. However, due to concerns about payment transparency identified in the proposal, consumers would benefit from obtaining an upcoming payment notice for the first payment. This revision would incentivize lenders to stick to the payment schedule and would only impose costs—which commenters pointed out may be more significant for paper notices—if they deviate from the consumer’s authorization. This change would eliminate the need for a consumer opt-out regime, because after the first payment consumers would only receive notices if something unusual was happening. It also may make the unusual payment notices more salient for consumers, who otherwise could become desensitized to notices that are delivered before every payment. Accordingly, the Bureau decided that if a borrower is given a disclosure before the first withdrawal, and there are future withdrawals that are not unusual—meaning they do not vary in amount, are not on a date other than the date of regularly scheduled payment, are not processed through a different payment channel, and are not for purposes of re-initiating a previous failed transfer—then that first payment notice should suffice to give borrowers notice of payment characteristics. Also in response to burden concerns, the Bureau has adjusted the timing requirements so that the first payment withdrawal notice could be provided earlier, such as during origination. Of course, under this new notice regime, the requirement that the initial notice be retainable is even more important. To further limit burden and allow flexibility as consumer preferences and technologies change, the Bureau is finalizing additional ways to 1128 deliver the notices electronically, including by providing the full text of the notice in the email and providing a PDF attachment of the full notice rather than a web link. To implement these revisions, the Bureau has restructured the regulatory text. At a high level, in the proposal the Bureau structured paragraph (b) as the requirement to provide notices before all withdrawals (including various requirements depending on whether the payments were unusual), and paragraph (c) set forth the ability to provide an electronic short notice instead. In the final rule, paragraph (c) has been built into paragraph (b), at paragraph (b)(4). Additionally, the Bureau has restructured paragraph (b) by splitting up the requirements for first payment withdrawal notices and unusual withdrawal notices—in paragraph (b)(2) and (b)(3) respectively—as separate sections. To clarify situations when the notices are required under this more limited frequency, definitions were added for the terms first payment withdrawal and unusual withdrawal under § 1041.9(b)(1)(i) and (b)(1)(ii), respectively. To ease readability, provisions are now repeated in (b)(2) and (b)(3) such that the requirements for each type of notice are self-contained in their respective sections. The commentary has been revised to incorporate these changes as well. In finalized paragraph (b)(2)(i), the Bureau has changed how early a first payment withdrawal notice can be provided by mail, electronically, or in person. Specifically, lenders can now provide the notice as early as when the lender obtains payment authorization. This change was intended to further reduce burden to lenders, as now lenders, if they wish, may provide the first payment withdrawal notice at origination, when they are already interacting with the consumer and providing other loan materials. Although the information would not be as timely for consumers, consumers would receive the information in retainable form and there are 1129 transparency benefits to incentivizing lenders to commit to a particular payment date, channel, and amount at the time of origination. The Bureau did not finalize proposed paragraph (b)(2)(i), which would have exempted payment transfers in connection with loans made under proposed § 1041.11 or § 1041.12 because the Bureau is not finalizing either of those sections here. The Bureau is also not finalizing the requirement to disclose APR. Although the Bureau received some comments supporting its inclusion, it agrees with other commenters that APR disclosures may be duplicative of the disclosures provided under Regulation Z, especially with regard to the first payment withdrawal notice that might be provided at origination, which the Bureau believes will now make up the majority of the notices provided under this rule. The Bureau is not changing the term “principal” to “balance.” Balance seems misleading in this context because the notice breaks out principal from interest and fees, and “balance” might lead consumers to believe that the interest and fees are not outstanding in addition to the principal amount. The Bureau is finalizing the requirement that lenders only include a truncated account number in the notices. It is concerned that full account number is sensitive information given that a lender or fraudster could use it in conjunction with a bank routing number to initiate an ACH or RCC transfer. Truncated account number (such as the last four digits) would still allow consumers to identify the account. The Bureau continues to believe that the account information is important for consumers to track which account is being debited. However, despite disclosure of this information on the notice, the Bureau has concerns that lenders at times debit accounts that the consumer did not provide authorization for. It will continue to monitor these 1130 unauthorized transfer practices related to account switching, and maintains that requiring a lender to commit to a specific account number, via notice, may assist in that effort. The Bureau is adding provisions to address overlap of the unusual withdrawal notices with disclosures required under Regulation Z for open-end credit plans. Under paragraph (b)(3)(i)(D), the unusual withdrawal notices may be provided in conjunction with the periodic statement required under Regulation Z, 12 CFR 1026.7(b). The Bureau added this provision to reduce burden on open-end lenders, which already must provide periodic statements under Regulation Z—which provides its own timing requirements—and may prefer to provide the notices at the same time; also, the Bureau believes that consumers of open-end credit would benefit, for comparison purposes, from receiving an unusual withdrawal notice in conjunction with or close in time to the periodic statement. It is further aware that minimum payments due for open-end credit plans may fluctuate depending on the outstanding balance. Under paragraph (b)(3)(ii)(C)(1)(ii), that unusual withdrawal notice need only include content about varying amount when the amount deviates from the scheduled minimum payment due as disclosed in that periodic statement required under Regulation Z.1059 The Bureau believes consumers would benefit from receiving an unusual withdrawal notice when an open-end credit lender deviates from the scheduled payment amount due. As the first payment withdrawal notice contains information that is not on the periodic statement (e.g., payment channel) and that it is a one-time notice that can be provided at origination, the Bureau believes that open-end credit consumers would benefit from receiving the first payment withdrawal notice. 1059 12 CFR 1026.7(b). 1131 The Bureau adjusted the electronic delivery provisions to allow for options beyond the two-step short notice plus link process. Under paragraph (b)(4)(i), there is an exception to the electronic short notice requirement if a lender is using email delivery as provided in paragraph (b)(4)(iii). Under (b)(4)(iii), when the consumer has consented to receive disclosures through electronic delivery, and the method of electronic delivery is email, the lender may either deliver the full notice required by paragraph (b)(1) in the body of the email or deliver the full notice as a linked URL webpage or PDF attachment along with the electronic short notice as provided in (b)(4)(ii). The revision is meant to address burden concerns raised by lenders and access concerns raised by consumer advocates. The Bureau has made corresponding changes in the commentary, and added a number of comments providing additional clarification about the meaning of first payment withdrawal. Comment 9(b)(1)(i)-1 explains that the term encompasses the first payment initiated by the lender, so it is not necessarily the first payment on a covered loan; for example, a lender that obtains payment authorization after a few payments have been made by the consumer in cash would deliver the notice later in the loan term. Comment 9(b)(1)(i)-2 explains that when an open-end credit plan is not a covered loan at origination, but becomes one later, the first payment withdrawal after the loan becomes a covered loan would qualify as the first payment withdrawal. Comment 9(b)(2)(i)-1 specifies that the earliest point at which a lender may provide the first payment withdrawal notice is when the lender obtains the payment authorization. It also specifies that the notice can be provided simultaneously with receiving payment authorization, which could be at origination. The Bureau did not finalize comment (b)(3)(i)(B)-3 because it implicated regular payment notices that are now not contemplated in the final rule. 1132 The Bureau added comments 9(b)(3)(ii)(C)-1 and -2 to provide further guidance on unusual withdrawal notices, with the latter providing an example of a payment that is unusual because the payment channel has changed. The Bureau added a paragraph to comment 9(b)(3)(ii)-3 describing how circumstances that trigger an unusual withdrawal for open-end credit plans are more limited according to § 1041.9(b)(3)(ii)(C)(1)(ii). It now says that since the outstanding balance on open-end credit plans may change over time, the minimum payment due on the scheduled payment date may also fluctuate. However, the minimum payment amount due for these open-end credit plans would be disclosed to the consumer according to the periodic statement requirement in Regulation Z. The payment transfer amount would not be considered unusual with respect to an open-end credit plan unless the amount deviates from the minimum payment due as disclosed in the periodic statement. Furthermore, the requirement for a first payment withdrawal notice under § 1041.9(b)(2) and the other circumstances that could trigger an unusual withdrawal notice under § 1041.9(b)(3)(ii)(C)(2) through (4), continue to apply. Lastly, the Bureau added comment 9(b)(4)-1 to clarify that an electronic short notice must be used for electronic delivery other than email, but that the lender can choose whether to use the electronic short notice or the full text when using email. The Bureau has determined that many of the extensive changes it made to the final rule largely incorporate and address the critical feedback received from commenters. While it does not share the fear that a borrower might choose not to pay if given a more informed choice, commenters’ concerns about the notices making collections more difficult are largely addressed by the fact that consumers will no longer receive notices before every payment. The Bureau also made changes to address concerns about overlapping Regulation Z requirements by adding 1133 caveats for open-end credit and taking APR off the notices. And as stated above, the compliance burden associated with payment notices should be reduced significantly now that lenders will only need to provide notices on the first payment withdrawal, and before unusual withdrawals. The Bureau does not agree that it needs to enact an opt-out provision for these notices. It has addressed concerns about consumers becoming desensitized to multiple identical notices by eliminating the need to send multiple identical notices. As lenders will only be sending notices upon infrequent events (the first payment, an unusual payment, or when the payment attempt cap is met), the risk of overloaded consumers is minimized; additionally, the Bureau wants to ensure that borrowers are aware of these rare events, and an opt-out regime might undermine that goal—including by allowing lenders to use the opt out feature to surreptitiously initiate payments that fall outside of consumers’ expectations. Credit union lenders making loans under the PAL program will not have to comply with any parts of this rule, including the payment notices. To the extent commenters believed that the Bureau’s exclusion did not fully capture all PAL program loans, the Bureau has added a clarification in § 1041.3(e) to explicitly exclude all PAL program loans. The Bureau does not see a basis for requiring lenders to identify the ODFI on the notices. Borrowers do not have a relationship with the ODFI, and would not need that information to understand any of the triggering events for which notices are required. Nor would borrowers need that information to enact a stop payment or revoke an authorization. The Bureau also knows from its experience in disclosures and consumer testing about the value of keeping the content of the notices limited so as not to crowd out or distract from the most important content. 1134 The Bureau maintains its view that a check number should be on the first payment withdrawal notices. As described above in Market Concerns—Payments, borrowers may need that information to enact a stop payment. Contrary to one commenter’s suggestion, the Bureau believes that this information will be useful to consumers. The Bureau is not aware of any State laws that would directly conflict with the notice requirements set forth in the proposal or this final rule. It believes it is important that all consumers in all States receive these notices, and trusts that State officials will find an appropriate way to ensure that improved disclosures required by State laws are helpful to consumers in their State, in accordance with their independent judgment. 9(c) Consumer rights notice. The Bureau has decided to finalize proposed § 1041.15(d) and 1041.15(e) as combined into § 1041.9(c) of the final rule. Other than adding some additional options for electronic delivery – which were also added to the notices in § 1041.9(b)—the Bureau is finalizing the consumer rights notice as proposed. Its reasons for doing so are set out below. Proposed Rule Proposed 15(d)(1) General. Proposed § 1041.15(d) required lenders to provide consumers with a consumer rights notice after a lender has initiated two consecutive or concurrent failed payment transfers and triggered the protections provided by the proposed rule. It also would provide timing and content requirements for this consumer rights notice, which would be triggered when the lender received information that its second consecutive payment attempt has failed. As described above, proposed § 1041.14 would have limited a lender’s ability to initiate a payment transfer 1135 after two consecutive attempts have failed, allowing the lender to initiate another payment attempt from the consumer’s account only if the lender had received the consumer’s consent under proposed § 1041.14(c) or authorization to initiate an immediate one-time transfer at the consumer’s request under proposed § 1041.14. 15(d)(2) Timing. The proposed rule would require a lender to send the consumer rights notice no later than three business days after the lender received information that the second consecutive attempt had failed, which proposed comment 15(d)(2) clarified would be triggered whenever the lender or its agent, such as a payment processor, received information that the second attempted payment transfer had failed. The Bureau believed that when a lender had initiated two consecutive failed payment transfers and triggered the protections provided by proposed § 1041.14(b), a consumer might not be aware that the lender was no longer permitted to initiate payment from the consumer’s account. In the meantime, some loans might accrue interest or fees while the balance would remain unpaid. For these reasons, the Bureau stated that the consumer rights notice should be provided shortly after the second attempt fails. However, the Bureau was aware that, depending on the payment method, there may be a delay between the lender’s initiation of the payment transfer and information that the payment transfer has failed. Accordingly, the Bureau proposed to require the lender to send the consumer rights notice within three business days after the lender received information that the payment transfer has failed. 15(d)(3) Content requirements. The proposal would also specify the content requirements for the consumer rights notice. The Bureau believed that a consumer should know that a lender has triggered the provisions in 1136 proposed § 1041.14 and was no longer permitted to initiate payment from the consumer’s account. It also considered it important to inform consumers that Federal law prohibits the lender from initiating further payment withdrawal attempts. Given that proposed § 1041.14 would prohibit the lender from initiating another payment attempt without a new consumer authorization, the Bureau proposed it would also be useful for the consumer to be aware that the lender may be contacting the consumer to discuss payment choices. Consistent with the Bureau’s authority under section 1032(a) of the Dodd-Frank Act, this content would inform consumers of the payment status on their covered loans. It also might help prevent consumer confusion or misinformation about why the lender cannot initiate another payment, by helping to ensure that this information about the situation is effectively, accurately, and fully disclosed to the consumer. The proposed rule specified that this content would include an identifying statement, a statement that the lender’s last two attempts to withdraw payment had failed, information about the consumer account and loan identification information, a statement on the Federal law prohibiting the lender from initiating further transfers without the consumer’s permission, a statement that the lender could contact the consumer to discuss payment choices going forward, the circumstances of why the lender could no longer withdraw payments from the consumer’s account, and information about the Bureau. 15(e) Electronic short notice. For lenders to deliver the required consumer rights notice through an electronic delivery method, the proposed rule would require the lenders to provide an electronic short notice that contains a link to the full consumer rights notice; a truncated version of the content specified in the proposal; an email subject line, if applicable; and a unique website URL that links to the full 1137 consumer rights notice. For many of the same reasons discussed above in connection with proposed § 1041.15(c), the Bureau believed that the electronic short notice should contain limited content to maximize the utility of notices for consumers and minimize the burden on lenders. Consistent with the Bureau’s authority under section 1032 of the Dodd-Frank Act, these proposed requirements would help ensure that consumer rights under proposed § 1041.14 are effectively disclosed to consumers. Proposed § 1041.15(e)(2) specified that the electronic short notice must contain an identifying statement, a statement that the last two attempts were returned, consumer account identification information, and a statement of the prohibition under Federal law, using language substantially similar to the language set forth in the proposed model form. These terms were described for the full consumer rights notice in proposed § 1041.15(d)(3)(i), (ii), (iii), and (v). Proposed comment 15(e)(2)-1 clarified that when a lender provides the electronic short notice by email, the email had to contain this identifying statement in both the subject line and the body of the email. In order to provide consumers access to the full consumer rights notice, proposed § 1041.15(e)(2)(v) would also require the electronic short notice to contain the unique URL of a website that the consumer may use to access the consumer rights notice. The Bureau understood that the unique website URL contains limited privacy risks because it would be unlikely that a third party will come across a unique URL. Even if a third party did discover this URL, the notice would not contain identifying information such as the consumer’s name or full account number. Comments Received 1138 Many of the comments relating to the notices were aimed more generally at all of the notice requirements, and not specifically at the consumer rights notice. For example, some commenters repeated the concern that these provisions would create additional regulatory requirements for loans made under the NCUA’s PAL program, which is not correct because those loans are not subject to the notice requirements. Others raised general concerns about the total compliance burden, which has been substantially lessened due to various changes in the final rule, including a significant scaling back of the frequency of the notices. Those comments are all addressed in the earlier discussions of comments above. Lastly, the Bureau did not receive any comments about the specific timing or content of the consumer rights notices. Final Rule The Bureau is now finalizing proposed § 1041.15(d) and 1041.15 (e) as § 1041.9(c) of the final rule. It has concluded that consumers should be informed when a lender has triggered the threshold of two consecutive failed payment withdrawal attempts so that they are made aware of the failed attempts and of the fact that, by operation of law, further attempts will cease even though they remain obligated to make continuing loan payments. The Bureau is also concerned that some lenders may pressure consumers to provide affirmative consent and could present the reasons behind the re-initiation limit in an incomplete manner. It has made the judgment that requiring disclosure of information about prior failed payments and consumer rights under § 1041.8 of the final rule would help ensure that the costs, benefits, and risks of the loan and associated payments are effectively disclosed to consumers, consistent with its authority under section 1032 of the Dodd-Frank Act. Due to these policy considerations, the Bureau has 1139 determined that a lender should be required to provide a standardized consumer rights notice after it has initiated two consecutive attempted payment withdrawals have failed. The Bureau has made a few technical changes to reconcile the numbering changes, but otherwise is finalizing these paragraphs as proposed with only one substantive change to the rule and a corresponding change to the commentary. To ease burden and provide lenders with additional options—which may be beneficial to consumers giving changing preferences and privacy concerns in an evolving technological world—the Bureau is explicitly stating that when making electronic delivery of the consumer rights notices via email, lenders can, if they choose and the consumer has provide required consent, provide the full notice in the text of the email instead of the electronic short notice, or provide the full notice in a PDF attachment instead of through a linked URL webpage. Lastly, the Bureau notes that the exclusions and exemptions listed in § 1041.3, including that for PAL loans, applies to all sections of part 1041, including this section. Subpart D—Information Furnishing, Recordkeeping, Anti-Evasion, and Severability Sections 1041.10 Information Furnishing Requirements and 1041.11 Registered Information Systems Overview of the Proposal As described earlier, the Bureau proposed that it is an unfair and abusive practice to make a covered short-term loan without reasonably determining that the consumer has the ability to repay the loan. The Bureau proposed to prevent this abusive and unfair practice by, among other things, including in the proposal requirements for how a lender could reasonably determine that a consumer has the ability to repay a loan. 1140 The Bureau stated that, in order to achieve these consumer protections, a lender must have access to reasonably comprehensive information about a consumer’s current and recent borrowing history, including covered loans made to the consumer by other lenders, on a realtime or close to real-time basis. As discussed above, online borrowers appear especially likely to move from lender to lender. This makes it particularly important for online lenders to have access to information about covered loans made by other lenders in order to assess properly a consumer’s eligibility for a loan under the proposal. The Bureau proposed § 1041.16 to require lenders to furnish certain information about most covered loans to each information system registered with the Bureau pursuant to proposed § 1041.17.1060 This requirement was intended to be in addition to any furnishing requirements existing under other Federal or State law. The proposed registered information systems would be consumer reporting agencies within the meaning of § 603(f) of the Fair Credit Reporting Act (FCRA).1061 Accordingly, lenders furnishing information to these systems under proposed § 1041.16 would be required to comply with the provisions of the FCRA and its implementing regulations applicable to furnishers of information to consumer reporting agencies.1062 The furnishing requirement under proposed § 1041.16 would enable a registered information system to generate a consumer report containing relevant information about a consumer’s borrowing history, regardless of which 1060 The proposal required entities seeking to become registered information systems after the effective date of proposed § 1041.16 to first be provisionally registered for a period of time. 1061 15 U.S.C. 1681a(f). 1062 These provisions include a number of requirements relating to the accuracy of information furnished, including the requirement to investigate consumer disputes and to correct and update information. See, e.g., 15 U.S.C. 1681s2(a) through (b); 12 CFR 1022.42 through 1022.43. Compliance with the FCRA may require that information in addition to that specified in the proposal is furnished to information systems registered with the Bureau. The proposed furnishing requirements aimed to ensure that lenders making most loans covered under the proposal would have access to information necessary to enable compliance with the provisions of the proposal, but would not supersede any requirements imposed upon furnishers by the FCRA. 1141 lender had made a covered loan to the consumer previously. A lender contemplating making most covered loans to a consumer would be required to obtain a consumer report from a registered information system and consider such a report in determining whether the loan could be made to the consumer, in furtherance of the consumer protections of proposed part 1041.1063 In developing the proposal, the Bureau considered an alternative approach to ensure that lenders could obtain reasonably comprehensive information about a consumer’s borrowing history across lenders. Under this alternative approach, lenders would furnish information about covered loans to only one of the entities registered with the Bureau, but would be required to obtain a consumer report from each such entity.1064 However, the Bureau preliminarily believed that this approach would be costlier for lenders than the proposed approach because lenders potentially would need to obtain several consumer reports for every application for a covered short-term loan made under proposed § 1041.5 or § 1041.7.1065 The Bureau recognized the costs involved in furnishing to multiple entities but anticipated that those costs could be substantially reduced with appropriate coordination concerning data standards. The Bureau considered an alternative under which lenders would be required to furnish information to the Bureau or a contractor designated by the Bureau, and to obtain a report from the Bureau or its contractor. 1063 The proposal explained that such lenders would be subject to the provisions of the FCRA and its implementing regulations applicable to users, including the requirement to provide a consumer a notice of taking an adverse action based in whole or in part on information contained in a consumer report. See, e.g., 15 U.S.C. 1681m(a). 1064 If lenders were required to furnish information to only one consumer reporting agency, the Bureau identified a substantial risk that, for many consumers, no consumer reporting agency would be able to provide a reasonably comprehensive report of the consumer’s current and recent borrowing history with respect to covered loans across lenders. 1065 Under the proposal, a lender would have had to review a consumer report in connection with loans made pursuant to proposed §§ 1041.5, 1041.6, 1041.7 and 1041.9. For ease of reference, this section-by-section analysis only refers to proposed § 1041.5 and/or § 1041.7 because the Bureau is adopting these proposed sections in the final rule (as §§ 1041.5 and 1041.6) and is not adopting proposed §§ 1041.6 and 1041.9. 1142 The Bureau believed that these functions would be better performed by the private sector and that the proposed approach would permit faster implementation of the rule. Further, it noted there may be legal or practical obstacles to this alternative approach. The proposal would have required the Bureau to identify the particular consumer reporting agencies to which lenders were required to furnish information pursuant to proposed § 1041.16, and from which lenders would obtain consumer reports pursuant to proposed § 1041.5 and § 1041.7. Specifically, under proposed § 1041.17, the Bureau would have registered these consumer reporting agencies with the Bureau as information systems. Lastly, proposed § 1041.17 set forth processes for registering information systems before and after the effective dates of the furnishing obligations under proposed § 1041.16, and established the conditions that an entity had to satisfy to become a registered information system. Legal authority for subpart D A. Section 1031(b) Section 1031(b) of the Dodd-Frank Act authorizes the Bureau to prescribe rules for the purpose of identifying unfair or abusive acts or practices, which rules may include requirements for the purpose of preventing such acts or practices.1066 As discussed above, the Bureau determined that it is an unfair and abusive practice to make a covered loan without determining that the consumer has the ability to repay the loan. Accordingly, consistent with aspects of the proposed rule, this final rule requires lenders to determine the consumer’s ability to repay a covered loan, including by reviewing the consumer’s borrowing history and any current difficulty with repaying an outstanding loan. 1066 12 U.S.C. 5531(b). 1143 The provisions of proposed §§ 1041.16 and 1041.17 were designed to ensure that lenders would have access to information to achieve the consumer protections of proposed §§ 1041.5 and 1041.7. The Bureau believed that to prevent the abusive or unfair practices identified in the proposed rule, it would be necessary or appropriate to require lenders to obtain and consider relevant information about a borrower’s current and recent borrowing history, including covered loans made by all lenders. Requiring lenders to furnish relevant information concerning most covered loans pursuant to proposed § 1041.16 would ensure that lenders have access to a reliable and reasonably comprehensive record of a consumer’s borrowing history when considering extending the consumer a loan. In turn, this would ensure that consumers receive the benefit of the protections imposed by proposed §§ 1041.5 and 1041.7. B. Section 1024(b) Section 1024(b)(7) of the Dodd-Frank Act provides that the Bureau may: (A) “prescribe rules to facilitate supervision of persons described in subsection (a)(1) and assessment and detection of risks to consumers;” (B) “require a person described in subsection (a)(1), to generate, provide, or retain records for the purposes of facilitating supervision of such persons and assessing and detecting risks to consumers;” and (C) “prescribe rules regarding a person described in subsection (a)(1), to ensure that such persons are legitimate entities and are able to perform their obligations to consumers.”1067 The provisions in proposed § 1041.17—including the criteria governing when the Bureau may register or provisionally register information systems, suspend or revoke such registration or provisional registration, or deny applications for registration or provisional registration—were proposed to facilitate supervision, enable the 1067 12 U.S.C. 5514(b)(7)(A)-(C). 1144 assessment and detection of risks to consumers, and ensure that registered information systems are legitimate entities able to perform their obligations to consumers. Proposed § 1041.17 permits the Bureau to provisionally register or register an information system only if the Bureau determines, among other things, that the information system acknowledges that it is, or consents to being, subject to the Bureau’s supervisory authority. Section 1024 of the Dodd-Frank Act grants the Bureau supervisory and enforcement authority over, among other non-bank persons, “larger participant[s] of a market for other consumer financial products or services,” as the Bureau defines by rule.1068 In 2012, the Bureau promulgated a final rule defining larger participants of the market for consumer reporting.1069 As noted in the proposal, the Bureau believes that entities that are registered information systems would be non-depository institutions that qualify as larger participants in the market for consumer reporting, and their acknowledgment would reflect that status. To the extent such an entity is not a larger participant, or if there is any ambiguity concerning that status, the proposal would require that an entity consent to the Bureau’s supervisory authority to be eligible for registration as an information system.1070 C. Sections 1022(b), 1022(c), and 1021(c)(3) Section 1022(b)(1) of the Dodd-Frank Act authorizes the Bureau to prescribe rules “as may be necessary or appropriate to enable the Bureau to administer and carry out the purposes 1068 12 U.S.C. 5514(a)(1)(B) and (a)(2). 77 FR 42873 (July 20, 2012). 1070 For example, 12 CFR 1091.110(a) provides that, “[n]otwithstanding any other provision, pursuant to a consent agreement agreed to by the Bureau, a person may voluntarily consent to the Bureau’s supervisory authority under 12 U.S.C. 5514, and such voluntary consent agreement shall not be subject to any right of judicial review.” 1069 1145 and objectives of the Federal consumer financial laws, and to prevent evasions thereof.”1071 The criteria defined in proposed § 1041.17 would ensure that registered information systems provide information to the Bureau about their activities and compliance systems or procedures. In addition to helping to achieve the purposes and objectives of the proposed rule, these provisions were proposed to ensure that “consumers are protected from unfair, deceptive, or abusive acts and practices,” and that “markets for consumer financial products and services operate transparently and efficiently to facilitate access and innovation.”1072 Section 1021(c)(3) of the Dodd-Frank Act provides that it is a function of the Bureau to “publish[] information relevant to the functioning of markets for consumer financial products and services to identify risks to consumers and the proper functioning of such markets.”1073 Section 1022(c)(7) further authorizes the Bureau to “prescribe rules regarding registration requirements applicable to a covered person, other than an insured depository institution, insured credit union, or related person.”1074 Pursuant to the authorities described above, the Bureau is thus finalizing subpart D.1075 Effective date of §§ 1041.10 and 1041.11 of the final rule Although the effective dates of the various sections of the rule are discussed in part VI, it is necessary to address them here also, as the imposition of information furnishing requirements 1071 12 U.S.C. 5512(b)(1). 12 U.S.C. 5511(b)(2) and (b)(5). 1073 12 U.S.C. 5511(c)(3). 1074 12 U.S.C. 5511(c)(7). 1075 See also 12 U.S.C. 5514(b)(1)(A) through (C) (authorizing, with respect to persons described in section 1024, the Bureau to “require reports and conduct examinations . . . for purposes of—(A) assessing compliance with the requirements of Federal consumer financial law; (B) obtaining information about the activities and compliance systems or procedures of such person; and (C) detecting and assessing risks to consumers and to markets for consumer financial products and services”). 1072 1146 and the registration of information systems involve operational issues where timing is a significant factor. Proposed Rule As discussed in the proposal, the Bureau believed that building a reasonably comprehensive record of recent and current borrowing would take some time and raises a number of transition issues. For entities that wanted to become registered information systems before the furnishing requirements under proposed § 1041.16 take effect, the Bureau proposed a process that would generally work in the following sequence: proposed § 1041.17 would take effect 60 days after publication of the final rule in the Federal Register so that the standards and process for registration would become operative. Interested entities would submit to the Bureau an application for preliminary approval for registration and, after receiving preliminary approval and obtaining certain written assessments from third parties concerning their compliance programs, a full application for registration. After an entity became a registered information system, the Bureau proposed to provide at least 120 days for lenders to onboard to the information system and prepare for furnishing before proposed § 1041.16 began to require furnishing. As detailed in the section-by-section analysis of proposed § 1041.17, the Bureau proposed a timeline for these steps that it believed would ensure that information systems would be registered, and lenders ready to furnish, on the date that the furnishing obligation in proposed § 1041.16 becomes effective. Ultimately, the Bureau proposed allowing approximately 15 months after publication of the final rule in the Federal Register for information systems to complete the registration process described above, and for lenders to onboard to registered information systems and prepare to 1147 furnish. The Bureau also considered whether an additional period was needed between the date that furnishing to registered information systems would begin and the effective date of the requirements to obtain a consumer report from a registered information system under proposed §§ 1041.5 and 1041.7. Comments Received A number of industry commenters and trade associations objected to the Bureau’s proposed timeline to implement §§ 1041.16 and 1041.17 as being too short. In particular, commenters argued that, given the proposal to require furnishing to each provisionally registered and registered information system (“furnish-to-all”), the sheer mechanics necessary to create furnishing relationships between all of the lenders making covered loans and all of the provisionally registered and registered information systems could not be accomplished in the allotted time frame. One commenter noted that in addition to common data standards, other standards would need to be established as well, which could take additional time. Pointing to the complexities of the proposal, one commenter urged the Bureau to delay the final rule’s effective date, including proposed § 1041.17, which the Bureau proposed to become effective 60 days after publication of the final rule. The commenter recommended that the furnishing requirement in proposed § 1041.16 become effective sometime between 18 and 24 months after publication of the final rule. Two others suggested an implementation period of 24 months or longer. As precedent, one commenter cited the Bureau’s TILA-RESPA Integrated Disclosure Rule, which became effective almost 24 months after the final rule was published. One commenter said delaying the effective date of the rule beyond the proposed 15 months would have two advantages. First, it would allow the Bureau to develop a contingency plan if no entity had 1148 applied or qualified for registration before the effective date. Second, if the Bureau experienced delays in registering information systems, the additional time would provide that lenders still had sufficient time to onboard. One industry commenter requested a 26-month implementation period and asserted that, in developing its timeline for implementation, the Bureau did not consider the time necessary for developing, testing, and deploying the infrastructure needed to comply with the proposal’s onboarding and furnishing requirements. Final Rule The Bureau has considered the points made in the comments regarding the time frames related to provisionally registered and registered information systems in proposed §§ 1041.16 and 1041.17 and engaged in further analysis of the operational aspects of this process in light of those comments. As a result, the Bureau has decided to extend some of the proposed time frames in final §§ 1041.10 and 1041.11 (proposed §§ 1041.16 and 1041.17 as adopted and renumbered), including the time frame for submitting an application for preliminary approval for registration, the time frame for submitting an application to become a registered information system, the time frame for provisional registered information systems to automatically become fully registered information systems, and the time frame within which furnishing to a particular provisionally registered or registered information system must begin (the onboarding period). The Bureau is also extending the overall general implementation period for the final rule. Nonetheless, the Bureau is adopting the proposed effective date for the registration provisions in § 1041.11. Accordingly, the standards and processes for becoming registered information systems will become operative 60 days after the final rule’s publication. However, based on the comments it received, the Bureau is persuaded that other time frames should be 1149 extended. In particular, the Bureau concluded that potential registered information systems needed more time than originally proposed to submit applications for registration before the effective date of the furnishing obligation. Final § 1041.11(c)(3)(i) extends the proposed time frame for entities to submit applications for preliminary approval for registration from 30 days to 90 days. In addition, final § 1041.11(c)(3)(ii) extends the proposed time frame from 90 days to 120 days for entities that have received preliminary approval to submit applications to become registered information systems. The Bureau is also extending from 180 to 240 days the proposed time frame for entities provisionally registered on or after the effective date of the furnishing obligation to automatically become registered information systems. Like the proposal, the process for registration on or after the effective date of § 1041.10 involves two steps: an entity will be required to apply to become a provisionally registered information system pursuant to § 1041.11(d)(1) and then, after it is provisionally registered for a period of time, it automatically will become a fully registered information system. Under the final rule, once an information system is provisionally registered for 180 days, lenders must furnish to it but cannot rely on reports from it to satisfy their obligations under the final rule until the system has become fully registered, 240 days after the date it was provisionally registered, pursuant to § 1041.11(d)(2). Like the proposal, the final rule provides 60 days for lenders to furnish to a provisionally registered information system before it becomes a fully registered information system. The Bureau also extended the time frames associated with the registered information systems to which information must be furnished. The proposed rule would require lenders to furnish to each information system that, as of the date of consummation of the loan, had been 1150 registered with the Bureau pursuant to § 1041.17(c)(2) for 120 days or more, or had been provisionally registered with the Bureau pursuant to § 1041.17(d)(1) for 120 days or more, or subsequently had become registered with the Bureau pursuant to § 1041.17(d)(2). The Bureau is extending these 120-day time frames to 180 days under final § 1041.10(b)(1) in order to allow additional time for provisionally registered and registered information systems to “onboard” lenders. Similarly, as noted above, the Bureau is extending the implementation period for all provisions of the rule except § 1041.11, concerning registration of information systems, from 15 to 21 months. The obligation, therefore, to start furnishing information to registered information systems pursuant to § 1041.10 will not take effect until 21 months after publication in the Federal Register. This extension will allow for additional time to register information systems and additional time for lenders to onboard to registered information systems before the effective date of the rest of the rule. The Bureau is extending the deadline to submit an application for preliminary approval for registration by 60 days in response to comments raising concerns about time needed to prepare such applications, but § 1041.11 will become effective 60 days after publication of the final rule in the Federal Register, as proposed. The Bureau is not modifying the procedures for registration on or after the effective date of the rule. If no entity is registered as an information system under § 1041.11 sufficiently in advance of the effective date of § 1041.10 so as to allow furnishing to begin as of that date, lenders will not be able to make a loan under § 1041.6 until such furnishing begins, as explained in comment 1041.6(a)-2. Lenders will be able to make loans under § 1041.5 in the event that no entity is registered as an information 1151 system under § 1041.11 or registered sufficiently in advance of the effective date of § 1041.10 so as to allow furnishing to begin as of that date. 10(a) Loans subject to furnishing requirement Proposed Rule In proposed § 1041.16(a), the Bureau proposed to require lenders making most types of covered loans to furnish to each information system described in proposed § 1041.16(b) the information concerning the loans as described in proposed § 1041.16(c). As described in the proposal, the purpose of the furnishing requirement was to enable a registered information system to generate a consumer report containing relevant information about the consumer’s borrowing history, regardless of which lender made a covered loan to the consumer previously. The Bureau believed that requiring lenders to furnish information about most covered loans would help achieve this result and, accordingly, help fulfill the consumer protections of proposed part 1041. The Bureau also stated that the development of common data standards across registered information systems would benefit lenders and registered information systems, and that the Bureau intended to foster the development of such common data standards where possible to minimize burdens on furnishers. Comments Received The Bureau received a wide range of comments about the furnishing requirements proposed under § 1041.16. Some comments supported the proposal to subject covered shortterm loans and covered longer-term loans to the furnishing requirements. A consumer reporting agency stated that the proposal would allow the registered information systems to collect more 1152 comprehensive credit information on consumers who sought covered loans. Likewise, various commenters—including a consumer reporting agency, two consumer advocates, a credit union, and another industry commenter—approved of the proposed registered information systems and the requirement that lenders furnish information concerning consumers’ borrowing histories. Consumer groups and others maintained that mandating the furnishing of information to registered information systems was critical to enabling compliance with the proposed regulation, including the restrictions on rollover transactions, back-to-back loans, and re-borrowing within a short period after paying off a prior loan. One industry commenter wrote that the furnishing requirements could potentially have a positive impact on consumers who make regular payments by helping them gain greater access to other types of credit. Another agreed with the Bureau’s proposed furnishing requirements, but stated it would be difficult to implement in a timely manner the requirements for the registered information systems, which it considered burdensome. Several commenters opposed either mandating the proposed furnishing requirements altogether, or suggested that the rule should only require certain kinds of lenders to furnish. Several commenters requested that the rule not require credit unions and other lenders to furnish to registered information systems at all, suggesting that their current furnishing to consumer reporting agencies is sufficient. Other commenters representing credit unions and auto lenders objected to the furnishing requirements on the basis that they do not generally furnish information to, or obtain information from, consumer reporting agencies. One consumer reporting agency contended that mandatory furnishing would stifle innovation among registered information systems, including among some specialty consumer reporting agencies, by 1153 diminishing their incentives to develop better risk-management products and services, which in turn would likely reduce the quality of products and services. A trade association asserted that the furnishing provisions were overly prescriptive and disproportionate to any consumer benefit. One industry commenter asked the Bureau to consider restricting access to any registered information system to properly licensed lenders, citing Statelicensed lenders as an example, to ensure that lenders were properly licensed in the State in which a consumer resided. Another group of commenters generally argued that the registered information requirements, including the furnishing provisions, would impose costs that would prevent lenders from providing small-dollar loans. Commenters criticized the furnishing requirements for other reasons. One anticipated that lenders would not comply with the furnishing requirements, including what they understood to be the obligation to furnish information in real time, and warned of the compliance risk this would create for lenders. A trade association noted that the furnishing requirements could have a negative effect on Veritec’s systems, which it thought are currently in use by most States that track payday loans. This commenter asserted that the proposal was silent on mechanisms to independently verify and secure the confidentiality of the data in the registered information systems. Other commenters expressed concerns about the monetary, operational, and accessrelated burdens imposed by the furnishing requirements. One State government entity anticipated that the costs of creating the infrastructure related to the furnishing requirements would be passed on to consumers in the form of higher costs for obtaining small-dollar loans. A number of industry commenters stressed the impact that the requirements would have on lenders 1154 such as online lenders and other small-volume lenders, especially additional costs and burdens. Another argued that larger lending entities would be at a competitive advantage because the scale of their operations would allow them to spread the costs of integration more easily. At least two of the industry commenters argued that the provisions related to the registered information systems would make it less profitable for banks and most credit unions to make small-dollar loans. One cited the high costs of investing in systems with furnishing capabilities and obtaining reports from registered information systems. Another claimed that obtaining consumer reports would increase the expense of making small-dollar loans for community banks, and that small-volume lenders would have to pay more for such reports than other lenders. One industry commenter stated that for lenders, the costs of hiring and training staff, along with the operational risks associated with data security and data integrity, would be significant. An industry commenter and a Tribal-entity commenter identified as burdensome the requirement to report information at various stages in the life of a covered loan. One commenter observed that many lending entities with Tribal affiliation have limited access to consumer reporting agencies, and could be unable to comply with the rule if registered information systems refused to work with them, unless the Bureau took action to address the problem. The Tribalentity commenter also asserted that satisfying the furnishing requirements would be more challenging for Tribes. Some commenters recommended changes that they thought would facilitate the implementation of the furnishing requirements. One trade association proposed that lenders only be required to furnish information on a monthly basis. A trade association whose membership 1155 includes vehicle title lenders commented that the Bureau should permit such lenders to comply with a simplified alternative process in lieu of the proposed furnishing requirements. Some commenters expressed concern about the impact of the furnishing requirements on the availability and cost of credit. One conveyed the importance of enabling consumers to build credit while they rely on covered short-term loans. This commenter suggested that the final rule should prohibit the use of furnished information to harm the score or profiles of less financially capable borrowers. One trade association speculated that the proposed rule could greatly restrict the availability of credit by discouraging community banks and other depository lenders from developing small-dollar lending programs and providing small-dollar loans as an accommodation to existing customers. This commenter asserted that restricted credit availability could fuel the growth of unlawful offshore lending from individuals and entities that are difficult to identify or regulate. An industry commenter stated that the registered information system framework creates a unique category of non-prime consumer reporting agencies, which the commenter cautioned could prevent consumers from accruing the credit benefits that result when lenders furnish repayment information to mainstream consumer reporting agencies. One trade association stated that without an overhaul of the existing credit reporting structure, the proposal would dramatically increase the potential for errors and inaccuracies on consumer credit reports, and thereby decrease access to credit for consumers with negative or insufficient credit history. Final Rule As explained below, the Bureau is adopting § 1041.10(a) (as renumbered from proposed § 1041.16(a) for the reasons discussed earlier) with the following modifications. The proposal’s coverage regarding the furnishing requirements included each covered loan, except covered 1156 loans made pursuant to proposed § 1041.11 or § 1041.12. Because proposed §§ 1041.11 and 1041.12 are not included in the final rule, as discussed above, the final rule no longer references loans made pursuant to those proposed provisions and thus, the Bureau has deleted the phrase “other than a covered loan that is made under § 1041.11 or § 1041.12.” Further, the final rule clarifies that a lender must furnish not for “each covered loan” as proposed but rather for “each short-term and covered balloon-payment loan” under the final rule. Thus the scope of the furnishing requirement is narrower than proposed and excludes a requirement that lenders furnish information regarding covered longer-term loans. The Bureau concluded that excluding such loans from the furnishing requirements would lessen the burden on lenders, especially in terms of the requirements to update loan information. Although this may create a gap in the information in the registered information systems to the extent an applicant has a prior or outstanding covered longer-term loan, lenders will still need to consider other sources of information concerning covered longer-term loans when performing the ability-to-repay analysis required by § 1041.5, as discussed in that section. Proposed comment 1041.16-1 is not adopted in the final rule because it pertained to proposed §§ 1041.11 and 1041.12 and the conditional exceptions to longer-term loans, which the Bureau is not adopting in the final rule. The Bureau is including in the final rule two new comments to § 1041.10(a). The first comment explains the application of the furnishing requirements to rollover loans. Comment 1041.10(a)-1 was added to align with the treatment of rollovers in comments 5(d)-2, 6(b)(1)-3, 6(b)(1)-4 and 6(c)(2)-1, and provide greater clarity regarding their treatment in the context of the furnishing requirements in § 1041.10(a). In sum, it clarifies that if a State permits lenders to rollover (or renew) covered short-term loans or longer1157 term balloon payment loans, then the rollover or renewal loan must be treated as a new loan for the purposes of the furnishing requirements in § 1041.16(a). It further offers an example that illustrates that if a lender rolls over a covered short-term loan, as allowed by State law, after determining that the consumer has the ability to repay the loan, then the lender must report the original loan as no longer outstanding and report the rollover as a new covered loan. Final comment 1041.10(a)-2 pertains to lenders’ furnishing through third parties. The Bureau added this comment in order to address concerns raised by commenters about the potential that, under the proposed rule, lenders may be required to furnish to multiple registered information systems with different interfaces and data standards. The comment clarifies that a lender may furnish information to a registered information system directly or through a third party acting on its behalf, including a registered information system. Accordingly, a lender could enter into an arrangement with one registered information system to allow that registered information system to furnish the lender’s information to the other registered information systems on its behalf. Under such an arrangement, the lender would not have to furnish to multiple registered information systems—it would furnish to just one. The Bureau anticipates that some registered information systems will provide such services to lenders. Accordingly, it included comment 1041.10(a)-2 in the final rule to clarify that direct furnishing to registered and provisionally registered information systems by lenders is not necessary, and to encourage registered information systems and service providers to provide services to reduce the potential challenges of a variety of different interfaces and data standards. As noted below, however, the Bureau anticipates that the market will create incentives for registered information systems to develop common data standards and interfaces. 1158 The Bureau declines to eliminate the proposed mandatory furnishing obligation, as some commenters suggested. As many other commenters recognized, the proposed furnishing requirement is important to allow the underwriting and other provisions in the rule to function properly. The Bureau believes that lenders making covered loans will benefit significantly from comprehensive information about the consumer’s recent borrowing history with respect to covered loans when making a reasonable assessment of a consumer’s ability-to-repay. Generally, lenders either do not furnish information regarding loans that will be covered under this rule at all or furnish information about such loans to specialty consumer reporting agencies only. The registered information system provisions of the final rule are designed to allow lenders to access information regarding the consumer’s borrowing history concerning short-term and covered longer-term balloon loans, beyond their own records and those of their affiliates. As described above, § 1041.5(d)(2) prohibits lenders from making the fourth loan in a loan sequence of covered short-term loans, covered longer-term balloon-payment loans, or a combination of those types of loans that are made under § 1041.5; and § 1041.5(d)(3) prohibits lenders from making a covered short-term or covered longer-term balloon loan under § 1041.5 concurrently or within a 30-day period following a loan made pursuant to the § 1041.6 conditional exception. To determine whether either prohibition applies to a contemplated loan, § 1041.5(d)(1) of the final rule requires a lender to obtain and review information about a consumer’s borrowing history from its own records, its affiliates’ records, and from a consumer report obtained from a registered information system, if available. These provisions require a cooling-off period of 30 days between the third and fourth loans in a § 1041.5 sequence, and before a consumer borrows a § 1041.5 loan following a § 1041.6 loan. These cooling-off periods are an integral component of 1159 the final rule’s ability-to-repay intervention that the registered information system fosters. Namely, the existence of a registered information system allows the underwriting provisions in the rule to function properly by enabling a lender to see the borrower’s previous and current use of covered short-term loans and covered longer-term balloon loans to determine the borrower’s eligibility for a new covered short-term loan or covered longer-term balloonpayment loan subject to § 1041.5. Importantly, the registered information system will ensure that lenders are aware whether a potential borrower is subject to a cooling-off period. That knowledge also may deter lenders from seeking to enter into referral arrangements to evade the cooling-off period requirements. Without a framework to ensure that information about a potential borrower’s previous and current use of covered short-term loans and covered longerterm balloon loans is provided and collected in an organized and accessible manner, it would be much less likely that the goals of the lending limits, conditions, or restrictions contained in the rule would be achieved. Accordingly, the Bureau continues to believe that furnishing requirements play an important role in ensuring that lenders have the information they need to comply with the rule and achieve the consumer protections that are the goal of this part. As discussed at great length above in Market Concerns—Underwriting, the market for covered short-term loans and covered longer-term balloon-payment loans is one where consumers who take out unaffordable loans confront considerable potential risks and harms. These risks and harms stem from default, delinquency, repeat re-borrowing, and the collateral consequences of having to make unaffordable payments, including forgoing basic living expenses or payments on major financial obligations. The underwriting requirement, that a lender must first make a reasonable assessment of the borrower’s ability to repay the loan 1160 according to its terms, is being imposed in this rule to prevent the identified unfair and abusive practice of failing to engage in such underwriting for such loans. The furnishing requirement is an important component of the approach taken in the final rule to address these harms and protect consumers by preventing the identified unfair and abusive practices, pursuant to the Bureau’s statutory authority to write such rules under section 1031(b) of the Dodd-Frank Act. The furnishing requirements also allow lenders to make loans under final § 1041.6, which provides an exemption from the ability-to-repay determination requirements in final § 1041.5. The information furnished to a registered information system allows lenders to review a consumer’s borrowing history, reflected in a consumer report from the registered information system, to determine the potential loan’s compliance with the requirements of final § 1041.6 (b) and (c). If no entity is registered as an information system or a registered information system has not been registered for a period of at least 180 days on the effective date of § 1041.6, the exemption under § 1041.6 will not be available. The Bureau anticipates that there will be at least one registered information system by the effective date for § 1041.6. The Bureau is not persuaded that requiring furnishing to registered information systems in this rule will exclude borrowers from nationwide consumer reporting agencies, as some commenters asserted. As noted in the proposal, for the most part, lenders currently making loans that would be covered under § 1041.10(a) do not currently furnish information concerning such loans to consumer reporting agencies consistently, if at all. Nothing in the final rule precludes lenders from furnishing to entities other than registered information systems, including nationwide consumer reporting agencies that do not seek to register as registered information systems. 1161 As noted elsewhere, databases, such as Veritec, contract with various States that have statutory caps on short-term loans; these States impose requirements that lenders provide loan information to the databases and check the databases before approving borrowers for loans. Such databases are useful tools in policing State requirements. If any database, including Veritec, were to become a registered information system, it would have to make adjustments to the services it provides to facilitate lenders’ compliance with part 1041’s furnishing requirements. As discussed in the Section 1022(b)(2) Analysis in part VII below, lenders that already report information to databases to comply with State laws will likely face lower costs to come into compliance with the furnishing requirements in § 1041.10. The Bureau expects that provisionally registered and registered information systems will find it in their competitive interests to develop common data standards and interfaces to facilitate accurate and timely reporting. Given the likelihood that standards for data will be established in this market, the Bureau is not persuaded that having more than one provisionally registered or registered information system will negatively impact the accuracy or quality of the data furnished to systems, as some commenters have suggested. As noted elsewhere, the FCRA and Regulation V will impose obligations with respect to data accuracy on lenders furnishing information to provisionally registered and registered information systems and on the information systems themselves. One commenter expressed concern that a registered information system may not “work with” Tribal lenders. However, this commenter did not indicate what it believed the bases for such refusal might be. To be eligible for provisional registration or registration, § 1041.11(b)(3) requires that an entity must perform in a manner that facilitates compliance with and furthers the 1162 purposes of part 1041. This includes facilitating lender compliance with obligations to furnish information to provisionally registered and registered information systems and to obtain consumer reports from registered information systems. The Bureau notes that, as explained in proposed comment 1041.17(b)(3)-1 (finalized as comment 1041.11(b)(3)-1), this requirement does not supersede consumer protection obligations imposed upon a provisionally registered or registered information system by other Federal law or regulation. For example, if receiving data furnished by a particular lender pursuant to this rule, or providing a consumer report to a particular lender pursuant to this rule, would cause a provisionally registered or registered information system to violate a Federal law or regulation, then § 1041.11(b)(3) would not require the provisionally registered or registered information system to do so. However, absent such a circumstance, provisionally registered and registered information systems will be required to receive furnished data and provide consumer reports required under the rule, and to generally perform in a manner that facilitates compliance with and furthers the purposes of part 1041, in order to maintain their eligibility for provisional registration or registration. The Bureau notes that § 1041.11(h) will permit the Bureau to suspend or revoke the provisional registration or registration of an information system that has not satisfied, or no longer satisfies, the eligibility conditions set forth in § 1041.11(b). The Bureau believes that, together, these provisions will ensure that lenders are only denied service by registered information systems for reasons authorized under the rule. The Bureau is not persuaded by the objection that commenters made to applying proposed § 1041.16 to vehicle title lenders. As explained in the proposal and above in Market Concerns—Underwriting, the Bureau has found a recurrence of high re-borrowing and high 1163 default rates among consumers who obtain short-term vehicle loans, which can result in severe harms to many consumers. Therefore, the Bureau remains convinced that it is in the public interest to require lenders that make such loans under § 1041.5 to furnish information to registered information systems pursuant to §§ 1041.10 and 1041.11 of the final rule. With respect to concerns about burdens on lenders associated with the furnishing requirements that some commenters have raised, the Bureau recognized in the proposal and further acknowledges that that the furnishing requirements will result in some added costs to lenders, especially those related to setting up furnishing arrangements with the registered information systems, but continues to believe that these costs are justified by the important benefits of the furnishing requirement. Commenters expressed concern about lenders having to furnish to and set up arrangements with multiple registered information systems. As discussed in greater detail in the Section 1022(b)(2) Analysis, furnishing information to registered information systems will require lenders to incur one-time and ongoing costs, including those associated with establishing a relationship with each registered information system, developing procedures for furnishing the loan data, and developing procedures for compliance with applicable laws. The Bureau also anticipates that lenders will face ongoing costs to furnish the data, although the Bureau estimates that the time costs for lending staff will be modest, particularly if one or more registered information systems or service providers offer a service of providing furnished information to some or all of the other registered information systems on behalf of lenders. The Bureau recognizes, however, that if multiple registered information systems exist and no such service is made available, then lenders will have to incur these costs multiple times. As noted in the proposal and in the Section 1022(b)(2) Analysis, the Bureau will 1164 encourage the development of common data standards for registered information systems in order to reduce the costs of providing data to multiple services where possible. The Bureau recognizes that these additional costs may flow to consumers, though in some cases, lenders may not be able to pass all, or any, of the additional costs on by increasing product pricing, given that many covered short-term loans are already priced at their maximum allowable level under different State laws, as discussed above in part II. For the reasons stated in the proposal, in Market Concerns—Underwriting above, and described herein, the Bureau continues to maintain that the furnishing requirement and related costs are important components of the rule that will assist with effectively addressing the identified unfair and abusive practice of making unaffordable covered loans to consumers without reasonably assessing their ability to repay these loans. Moreover, as stated above, the Bureau expects that the registered information systems will find it in their interests to develop common data standards and interfaces to facilitate accurate and timely reporting. Specifically, if registered information systems take such steps and furnishing becomes more automated over time, it will make compliance with the rule easier and cheaper. In addition, because the rule, as described in the above discussion of comment 1041.10(a)-2, allows a lender to rely on a third party to furnish on behalf of the lender, the Bureau anticipates that registered information systems and other providers will offer services that include furnishing to registered information systems, and will compete to offer such a service. The availability of such a service will mean that lenders can minimize any challenges of furnishing to all of the registered information systems and furnish to one who acts on its behalf to furnish data to the others. The Bureau anticipates that these arrangements will also result in cost-savings. 1165 Nonetheless, the Bureau also notes that the final rule reflects two modifications that are likely to alleviate some of the burden stemming from complying with the furnishing requirement under § 1041.10. First, the Bureau has narrowed the scope of loans required to be furnished under final § 1041.10(a) to exclude covered longer-term loans (other than covered longer-term balloon-payment loans). As a result of this change, lenders will be required to furnish information about fewer loans than would have been required under the proposed rule. Second, as explained further below, the Bureau has also eliminated some of the information that it proposed to require lenders to furnish when a loan ceases to be an outstanding loan. Again, the Bureau anticipates that this modification will reduce burdens for lenders to satisfy their furnishing obligations under § 1041.10 of the final rule. 10(b) Information systems to which information must be furnished 10(b)(1). Proposed Rule Proposed § 1041.16(b)(1) stated that a lender had to furnish the information required in proposed § 1041.16(a) and (c) to each information system registered pursuant to proposed § 1041.17(c)(2) and (d)(2) or provisionally registered pursuant to proposed § 1041.17(d)(1). Proposed comment 16(b)-2 clarified that lenders were not, however, required to furnish information to entities that had received preliminary approval for registration pursuant to § 1041.17(c)(1) but were not registered pursuant to § 1041.17(c)(2). To allow lenders and provisionally registered and registered information systems time to prepare for furnishing to begin, the proposal delayed the furnishing obligation for newly registered and provisionally registered systems by requiring that lenders furnish information about a loan to such systems 1166 only if the system had been provisionally registered or registered for 120 days or more as of the date the loan was consummated. The Bureau believed that this 120-day period would allow lenders sufficient time to prepare for compliance with proposed § 1041.16, while giving provisionally registered or registered information systems sufficient time to onboard all of the lenders required to furnish to the information system. Comments Received Various consumer reporting agencies and consumer advocates approved of the proposal to require lenders to furnish information to each registered information system. An academic commenter stated that a more coordinated reporting of loans across lenders and States could matter in protecting consumers, many of whom had been harmed when they incurred large debts by borrowing from multiple lenders simultaneously. One consumer reporting agency asserted that proposed § 1041.16(b)(1) was a practical solution for the industry. Another claimed that the proposal to have lenders report to each registered information system would improve the industry’s understanding of small-dollar loan usage among consumers and, combined with the data proposed to be furnished, this framework could lead to better and cheaper loan products. A group of consumer advocates urged the Bureau to adopt the requirement that lenders must furnish to each of the registered information systems because, they argued, giving lenders the discretion to furnish to only one registered information system would incentivize the systems to be more responsive to lender concerns than to consumer concerns. These commenters also believed that permitting lenders to furnish to only one registered information system would be more cumbersome because it would be more difficult to guarantee access to a comprehensive borrowing history; doing so either would require lenders to obtain reports from all registered 1167 information systems, or would necessitate all of the registered information systems to complete data-sharing agreements with each other. One industry commenter approved of the proposed rule generally, but recommended that lenders should also be required to register with the Bureau. One consumer reporting agency believed that the proposed approach requiring furnishing to all of the registered information systems was realistic because in its view the industry norm for information furnishing already has creditors furnishing information to multiple nationwide consumer reporting agencies. It advocated for a single platform or gateway to accomplish the “furnish to all” approach, through which lenders would furnish information to each registered information system while being able to obtain the required consumer reports from this same single platform. At least two industry commenters supported the single-platform approach, one of which suggested that the single platform to which the lenders would furnish could coordinate furnishing and dispute resolution with the registered information systems. One consumer reporting agency otherwise in support of the Bureau’s proposal opposed the single-platform approach. This commenter argued that the mechanics of such an approach could not be accomplished on a reasonable timeline, and that such an approach would increase the infrastructure costs for registered information systems. It believed the single-platform approach was likely to be inadequate for other reasons also. This commenter argued that it would be difficult for the Bureau to select the single-platform provider and ascertain reasonable cost for the service. It further submitted that such an approach would reduce competition to improve the performance of the registered information systems, and any service interruption or disruption would affect the entire industry. This commenter suggested that, even with a single platform, lenders may still choose to obtain multiple reports to obtain a comprehensive 1168 understanding of a consumer’s borrowing history, and establishing the contracting requirements for each registered information system would be a complex undertaking. At least two commenters opposed the requirement to furnish to multiple registered information systems altogether. One trade association stated that for lenders, the costs of hiring and training staff, along with the operational risks associated with data security and data integrity, would be significant. One industry commenter echoed that the furnishing provisions were cumbersome, expensive, and presented the risk that inaccurate data would be furnished and that data would be disputed or handled improperly. Citing the potential high costs of compliance, one industry commenter criticized the Bureau’s efforts for not sufficiently researching the impact of this approach on small businesses. Several commenters responded to the Bureau’s request in the proposal for ideas about alternatives to requiring lenders to furnish to each information registered system. One was concerned about the complexity of reporting to multiple systems with unique interfaces, credentialing, and the increased risks of errors. Two credit union commenters encouraged the Bureau to require lenders to furnish to the nationwide consumer reporting agencies only. An industry commenter recommended that, in lieu of the proposed registered information system approach, the Bureau require nationwide consumer reporting agencies to accept information furnished under the rule and share the information with other nationwide consumer reporting agencies. Some nationwide consumer reporting agencies advocated they are in the best position to act as registered information systems. A mix of commenters recommended that the Bureau amend the proposal to allow lenders to furnish to one registered information system, and obtain from the system a merged report that 1169 would contain all the data furnished about the consumer. They noted that this “furnish to one, pull a merged report” approach was akin to the consumer reporting approach that typically is used in mortgage and certain other credit markets. A consumer reporting agency suggested that in order to enable the “merge report” concept to work, the Bureau would need to require each registered information system to agree to provide to other registered information systems, upon request, any furnished data concerning a loan applicant. One trade association and another industry commenter favored a single, nationwide registered information system hosted by the Bureau or its contractor. A commenter with the capability to develop such a database asserted that this approach would create a unitary set of standards for data capture and electronic communication, while providing lenders with a single provider for assistance. This commenter stated that other advantages of a singular system included minimized costs and burdens for furnishing and maintaining information, increased compliance from lenders, improved regulatory oversight of lenders and the registered information system by the Bureau, more restricted access to the database and corresponding privacy protections for consumers, increased accuracy and consistency for both consumer and product data, reduced costs on the basis of scale, faster implementation, and improved ability to innovate and adapt to regulatory change. A group of consumer advocates also supported a single registered information system on the condition that the Bureau consider housing the database either itself or with a contractor hired by the Bureau. These commenters believed this approach would improve protections for consumers while generating fewer data errors. One trade association listed as precedents for this approach the sanctions list hosted by the Department of Treasury’s Office of Foreign Assets 1170 Control, and the list of active-duty servicemembers that the Department of Defense has developed to help implement the Servicemembers Civil Relief Act and the Military Lending Act. Other commenters noted the experience of the 14 States that have State-mandated databases containing information about short-term, small-dollar loans. Commenters said that most of those regulatory regimes include a sole source contract with a single State-selected contractor that collects and discloses limited information about eligibility to lenders seeking to make loans. Some commenters noted that these systems lack market incentives to increase value and service while reducing costs and that the system as proposed by the Bureau will lead to better, less expensive products for lenders. Some commenters pointed to those State-mandated databases as success stories in terms of efficiencies and noted the experiences of two States that started out with multi-database reporting systems but, because of the challenges associated with such an approach, ultimately developed a single database reporting system. One commenter noted that there were at least nine firms that would have the technical capability to act as registered information systems. Several noted that consistent data standards should be established, with many recommending the Metro 2 format but with others requesting that no standard be established. As described above, the Bureau also received numerous comments about the amount of time provided under the proposed rule for lenders to onboard to registered information systems. Proposed § 1041.16(b)(2) provided that a lender must furnish information as required in paragraphs (a) and (c) to each information system that, as of the date the loan is consummated: had been registered with the Bureau pursuant to § 1041.11(c)(2) for 120 days or more; or had been provisionally registered with the Bureau pursuant to § 1041.11(d)(1) for 120 days or more 1171 or subsequently had become registered with the Bureau pursuant to § 1041.11(d)(2). This would have provided lenders with 120 days to onboard to a provisionally registered information system and an information system registered before the effective date of § 1041.10 and prepare to furnish. At least two consumer reporting agencies suggested that they could onboard all covered lenders within this proposed time frame. Referring to the process of credentialing and onboarding potential furnishers, one consumer reporting agency estimated that it could onboard the lenders in a matter of months with the appropriate technical expertise and support. Another consumer reporting agency estimated that in its current capacity as a consumer reporting agency, credentialing and onboarding a new lender could take the commenter around four weeks. However, the commenter cautioned that if more extensive requirements than were proposed were included in the final rule, including additional or longer data fields, or a requirement to furnish using a data standard other than Metro 2, it could take longer to implement. Several commenters argued that the 120-day period would be insufficient to permit onboarding of all lenders to all registered information systems. One industry commenter cautioned that the proposed timeline did not appear to contemplate the burdens lenders could face while working with the unique onboarding requirements of each registered information system. One commenter argued that the Bureau was underestimating the effort and time required to enroll and onboard lenders, and speculated that it would take years to implement the proposed furnishing provisions. It noted that the onboarding process at registered information systems could be unique because of variations in technology platforms, interfaces, and reporting formats. Additionally, this commenter explained that storefront lenders could face more 1172 difficulties than online lenders in integrating with consumer reporting agencies, which could delay such lenders’ ability to onboard to a registered information system. Final Rule The Bureau has reviewed and analyzed the comments, and now adopts (renumbered) § 1041.10(b)(1) to require that a lender furnish the information as required in § 1041.10(a) and (c) to each information system registered pursuant to (renumbered) § 1041.11(c)(2) and (d)(2), and to provisionally registered information systems pursuant to § 1041.11(d)(1), as proposed. Of note, final §§ 1041.5 and 1041.6 require lenders to obtain a report from only one registered information system, also as proposed. The Bureau is responding to commenters that suggested extending the 120-day time period registered information systems need to be registered or provisionally registered before the furnishing requirements are applicable (onboarding period) by extending the onboarding period by 60 days. The final rule sets the onboarding period at 180 days. Other changes to the rule text reflect the renumbering from the proposal to the final rule. Likewise, comment 1041.10(b)-1 is modified from the proposal to reflect the final rule’s renumbering and adoption of the 180 day time frame described above. The illustrative example contained in the comment is also updated to reflect that lenders are not required to furnish to an information system that was provisionally registered 179 days before a loan was consummated. Comment 1041.10(b)-2 is likewise altered to reflect the final rule’s renumbering. Commenters expressed concerns regarding the potential for inconsistencies in the furnished data and potential burdens on lenders they anticipated as a result of the proposal’s requirement that lenders furnish to multiple registered information systems. Some commenters suggested that the Bureau register only one information system under proposed § 1041.17 while 1173 others suggested that the Bureau contract with a single provider or house the system within the Bureau. The Bureau recognizes that a single registered information system approach—whether administered by the Bureau, its contractor, or another entity— may provide benefits in terms of the uniformity and consistency of data and the expenditure of fewer lender resources initially, as lenders would not have to furnish to multiple systems. However, there are also risks to a single registered information system approach. With respect to the suggestion that the Bureau house information concerning covered loans itself or through the use of a contractor, it continues to believe that the private sector is better equipped to implement the requirements for registered information systems in a timely manner. The Bureau also continues to believe that there may be legal or significant practical obstacles to the Bureau contracting with or maintaining the single system. Further, the Bureau is concerned that, if it registered only one information system where more than one entity has applied to be a registered information system and satisfies the eligibility requirements, the single registered information system would likely lack the market incentives to increase value and service while reducing costs on lenders. The Bureau is thus convinced that registering a single information system where others are available would stifle innovation and, as some commenters noted, competition to improve the performance of the registered information system. The Bureau is confident that the market will adequately respond to challenges that may arise in connection with the final rule’s furnish to all approach, and has determined that this approach is better than the single registered information system approach some commenters have suggested. Some commenters suggested that the Bureau establish common data standards or require the use of an existing credit reporting standard. The Bureau decided not to create or require a 1174 particular data standard. As described above, the Bureau concluded that the market will provide incentives for the development of appropriate data standards. The Bureau is concerned that requiring the use of a specific data standard would stifle innovation. The Bureau believes that registered information systems will be incentivized to work together to develop common data standards and create efficiencies, especially in light of the ability of registered information systems or service providers, clarified under the final rule, to furnish information on behalf of lenders. As noted in the proposal, the Bureau intends to help foster the development of such coordinated data standards. Some commenters advocated for an alternative that would require lenders only to furnish to one of the registered information systems and to obtain a “merged” report from only one registered information system. In order to facilitate that approach, commenters recommended that the Bureau require each registered information system to agree to provide information in its system concerning a specific loan applicant to each other registered information system in response to a request for such information and that each agree to charge no more than a reasonable fee for doing so. The Bureau chose not to pursue that alternative for a variety of reasons. The Bureau is particularly concerned that if lenders only furnished to one of the registered information systems, the unique data that rest at a particular registered information system would be unavailable to other lenders if the registered information system experienced a problem, such as temporary system outage, or had its registration revoked. However, if lenders are obligated to furnish to all registered information systems, then an outage or revocation at one registered information system would not impact the comprehensiveness of the consumer report provided to a lender by any other registered information system pursuant to the rule. In addition, 1175 an approach that relied on registered information systems sharing unique information to produce a merged report could create incentives for individual registered information systems to leverage their (perhaps limited) data to extract a high price from other registered information systems for access. Although the imposition of a limitation on what a registered information system may charge another registered information system for data could ameliorate that concern, the Bureau ultimately concluded that it did not want to engage in the policing of pricing practices of registered information systems related to the sale of data and, overall, the furnish to all requirement reflected in the final rule is the better approach. Other commenters suggested another approach as an alternative that would involve reporting to all systems, but would also entail a centralized gateway or platform through which lenders could furnish. Some noted that some specialty consumer reporting agencies currently provide such a service. The Bureau believes that there is no need to mandate the creation of such a platform or gateway. If there is a demand for such a service, the Bureau believes the registered information systems or other market actors will respond to the demand. Commenters encouraged the Bureau to require lenders to furnish to the nationwide consumer reporting agencies and to require such consumer reporting agencies to accept the information furnished under the rule. Based on its market outreach and experience, as well as the comments it received, the Bureau believes that there are firms capable of taking on the task of acting as a registered information system under the final rule. Accordingly, the Bureau has concluded that it is more appropriate to grant players in the market who satisfy the eligibility criteria set forth in § 1041.11 the choice of whether to become a registered information system. 1176 Nothing precludes nationwide consumer reporting agencies from seeking to become registered information systems, and the Bureau would welcome their participation in this area. Many commenters expressed concern about the length of time allotted in the proposal for registered information systems to onboard lenders. Under the proposal, lenders would be required to furnish to registered information systems that had been registered with the Bureau pursuant to § 1041.17(c)(2) for 120 days or more, or had been provisionally registered with the Bureau pursuant to § 1041.17(d)(1) for 120 days or more or subsequently had become registered with the Bureau pursuant to § 1041.17(d)(2). Commenters noted that the amount of time it would take for registered information systems to onboard lenders could be significant. One suggested that from its experience, it could even take years to onboard all of the lenders that would be required to furnish under the proposal. Others anticipated that the process would only take several months. The Bureau attempted to balance these concerns against the need for the systems to be operational as soon as possible so as to permit timely implementation of the rule. Accordingly, in the final rule, the Bureau is extending the onboarding period by 60 days, such that a lender now has 180 days to onboard to a provisionally registered information system and an information system registered pursuant to 11(c)(2). However, depending on how far in advance of the effective date of the furnishing obligations information systems are registered, the onboarding period for information systems registered pursuant to § 1041.11(c)(2) could exceed 180 days. For example, if an information system is registered 210 days before the effective date, then lenders will have 210 days to onboard to that registered information system before they are required to furnish to it. No lender would be obligated to start furnishing before the effective date of § 1041.10. The Bureau concludes that the revised time frame provides sufficient time for 1177 lenders to onboard and prepare to furnish, and for registered or provisionally registered information system to prepare to receive, information pursuant to §§ 1041.10 and 1041.11 of the final rule. 10(b)(2). Proposed Rule Proposed § 1041.16(b)(2) would require the Bureau to publish on its website and in the Federal Register notice of the provisional registration of an information system pursuant to proposed § 1041.17(d)(1), registration of an information system pursuant to proposed § 1041.17(c)(2) or (d)(2), and suspension or revocation of the provisional registration or registration of an information system pursuant to proposed § 1041.17(g). Proposed § 1041.16(b)(2) provided that, for purposes of proposed § 1041.16(b)(1), an information system was provisionally registered or registered, and its provisional registration or registration suspended or revoked, on the date that the Bureau published notice of such provisional registration, registration, suspension, or revocation on its website. Proposed § 1041.16(b)(2) further required the Bureau to maintain on its website a current list of information systems provisionally registered pursuant to § 1041.17(d)(1) and registered pursuant to § 1041.17(c)(2) and (d)(2). Under the proposal, the date that a particular information system becomes provisionally registered pursuant to proposed § 1041.17(d)(1) or registered pursuant to proposed § 1041.17(c)(2) is the date that would trigger the 120-day period at the end of which lenders would be obligated to furnish information to that particular registered information system pursuant to proposed § 1041.16. The general furnishing requirement would commence at the 1178 effective date of proposed § 1041.16, namely, 15 months from publication of the final rule in the Federal Register. An information system’s automatic change from being provisionally registered pursuant to proposed § 1041.17(d)(1) to being registered pursuant to proposed § 1041.17(d)(2) would not have triggered an additional obligation on the part of a lender; rather the significance of the full registration of a provisionally registered system was that lenders could, once fully registered, rely on a consumer report from the system to comply with their obligations under proposed §§ 1041.5 and 1041.7.1076 Under the proposal, suspension or revocation of an entity’s provisional registration or registration pursuant to proposed § 1041.16(g) would relieve lenders of their obligation to furnish information to the information system pursuant to proposed § 1041.16 and lenders would no longer be permitted to rely on a consumer report generated by the entity to comply with their obligations under proposed §§ 1041.5 and 1041.7. The Bureau believed that publication of a notice on its website would be the most effective way to ensure that lenders received notice of an information system’s provisional registration or registration, or of a suspension or revocation of its provisional registration or registration. Accordingly, for purposes of proposed § 1041.16(b)(1),1077 the Bureau proposed to tie the dates of provisional registration, registration, and suspension or revocation of provisional registration or registration, as applicable, to publication of a notice on its website. The proposal also would have required the Bureau to maintain on its website a current list of information 1076 The proposal required lenders to furnish to such a system beginning 120 days from the date of the system’s provisional registration and to continue to do so after the system becomes registered. 1077 For purposes of proposed §§ 1041.5 and 1041.7, which would require a lender to obtain a consumer report from a registered information system, the Bureau proposed that a suspension or revocation of registration would be effective five days after the Bureau published notice of the suspension or revocation on its website. 1179 systems that were registered pursuant to § 1041.17(c)(2) and (d)(2) and provisionally registered pursuant to § 1041.17(d)(1). Final Rule The Bureau did not receive any comments addressing this provision. The Bureau has added language to clarify that, if it suspends the provisional registration or registration of an information system, it will provide instructions to lenders concerning the scope and terms of such suspension. For example, depending on the facts and circumstances of a particular determination that suspension is appropriate, the Bureau may suspend registration of a provisionally registered information system or registered information system for purposes of final §§ 1041.5 and 1041.6 but still require lenders to furnish to the suspended system pursuant to § 1041.10. The Bureau may also determine that suspension is only appropriate for a certain period of time. Other than those clarifications, the Bureau is finalizing this provision substantially as proposed except that it is renumbering it as § 1041.10(b)(2). 10(c) Information to be furnished Proposed Rule Proposed § 1041.16(c) would have identified the information a lender had to furnish concerning each covered loan as required by proposed § 1041.16(a) and (b). This provision would require lenders to furnish information when the loan was consummated and again when it ceased to be an outstanding loan. If there was any update to information previously furnished pursuant to proposed § 1041.16 while the loan was outstanding, then proposed § 1041.16(c)(2) required lenders to furnish the update within a reasonable period of the event that caused the information previously furnished to be out of date. However, the proposal did not require a 1180 lender to furnish an update to reflect that a payment was made unless the payment caused the loan to cease to be outstanding. A lender was only required to furnish an update if such payment caused information previously furnished to be out of date. Proposed § 1041.16(c)(1) and (3) required lenders to furnish information no later than the date of consummation, or the date the loan ceased to be outstanding, as applicable, or as close in time as feasible to the applicable date. Proposed comment 1041.16(c)-1 clarified that under proposed § 1041.16(c)(1) and (3), if it was feasible to report on the applicable date, then the applicable date was the date by which the information had to be furnished. Under the proposal, the Bureau would have encouraged lenders to furnish information concerning covered loans on a real-time basis, but permitted lenders to furnish the required information on a daily basis or as close in time to consummation as feasible. Proposed § 1041.16(c) also stated that a lender had to furnish the required information in a format acceptable to each information system to which it was required to furnish information. This requirement was complemented by proposed § 1041.17(b)(1), discussed further below, which conditioned an entity’s eligibility for provisional registration or registration as an information system on its capability to use reasonable data standards that would facilitate the timely and accurate transmission and processing of information in a manner that would not impose unreasonable costs or burdens on lenders.1078 Final Rule The introductory paragraph of § 1041.10(c) of the final rule is being finalized as proposed (aside from being renumbered), and comments directed at the substance of this 1078 Among other things, these standards had to facilitate lender and registered information system compliance with the provisions of the FCRA and its implementing regulations concerning the accuracy of information furnished. 1181 provision are addressed in the analysis for § 1041.10 (c)(1) to (c)(3) below. The introductory paragraph summarizes the main thrust of § 1041.10(c), which addresses what information must be furnished with respect to covered loans as required in § 1041.10(a) and (b), and when it must be furnished. It also specifies that a lender must furnish the information in a format acceptable to each information system to which it must furnish information. 10(c)(1) Information to be furnished at loan consummation Proposed Rule Proposed § 1041.16(c)(1) would have required that at the time a loan was made, or as close in time as feasible to that date, lenders must furnish eight pieces of information about the loan to each registered and provisionally registered information system. The specified pieces of information would be as follows: Proposed § 1041.16(c)(1)(i) would have required information that is necessary to uniquely identify the covered loan. This would likely be the loan number assigned to the loan by the lender, but the proposal deferred to lenders and provisionally registered and registered information systems to determine what information is necessary or appropriate for this purpose. Proposed § 1041.16(c)(1)(ii) would have required information necessary to identify the specific consumer(s) responsible for the loan. The proposal deferred to each provisionally registered and registered information system the determination of the specific items of identifying information necessary for this purpose. Proposed § 1041.16(c)(1)(iii) would have required information stating whether the loan was a covered short-term loan, a covered longer-term loan, or a covered longer-term balloonpayment loan, as those terms were defined in proposed § 1041.2. Proposed comment 1182 1041.16(c)(1)-1 would clarify that compliance with proposed § 1041.16(c)(1)(iii) required a lender to identify the covered loan as one of these types of loans, and provided an example. Proposed § 1041.16(c)(1)(iv) would have required information concerning whether the loan was made under proposed §§ 1041.5 or 1041.7, as applicable. Proposed comment 1041.16(c)(1)-2 would clarify that compliance with proposed § 1041.16(c)(1)(iv) required a lender to identify the covered loan as made under one of these sections, and provided an example. Proposed § 1041.16(c)(1)(v) would require the furnishing of information about the loan consummation date for a covered short-term loan. Proposed § 1041.16(c)(1)(vi) would require the furnishing of information about the principal amount borrowed for a loan made under proposed § 1041.7. Proposed § 1041.16(c)(1)(vii) would require the furnishing of the following information about a loan that is closed-end credit: (a) the fact that the loan is closed-end credit, (b) the date that each payment on the loan is due, and (c) the amount due on each payment date. This information was intended to reflect the amount and timing of payments due under the terms of the loan as of the loan’s consummation. Proposed comment 1041.16(c)(2)-1 explained that, for example, if a consumer made a payment on a closed-end loan as agreed and the loan was not modified to change the dates or amounts of future payments on the loan, then the lender was not required to furnish an update to information previously furnished. If, however, the lender extended the term of the loan, then the lender would be required to furnish an update to the date that each payment on the loan was due and the amount due on each payment date, to reflect the updated payment dates and amounts. 1183 Finally, proposed § 1041.16(c)(1)(viii) would require the furnishing of the following information for a loan that is open-end credit: (a) the fact that the loan is open-end credit, (b) the credit limit on the loan, (c) the date that each payment on the loan is due, and (d) the minimum amount due on each payment date. As discussed further below, lenders would be required to furnish an update to information previously furnished within a reasonable period after the event that caused the prior information to be out of date. Comments Received As noted above, the proposal required lenders to furnish the information no later than the date on which the loan was consummated or as close as feasible to the date the loan was consummated. Several commenters opposed what they deemed the “real-time” furnishing requirement of proposed § 1041.16(c). Other commenters recognized that the Bureau was not requiring real-time furnishing and advocated that the Bureau adopt such a requirement as a reasonable means of ensuring compliance. One trade association suggested that some lenders would not comply with the furnishing requirements on a real-time basis, if at all. Several commenters said this requirement would add costs and operational complexity that would hinder lenders from providing small-dollar credit. One consumer reporting agency expressed concern that without a system to facilitate the sharing of the updated account information between the registered information systems, correcting a consumer report across all registered information systems would involve substantial burden and expense. A commenter also asserted that potential lags in the timing of furnishing to a registered information system could result in a “window of invisibility” with respect to a consumer report produced by the registered information system. For example, if a consumer 1184 secured a loan from a lender but the lender did not furnish information about the loan to a registered information system until later that day, then the loan would not be reflected in a consumer report obtained from that registered information system by another lender immediately after the loan was made, and therefore would be invisible to the second lender unless the loan was made by an affiliate of that lender. This commenter also appeared to suggest that if a loan was furnished to registered information systems after the disbursement of funds, then the potential window of invisibility would be shorter for storefront lenders as these lenders disburse funds immediately, and longer for online lenders as these lenders may have a lag period between the loan’s approval and the disbursement of funds. The commenter expressed concern that a consumer could obtain multiple loan approvals during this window of invisibility. Relatedly, several commenters requested a safe harbor from liability to account for circumstances in which a lender checks a registered information system and finds no outstanding loan, but later discovers that a borrower did have another covered loan outstanding. The Bureau has addressed these concerns in comments 1041.5(c)(2)(ii)(B)-3 and 1041.6(a)-3, as discussed in more detail below. A set of consumer advocates generally supported the elements of proposed § 1041.16(c) but urged the Bureau also to require lenders to report more information, such as the all-in APR at consummation and a summary of collection efforts. They also suggested that whether a loan is short-term or long-term should be supported by the underlying information, such as the loan’s date of consummation, due date, and amount and timing of payment, rather than by merely checking a box. Several commenters criticized the Bureau’s inclusion in proposed § 1041.16(c)(1) of the phrase “as close in time as feasible to the date the loan is consummated.” Consumer advocates urged the Bureau to remove the above phrase to ensure the timelier 1185 furnishing of data, which would improve the determinations made by lenders considering consumer reports from registered information systems when making a covered loan. An industry commenter stated that this standard would thwart the provisions of the proposed rule that were intended to prevent repeat borrowing. Focusing on proposed § 1041.16(c)(1)(i), an industry commenter suggested that the unique loan identifier should be consistent across all lenders and registered information systems. This commenter contended that the lack of a unique loan identifier would create substantial issues related to preserving data integrity with respect to data furnished under proposed § 1041.16. With respect to proposed § 1041.16(c)(1)(ii), a group of consumer advocates urged the Bureau to require lenders to furnish the borrower’s full name, address, phone number, date of birth, and all nine digits of the borrower’s Social Security number. They further requested that the Bureau mandate a set of strict matching criteria to be used to properly match borrowers to the correct file at a registered information system. The commenters suggested this was essential to protect consumers against the risk of “mixed files” (i.e., the inclusion, in a consumer report concerning one consumer, of information concerning another consumer). One industry commenter noted that proposed § 1041.16(c)(1)(ii) would create a Federal mandate for Statelicensed providers to furnish personally identifying information that is otherwise protected under several State laws. It also stated that the Bureau should combine proposed § 1041.16(c)(1)(iii) and (c)(1)(iv) together in the final rule. Regarding proposed § 1041.16(c)(1)(v), a group of consumer advocates suggested that the Bureau require the loan consummation date for all loans required to be furnished, not just for 1186 covered short-term loans. They also urged the Bureau to modify proposed § 1041.16(c)(1)(vi) to require that the principal amount borrowed for all loans be furnished, not just for loans made under proposed § 1041.7. Similarly, an industry commenter suggested that this requirement should be extended to all loans made under proposed §§ 1041.5 and 1041.7. A group of consumer advocates supported proposed § 1041.16(c)(1)(vii) and (viii), but urged the Bureau to require lenders to report at the time these loans are consummated the loan consummation date, the total number of payments required, and the loan due date. They also noted that lenders should be required to report loans outstanding on the effective date of the furnishing requirements. They believed this addition was critical to limiting a borrower’s days of indebtedness in a 12-month period. An industry commenter stated that lenders should be required to furnish to registered information systems the following additional information to enable compliance. First, the lender should provide information to uniquely identify itself and the store location that issued the loan. The commenter stated that the identifier should be verified to ensure that the lender was actively licensed to conduct business with the borrower in the borrower’s State, but did not specify whether the party responsible for conducting the verification should be the furnisher or the registered information system, and what a registered information system or lender using a consumer report containing such information would do with the information. The same commenter also suggested that lenders should report whether the loan was provided at the physical location of the entity that issued the loan or elsewhere, including electronically. Three consumer reporting agencies commented on the format of the data to be furnished pursuant to proposed § 1041.16. One stated that a robust set of registration requirements— 1187 including mandating a standardized format for furnishing the data required under the rule— would minimize variation and inconsistencies in the consumer reports provided to lenders across different registered information systems. This commenter acknowledged that in the short run, some entities could face challenges in implementing any standardized data format, but argued that this approach would reduce the burden on furnishers and be more efficient in the long run. It argued that requiring use of the Metro 2 format would standardize the small-dollar lending market and ensure greater data integrity and consistency, which it said would benefit both lenders and consumers. Another consumer reporting agency likewise encouraged the Bureau to require uniformity across furnishing formats in order to ensure that lenders are able to furnish accurate, complete, and timely information. Conversely, one consumer reporting agency urged the Bureau to give registered information systems flexibility rather than mandating data furnishing standards in the rule. However, this commenter agreed that a single standard would support consistency. It also said that though developing a uniform data standard would be costly for registered information systems, software companies could help new furnishers comply with Metro 2 standards, which would allow for faster onboarding. It cited Metro 2 as an example of a best practice and stated that this format was a good model for enabling entities to furnish to registered information systems. This commenter said it did not believe lenders pay dues to use Metro 2. Relatedly, this commenter asked the Bureau to stress to lenders the importance of adequate staffing and of designing their furnishing systems with the appropriate speed and quality. It also asked the Bureau to clarify to lenders that registered information systems would not be responsible for deficiencies in the lenders’ furnishing capabilities. 1188 One consumer reporting agency stated that common standards to ensure equal access to data were in the interest of every registered information system, and emphasized the utility of a standardized electronic data reporting format akin to Metro 2, which the commenter believed would decrease operational burdens for lenders. This commenter speculated that, to the extent the industry could leverage the existing Metro 2 infrastructure to develop a standard appropriate for furnishing data required under the rule, the onboarding process would be relatively quick and simple, whereas a registered information system based on a brand-new data furnishing standard would delay the prospective timeline. Final Rule For the reasons set forth herein, the Bureau is finalizing § 1041.10(c)(1) as proposed, with two revisions and as renumbered in light of other structural changes made in the rule. First, the Bureau has removed from § 1041.10(c)(1)(iii) the phrase “a covered longer-term loan,” and from § 1041.10(c)(1)(iv) the corresponding reference to proposed § 1041.9, to reflect that the final rule does not require furnishing of information about covered longer-term loans (other than covered longer-term balloon-payment loans). Second, § 1041.10(c)(1)(v) of the final rule now requires lenders to provide the loan consummation date for covered longer-term balloonpayment loans in addition to covered short-term loans. As discussed in the section-by-section analysis to the proposal, this information will enable a registered information system to generate a consumer report that will allow a lender to determine whether a contemplated loan is part of a loan sequence and the chronology of prior loans within a sequence, which will enable the lender to meet its obligations under final §§ 1041.5 and 1041.6. Because the definition of loan sequence in the final rule includes covered longer-term balloon-payment loans, the Bureau is 1189 requiring lenders to furnish loan consummation date for all covered loans required to be furnished. Accordingly, the Bureau has deleted the phrase “For a covered short-term loan” from proposed § 1041.16(c)(1)(v). The Bureau is making adjustments to comments 1041.10(c)(1)-1 and 1041.10(c)(1)-2, in order to reflect that § 1041.10(c)(1)(iii) and (iv) relate only to covered short-term loans and covered longer-term balloon loans. As finalized, § 1041.10(c)(1) requires lenders to furnish the specified information no later than the date on which the loan is consummated or as close in time as feasible after that date. The Bureau recognized in the proposal, and acknowledges here, that some installment lenders currently furnish loan information to consumer reporting agencies in batches on a periodic basis. However, the Bureau is not persuaded that batch reporting less frequently than daily would provide information sufficiently timely to serve the purposes of this rule. On the contrary, the Bureau maintains that the proposed timing requirement is needed to further the consumer protections envisioned for part 1041. With respect to the concern some commenters stated—that there would be no way to ensure that data furnished and updated by lenders is consistent across all registered information systems because of the possible delays in the availability of loan data from each individual registered information system—the Bureau is aware of the potential for gaps in information. It further agrees that there exists the potential for a window of invisibility for some loans, as the rule does not require true “real-time” furnishing. Instead, it requires that information must be furnished no later than the date on which the loan is consummated, or as close in time as feasible to the date the loan is consummated. The Bureau has weighed the risk of potential gaps in the available information against the burden on lenders of imposing a realtime furnishing requirement. Ultimately, the Bureau concluded that the incremental benefit of a 1190 real-time furnishing requirement would not justify the burden that would result from such a requirement. In the event that lenders exploit timing delays with the intent to evade the requirements of the rule, the Bureau may address the behavior by relying on its anti-evasion authority, as outlined in final § 1041.13. A commenter expressed concerns about consumer disputes not being adequately conveyed to all registered information systems because of concerns about the systems’ ability to communicate with each other. The Bureau notes that the FCRA and Regulation V impose obligations on furnishers to convey corrections to data previously furnished identified by a consumer dispute. The Bureau expects that lenders will comply with their obligations under the FCRA and Regulation V with respect to updating information at each registered information system to which it previously furnished information about a loan. The Bureau recognizes the concern that commenters have expressed about a lender incurring liability for making a covered short-term loan or covered longer-term balloon-payment loan based on an incomplete or inaccurate consumer report obtained from a nationwide consumer reporting agency or registered information system. The Bureau has added commentary to both §§ 1041.5 and 1041.6 to allay such concerns.1079 Relatedly, the Bureau expects that lenders will furnish the specified information no later than the date on which the loan is consummated. It includes the phrase “or as close in time as feasible to the date the loan is consummated” not to undercut this expectation or to create, as some commenters fear, a loophole. The Bureau includes this phrase because it recognizes that there may be certain circumstances under which it may not be feasible to furnish information on 1079 See Comments 1041.5(c)(2)(ii)(B)-3 and 1041.6(a)-3. 1191 the date the loan is consummated, such as the temporary unavailability of a furnishing system. Final comment 1041.10(c)-1, unchanged from the proposal except for numbering changes, clarifies that “if it is feasible to report on a specified date (such as the consummation date), the specified date is the date by which the information must be furnished.” The Bureau concludes that the expectation under the rule regarding the timing of furnishing information regarding consummation is reasonable and clear and thus it declines to remove from proposed § 1041.16(c) the phrase “as close in time as feasible to the date the loan is consummated” and thus adopts § 1041.10(c)(1) as described above. Final Rule § 1041.10(c)(1)(i)-(vii) also sets out the types of information that lenders must furnish at loan consummation. After carefully evaluating the comments it received regarding increasing the number of data points lenders should be required to furnish, the Bureau has decided to adopt § 1041.10(c)(1) as proposed. Regarding proposed § 1041.16(c)(1)(ii), the Bureau weighed the utility of requiring furnishing of more extensive identifying information (e.g., identifying specific consumers responsible for the loan), as suggested by a group of consumer advocates, against the potential burdens on furnishers associated with such a requirement and the potential privacy and data security concerns associated with the collection and furnishing of more identifying information than is necessary, and concluded that the proposed approach strikes the right balance. Under this approach, rather than prescribing specific identifying information that could, in practice, prove to be under-inclusive, over-inclusive, or both, the Bureau instead concludes that it is preferable for individual provisionally registered and registered information systems to identify the identifying information needed to avoid errors. This approach will also ensure that lenders and provisionally 1192 registered and registered information systems collect no more identifying information from applicants and borrowers than is necessary, consistent with best data security practices. Thus, the Bureau defers to each provisionally registered and registered information system concerning the specific items of identifying information they deem necessary to identify the particular consumer responsible for the loan. The Bureau also decided not to modify proposed § 1041.16(c)(1)(vi) to require lenders to furnish the principal amount borrowed for all loans required to be furnished. The proposal required lenders to furnish the principal amount borrowed only for loans made under proposed § 1041.7(b)(1). The express purpose of this requirement was to allow lenders to determine whether a contemplated loan satisfied the limitations on principal amount set in proposed § 1041.7(b)(1). Under the corresponding provision in the final rule (now renumbered as § 1041.6), the lender must first obtain and consider a consumer report from a registered information system to make covered loans under that framework. However, lenders are permitted to make loans pursuant to proposed § 1041.5 without first obtaining a consumer report from a registered information system if such consumer reports are not available because there are no registered information systems, or none have been registered for the required length of time. While a record of the principal amount is crucial to a lender’s review for a loan made under final § 1041.6, it is not essential for registered information systems to collect and provide this information for loans made pursuant to § 1041.5. After carefully considering the potential burdens that the suggested approach would pose on lenders that furnish to registered information systems, the Bureau declines to adopt the additional data points that some commenters recommend requiring from furnishers in § 1041.10(c) of the final rule. The Bureau finds instead 1193 that § 1041.10(c) will provide sufficient information for lenders to make ability-to-repay determinations that can achieve the consumer protections intended in part 1041. The Bureau is also finalizing § 1041.10(c)(1)(vii) and (viii) as proposed, except for numbering adjustments for internal consistency. These provisions outline the specific information required to be furnished depending on whether the loan is closed or open credit. The Bureau continues to believe these data points will assist with ability-to-repay determinations under the final rule. 10(c)(2) Information to be furnished while loan is an outstanding loan Proposed Rule Proposed § 1041.16(c)(2) would have required lenders to furnish, while a loan is an outstanding loan, any update to information previously furnished pursuant to proposed § 1041.16 within a reasonable period of the event that caused the information previously furnished to be out of date. Proposed comment 1041.16(c)(2)-1 provided examples of scenarios under which proposed § 1041.16(c)(2) required a lender to furnish an update to information previously furnished. Proposed comment 1041.16(c)(2)-2 clarified that the update requirement extended to information furnished pursuant to proposed § 1041.16(c)(2). The Bureau believed that each item of information that the proposal required lenders to furnish under § 1041.16(c)(1) strengthened the consumer protections of proposed part 1041. Updates to these items of information could affect a consumer’s eligibility for covered loans under the proposal and, thus, the achievement of those protections. The Bureau concluded that such updates should be reflected in a timely manner on a consumer report that a lender obtains from a registered information system. However, the Bureau also believed that, to the extent 1194 furnishing updates would impose burden on lenders, a more flexible timing requirement was appropriate for furnishing an update. The Bureau thus proposed that when a covered loan was outstanding, lenders had to furnish updates pursuant to proposed § 1041.16(c)(2) within a reasonable period after the event that caused this type of information previously furnished to be out of date. Comments Received One group of commenters supported the proposed requirement that a lender be required to furnish updates regarding any changes to a loan’s due date, payments, and payment amount. However, they urged the Bureau to require furnishing of more information about a loan while it was outstanding, including information about the payments made, principal and charges owed after each payment, the number of days that a borrower was delinquent on a payment, and whether the loan was refinanced or renewed. These commenters stated that if the loan was refinanced or renewed, then the lender should have to report the amount of principal paid down on the original loan at the time of renewal, the amount of principal owed after renewal, and lastly, all the other requirements for a loan at consummation. They believed the proposed additional information would be important to a lender’s ability-to-repay calculation, and would improve compliance with the proposed provisions addressing repeat re-borrowing of longer-term loans. Other commenters recommended that furnishing updates include any changes to balance amount, credit limit, high credit, minimum payment due, actual payment made, past due amount, delinquency status, and all dates associated with those updates. One industry commenter submitted that the lack of a consistent means for loan identification across lenders and registered information systems could create disparities in the 1195 application of updates to borrower loan records. Some commenters expressed concerns about the required frequency of the furnishing updates and that lenders may need to furnish updates more often than once a month because of the short billing cycle for small-dollar loans. In addition, a group of consumer advocates opposed a timing requirement that would be any more flexible than that contained in proposed § 1041.16(c)(1) and (c)(3), and asked the Bureau to require lenders to furnish updates to information previously furnished no later than the date on which the changes to the terms of the outstanding loan are made. Another industry commenter likewise urged a real-time furnishing requirement. Final Rule The Bureau is adopting § 1041.16(c)(2) as proposed, other than renumbering it as § 1041.10(c)(2). It declines to expand this furnishing requirement as proposed by some commenters. Ultimately, the Bureau has concluded that the information lenders must provide pursuant to § 1041.10(c)(2) strikes the right balance between permitting lenders to conduct a precise assessment for purposes of the proposed rule, and limiting the furnishing burdens that the rule imposes on lenders. These requirements, and the resulting balance struck between demanding either more or less information, are in service of the core principle of the underwriting provisions, which require lenders that contemplate making a covered short-term loan or a covered longer-term balloon-payment loan to make a reasonable assessment of the borrower’s ability to repay the loan according to its terms. Thus, they generally further the consumer protections advanced by part 1041. The Bureau does not agree with the commenter that suggested that a loan identifier that is unique across all lenders and registered information systems would be needed to ensure that 1196 updates are properly applied to the correct loan. Even if two lenders assigned the same loan number to a loan that each furnished, since each lender will be updating its own loan, a registered information system will be able to distinguish the loans. Further, the Bureau does not believe that such a requirement is feasible in the context of this rule, which would require thousands of unaffiliated lenders to develop and use a system to generate a unique number at the consummation of every covered short-term and longer-term balloon payment loan for use when furnishing information to each registered information system. The Bureau disagrees that the proposed requirement to update information previously furnished did not adequately describe the loans for which updates would be required or the timing of the required reporting. As described above, final § 1041.10(c)(2) requires lenders to furnish—for all outstanding covered short-term loans and covered longer-term balloon-payment loans—updates within a reasonable period after the event that causes the information that was previously furnished to be out of date. For the reasons described in the proposal, the Bureau also maintains that granting lenders a more flexible timing requirement for furnishing updates is an appropriate component in drawing the balance between the burdens and the benefits of this provision. The Bureau adopts the commentary related to § 1041.10(c)(2) as proposed, other than to make updates regarding numbering. Final comment 1041.10(c)(2)-1 sets out an example of the types of updates lenders must furnish while loans are outstanding. 10(c)(3) Information to be furnished when loan ceases to be an outstanding loan Proposed Rule 1197 Proposed § 1041.16(c)(3) would have required lenders to furnish specified information no later than the date the loan ceased to be an outstanding loan, or as close in time as feasible to the date that the loan ceased to be an outstanding loan. The Bureau believed that a real-time or close-to-real-time furnishing requirement for when a loan ceased to be an outstanding loan was appropriate to achieve the consumer protections of proposed part 1041. The proposed requirement sought to give lenders that use consumer reports from a registered information system timely information about most covered loans made by other lenders to a consumer. Although the Bureau would have encouraged lenders to furnish information about covered loans on a real-time or close-to-real-time basis, the proposal permitted lenders to furnish the required information on a daily basis or as close in time to the date the loan ceased to be outstanding as would be feasible. Proposed § 1041.16(c)(3)(i) would have required lenders to furnish the date as of which the loan ceased to be an outstanding loan. Proposed § 1041.16(c)(3)(ii) would require lenders to furnish for a covered short-term loan that had ceased to be an outstanding loan whether all amounts owed in connection with the loan were paid in full including the amount financed, charges included in the total cost of credit, and charges excluded from the total cost of credit. If all amounts owed in connection with the loan were paid in full, then this provision would further require lenders to specify the amount paid on the loan, including the amount financed and the charges comprised in the total cost of credit, but excluding any charges excluded from the total cost of credit. Comments Received 1198 Very few commenters specifically addressed the requirements listed under proposed § 1041.16(c)(3). A group of consumer advocates asserted that the Bureau’s furnishing requirements when a loan ceases to be outstanding were lacking, and made recommendations intended to strengthen the requirements applicable to both covered short-term loans and covered longer-term loans. They contended that the Bureau should require lenders to furnish charges excluded from the total cost of credit even if a loan was paid in full, and to furnish the amount financed and charges included and excluded from the total cost of credit separately from one another. They also urged the Bureau to clarify that charges not included in the total cost of credit include any fees associated with late payment on the loan, including both late fees and returned item fees. These commenters advised the Bureau to require lenders to furnish any date on which the borrower became delinquent, or the lender determined the loan to be in default, or the lender charged off the loan. They also urged the Bureau to require furnishing of information related to collection activity, including the date that the collection activity began, and records of any failed payment transfer such as transfers that trigger a prohibition on further payment transfer attempts and the reauthorization requirement. They considered this information to be relevant to a consumer’s borrowing history and a subsequent lender’s ability-to repay determination, and stated that the availability of such information in a consumer report provided by a registered information system would help protect consumers against unaffordable longer-term refinancings. An industry commenter urged that the Bureau adopt a real-time furnishing requirement. Final Rule 1199 The Bureau is finalizing § 1041.10(c)(3) as proposed and renumbered with two substantive modifications and a minor technical edit. First, final rule § 1041.10(c)(3)(ii) now requires the information described in proposed § 1041.16(c)(3)(ii)(A) to be furnished for all loans for which information is required to be furnished under the rule, not only covered short-term loans. The information that must be furnished under this section is whether the borrower paid in full all amounts owed in connection with the loan, including the amount financed, charges included in the cost of credit, and charges excluded from the cost of credit. Under the proposal, this information was necessary to establish whether an exception to the presumption against a consumer’s ability to repay in proposed § 1041.6 applied. Because of the narrowing of the scope of the rule, this information is no longer necessary for that purpose. However, the Bureau believes that this information will be useful to lenders’ underwriting of subsequent loans. Although this change will slightly increase furnishing burden, the Bureau believes the increased burdens are outweighed by the insights this information would provide about actual prior loan performance. The Bureau is not finalizing proposed 1041.16(c)(3)(ii)(B), which would have required furnishers to furnish the actual amounts paid in instances where borrower successfully paid in full all amounts connected with loans. This also was proposed to allow lenders to establish whether an exception to the presumption against a consumer’s ability to repay in proposed § 1041.6 applied. Because the Bureau is not adopting proposed § 1041.6, this information is no longer needed. Additionally, this section now references “cost of credit,” rather than “total cost of credit,” consistent with the Bureau’s adoption of the former term. 1200 Commenters had suggested the inclusion of several other data points in the furnishing requirements applicable to loans that are no longer outstanding, as they suggested that this information would be helpful for lenders in evaluating the borrowers’ ability to repay loans or refinanced loans. Although the additional information indeed might be helpful to lenders in their ability-to-repay evaluations, the Bureau finds that this benefit is outweighed by the burden on lenders that would result from requiring the additional information. Likewise, for reasons described above, the Bureau chose not to require real-time furnishing. Section 1041.11 Registered Information Systems. As discussed in more detail in the overview of proposed § 1041.16 and § 1041.17, the Bureau sought to ensure that lenders making most covered loans would have access to timely and reasonably comprehensive information about a consumer’s current and recent borrowing history with other lenders. Proposed § 1041.16 would require lenders to furnish information about most covered loans to each information system that was either provisionally registered or registered with the Bureau pursuant to proposed § 1041.17. This requirement would allow a registered information system to generate a consumer report containing relevant information about a consumer’s borrowing history, regardless of which lender or lenders had made a covered loan to the consumer previously. A lender that was contemplating making most covered loans would obtain a consumer report from a registered information system and consider such a report in determining whether the loan could be made, in furtherance of the consumer protections of proposed part 1041. The proposal also would have required the Bureau to identify the particular consumer reporting agencies to which lenders had to furnish information pursuant to proposed § 1041.16, 1201 and from which lenders could obtain the consumer reports needed to satisfy their obligations under proposed §§ 1041.5 and 1041.7. Proposed § 1041.17 would require the Bureau to identify these consumer reporting agencies by registering them with the Bureau as “information systems.” As described in more detail below, proposed § 1041.17 set forth proposed processes for registering information systems before and after the furnishing obligations under proposed § 1041.16 take effect and it stated the proposed conditions that an entity would have to satisfy in order to become a registered information system. 11(a) Definitions. 11(a)(1) Consumer report. Proposed Rule Proposed § 1041.17(a)(1) would have defined consumer report by reference to the definition of consumer report in the FCRA.1080 The Bureau explained that this definition accurately reflected how the FCRA would apply to provisionally registered and registered information systems, to lenders that furnish information about covered loans to provisionally registered and registered information systems pursuant to proposed § 1041.16, and to lenders that use consumer reports obtained from registered information systems. The proposal would require a lender that contemplated making most covered loans to obtain a consumer report about the consumer from a registered information system, which would enable the lender to determine the consumer’s eligibility for most covered loans. The proposal clarified that registered information systems providing consumer reports to such lenders would be consumer reporting agencies 1080 15 U.S.C. 1681a(d). 1202 within the meaning of the FCRA1081 and would be subject to its applicable provisions and implementing regulations. Moreover, lenders that obtained consumer reports from registered information systems and those required to provide information to provisionally registered and registered information systems under proposed § 1041.16 also would be required to comply with the provisions of the FCRA applicable to users of consumer reports and to furnishers of information to consumer reporting agencies. Comments Received One consumer reporting agency expressed general support for the proposed definition of consumer report and agreed that the FCRA is applicable. A few commenters disagreed with the definition of consumer report proposed in § 1041.17(a)(1). One industry commenter stated that the definition was not consistent with the purposes of a registered information system and a consumer report issued under the proposed rule. The commenter posited that information communicated is only a consumer report within the definition in the FCRA if the information is used by a lender to answer the question of whether a lender should make a loan to a borrower. The commenter suggested that consumer reports under the rule would not qualify as consumer reports under the FCRA because the purpose of the reports under the rule would be to determine if a lender could lend to a consumer in compliance with the regulation, not whether they should lend to the consumer. The commenter asserted that a consumer report obtained from a registered information system is not sufficient, and not intended to determine whether a lender should make a loan to the borrower. The commenter indicated that consumer reports provided by nationwide consumer reporting agencies were more appropriate to this purpose than consumer reports 1081 See 15 U.S.C. 1681a(f). 1203 provided by a registered information system. One consumer reporting agency stated that the proposed registered information systems would be in conflict with the FCRA’s definitions and requirements for consumer reporting agencies, but did not elaborate further. Final Rule The Bureau is finalizing proposed § 1041.17(a)(1), renumbered as § 1041.11(a)(1) of the final rule, without any modifications. The Bureau remains persuaded that it is appropriate to define consumer report by reference to the FCRA’s definition of consumer report. The FCRA defines consumer report to mean “any written, oral, or other communication of any information by a consumer reporting agency bearing on a consumer’s credit worthiness, credit standing, credit capacity, character, general reputation, personal characteristics, or mode of living which is used or expected to be used or collected in whole or in part for the purpose of serving as a factor in establishing the consumer’s eligibility for,” among other permissible purposes, credit.1082 Under the final rule, information contained in a consumer report obtained from a registered information system will bear on the aspects listed in section 603(d)(1) of the FCRA, and will be used in whole or in part to serve as a factor in establishing the consumer’s eligibility for a covered short-term or covered longer-term balloon loan. The Bureau does not agree with the comment suggesting that, because the information in a consumer report from a registered information system will be used to determine whether a loan would comply with this regulation, such information will not be used in whole or in part as a factor in establishing the consumer’s eligibility for credit. 1082 Section 603(d) of the Fair Credit Reporting Act, 15 U.S.C. 1681(d). 1204 11(a)(2) Federal consumer financial law. Proposed Rule Proposed § 1041.17(a)(2) would have defined Federal consumer financial law by reference to the definition of Federal consumer financial law in the Dodd-Frank Act, 12 U.S.C. 5481(14). This term is defined in the Dodd-Frank Act to include several laws that would apply to registered information systems, including the FCRA. Comments Received A set of comments generally addressed the applicability of the FCRA or other Federal laws such as the FTC’s Standards for Safeguarding Customer Information,1083 16 CFR part 314, to provisionally registered and registered information systems and covered lenders and the scope of the applicability of those laws. One consumer reporting agency agreed that registered information systems and furnishers are subject to the FCRA. A group of consumer advocates believed it was important and only fair that the FCRA applies to information that is furnished to registered information systems. The commenters said that the FCRA requirements were basic, fundamental principles of fair information use. 1083 Generally known as the Safeguards Rule, part 314 sets forth standards for developing, implementing, and maintaining safeguards to protect the security, confidentiality, and integrity of customer information. The Safeguards Rule was promulgated and is enforced by the FTC pursuant to the Gramm-Leach-Bliley Act (GLBA), 15 U.S.C. §§ 6801 through 6809. The data security provisions of the GLBA direct the prudential regulators, the SEC, and the FTC to establish and enforce appropriate standards for covered entities relating to administrative, technical and physical safeguards necessary to protect the privacy, security, and confidentiality of customer information. Congress did not provide the Bureau with rulemaking, enforcement, or supervisory authority with respect to the GLBA’s data security provisions. 15 U.S.C. 6801(b), 6804(a)(1)(A), and 6805(b). The portion of the GLBA concerning data security is not a Federal consumer financial law under the Dodd-Frank Act. However, data security practices that violate those GLBA provisions and their implementing regulations may also constitute unfair, deceptive, or abusive acts or practices under the Dodd-Frank Act. 1205 Conversely, some commenters argued that registered information systems would not fit well within the scope of the FCRA and the FACT Act. One of them added that the rule’s provisions would be subject to misinterpretation, litigation, and unpredictable regulatory examination and oversight. Another commenter stated that requiring credit unions to comply with the FCRA, when such entities do not typically furnish loan information to specialty consumer reporting agencies, would greatly increase operational costs for such lenders. Some commenters requested clarification about the scope of the FCRA’s applicability to the proposed rule. One asked the Bureau to clarify whether lenders would be required to provide a notice of adverse action. Another asked the Bureau to formalize certain best practices with respect to consumer report disputes as requirements in the final rule, saying that it was essential for the registered information systems to have the capacity to coordinate with lenders in real time in order to handle consumer disputes effectively while complying with FCRA requirements and deadlines. One commenter noted that the FCRA imposes duties on furnishers to provide accurate information and investigate disputes, and encouraged the Bureau to state in the final rule whether the registered information systems would be expected to monitor furnishers and take corrective action. At least two commenters sought clarification about the extent to which consumers would have access to the consumer protections available to them under the FCRA. One stated that consumers should have the right to review the information pertaining to them in a provisionally registered or registered information system, and to dispute those records. This commenter explained that the FCRA entitles consumers to receive information about adverse credit determinations, and stated that such a consumer right would be useful in instances where some 1206 borrowers are denied credit. One commenter encouraged the Bureau to evaluate and clearly state any requirement permitting a consumer to freeze, block, or place a fraud alert on their registered information system consumer report. It also asked the Bureau to clarify any requirement that a registered information system place an address discrepancy notation on a consumer's file with a registered information system. Lastly, this commenter also noted that it was possible that some registered information systems subject to the final rule would not be nationwide consumer reporting agencies within the FCRA’s definition. Numerous commenters were concerned about the possibility of provisionally registered and registered information systems using the furnished data for purposes other than in furtherance of part 1041. One industry commenter encouraged the Bureau to consider further restricting access to furnished information in order to protect borrower information in a manner that is consistent with applicable State law. It argued that registered information systems that supplied reports containing information furnished under the rule would not be subject to the Bureau’s supervisory authority. It further argued that permitted uses of furnished information were more permissive under the FCRA than under State requirements, and contended that the FCRA would enable registered information systems to exploit the private information of consumers in ways detrimental to borrowers, including for the purposes of generating marketing leads and advertising. Likewise, one consumer advocate opposed allowing provisionally registered and registered information systems to generate lead lists based on information furnished under the proposed rule. The commenter believed that the history of the payday lending industry showed that new supplies of debt competition would not reduce prices and pointed out that it was a 1207 standard practice of the payday industry to set interest rates at the maximum level allowed by law. It suggested that consumers would be unlikely to benefit if lenders had the ability to purchase prescreened lists from a provisionally registered or registered information system and then make pre-screened offers of credit, and submitted that the FCRA grants consumers the right to control where and how their personal information is disseminated. Consumer advocates urged the Bureau to limit the use of information furnished pursuant to part 1041 to credit purposes. Specifically, they requested that the Bureau prohibit use of the furnished information for prescreening and non-credit permissible purposes like determinations related to employment or insurance. One commenter stated that permitting use of the data for other purposes would expose consumers to negative consequences that could resulting from employers or other creditors learning that they had applied for a payday loan. One commenter stated that the FCRA and FTC Safeguards Rule would protect the security, confidentiality, and integrity of the consumer information, but cautioned that to better protect consumer privacy, the Bureau should impose additional limitations on the information collected, and should further restrict access to and use of consumer information held by registered information systems. Some consumer reporting agencies disagreed with recommendations to restrict additional uses of information furnished to provisionally registered and registered information systems pursuant to proposed §§ 1041.16 and 1041.17. One asserted that prescreening consumers for firm offers of credit would help them transition into traditional credit products by giving them targeted information on credit alternatives for which they qualify, expanding their options. It 1208 stated that consumer unawareness of these products could limit people’s access to lower cost loans. One consumer reporting agency argued that in certain contexts—including during the underwriting process—underbanked consumers, unbanked consumers, and consumers with little to no traditional credit history could benefit from the alternative use of their furnished data. It said that registered information systems would be obligated to comply with the FCRA, including the provisions that restrict access to credit reports for permissible purposes. It also noted that the Bureau, pursuant to its supervisory and enforcement authority over registered information systems, could monitor compliance with the FCRA and bring enforcement actions against registered information systems as applicable. Final Rule The Bureau has carefully considered the comments on proposed § 1041.17(a)(2). For the reasons discussed in the proposal and further below, the Bureau is finalizing this section as proposed, except for renumbering it as § 1041.11(a)(2) of the final rule, along with conforming internal references to other renumbered sections of the final rule. Registered information systems performing as required under the rule will be consumer reporting agencies within the meaning of the FCRA. Regarding the comments seeking clarification about applicability of various sections of the FCRA, the Bureau concludes that it is beyond the scope of this rulemaking to clarify the scope of other rules or statutes. Specifically, it declines to provide in this rulemaking guidance concerning how registered information systems and lenders comply with the FCRA. 1209 It should be noted that the Bureau included in § 1041.11(b)(4) and (5) eligibility requirements for becoming a registered or provisionally registered information system that include specific requirements for an applicant to have a Federal consumer financial law compliance program and for it to provide the Bureau with an independent assessment of its compliance program as part of its application for provisional registration or registration. Accordingly, it is the Bureau’s expectation that registered information systems will determine their rights and obligations under the applicable Federal consumer financial laws. The Bureau declines to impose restrictions on the use of information furnished to registered information systems pursuant to this rule beyond the restrictions contained in the FCRA. The Bureau recognizes that a provisionally registered or registered information system’s provision of prescreened lists based on information furnished pursuant to this rule may create a risk that an unscrupulous provider of risky credit products could use such a list to target potentially vulnerable consumers. At the same time, however, the Bureau believes that prescreening could prove useful to certain consumers to the extent they needed credit and received firm offers of affordable credit. Commenters also sought clarity regarding the applicability of the Safeguards Rule; again, the Bureau concludes that it is beyond the scope of this rulemaking to clarify the scope of other rules or statutes. The Bureau also notes that, as explained above, it does not have authorities with respect to the Safeguards Rule. The Bureau notes it is including in § 1041.11(b)(6) and (7) eligibility requirements for becoming a registered or provisionally registered information system that include specific requirements for an applicant to have developed, implemented, and maintain a comprehensive information security program that complies with the Safeguards Rule and for it 1210 to provide the Bureau with an independent assessment of its information security program as part of its application for provisional registration or registration and on at least a biennial basis thereafter. 11(b) Eligibility criteria for information systems. Proposed Rule The subparts of proposed § 1041.17(b) set forth the conditions the Bureau would consider in determining whether an entity is eligible to become a registered or provisionally registered information system pursuant to proposed § 1041.17(c) or (d). As with other portions of the proposed rule that are being renumbered in light of changes made to their provisions, proposed § 1041.17(b) is ultimately being renumbered as § 1041.11(b) of the final rule. Proposed § 1041.17(b)(1) would have required the Bureau to determine that an entity possesses the technical capability to immediately receive information furnished pursuant to proposed § 1041.16, and that the entity uses reasonable data standards that facilitate the timely and accurate transmission and processing of information in a manner that does not impose unreasonable cost or burden on lenders.1084 Proposed § 1041.17(b)(2) would require the Bureau to determine that the entity possessed the technical capability to generate a consumer report containing, as applicable for each unique consumer, all information described in proposed § 1041.16 substantially simultaneous to receiving the information from a lender. Proposed § 1041.17(b)(3) would require the Bureau to determine that the entity would perform in a manner that facilitates compliance with, and furthers the purposes of, proposed part 1041. 1084 Among other things, these standards must facilitate lender and registered information system compliance with the provisions of the FCRA and its implementing regulations concerning the accuracy of information furnished. 1211 Proposed § 1041.17(b)(4) would require the Bureau to determine that the entity had developed, implemented, and maintains a program reasonably designed to ensure compliance with all applicable Federal consumer financial laws. This compliance program would have to include written policies and procedures, comprehensive training, and monitoring to detect and promptly correct compliance weaknesses, as described in more detail in the proposed commentary. Proposed § 1041.17(b)(5) required the entity to provide to the Bureau in its application for registration or provisional registration a written assessment of the Federal consumer financial law compliance program just described. The assessment would have to set forth a detailed summary of the Federal consumer financial law compliance program that the entity had implemented and maintained, and explain how that compliance program was appropriate for the entity’s size and complexity, the nature and scope of its activities, and risks to consumers presented by such activities. The assessment also would have to certify that, in the opinion of the independent assessor, the Federal consumer financial law compliance program was operating with sufficient effectiveness to provide reasonable assurance that the entity was fulfilling its obligations under all Federal consumer financial laws. In addition, the assessment would have to certify that it had been conducted by a qualified, objective, independent thirdparty individual or entity that used procedures and standards generally accepted in the profession, adhered to professional and business ethics, performed all duties objectively, and was free from any conflicts of interest that might have compromised the assessor’s independent judgment in performing the assessment. The written assessment of an entity’s Federal consumer financial law compliance program required under proposed § 1041.17(b)(5) would have to be included in the entity’s 1212 application for registration pursuant to proposed § 1041.17(c)(2) or for provisional registration pursuant to proposed § 1041.17(d)(1). However, this written assessment would not be required in an entity’s application for preliminary approval for registration pursuant to proposed § 1041.17(c)(1), and would not have to be provided to the Bureau when a provisionally registered information system became registered pursuant to proposed § 1041.17(d)(2). With respect to entities seeking to become registered prior to the effective date of proposed § 1041.16, the proposal would have provided an entity 90 days from the date that preliminary approval was granted to prepare its application for registration, including obtaining the written assessment required under proposed § 1041.17(b)(5). Proposed § 1041.17(b)(6) would have required the Bureau to determine that an applicant had developed, implemented, and maintained a comprehensive information security program that complied with the Safeguards Rule. Proposed § 1041.17(b)(7)(i) would require the entity to provide to the Bureau in its application for provisional registration or registration, and on at least a biennial basis thereafter, a written assessment of the information security program described in proposed § 1041.17(b)(6). Each assessment had to set forth the administrative, technical, and physical safeguards that the entity had implemented and maintained; explain how such safeguards were appropriate to the entity’s size and complexity, the nature and scope of its activities, and the sensitivity of the customer information at issue; explain how the safeguards that were implemented met or exceeded the protections required by the Safeguards Rule; and certify that, in the opinion of the assessor, the information security program was operating with sufficient effectiveness to provide reasonable assurance that the entity was fulfilling its obligations under the Safeguards Rule. The assessment also had to certify that it had been 1213 conducted by a qualified, objective, independent third-party individual or entity that used procedures and standards generally accepted in the profession, adhered to professional and business ethics, performed all duties objectively, and was free from any conflicts of interest that might have compromised the assessor’s independent judgment in performing assessments. The proposed commentary clarified the timing of the assessments, provided examples of individuals and entities qualified to conduct the assessment, and addressed matters of format and style. With respect to entities seeking to become registered prior to the effective date of proposed § 1041.16, the Bureau proposed to allow 90 days from the date that a preliminary approval for registration was granted for the entity to prepare its application for registration, including obtaining the written assessment required pursuant to proposed § 1041.17(b)(7). Proposed § 1041.17(b)(7)(ii) required each written assessment produced pursuant to proposed § 1041.17(b)(7)(i) to be completed and provided to the Bureau within 60 days after the end of the period to which the assessment applies. Proposed § 1041.17(b)(8) required that to become a registered information system, the entity had to have acknowledged that it was, or consented to being, subject to the Bureau’s supervisory authority. Comments Received The Bureau received a broad range of comments about the adequacy of the eligibility requirements applicable to entities seeking to become registered information system pursuant to proposed § 1041.17(b). One set of commenters was generally apprehensive about the potential lack of interest from eligible entities in serving as registered information systems. One trade association questioned the Bureau’s support for establishing the measures, and stated that it doubted any entities would register as information systems. This commenter predicted that 1214 consumer access to the covered loan products would turn more on registration compliance than lender compliance. Another commenter speculated that there would be little interest from entities to become registered information systems because it viewed the proposed independent assessment of the information security program as exceeding the scope of the Safeguards Rule. It criticized the Bureau for lacking a contingency plan to ensure continuity in the market in the event that no entities chose to become registered information systems. Some comments addressed proposed § 1041.17(b)(1), concerning the requirement that a registered information system be able to use reasonable data standards in a manner that does not impose unreasonable costs or burdens on lenders. One Tribal entity urged the Bureau to prevent registered information systems from engaging in price-gouging practices, particularly when transacting with parties that wholly depend on the ability to access the services to be provided by these systems. A consumer reporting agency argued that the heterogeneity of specialty consumer reporting agencies with respect to technology, data collected, business model, and business practices, would make it challenging for the Bureau to assess whether any costs meet the reasonableness standard of proposed § 1041.17(b)(1). Furthermore, this commenter cautioned that some entities applying for registration could be regulatory monopolists and could charge high costs for access to their data. This commenter believed that registered information systems should agree to data interchange standards in order to keep prices down. In addition, it recommended that any fee charged to lenders should be conditioned on the provision of actual data, such that a result of no data would not incur a fee. The commenter believed this approach would prevent a registered information system from being compensated simply for inquiries that generate no hits. On the other hand, one industry commenter stated that the Bureau should 1215 consider several factors before restricting fees and charges in connection with the proposed furnishing requirements. It argued that fees and charges should permit a registered information system to maintain financially sound business operations while enabling lenders to use these compliance services at a reasonable business-friendly cost. With respect to an entity’s general capability to receive information, one consumer reporting agency stated that a registered information system would need access to data about outstanding loans as of the effective date of the furnishing requirement, along with historical data on loans originated and closed in the six months leading up to the requirement to furnish data. Another commenter agreed with this suggestion, stating that it was necessary for lenders to upload historical loan data by the effective date of the furnishing requirement. Other commenters encouraged requiring registered information systems to be able to receive information furnished in the Metro 2 format, explaining that, in their view, Metro 2 fully complies with Federal requirements, is publicly available and time-tested, and would ensure proper classification of loans and loan statuses. Others agreed that standardizing how data is furnished is important but requested that the Bureau not designate a specific standard. Proposed § 1041.17(b)(2) requires entities to have the capability to generate a consumer report substantially simultaneous to receiving information from a lender. One trade association doubted that entities seeking to act as registered information systems would be able to generate reports substantially simultaneous to their receipt of the information. Commenters who urged requiring provisionally registered and registered information systems to be able to receive information furnished in the Metro 2 format also requested that registered information systems have the capability to generate a consumer report containing information furnished in the Metro 1216 2 format. Others asked the Bureau to clarify provisionally registered and registered information systems’ responsibility to perform quality assurance assessments on furnished information received pursuant to proposed §§ 1041.16 and 1041.17. As an example of what such potential responsibilities might entail, the commenter described the process that it follows to analyze its portfolio of records for data quality and consistency, and to monitor the frequency of updates to its records. Some commenters raised concerns about the feasibility of developing within the proposed time frames the standards necessary to meet the requirement that registered information systems generate reports “substantially simultaneous” to receipt of the information from the lender. Other commenters indicated that some consumer reporting agencies have that capability now. The Bureau received several comments on proposed § 1041.17(b)(3), which requires an entity to be able perform its obligations as a registered information system in furtherance of the purposes of part 1041. A number of consumer groups noted their support for proposed comment 1041.17(b)(3)-1, which clarifies that part 1041 does not supersede the consumer protection obligations imposed under other Federal law or regulation and provides a specific example concerning an obligation under the FCRA. One commenter regarded it as a fundamental condition of eligibility for registered information systems. One consumer reporting agency urged the Bureau to condition an entity’s eligibility to become a registered information system on certain financial stability requirements, to subject the systems to oversight, and to apply standards of ownership and management that would exclude inexperience or criminal backgrounds. It also urged the Bureau to require entities to demonstrate a proven record of core competencies, compliant market-place behavior, and an effective 1217 dispute-handling system. Another commenter agreed that an entity should be required to show a proven history of successfully implementing and maintaining a compliance management system. A trade association suggested that the Bureau mandate the lender’s submission of net worth requirements, a bond for performance, background checks on the owners, and anti-sale provisions of the company without notice or approval elements. Another commenter recommended that the Bureau require entities to provide evidence of their relationships with lenders that would furnish data to the entities pursuant to proposed § 1041.16. It believed that the existence and nature of such relationships could help maximize the effectiveness of efforts to preserve and produce high-integrity data. One industry commenter argued that, generally, consumer reporting agencies were not well-suited to satisfy the proposed conditions to become registered information systems because they were not designed for real-time data capture and reporting, and in the past had not been required to perform in the manner required by proposed § 1041.17 to meet requirements under the FCRA. This commenter asserted that consumer reporting agencies had a poor track record in maintaining the accuracy of furnished information, among other obligations. Very few commenters disagreed with the substance of proposed § 1041.17(b)(4). One industry commenter argued the proposal is vague, and does not provide enough information to adequately determine the applicability of the referenced Federal consumer financial laws. A consumer reporting agency suggested that entities should have to demonstrate their capability to reasonably reinvestigate a consumer dispute, based on the circumstances. It urged the Bureau to retain exclusive jurisdiction over the enforcement and oversight of the registered information systems. It speculated that fear of private litigation could constrain new registered information 1218 systems. It also raised the possibility that State actions and plaintiff litigation would risk the development of inconsistent or conflicting law, which could restrain future rulemaking relating to registered information systems. The Bureau received several comments on the requirement in proposed § 1041.17(b)(6) that an entity would have to develop an information security program that is compliant with the Safeguards Rule and submit it to the Bureau. One commenter praised the Bureau for acknowledging that registered information systems must comply with the Safeguards Rule. Another stated that registered information systems should be required to monitor data furnishing and generally take an active role in working with lenders to reduce compliance burdens and streamline reporting systems. Yet another commenter said that the required independent assessment of the information security program exceeded the scope of the Safeguards Rule, which would increase the costs of obtaining reports and eventually shut down small businesses and hinder innovation. One commenter requested that the Bureau explicitly restrict the access to information furnished to registered information systems to authorized users exclusively and on an as-needed basis only.1085 A trade association argued that the proposal did not address mechanisms to independently verify the data in the registered information systems and to secure the data’s confidentiality. This commenter generally asked the Bureau for more details about the registered information systems. A consumer reporting agency asked the Bureau how consumer disputes were to be accurately communicated to all registered information systems to ensure that each had identical data. 1085 It should be noted that the FCRA limits access to consumer reports to those with a permissible purpose. 1219 With respect to the requirements under proposed § 1041.17(b)(5) and (b)(7), a consumer reporting agency expressed concern that requiring all registered information systems to conduct independent assessments would substantially increase the costs of compliance, which would then pass through to consumers in the form of higher-cost credit. It suggested that a sufficiently independent internal audit process could provide the appropriate balance and oversight. Lastly, the Bureau did not receive any comments about proposed § 1041.17(b)(8). Final Rule After carefully considering the comments received, the Bureau is finalizing § 1041.11(b) of the final rule—including subsections (b)(1) through (b)(8)—in substantially the same form as proposed § 1041.17(b), aside from renumbering the sections and conforming the internal references from the proposal, and it is also adding to the commentary relating to § 1041.11(b)(3) as described below. In general, the Bureau disagrees with the prediction that no entity would be interested in registering as an information system under the rule. During its market outreach, several firms have expressed interest in serving as registered information systems pursuant to the rule. Several commenters emphasized the importance of moderating any costs to furnish information pursuant to § 1041.10 of the final rule. Section 1041.11(b)(1) requires that registered information systems use reasonable standards with respect to furnishing that, among other things, do not impose unreasonable costs or burdens on lenders. The Bureau considered the comments regarding moderating costs associated with furnishing and the related concern that registered information systems are able to cover their costs (and earn a return) in satisfying their obligations pursuant to § 1041.11 of the final rule. It agrees with commenters who suggest that 1220 fees and charges should permit a registered information system to maintain financially sound business operations while enabling lender to use these compliance services at a reasonable business-friendly cost. However, in finalizing final § 1041.11(b)(1), the Bureau concludes that in connection with furnishing, lenders must not impose unreasonable costs or burdens on lenders. Several commenters suggested that lenders should be able to access historical data on loans made prior to the effective date of the rule when contemplating making a covered loan under the rule. As described elsewhere, the final rule does not require any furnishing until the effective date of § 1041.10, which will be 21 months after publication of the rule in the Federal Register. Because §§ 1041.5 and 1041.6 will be effective at the same time as § 1041.10, there will be some period of time during which reports obtained from information systems registered before the effective date of § 1041.10 will have little or no information. The Bureau weighed the risk of having little or no information in these registered information systems against the burdens related to requiring lenders to furnish information about loans made prior to the effective date of the rule’s substantive provisions. The Bureau has determined that such a requirement would impose significant burden on lenders and that such burden would not be justified by the benefits. For example, under such a requirement, lenders would have to determine whether loans made prior to the effective date would qualify as “covered short-term loans” or “covered longer-term balloon payment loans” if they had been made after the rule’s effective date. Further, lenders would not be able to furnish some of the required fields, reducing the utility of the data to further the purposes of the rule. Finally, requiring the furnishing of historical loan data would require additional time for onboarding lenders to registered information systems, delaying the implementation of the rule. 1221 The Bureau also considered whether, in order to increase the amount of data held by registered information systems when lenders begin obtaining consumer reports as required under the rule, it should stagger the effective dates of the furnishing obligation under § 1041.10 and the obligations to obtain a consumer report from a registered information system under §§ 1041.5 and 1041.6. Staggering effective dates may increase to some degree the utility of the consumer reports that lenders would be required to obtain at first, but may add complexity to implementation of the rule and would involve other tradeoffs, as discussed in the proposal. The Bureau has determined that not staggering the effective dates of §§ 1041.10, 1041.5 and 1041.6, and requiring furnishing on a going forward basis, is the better approach. The Bureau agrees with commenters who suggest that requiring provisionally registered and registered information systems to agree to use a common data standard would have the potential to keep costs incurred by lenders in connection with furnishing down. However, it declines to require that provisionally registered and registered information systems agree to use a common data standard. The Bureau is not convinced that requiring such agreement as a condition of eligibility for registration is necessary. The Bureau has concluded that it will be in the interest of the registered information systems to use a common data standard. The Bureau also declines to require that provisionally registered and registered information systems use a particular data standard, such as Metro 2, for purposes of receiving furnished information from lenders. As explained elsewhere, the Bureau believes that the development of common data standards across provisionally registered and registered information systems would benefit lenders and the information systems and intends to foster the development of such common data standards where possible. However, the Bureau believes that 1222 development of these standards by market participants would likely be more efficient and offer greater flexibility and room for innovation than if the Bureau prescribed particular standards in this rule. With respect to Metro 2 in particular, the Bureau notes that it believes the standard would need to be modified in order to allow furnishing as required under this rule. Though Metro 2 may be useful as a starting point for development of a common data standard, especially to the extent that the entities that become provisionally registered or registered information systems already use Metro 2 to receive data, the Bureau declines to condition an entity’s eligibility to become a registered information system on its use of Metro 2. With respect to the requirement that registered information systems generate a consumer report substantially simultaneous to receiving the information from a lender, the Bureau is finalizing proposed § 1041.17(b)(2) as § 1041.11(b)(2). Comment 1041.11(b)(2)-1 clarifies that technological limitations may cause some slight delay in the appearance of a consumer report of information furnished pursuant to § 1041.10, but that any delay must be reasonable. The Bureau concludes that this expectation is reasonable. Under final § 1041.11(b)(3), as proposed, an entity seeking to become a provisionally registered or registered information system must be able to perform in a manner that facilitates compliance with and furthers the purposes of this part. The Bureau disagrees with the comment recommending that it seek to override other existing Federal consumer financial laws that would, example, permit States to bring enforcement actions pursuant to the Dodd-Frank Act, or private individuals to bring an action pursuant to a private cause of action created by the FCRA. The Bureau maintains the position that the consumer protections conferred by part 1041 will best be furthered if the final rule does not supersede the obligations imposed by other Federal laws or 1223 regulations. Accordingly, it is finalizing comment 1041.11(b)(3)-1, as proposed, which clarifies that the requirement that to be eligible for provisional registration or registration as an information system, an entity must perform in a manner that facilitates compliance with the purposes of the final rule, does not supersede consumer protection obligations imposed on the entity by other Federal law or regulation. Several commenters expressed concern that the Bureau would consider registering entities with no demonstrated experience with compliance management systems, FCRA compliance, or with the types of lenders that will be furnishing data under the rule. In response, the Bureau has added comment 1041.11(b)(3)-2 to clarify that in evaluating whether an applicant is reasonably likely to satisfy or does satisfy the requirement set forth in § 1041.11(b)(3) of the final rule, the Bureau will consider any experience the applicant has in functioning as a consumer reporting agency. In addition, the Bureau declines to prescribe in this rule a provisionally registered or registered information system’s responsibility to perform quality assurance assessments on furnished information received pursuant to § 1041.10 of the final rule. As described in the proposal, the Bureau’s general approach is to seek to preserve more latitude for market participants that are interested in becoming registered information systems, with the understanding that other regulations and laws already apply or will apply to them, such as the FCRA and the Safeguards Rule, providing additional consumer protections. The final rule confers on provisionally registered and registered information systems the discretion to develop and refine their policies and procedures to satisfy the requirements of §§ 1041.10 and 1041.11. The Bureau has concluded that it is more efficient and effective to allow a market entity to 1224 determine its individual approach to complying with § 1041.11(b)(1), (b)(4) and (b)(6) and other regulatory requirements, including potentially designing a quality assessment process in a manner that accounts for features that may be unique to that entity, such as its technology, infrastructure, or business model. As noted in comment 1041.11(b)(3)-1, the FCRA would obligate any registered information system preparing a consumer report to “follow reasonable procedures to assure maximum possible accuracy of the information concerning the individual about whom the report relates.”1086 The central point in § 1041.11(b)(4) of the final rule is to ensure that provisionally registered and registered information systems have appropriate Federal consumer financial law compliance programs in place, including written policies and procedures, comprehensive training, and monitoring to detect and to promptly correct compliance weaknesses. As described in the proposal and in the discussion below, the commentary to this section provides examples of the policies and procedures, training, and monitoring that are required here. The proposal explained that these examples were modeled after the Compliance Management Review examination procedures contained in the Bureau’s Supervision and Examination Manual. Moreover, the final rule refers to the Dodd-Frank Act’s definition of Federal consumer financial law which includes several laws that the Bureau sees as applicable to registered information systems, including the FCRA, as discussed in greater detail in the proposal. The required Federal consumer financial law compliance program in § 1041.11(b)(4) of the final rule is reinforced by the provision requiring an independent assessment of that compliance program in § 1041.11(b)(5) of the final rule. To summarize, as noted in the 1086 15 U.S.C. 1681e(b). 1225 proposal, an entity’s application for registration pursuant to § 1041.11(c)(2) or provisional registration pursuant to § 1041.11(d)(1) is required to contain this written assessment, which includes a detailed summary of the entity’s compliance program, an explanation of how the program is appropriate to the entity’s size and activities, certification by an assessor that the program is effective in assuring that the entity is fulfilling its legal duties, and certification of the assessor’s qualifications, objectivity, and independence. The Bureau received comments suggesting that § 1041.11(b)(5) would add costs to the preparation of an application to be a registered information system, which the Bureau agrees is likely. However, with respect to entities seeking to become registered information systems before the effective date of § 1041.10, the Bureau purposefully staggered the requirement for submitting such an assessment to the Bureau until after the entity receives preliminary approval to become a registered information system. The applicants will incur such costs only after they receive preliminary approval. The costs of having an actual compliance management program are ones that responsible companies already budget for and are not imposed by this requirement. It should also be noted that effective programs often tend to reduce costs by minimizing legal, regulatory, and reputational risk for the entity. The Bureau is including the requirement in § 1041.11(b)(5) so that the Bureau can be reasonably assured that the entity has developed, implemented, and maintains a program reasonably designed to ensure compliance with all applicable Federal consumer financial laws until such time as the Bureau itself can evaluate the entity’s compliance program under its supervisory authority. Thus, the Bureau is finalizing § 1041.11(b)(4) and (b)(5) as proposed and renumbered. The Bureau is also finalizing the related commentary related to those provisions, as proposed. 1226 The Bureau also adopts § 1041.11(b)(6) as proposed and renumbered. The Bureau acknowledges that, as one commenter stated, the rule does not prescribe how provisionally registered and registered information systems comply with the Safeguards Rule. As mentioned above, the Bureau declines to provide in this rulemaking guidance concerning how provisionally registered and registered information systems comply with other applicable laws. The Bureau concludes that it is beyond the scope of this rulemaking to do so. And for essentially the same reasons that were discussed above with respect to § 1041.11(b)(4) and (b)(5), the Bureau adopts § 1041.11(b)(7) as proposed. The information security program required under § 1041.11(b)(6) is reinforced by the provision requiring an independent assessment of the program in § 1041.11(b)(7) of the final rule. Here too, commenters stated that the independent assessment requirement would add cost to the preparation of an application to be a registered information system, which the Bureau agrees is likely. However, with respect to entities seeking to become registered information systems before the effective date of § 1041.10, the Bureau purposefully staggered the requirement for submitting such an assessment to the Bureau until after the entity receives preliminary approval to become a registered information system. The Bureau is finalizing § 1041.11(b)(7) and its related commentary, as proposed and renumbered. Several commenters sought to condition the Bureau’s approval of an entity as a provisionally registered or registered information system upon it meeting certain additional criteria, including, among other things, financial stability criteria, background checks, net worth thresholds, criminal background checks, and performance bonds. The Bureau declines to add additional eligibility requirements. The Bureau takes the view that its expertise and experience 1227 with this market, together with its consumer protection obligations under the Dodd-Frank Act, this final rule, and other applicable Federal consumer financial laws and regulations, provide sufficient sources to guide it in evaluating an applicant’s eligibility to become a registered information system. It should be noted that several of the additional criteria suggested by commenters are already addressed by the eligibility requirements in final § 1041.11(b). For example, one commenter suggested that the Bureau condition eligibility on a company having an established compliance management system designed to ensure adherence with Federal consumer financial laws. Final § 1041.11(b)(4) requires that registered information systems have developed, implemented, and maintain a program reasonably designed to ensure compliance with all applicable Federal consumer financial law. Such a program is a key component of an adequate compliance management system; other components of such a system include Board and management oversight, consumer complaint response monitoring, compliance audit, and service provider oversight. The Bureau expects that all supervised entities (which under § 1041.11(b)(8) will include all provisionally registered and registered information systems) will have adequate compliance management systems. Proposed § 1041.16(b)(8) would have required that an entity seeking to become a provisionally registered or registered information system must acknowledge it is or consents to be subject to the Bureau’s supervisory authority. This provision received no comments and thus the Bureau is finalizing § 1041.11(b)(8) as proposed and renumbered. 11(c) Registration of information systems prior to the effective date of § 1041.10. 1228 Proposed Rule Proposed § 1041.17(c) described the process that the Bureau proposed for the registration of information systems before the effective date of proposed § 1041.16. Once proposed § 1041.16 was in effect, lenders would have to furnish information to an information system that was registered pursuant to proposed § 1041.17(c)(2) for 120 days or more. The Bureau proposed a two-stage process to become registered prior to the effective date of proposed § 1041.16. First, interested entities would submit to the Bureau an initial application for preliminary approval for registration. Second, the entities would submit a full application for registration after receiving preliminary approval and obtaining certain written assessments from third parties concerning their compliance programs. 11(c)(1) Preliminary approval Proposed § 1041.17(c)(1) provided that, prior to the effective date of proposed § 1041.16, the Bureau could preliminarily approve an entity for registration only if the entity submitted an application for preliminary approval to the Bureau by the deadline set forth in proposed § 1041.17(c)(3)(i). The application had to contain information sufficient for the Bureau to determine that the entity was reasonably likely to satisfy the conditions set forth in proposed § 1041.17(b) by the deadline set in proposed § 1041.17(c)(3)(ii). The proposed rule and comments outlined further details about the process, including that the entity’s application would need to describe the steps the entity plans to take to satisfy the conditions and the entity’s timeline for such steps and that the entity’s plan would need to be reasonable. 11(c)(2) Registration 1229 Proposed § 1041.17(c)(2) allowed the Bureau to approve the application of an entity seeking to become a registered information system prior to the effective date of proposed § 1041.16 only if the entity had received preliminary approval pursuant to proposed § 1041.17(c)(1), and applied to be a registered information system by the deadline proposed in § 1041.17(c)(3)(ii) by submitting information sufficient for the Bureau to determine that the conditions set forth in proposed § 1041.17(b) were satisfied. Proposed § 1041.17(c)(2) further provided that the Bureau could require additional information and documentation to facilitate this determination or otherwise to assess whether registration of the entity would pose an unreasonable risk to consumers. Its related commentary clarifies that the entity seeking to become a registered information system would have to submit the application by the deadlines, and that the application would need to contain information and documentation adequate for the Bureau to determine the required conditions are satisfied, and succinctly and accurately convey the required information, including the required written assessments. 11(c)(3) Deadlines Proposed § 1041.17(c)(3)(i) and (ii) provided that the deadline to submit an application for preliminary approval for registration pursuant to proposed § 1041.17(c)(1) would be 30 days from the effective date of proposed § 1041.17, and the deadline to submit a registration application pursuant to proposed § 1041.17(c)(2) would be 90 days from the date that preliminary approval for registration is granted. Proposed § 1041.17(c)(3)(iii) would permit the Bureau to waive these deadlines. Comments Received 1230 Few commenters objected to the time frames that were proposed in § 1041.17(c). One commenter interested in registering as an information system under proposed § 1041.17 stated that its existing infrastructure could allow it to implement the requirements within four months to a year. The commenter stated that the factors that could delay implementation toward the longer side of that range were the historical data component, the complexity of products, the number of products, and interfaces and rules as yet unknown. One consumer reporting agency stated that if the Bureau did not announce the eligibility criteria for registration until it published the final rule, the proposed 30-day period after § 1041.17’s effective date to apply for preliminary approval would be insufficient to allow applicants to conduct a business analysis and the technical planning necessary to prepare their applications for preliminary approval. This commenter urged the Bureau to signal its views on configuration issues far ahead of the formal application period for registration. Alternatively, it proposed that the Bureau extend the period to prepare an application for preliminary approval to at least six months. Another industry commenter argued that the deadlines under proposed § 1041.17(c)(3) did not allow adequate time for a preliminary approval application, technical development, operational development, incorporation of common data standards, and completion of written assessments. That commenter asked the Bureau to reconsider the timeline required to meet eligibility criteria and foster common data standards, and for prospective applicants to integrate these standards with their service offerings. It urged the Bureau to initiate the common data standards process prior to publication of the rule, if possible, to facilitate completion of the registered information system’s environment prior to the effective date of the final rule. Final Rule 1231 The Bureau is finalizing proposed § 1041.17(c) as § 1041.11(c) of the final rule in accordance with the renumbering of sections within the rule described earlier. As described above, the Bureau is doing so with one minor modification to the proposed rule, along with substantive changes to the proposed deadlines and technical revisions. The Bureau is finalizing § 1041.11(c)(1) as proposed, except that the provision now permits the Bureau to require additional information and documentation to facilitate its determination of whether to grant an applicant preliminary approval. The Bureau has determined that this modification will facilitate its engagement with entities seeking registration before the effective date at an earlier stage in the registration process, while granting entities additional opportunities to supplement their applications and ensuring the Bureau has received all the information necessary to make a wellinformed determination. The Bureau is also finalizing proposed § 1041.17(c)(2) as § 1041.11(c)(2). As described above, the section allows the Bureau to approve the application of an entity seeking to become a registered information system prior to the effective date of § 1041.10 only if the entity received preliminary approval pursuant to § 1041.11(c)(1), and applied to be a registered information system by the deadline in § 1041.11(c)(3)(ii) by submitting information sufficient for the Bureau to determine that the conditions set forth in § 1041.11(b) are satisfied. Section 1041.11(c)(2) further provides that the Bureau can require additional information and documentation to facilitate this determination or otherwise to assess whether registration of the entity would pose an unreasonable risk to consumers. In addition, the Bureau is finalizing the commentary related to § 1041.11(c)(1) and (c)(2). 1232 In response to concerns that commenters raised about the proposed deadlines, the Bureau is finalizing § 1041.11(c)(3)(i) as proposed, except that it is extending the deadline to submit an application for preliminary approval by 60 days—which now establishes a deadline of 90 days after the effective date of § 1041.11. The Bureau is adopting § 1041.11(c)(3)(ii) as proposed, except that it is extending the deadline to submit an application for registration by 30 days— which now establishes a deadline of 120 days from the date that preliminary approval for registration is granted. The Bureau has concluded that the revised deadlines will provide interested entities with adequate time to prepare their applications, and will provide the Bureau with adequate time to review applications, while still allowing entities to register sufficiently in advance of the effective date of § 1041.10 so that furnishing may begin upon that effective date. The proposed deadlines complement the final rule, which extends the implementation period for all provisions other than § 1041.11 by six more months—moving it from 15 months to 21 months, as described above – and which provides for a 180-day period (rather than the 120-day period that was proposed) before lenders are obligated to begin furnishing to an information system registered prior to the effective date of the rest of the rule. The Bureau is not requiring that registered information systems use a common data standard for receiving information from lenders. The Bureau will welcome suggestions regarding how it can foster the development of such standards with applications for preliminary approval as registered information systems. 11(d) Registration of information systems on or after the effective date of § 1041.10 Proposed Rule 1233 Proposed § 1041.17(d) set forth the process that the Bureau proposed to be used for the registration of information systems on or after the effective date of proposed § 1041.16. The process involved two steps: first, an entity had to apply to become a provisionally registered information system; second, after it had been provisionally registered for a period of time, the entity automatically would become a fully registered information system. Under the proposal, lenders had to furnish information to a system that had been provisionally registered pursuant to proposed § 1041.17(d)(1) for 120 days or more, or that subsequently had become registered pursuant to proposed § 1041.17(d)(2). However, lenders could not rely on consumer reports from a provisionally registered system to satisfy their obligations under proposed §§ 1041.5 and 1041.7 until the system was fully registered pursuant to proposed § 1041.17(d)(2). The proposed period between provisional registration and full registration would be 180 days, to provide 120 days for onboarding and 60 days of furnishing before lenders could rely on consumer reports from the registered information system for purposes of the rule. 11(d)(1) Provisional registration Proposed § 1041.17(d)(1) would have provided that, on or after the effective date of proposed § 1041.16, the Bureau could only approve an entity’s application to be a provisionally registered information system if the entity’s application contained information sufficient for the Bureau to determine that the entity satisfied the conditions set forth in proposed § 1041.17(b). Proposed § 1041.17(d)(1) added that the Bureau could require more information and documentation to facilitate this determination or otherwise assess whether provisional registration of the entity would pose an unreasonable risk to consumers. 11(d)(2) Registration 1234 Proposed § 1041.17(d)(2) stated that an information system which is provisionally registered pursuant to proposed § 1041.17(d)(1) would automatically become a registered information system pursuant to proposed § 1041.17(d)(2) upon the expiration of the 180-day period commencing on the date the information system was provisionally registered. Once a system was registered pursuant to proposed § 1041.17(d)(2), lenders were permitted to rely on a consumer report generated by the system to satisfy their obligations under proposed §§ 1041.5 and 1041.7. Proposed § 1041.17(d)(2) would provide that, for purposes of proposed § 1041.17(d), an information system was provisionally registered on the date that the Bureau published notice of such provisional registration on the Bureau’s website. Final Rule The Bureau did not receive comments on proposed § 1041.17(d). In the proposal, the Bureau explained that it anticipated that, in order to permit lenders time to adjust to furnishing to information systems that are registered before the effective date of the furnishing obligation, proposed § 1041.16, it would not provisionally register any information systems during the first year that proposed § 1041.16 would be in effect. One consumer reporting agency expressed support for this proposed pause, which it believed would provide entities registered as information systems before the effective date with time to collaborate on data exchange standards. The Bureau now confirms that it plans to not provisionally register any information systems during the first year the final rule is in effect. The Bureau concludes that such a pause in registrations of information systems will allow lenders time to adjust to the furnishing to registered information systems that are registered pursuant to § 1041.11(c)(2). The Bureau adopts § 1041.17(d) as proposed, which is now renumbered as § 1041.11(d) of the final rule, 1235 with one modification. Under final § 1041.11(d)(2), as explained above, a provisionally registered information system under § 1041.11(d)(1) automatically becomes a fully registered information system upon the expiration of 240 days, not 180 days as proposed. This change is to preserve the 60-day “furnishing-only” stage proposed for entities provisionally registered on or after the effective date of § 1041.10. Under the final rule, once an information system is provisionally registered for 180 days, lenders must furnish to the system under § 1041.10. Lenders cannot rely on reports from the system to satisfy its obligations under § 1041.5 and § 1041.6 until the system becomes a fully registered information system, which will happen automatically 240 days after the system was provisionally registered. Thus, these registered information systems will receive furnished information for 60 days before lenders can rely on their reports to satisfy their obligations under the rule. This will ensure that at the point at which an information system becomes registered on or after the effective date of § 1041.10 and lenders can rely on its reports, such reports would include reasonably comprehensive information about consumers’ recent borrowing histories. The Bureau adopts comment 1041.11(d)(1)-1 as proposed, as well, which clarifies that the entity seeking to become a provisionally registered information system must submit an application to the Bureau containing information and documentation adequate for the Bureau to assess that § 1041.11(b) are satisfied. 11(e) Applications In § 1041.11 of the final rule, the Bureau has added a new provision, § 1041.11(e), for the purpose of ensuring more specifically that it receives from applicants the information necessary to evaluate applications pursuant to § 1041.11(c) and (d) of the final rule. The provision requires 1236 entities to submit their applications for preliminary registration, registration, and provisional registration in the form required by the Bureau. Applications must include the name of the entity, its business and mailing address as applicable, and the name and contact information of the person who is authorized to communicate with the Bureau on the applicant’s behalf concerning the application. The Bureau expects that applicants will be able to provide this information in their application to the Bureau without incurring unreasonable costs or burdens. 11(f) Denial of application Proposed Rule Proposed § 1041.17(e) would have provided that the Bureau deny the application of an entity seeking preliminary approval for registration pursuant to proposed § 1041.17(c)(1), registration pursuant to proposed § 1041.17(c)(2), or provisional registration pursuant to proposed § 1041.17(d)(1) if the Bureau made any of three determinations. First, if the Bureau determines that the entity did not satisfy the conditions set forth in proposed § 1041.17(b), or, in the case of an entity seeking preliminary approval for registration, was not reasonably likely to satisfy the conditions as of the deadline set forth in proposed § 1041.17(c)(3)(ii). Second, if the Bureau determines that the entity’s application was untimely or materially inaccurate or incomplete. Third, if the Bureau determines that preliminary approval, provisional registration, or registration would pose an unreasonable risk to consumers. Final Rule The Bureau did not receive comments on proposed § 1041.17(e). Therefore, the Bureau adopts § 1041.17(e) as proposed except that, as described above, the Bureau has renumbered this provision as § 1041.11(f) of the final rule. 1237 11(g) Notice of material change Proposed Rule Proposed § 1041.17(f) would have required a provisionally registered or registered information system to provide to the Bureau a written description of any material change to information contained in its application for registration submitted pursuant to proposed § 1041.17(c)(2) or provisional registration submitted pursuant to proposed § 1041.17(d)(1), or to information previously provided to the Bureau pursuant to proposed § 1041.17(f), within 14 days of any such change. Final Rule The Bureau did not receive comments on proposed § 1041.17(f). Therefore, the Bureau adopts § 1041.17(f) as proposed except that, as described above, the Bureau has renumbered this provision as § 1041.11(g) of the final rule. 11(h) Revocation Proposed Rule Proposed § 1041.17(g)(1) would have provided that the Bureau would suspend or revoke an entity’s preliminary approval for registration, provisional registration, or registration, if it determined either that the entity had not satisfied or no longer satisfied the conditions described in proposed § 1041.17(b); or that it had not complied with the requirement described in proposed § 1041.17(f); or that preliminary approval for registration, provisional registration, or registration of the entity posed an unreasonable risk to consumers. Proposed § 1041.17(g)(2) would allow the Bureau to require additional information and documentation from an entity if it had reason to believe suspension or revocation under proposed 1238 § 1041.17(g)(1) may be warranted. Proposed § 1041.17(g)(3) stated that, except in cases of willfulness or those in which the public interest required otherwise, prior to suspension or revocation under proposed § 1041.17(g)(1), the Bureau would issue written notice of the facts or conduct that could warrant the suspension or revocation and grant an opportunity for the entity to demonstrate or achieve compliance with proposed § 1041.17 or otherwise address the Bureau’s concerns. Proposed § 1041.17(g)(4) would allow the Bureau to revoke an entity’s preliminary approval for registration, registration, or provisional registration if the entity submitted a written request to the Bureau that its preliminary approval for registration, registration, or provisional registration be revoked. Proposed § 1041.17(g)(5) provided that for the purposes of §§ 1041.5 and 1041.7— which would require a lender making most covered loans to obtain and consider a consumer report from a registered information system—suspension or revocation of an information system’s registration would become effective five days after the date that the Bureau published notice of the suspension or revocation on its website. It also provided that, for purposes of proposed § 1041.16(b)(1), suspension or revocation of an information system’s provisional registration or registration would be effective on the date that the Bureau published notice of the revocation on its website. Finally, proposed § 1041.17(g)(5) provided that the Bureau would also publish notice of a suspension or revocation in the Federal Register. Final Rule The Bureau did not receive comments on proposed § 1041.17(g). However, the Bureau is finalizing it as § 1041.11(h) with one change. The Bureau has added § 1041.11(h)(6) to clarify that, if it suspends the provisional registration or registration of an information system, it will 1239 provide instructions to lenders concerning the scope and terms of such suspension. For example, depending on the facts and circumstances of a particular determination that suspension is appropriate, the Bureau may suspend registration of a provisionally registered information system or registered information system for purposes of §§ 1041.5 and 1041.6 only; lenders may still be required to furnish to the provisionally registered information system or registered information system pursuant to § 1041.10. The Bureau may also determine that suspension is only appropriate for a certain period of time. 11(i) Administrative Appeals The Bureau added § 1041.11(i), which provides a process for entities to submit to the Bureau an administrative appeal in certain circumstances. According to § 1041.11(i) of the final rule, an entity may appeal: a denial of its application for preliminary approval for registration pursuant to § 1041.11(c)(1), registration under § 1041.11(c)(2) or (d)(2), or provisional registration under § 1041.11(d)(1); and a suspension or revocation of its preliminary approval for registration pursuant to § 1041.11(c)(1), registration under § 1041.11(c)(2) or (d)(2), or provisional registration under § 1041.11(d)(1). The subsections of § 1041.11(i) of the final rule address other matters pertinent to administrative appeals. Section 1041.11(i)(1) sets out the grounds for administrative appeal while under § 1041.11(i)(2), an entity has 30 business days to submit an appeal from the date of the determination, although the Bureau may extend this time for good cause. Section 1041.11(i)(3) sets forth the form and content of the administrative appeal, which shall be submitted by electronic means as set forth on the Bureau’s website. Section § 1041.11(i)(4) establishes the appeals process and that the filing and pendency of an appeal does not by itself 1240 suspend the determination that is the subject of the appeal during the appeals process, but grants the Bureau discretion to suspend the determination that is the subject of the appeal during the appeals process. Lastly, § 1041.11(i)(5) specifies that the Bureau has the power to decide whether to affirm or reverse the determination in whole or in part, and requires the Bureau to notify the appellant of this decision in writing. The Bureau concluded that modifying the proposal to add § 1041.11(i) is consistent with the tenets of due process and administrative law and affords entities under its supervisory authority, including registered information systems, more clarity and transparency about their rights in the event that they receive an adverse determination from the Bureau pursuant to any of the provisions of § 1041.11. Section 1041.12 Compliance Program and Record Retention Overview of the Proposal The Bureau proposed § 1041.18 to require a lender that makes a covered loan to develop and follow written policies and procedures that are reasonably designed to ensure compliance with part 1041 and that are appropriate to the size and complexity of the lender and its affiliates and the nature and scope of their covered loan activities. The Bureau also proposed to require a lender to retain evidence of compliance with the requirements in part 1041 for 36 months after the date a covered loan ceases to be an outstanding loan. Specifically, the Bureau proposed to require a lender to retain several types of documentation and loan-level records. It proposed both requirements pursuant to its authority to prevent unfair or abusive acts or practices under section 1031 of the Dodd-Frank Act and for the reasons discussed below. 1241 The Bureau stated that the proposed requirement to develop and follow written policies and procedures would help foster compliance with proposed part 1041,1087 which would have prescribed detailed ability-to-repay and payment collection requirements that were generally more comprehensive than the requirements in States that permit lenders to make covered loans.1088 To make covered loans that comply with part 1041 when they are originated and when they are outstanding, proposed § 1041.18 would have required lenders to develop written policies and procedures to reasonably ensure that their staff understands the proposed requirements and conducts covered loan activities in accordance with the proposed requirements. In facilitating lender compliance with these requirements, the proposed compliance program requirements would have helped to prevent the identified unfair and abusive practices addressed in part 1041. As discussed above in part III, the Bureau has extensive experience to date in using its supervisory authority to examine the operations of certain payday lenders and its enforcement authority to investigate the acts or practices of payday lenders. Based on that experience, as well as through its general market outreach, the Bureau believed that it may be useful to provide greater specificity as to the record retention requirement than is typical in many other Federal consumer financial regulations, which are usually phrased in more general terms.1089 In the 1087 A written policies and procedures requirement is a requirement in other Bureau rules. See, e.g., Regulation E, 12 CFR 1005.33(g)(1). 1088 See discussion of the current regulatory environment by product type in part II above. 1089 The Bureau believed that record retention was necessary to prove compliance with a rule and was a common requirement across many of the Bureau’s rules. See, e.g., Regulation B, 12 CFR 1002.12; Regulation Z, 12 CFR 1026.25. In this context, the Bureau noted that it had found it necessary to levy a civil penalty of $5 million against a large payday lending company for engaging in the destruction of records around one of the Bureau’s initial supervisory examinations in this market, which had included continuing to shred documents for weeks, even after 1242 Bureau’s experience, current record retention practices vary widely across the industry, depending on lender business practices, technology systems, State regulatory requirements, and other factors, but often have proved to be problematic.1090 Particularly given that ability-to-repay determinations would likely involve different levels of automation and analysis from lender to lender, the Bureau believed that providing an itemized framework listing the nature and format of records that must be retained would help reduce regulatory uncertainty and facilitate supervision by the Bureau and other regulators. The Bureau also noted that the level of detail in the proposed record retention requirements was similar to the level of detail in the recordkeeping obligations in the small-dollar lending statutes and regulations of some States.1091 Given that part 1041 would have imposed requirements tied to, among other things, checking the records of the lenders and its affiliates regarding a consumer’s borrowing history and verifying a consumer’s income and major financial obligations, the Bureau believed that the record retention requirements proposed in § 1041.18(b) would assist a lender in complying with the requirements in part 1041. By providing a non-exhaustive list of records that would need to be retained in proposed § 1041.18(b)(1) to (b)(5), proposed § 1041.18(b) would help covered persons determine whether a contemplated covered loan would comply with the requirements in part 1041 and aid covered persons in complying with the record retention requirements. Furthermore, the proposed record retention requirements would support the external supervision Bureau examiners told employees to halt such activities. See Consent Order, In re Cash America Int’l, Inc., No. 2013-CFPB-0008 (Nov. 20, 2013). 1090 Bureau of Consumer Fin. Prot., “Supervisory Highlights,” at 16 (Spring 2014), available at http://files.consumerfinance.gov/201405_cfpb_supervisory-highlights-spring-2014.pdf (“At multiple lenders, policies and procedures for record retention either did not exist or were not followed, leading to incomplete record destruction logs and improperly destroyed records.”). 1091 See, e.g., Colo. Code Regs. sec. 902–1–10; Wash. Admin. Code sec. 208–630–610. 1243 of lenders for compliance with part 1041. In facilitating lender compliance and helping the Bureau and other regulators assess compliance with the requirements in part 1041, the proposed record retention requirements would help prevent and deter the identified unfair and abusive practices addressed in part 1041. Comments Received A number of industry commenters disagreed with the Bureau’s general approach in the proposal, describing the recordkeeping provisions as overly stringent, unnecessarily prescriptive, and disproportionate to any benefit for consumers. They also suggested that the Bureau should pursue less burdensome alternatives than requiring borrower information to be maintained electronically. By contrast, consumer groups recommended expanded record retention provisions, partly to ensure that lenders report to the Bureau sufficient information about loans and borrowers. They suggested twenty additional, non-exhaustive data points for the Bureau to analyze under an expanded requirement to retain more records. They also suggested that lenders should report aggregate data to the Bureau at least annually, that the Bureau should create a searchable public database of such information, and that the Bureau should publish an annual report—based on both retained and aggregate data—to demonstrate whether the rule is proving to be effective in achieving its purposes. Another commenter requested that the Bureau create a review process of lender practices for lender portfolios of covered loans that perform unusually poorly over time. This commenter also supported making more of the retained information available to the public for scrutiny. 1244 Several commenters urged that classes of lenders, such as State-regulated entities, should be exempted from compliance with the proposed rule, including its compliance program and record retention requirements. Trade associations, including those for credit unions, advocated for more sweeping exemptions of entire categories of lenders from coverage under the rule. A group of chief legal officers from certain States also supported exempting those lenders that are already covered by such State and local regulatory systems from coverage under the proposal, citing Alabama and Idaho as particular examples of State regulatory systems that they viewed as operating effectively. Some industry commenters were critical of the Bureau for not exempting small businesses and other small entities from coverage under the proposed rule’s compliance program and record retention requirements. One commenter acknowledged, but disagreed with, the Bureau’s stated rationale that small lenders are not engaged in meaningfully different practices from other lenders that offer the same types of loans. Others noted that the costs and burdens of meeting any new and additional requirements tended to fall disproportionately heavily on small entities. Commenters with experience in documenting loans in accordance with existing laws asserted that the recordkeeping requirements were not specific enough for lenders to determine accurately the associated costs, and advanced that to make such determinations, more information was needed about format, content, retention, among other factors. A few commenters noted that some of the recordkeeping requirements contained in the proposal could be satisfied if regulators could access the consistent, real-time information that lenders would furnish to registered information systems, which then could reduce costs and burdens to both 1245 lenders and regulators while being more conducive to review and analysis. They also noted that the proposal would cause the regulatory authorities themselves to incur substantial costs to compile, review, and analyze the records they receive from lenders, especially if they are maintained in different formats or contain different content. Several industry commenters noted that the practical effect of conditional exemptions from certain provisions of the rule was likely to be limited if compliance and records retention requirements still had to be met, as they believed would be the case. Some industry commenters cautioned that the record retention requirements could expose consumers and lenders to significant operational risks to the security of their data. Final Rule In § 1041.12 of the final rule, renumbered from proposed § 1041.18, the Bureau has decided to maintain the same general approach to the compliance and record retention requirements as was framed in the proposal. In particular, the final rule requires lenders that make covered loans to develop and follow written policies and procedures that are reasonably designed to ensure compliance with the rule’s requirements. Such policies and procedures must be appropriate to the size and complexity of the lender and its affiliates and the nature and scope of its covered loan activities. The final rule requires lenders to retain evidence of compliance and includes a non-exhaustive list of the types of loan-level records and documentation that lenders are required to retain. However, because the scope of coverage has changed from the proposed rule to the final rule to omit the underwriting requirements for covered longer-term loans other than covered longer-term balloon-payment loans, the compliance program and record retention requirements of the final rule are narrower as well. In addition, the final rule exempts 1246 from the compliance program and record retention requirements alternative loans pursuant to § 1041.3(e), and accommodation loans pursuant to § 1041.3(f), regardless of the type of lender. The Bureau notes, however, that lenders making alternative loans must maintain and comply with policies and procedures documenting proof of recurring income, as specified as a condition of the exemption in the final rule. The commentary to the final rule contains changes that conform to the modifications made in the final rule. Several commenters raised issues about the potential burden on lenders and the level of detail required by the proposal, yet the Bureau has determined that the record retention and compliance program requirements will foster compliance with the final rule and as such will benefit consumers. Although the record retention requirements are the same regardless of the size of the lender’s operation, the compliance program requirements are calibrated to the size and complexity of the lender and its affiliates, and the nature and scope of the covered lending activities of the lender and its affiliates. Lenders’ written policies and procedures must be reasonably designed to ensure compliance with the final rule but the Bureau’s regulatory expectation is for lenders to develop compliance programs that are commensurate with their size and complexity and the scope of their offered products. Accordingly, although the compliance program and record retention requirements may increase lenders’ regulatory responsibilities, the Bureau concludes that the requirements of the final rule will not be overly burdensome for such lenders. In the final rule, the Bureau has opted to continue to include detailed record retention requirements in order to reduce regulatory uncertainty and facilitate supervision by the Bureau and other regulators. It concludes that this level of detail is necessary because part 1041 is establishing a new regulatory regime, which includes flexible underwriting requirements and 1247 limitations on payment attempts. It is important that lenders are aware of what records they need to maintain to demonstrate compliance. In addition, it is important that the Bureau and other regulators are able to use those records to evaluate whether lenders are complying with the rule’s requirements. Some commenters noted that the record retention requirements may increase the costs incurred by regulatory authorities to compile, review, and analyze any records they receive from lenders, especially if the records are maintained in different formats or contain different content. The Bureau finds that the format and content differences in the materials retained by lenders will not impact the overall benefit of the compliance program and record retention requirements. The Bureau would prefer to bear the costs of reviewing such records in different formats rather than pass those costs on to lenders by imposing more specific format requirements. Several commenters suggested that whole categories of lenders should be exempted from compliance with the final rule’s compliance or record retention requirements because they are already subject to State or Federal regulation, such as credit unions or banks, or because they are small businesses. The Bureau’s approach to the final rule remains primarily focused on the kinds of loans lenders provide and how they impact consumers, not on the type or size of lenders. As noted above, the Bureau has concluded that it will exclude several categories of loans from coverage of the rule, in part, because they do not present the same kinds of consumer risks and harms as the covered loans addressed by part 1041. Providers of those excluded loans who do not also offer covered loans will not be subject to the compliance program and reporting requirements in § 1041.12 of the final rule. For providers of covered loans, the compliance program and record retention required by the final rule will assist them in complying with the 1248 substantive requirements of the rule, benefit supervisory and monitoring efforts, and thus help deter unfair and abusive practices. The Bureau thus has concluded that based on these benefits, the record retention and compliance program requirements in the final rule should apply to all lenders of covered loans, and that it should not exempt any particular class of lenders. The Bureau continues to observe that most small lenders are not engaged in meaningfully different practices from other lenders that offer the same types of loans. Accordingly, the Bureau has decided not to carve out any exceptions for small businesses from the compliance program and record retention requirements of the final rule. Several commenters recommended that the Bureau require lenders to retain additional specific information and that lenders periodically report to the Bureau about their loan data and lending practices. The Bureau is not requiring additional reporting requirements in the final rule at this time, based in part on the comments it received raising concerns about the perceived regulatory burden related to the existing components of the proposed compliance program and record retention requirements. In addition, the Bureau concludes that it is premature to establish a blanket reporting requirement for all lenders, given that regulators may want different information for different supervisory or monitoring purposes. In the same vein, the Bureau is not adopting the recommendation by some commenters to make the reported information available to the public. Likewise, the Bureau is not increasing lenders’ requirements to report to the registered information systems as a means of having real-time data available for review for compliance and monitoring purposes, as some commenters suggested. Although real-time access to such data might serve the supervisory purposes of the Bureau and other regulators, it would be contrary to 1249 the Bureau’s decision to ease some of the burdens of the reporting requirement to the registered information systems in the final rule, as discussed earlier. Many commenters discussed the increased costs associated with the proposed compliance program and record retention requirements, and several cautioned that the record retention requirements could expose consumers and lenders to significant operational risks for the security of their data. The Bureau has considered all of these concerns about the increase in costs to lenders and the industry as a whole and has concluded that the benefits to consumers and the marketplace outweigh concerns about the costs to industry, but those costs should not be exacerbated by adding further burdens at this time of initiating a new Federal regulatory framework. Finally, the Bureau disagrees that the compliance program and record retention requirements increase risks for the security of the consumer data. Providers of covered loans are already subject to legal obligations to secure the data of their consumers under the Safeguards Rule1092 and the final rule does not change those obligations. If lenders are meeting those obligations in their everyday operations, then the additional information that the rule requires them to retain should not affect the security of consumer data. 12(a) Compliance program. Proposed Rule 1092 Standards for Safeguarding Customer Information, 16 CFR part 314. This regulation was promulgated and is enforced by the FTC pursuant to its specific authority under the Gramm-Leach-Bliley Act, 15 U.S.C. 6801-6809. See earlier discussion regarding the requirements of the Safeguards Rule in the discussion of final rule section 11. In particular, Congress did not provide the Bureau with rulemaking, enforcement, or supervisory authority with respect to the GLBA’s data security provisions. 15 U.S.C. 6801(b), 6804(a)(1)(A), and 6805(b). The portion of the GLBA concerning data security is not a Federal consumer financial law under the Dodd-Frank Act; the Bureau does not have authority with respect to the GLBA data security provisions. However, data security practices that violate those GLBA provisions and their implementing regulations may also constitute unfair, deceptive, or abusive acts or practices under the Dodd-Frank Act. 1250 In proposed § 1041.18(a), the Bureau would have required a lender making a covered loan to develop and follow written policies and procedures that are reasonably designed to ensure compliance with part 1041 and that are appropriate to the size and complexity of the lender and its affiliates and the nature and scope of their covered loan activities. Proposed comments 18(a)1 and 18(a)-2 explained and provided examples of the proposed requirements. Comments Received One trade association noted that the proposal would require lenders to develop corresponding policies, which may then grow in complexity if multiple vendors provide the underlying hardware and software infrastructure for origination systems. A number of industry commenters stated that the compliance requirements would substantially increase the costs for providing covered loans, which will either restrict the availability of such credit or make it more costly as these higher compliance costs are passed on to consumers. Several commenters noted that this is particularly a problem for small entities, where the costs of compliance can feel especially heavy and disproportionate to their business operations that lack much scale. Final Rule After considering the many comments made on the proposal, the Bureau has decided to finalize § 1041.12(a) as it was proposed (and now renumbered from proposed § 1041.18(a)). The provision states that a lender making a covered loan must develop and follow written policies and procedures that are reasonably designed to ensure compliance with the final rule’s requirements. The written policies and procedures must be appropriate to the size and complexity of the lender and its affiliates, and the nature and scope of the covered loan activities. 1251 The commentary to § 1041.12(a) of the final rule differs from the proposed commentary because of technical changes to update the relevant references to the final rule, rather than to the proposed rule. Moreover, throughout, it deletes references to provisions in the proposed rule that would have covered the underwriting of all covered longer-term loans but were omitted from the final rule. By modifying the scope of the final rule from the proposed rule, the Bureau thereby has altered the compliance program requirements in the final rule. Comment 12(a)-2 of the final rule modifies the reference to “covered short-term loan” by replacing it with “covered loan” to align it more accurately with the terms of the final rule, which also applies the ability-to-repay underwriting requirements to covered longer-term balloon-payment loans. It also specifies that lenders who make such loans under § 1041.5 of the final rule have to develop and follow written policies and procedures to ensure compliance with the ability-to-repay requirements set out in modified form in § 1041.5 of the final rule. For instance, the example in the commentary no longer includes a discussion of the need for lenders to develop and follow policies and procedures regarding estimating housing expenses because under final § 1041.5(c)(iii), lenders can rely on borrower’s statements of rental expenses, rather than follow the proposal’s requirement that the lender estimate those expenses. And, as discussed above, the commentary to § 1041.12(a) of the final rule has been modified based on changes to the scope of the final rule declining to apply the ability-to-repay underwriting criteria to all covered longer-term loans. Commenters raised concerns about the complexity of the required policies and procedures, given the underlying complexity of the proposed rule’s requirements. They also expressed concern about the cost of developing compliance systems, especially for smaller lenders, and predicted that such costs are likely to be passed on to consumers. These general 1252 concerns have already been considered and addressed in the discussion above, yet they also militate in favor of maintaining a certain amount of flexibility. In this regard, it bears emphasis that this provision requires lenders to develop and follow policies and procedures that are reasonably designed to ensure compliance with the requirements of the final rule. The written policies and procedures must be appropriate to the size and complexity of the lender and its affiliates, and to the nature and scope of the covered loan activities. In short, the final rule is not a one-size-fits-all approach. And because of changes made in the scope of coverage under the final rule, the compliance costs highlighted by commenters that were reacting to the proposed rule will be less than they anticipated. The Bureau thus has determined at this time that the final rule appropriately takes into account the size and complexity of lenders’ operations and will not create unreasonable compliance costs or burdens on lenders. 12(b) Record retention Proposed Rule Proposed § 1041.18(b) would have required a lender to retain evidence of compliance with part 1041 for 36 months after the date a covered loan ceases to be an outstanding loan. The Bureau believed, in general, that the proposed period would be appropriate for purposes of record retention, and it would give the Bureau and other Federal and State enforcement agencies time to examine and conduct enforcement investigations in the highly fragmented small-dollar lending market that could help address and prevent the unfair and abusive practices that the Bureau had identified as a preliminary matter. The Bureau believed that the proposed requirement to retain records for 36 months after a covered loan ceases to be an outstanding loan also would not impose an undue burden on a lender. The Bureau believed that the proposed 1253 record retention requirements would have promoted effective and efficient supervision and enforcement of part 1041, thereby further preventing and deterring the unfair and abusive acts the Bureau proposed to identify. The Bureau also proposed to specify requirements as to the format in which certain records would have to be retained. In particular, the proposed approach would have provided flexibility as to how lenders could retain the loan agreement and documentation obtained in connection with a covered loan from the consumer or third parties, while requiring that the lender retain various other records that it generates in the course of making and servicing loans in an electronic tabular format such as a spreadsheet or database, so as to facilitate analysis both by the lender and by its external supervisors. Specifically, proposed § 1041.18(b)(1) would have required a lender of a covered loan either to retain the original version of the loan agreement or to be able to reproduce an image of it and certain documentation obtained from the consumer or third parties in connection with a covered loan. That additional documentation would include, as applicable, the following items: a consumer report obtained from a registered information system; verification evidence; any written statement obtained from the consumer; authorization of an additional payment transfer; and an underlying one-time electronic transfer authorization or underlying signature check. These matters were further described and clarified in the proposed commentary. Proposed § 1041.18(b)(2) would have required a lender to retain electronic records in tabular format of certain calculations and determinations that it would have been required to make in the process of making a covered loan. A lender would, at a minimum, have been 1254 required to retain the records listed in proposed § 1041.18(b)(2), as explained further in the proposed commentary. Proposed § 1041.18(b)(3) would have required a lender to retain electronic records in tabular format for a consumer who qualifies for an exception to or overcomes a presumption of unaffordability for a covered loan in proposed § 1041.6, § 1041.12(a), or § 1041.10. A lender would, at a minimum, have been required to retain the records listed in proposed § 1041.18(b)(3), as explained further in the proposed commentary. Proposed § 1041.18(b)(4) would have required a lender to retain electronic records in tabular format on a covered loan’s type and terms. A lender would, at a minimum, have been required to retain the records listed in proposed § 1041.18(b)(4), and as explained further in the proposed commentary. Proposed § 1041.18(b)(5) would have required a lender to retain electronic records in tabular format on payment history and loan performance for a covered loan. A lender would, at a minimum, have been required to retain the records listed in proposed § 1041.18(b)(5), and as explained in the proposed commentary. Comments Received Industry commenters asserted that the length of the proposed record retention period was excessive, unjustified, and not in line with existing Federal law, and several advocated for a shorter period. Many relied on the 25-month record retention requirements of the Equal Credit Opportunity Act as the basis for recommending a shorter period. Another commenter supported the proposed record retention period as an appropriate length of time, joined by others that pointed to the furnisher requirements under the FCRA to retain substantiation for 36 months. 1255 Some commenters contended that the 36-month period would not impose an undue burden on lenders. Consumer groups believed that an even longer retention period is justified in light of the requirements already imposed on lenders who typically must substantiate any information they report to consumer reporting agencies for 36 months or more. If the period is not lengthened, they urged the Bureau to specify that this rule does not affect any record retention requirement imposed under any other Federal or State law, including those for substantiating information furnished to a consumer reporting agency and Federal standards for safeguarding consumer information. Industry commenters also viewed the proposed formatting requirements and mandatory data points as too complex and onerous. They said the electronic tabular format as framed in the proposal was too specific and the data points were too detailed, and that compliance with these requirements would force lenders to develop new systems at substantial cost. Some of this discussion of cost was directed at covered longer-term loans made by traditional installment lenders, but much of it was directed at covered short-term loans. Many commenters claimed that the record retention provisions, including the electronic tabular format, would likely impose large operating costs that would either cause lenders to exit the market or be passed on to consumers. They suggested that the Bureau should pursue less burdensome alternatives than requiring borrower information to be maintained electronically. Commenters noted that that lenders maintain many of the records required under the proposal, but they often do not have one system of record and predicted that the required information would have to be manually entered into an electronic tabular format. 1256 Several industry commenters expressed concerns that the recordkeeping burden was the same for lenders who offered loans under the conditional exemption (proposed § 1041.7) as for those who offered loans subject to the underwriting requirements. Credit unions noted that PAL loans would also be subject to the record retention requirements and expressed concern about the attendant added costs. Industry commenters, including credit unions and banks, contended that they already follow certain recordkeeping requirements pursuant to existing regulatory oversight by other Federal and State authorities. They asserted that they can provide such information when requested and thus the electronic tabular format described in the proposal is unnecessary. They regarded the proposal’s requirements as more stringent than parallel rules applicable to lenders of other types of credit. One commenter supported the electronic tabular format as a reasonable approach to the kind of recordkeeping needed to monitor compliance with the proposed rule, and stated that lenders will save on costs by accepting and storing records electronically. Final Rule The Bureau is finalizing the opening paragraph of § 1041.12(b) unchanged from proposed § 1041.18(b), other than being renumbered to reflect other modifications made in the rule as discussed earlier. This provision requires a lender to retain evidence of compliance with the final rule for 36 months after the date on which a covered loan ceases to be an outstanding loan. In particular, the Bureau has concluded that the 36-month record retention period contained in the proposal is appropriate here for several reasons. First, it would provide the 1257 Bureau and other Federal and State enforcement agencies with an appropriate and practical amount of time to examine and conduct enforcement investigations in order to prevent and deter the unfair and abusive practices identified in the final rule. Record retention provisions are common in Federal consumer financial law to facilitate effective supervisory examinations, which depend critically on having access to the information necessary to assess operations, activities, practices, and legal compliance.1093 If the record retention period were reduced, it could be considerably more difficult to ensure that the necessary information and records would remain routinely available for proper oversight of the industry. The Bureau is in a position to evaluate such issues from its experience and perspective of exercising supervision and enforcement authority over this industry, as it has done now for the past several years, as described above in part III. That experience has led the Bureau to perceive that there are some special challenges of oversight in this industry, including around the topic of record retention.1094 Second, the 36-month time frame fits relatively comfortably within the other recordkeeping requirements provided under other consumer financial laws, paralleling the FCRA in particular. And though some statutes and regulations provide for shorter periods, the highly fragmented small-dollar lending market argues for a somewhat longer record retention period in order to facilitate the Bureau and other regulators in covering more of the industry while maintaining reasonably spaced examination cycles. 1093 As noted earlier, record retention is necessary to prove compliance with a rule and is a common requirement across many of the Bureau’s rules. See, e.g., Regulation B, 12 CFR 1002.12; Regulation Z, 12 CFR 1026.25. 1094 See, e.g., Bureau of Consumer Fin. Prot., Supervisory Highlights, at 16 (Spring 2014) (“At multiple lenders, policies and procedures for record retention either did not exist or were not followed, leading to incomplete record destruction logs and improperly destroyed records.”); Consent Order, In re Cash America Int’l, Inc., File No. 2013CFPB-0008 (Nov. 20, 2013) (levying civil penalty for ongoing destruction of records that were needed to conduct an examination), available at http://files.consumerfinance.gov/f/201311_cfpb_cashamerica_consent-order.pdf. 1258 Third, given that some record retention period is virtually inevitable in this market for all the reasons stated, the 36-month retention period would be unlikely to impose an undue burden on lenders, as some commenters noted, when viewed in light of the marginal difference in cost or burden between, say, a 24-month period or a 36-month period. That is especially so given that it is increasingly common even for smaller entities to maintain their lending records on computers. The commentary to § 1041.12(b) of the final rule was modified to consolidate references previously found in the proposed commentary for the individual subsections. New comment 1041.12(b)-1 now clarifies that items listed in final § 1041.12(b) – documentation and information in connection with the underwriting and performance of covered short-term loans and covered longer-term balloon-payment loans, as well as payment practices in connection with covered loans, generally – are non-exhaustive as to the records that may need to be retained as evidence of compliance with part 1041. The Bureau has finalized § 1041.12(b)(1) in a slightly reorganized form. Other than its organizational structure, it is in substantially the same form as proposed, except for changes that clarify that the loan agreement and documentation that lenders must retain relates to that which lenders obtained in connection with originating a covered short-term or covered longer-term balloon payment loan, not a “covered loan” as described in the proposal. Other changes are technical in nature to make references to the final rule accurate. In particular, the list of required documentation in final § 1041.12(b)(1)(i)-(iii) no longer references proposed § 1041.9(c)(3), which pertained to the ability-to-repay requirements for the covered longer term loans that were included in the proposal but have not been retained in the final rule. It continues to require retention of consumer reports from registered information systems (i), as well as verification 1259 evidence (ii) and written statements (iii) under § 1041.5. It clarifies that the consumer reports must be from an information system that has been registered for 180 days or more pursuant to final § 1041.11(c)(2) or is registered with the Bureau pursuant to § 1041.11(d)(2). However, the requirements in proposed (iv) and (v) that relate the requirements relating to proposed § 1041.14 (renumbered as final § 1041.8) are now found in a new § 1041.12(b)(4) regarding retention of certain records pertaining to payment practices for covered loans. To reflect the addition of comment 1041.12(b)-1, the proposed comment 1041.18(b)(1)-1 was deleted. New comment 1041.12(b)(1)-1 is substantially the same as 1041.18(b)(1)-2 in the proposal. It reflects technical changes, including those to clarify that the provision relates to covered short-term or covered longer-term balloon-payment loans and describes the methods of retaining loan agreement and documentation for short-term or covered longer-term balloon payment loans, including in original form or being able to reproduce an image of the loan agreement and documentation. In addition, the commentary to proposed § 1041.18(b)(ii) was deleted, as it referred to estimates of housing expenses. In light of other substantive changes to the final rule, § 1041.12(b)(2) is more streamlined than the proposed rule. As in the proposal, it requires lenders of covered loans to retain electronic records in tabular format that include specific underwriting information for covered loans under § 1041.5 of the final rule. The final rule clarifies that lenders must retain electronic records in tabular format regarding origination calculations and determinations for covered shortterm or covered longer-term balloon-payment loans under § 1041.5. The list of required information is reduced somewhat from the proposal because it no longer includes references to the timing of net income or of major financial obligations, and it no longer requires the retention 1260 of information about the underwriting of covered longer-term loans (other than covered longerterm balloon-payment loans). These changes to the record retention provisions thus mirror the corresponding changes made to the substantive underwriting requirements in § 1041.5 of the final rule. The information that lenders must retain under § 1041.12(b)(2)(i)-(iv) includes: the projection made by the lender of the amount of a consumer’s income; the projections made by the lender of the amounts of the consumer’s major financial obligations; calculated residual income or debt-to-income ratio; and, the estimated basic living expenses for the consumer. The Bureau also added new § 1041.12(b)(2)(v), which requires the retention of other information considered in making the ability-to-repay determinations to clarify that the enumerated list, as stated in the commentary, is non-exhaustive. The commentary to this provision is substantially similar to the proposal but reflects those other substantive and technical changes that were made to the final rule. Proposed comment 1041.18(b)(2)-1 was not finalized because its content is addressed in final comment 1041.12(b)-1, as discussed above. The Bureau finalized former comment 1041.18(b)(2)-2, as comment 1041.12(b)(2)-1. It discusses the requirement that lenders retain records in an electronic tabular format and clarifies, as was proposed, that a lender would not have to retain records under this section in a single, combined spreadsheet or database with the other records required by the provisions of 1041.12(b). It notes, however, that § 1041.12(b)(2) requires a lender to be able to associate the records for a particular covered short-term or covered longer-term balloon-payment loan with a unique loan and consumer identifiers in § 1041.12(b)(3). In § 1041.12(b)(3) of the final rule, the Bureau did not finalize the requirement to retain electronic records in a tabular format for a consumer that qualifies for an exception to or 1261 overcomes a presumption of unaffordability for a covered loan. It thus has eliminated this provision and renumbered the subsequent subsections. The Bureau did not include this provision in the final rule because the presumptions of unaffordability in proposed § 1041.6 have been eliminated from the rule. The commentary reflects these same changes. The Bureau has finalized in § 1041.12(b)(3) provisions proposed in § 1041.18(b)(4) with changes from the proposal that reflect some reorganization of provisions formerly found in subsection (b)(5) and technical changes to address the modification of references from the proposal to the final rule. In particular, this renumbered provision requires lenders to retain electronic records in tabular format regarding loan type, terms, and performance of covered short-term or covered longer-term balloon-payment loans. The final rule now includes the requirement that lenders of such loans retain: the applicable information listed in § 1041.10(c)(1) and (2) of the final rule; whether the lender obtained vehicle security from the consumer; the loan number in a sequence of covered short-term loan, covered longer-term balloon-payment loans, or a combination thereof; information regarding loans not paid in full by the due date; for a loan with vehicle security, whether repossession of the vehicle was initiated; the date of last or final payment received; and, the information listed in § 1041.10(c)(3). The Bureau also deleted language from the proposal that would have covered matters that are now treated elsewhere in the final rule. The related commentary reflects similar changes, including the reorganization of several subsections. Proposed comment 1041.18(b)(3)-1 was not finalized. Former comment 1041.18(b)(3)-2, now renumbered as 1041.12(b)(3)-1 explains the requirement for lenders to retain records regarding loan type, terms, and performance of covered longer-term balloon 1262 payment loans in an electronic tabular format and notes that the records are not required to be in a single, combined spreadsheet or database with the other records required by the provisions of 1041.12(b); however, it states that § 1041.12(b)(3) requires that the lender be able to associate a particular covered short-term or covered longer-term balloon-payment loan with unique loan and consumer identifiers in § 1041.12(b)(3). Of note, the requirements formerly outlined in proposed § 1041.18(b)(5)(iii) regarding retaining information about past due loans has been altered in final § 1041.12(b)(3)(iv). The proposal required that a lender retain information on the maximum number of days, up to 180, any full payment, as defined, was past due in relation to the payment schedule. The final rule § 1041.12 (b)(3)(iv) instead requires that lenders retain information “for any full payment on the loan that was not received or transferred by the contractual due date, the number of days such payment was past due, up to a maximum of 180 days.” Final comment 12(b)(3)(iv)-1 explains that under § 1041.12(b)(3)(iv), a lender that makes a covered loan must retain information regarding the number of days any full payment is past due beyond the payment schedule established in the loan agreement, up to 180 days. The comment defines “full payment” as principal, interest, and any charges and explains that if a consumer makes a partial payment on a contractual due date and the remainder of the payment 10 days later, the lender must record the full payment as being 10 days past due. If a consumer fails to make a full payment more than 180 days after the due date, the lender must only record the full payment as being 180 days past due. With the adjustments to other sections of § 1041.12(b), the Bureau is finalizing § 1041.12(b)(4) to focus on the retention of documents regarding payment practices generally, 1263 as they relate to all covered loans. It contains many of the provisions originally in proposed § 1041.18(b)(4) with some adjustments. It requires lenders to retain certain payment-related records for covered loans. Like final § 1041.12(b)(1), a lender must retain or be able to reproduce an image of the required records. Lenders do not need to retain these documents in an electronic tabular format, which for many of the required documents reflects a change from the proposal. The records include leverage payment mechanisms with respect to covered longerterm loans, authorizations of additional payment transfers, and underlying one-time electronic transfer authorizations. It reflects technical changes in the references and content of the final rule. The final commentary outlines methods of retaining documentation. In particular, as an example, comment 12(b)(4)-1 clarifies that a lender must either retain a paper copy of a leveraged payment mechanism obtained in connection with a covered longer-term loan or be able to reproduce an image of the mechanism. The Bureau is finalizing § 1041.12(b)(5) to require that lenders retain certain other records relating to payment practices for covered short-term or longer-term balloon-payment loans. However, unlike the records retained under § 1041.12(b)(4), these records must be retained in an electronic tabular format. The list of documents is the same as that proposed with one exception. Proposed § 1041.18(b)(5)(iii) has been rephrased and renumbered as § 1041.12(b)(3)(iv). The commentary related to the proposed section was moved to reflect this reorganization and any renumbering of provisions in the rule. The commentary explains that the lender does not have to retain the records required under § 1041.12(b)(3) in a single, combined spreadsheet or database with other records required by the provisions of 1041.12(b); however, it noted that § 1041.12(b)(5) requires a lender to be able to associate the records for a particular 1264 covered-short-term, or covered longer-term balloon-payment loan with a unique loan and consumer identifiers in § 1041.12(b)(3). With respect to § 1041.12(b) as a general matter, many commenters had objected to the scope of the information that lenders must retain under the proposal as complex, onerous, stringent, and burdensome. As noted above, the most major change in this regard is the change in the scope of coverage of the rule, which eliminated underwriting requirements for covered longer-term loans (other than covered longer-term balloon loans). Yet in light of the comments received, the Bureau has also lessened the record retention requirements in other respects. For example, the Bureau changed the method of retention required for some of the required records. In particular, it no longer is requiring lenders to retain certain records relating to payment practices in an electronic tabular format. Some commenters had expressed concern that even if loans were exempted from the ability-to-repay requirements, the lenders were still subject to the compliance program and record retention requirements. To address those concerns, the Bureau has exempted certain types of loans from coverage entirely—namely, alternative loans (§ 1041.3(e)), and accommodation loans (§ 1041.3(f))—including from the compliance program and record retention requirements. As a result, lenders that exclusively provide such loans will not be subject to the compliance program or record retention requirements. For lenders of covered loans, including loans that are conditionally exempted from § 1041.5 under § 1041.6, the Bureau concluded that retention of the documents and information enumerated in final § 1041.12(b)(1)(4) will suffice to facilitate lender compliance with the rule and the ability to examine for such 1265 compliance. As such, the retention of such documents will help prevent unfair and abusive practices. Some commenters objected to the application of the retention requirements to loans made pursuant to § 1041.6 of the final rule, arguing that the record retention requirements may deter lenders from making such loans. The Bureau believes that the record retention requirements are necessary to ensure that lenders are complying with the specific requirements of § 1041.6 which are designed to protect consumers in the absence of underwriting requirements. In addition, it notes that lenders of loans under § 1041.6 would not have to retain all of the information that relates to origination decisions for loans made under § 1041.5. The Bureau disagrees with the commenters that asserted records retention provisions are unnecessary because they already retain documents in accordance with other Federal consumer financial laws and can produce them when requested. The obverse of this argument is that it shows the supposed burdens of imposing these provisions are not significant for these entities. As outlined in the proposal, the Bureau’s experience is that current record retention practices vary widely across the industry, depending on lender business practices, technology systems, State regulatory requirements, and other factors. In addition, as mentioned above, the Bureau itself, in the context of its supervision and enforcement activities, has encountered difficulties at times with the industry’s handling of records. Accordingly, the Bureau has concluded that listing the specific nature and format of records to be retained will help reduce regulatory uncertainty and facilitate supervision by the Bureau and other regulators. That some lenders can easily produce these types of documents upon request does not undercut the Bureau’s conclusion that, based on its supervisory and enforcement experience, many lenders of covered loans do not have 1266 robust compliance management systems and would benefit from more guidance regarding compliance expectations. Indeed, as noted above, what it actually shows is that records retention is a functionality that can be managed successfully by these entities, especially as it is computerized and automated. The other principal objection that commenters made here concerned the requirement that much of the specified information is to be maintained in an electronic tabular format, which they claimed is complex, onerous, burdensome, and unnecessary. Other commenters, however, found this requirement to be a reasonable approach, and as outlined in the proposal, the Bureau sought to strike a balance that would allow lenders substantial flexibility to retain records in a way that would reduce potential operational burdens while also facilitating access and use by the lender itself and by the Bureau and other regulators. The Bureau has carefully considered the comments that it received and concludes that this requirement to retain records in an electronic tabular format should be relatively simple for lenders to carry out. That is especially so because lenders can create multiple spreadsheets or databases to capture the related sets of information, as long as they could cross-link materials through unique loan and consumer identifiers. As at least one commenter noted, these are documents that many lenders are already generating right now. That fact, coupled with the 21-month implementation period leading up to the effective date of the rule, indicates that the industry is relatively well positioned to comply with this component. The other complaint raised by some commenters was that the proposed compliance program and record retention requirements would increase lender costs in providing such loans and may result in some lenders leaving the small-dollar loan market. Other commenters noted 1267 that lenders would actually save on costs by accepting and storing records electronically, as is increasingly common with businesses of all kinds. The Bureau has concluded that any increased costs associated with developing a record retention system that is compliant with the final rule are likely to be offset by benefits that will flow to lenders, consumers, and the marketplace from lenders having systems in place that enable them more easily to track and monitor their compliance with the final rule. For example, lenders will be better able to review their loan performance metrics and identify the root causes of systemic problems while preventing violations of the final rule. The Bureau has also concluded that the record retention requirements would promote effective and efficient enforcement and supervision of the final rule, thereby deterring and preventing unfair and abusive practices that create risks and harms for consumers. Section 1041.13 Prohibition against Evasion Proposed Rule Proposed § 1041.19 would have provided that a lender must not take any action with the intent of evading the requirements of part 1041. It would have complemented the specific, substantive requirements of the proposed rule by prohibiting any lender from undertaking actions with the intent to evade those requirements. The Bureau proposed § 1041.19 based on its express statutory authority under section 1022(b)(1) of the Dodd-Frank Act to prevent evasions of “the purposes and objectives of the Federal consumer financial laws.”1095 The proposed commentary would clarify the meaning of this general provision by indicating when a lender action is taken with the intent of evading the requirements of the Federal consumer financial laws, including this rule. Specifically, the commentary noted that the 1095 12 U.S.C. 5512(b)(1). 1268 form, characterization, label, structure, or written documentation in connection with the lender’s action shall not be dispositive, but rather the actual substance of the lender’s actions, as well as other relevant facts and circumstances will determine whether the lender took action with the intent of evading the requirements of part 1041. It also clarified that if the lender’s action is taken solely for legitimate business purposes, then it is not taken with the intent of evading the requirements of part 1041, and that, by contrast, if a consideration of all relevant facts and circumstances reveals the presence of a purpose that is not a legitimate business purpose, the lender’s action may have been taken with the intent of evading the requirements of part 1041.1096 The commentary also clarified that action taken by a lender with the intent of evading the requirements of part 1041 may be knowing or reckless. Furthermore, it clarified that fraud, deceit, or other unlawful or illegitimate activity may be one fact or circumstance that is relevant to the determination of whether a lender’s action was taken with the intent of evading the requirements of the proposed rule, but fraud, deceit, or other unlawful or illegitimate activity is not a prerequisite to such a finding. The proposed comments also provided some illustrative examples of lender actions that, depending on the facts and circumstances, may have been taken with the intent of evading the requirements of the proposed rule and thus may be violations of the proposed rule, as well as one counter-example. The Bureau proposed § 1041.19 for two primary reasons. First, the provision would address future lender conduct that is taken with the intent of evading the requirements of the rule 1096 The proposal noted that even if a lender’s action can be shown to have been taken solely for legitimate business purposes—and thus was not taken with the intent of evading the requirements of the proposed rule—the lender’s action is not per se in compliance with the proposed rule because, depending on the facts and circumstances, the lender’s action may have violated specific, substantive requirements of the proposed rule. 1269 but which the Bureau may not, or could not, have fully anticipated in developing the rule. The proposed rule contained certain requirements that are specifically targeted at potential lender evasion and which rely on the Bureau’s authority to prevent evasion under section 1022(b)(1) of the Dodd-Frank Act.1097 However, the Bureau cannot anticipate every possible way in which lenders could evade the requirements of the proposed rule.1098 The Bureau was also concerned about the further complexity that would result from attempting to craft additional rule provisions designed to prevent other conduct taken with the intent of evading the proposed rule. Proposed § 1041.19 would provide flexibility to address future lender conduct that is taken with the intent of evading the proposed rule. By limiting avenues for potential evasion, proposed § 1041.19 would enhance the effectiveness of the proposed rule’s specific, substantive requirements, and thereby preserve the consumer protections of the proposed rule. Second, the Bureau believed that proposed § 1041.19 was appropriate to include in the proposed rule given the historical background of the markets for covered loans. As discussed in the proposal, over the past two decades many lenders making loans that would be treated as covered loans under the proposed rule have taken actions to avoid regulatory restrictions at both the State and Federal levels. For example, as discussed above in part II, some lenders have reacted to State restrictions on payday loans by obtaining State mortgage lending licenses and continuing to make short-term, small-dollar loans. In Delaware, a State court of chancery 1097 For example, proposed § 1041.7(d) was designed to prevent evasion of the requirements of proposed § 1041.7 through the making of a non-covered bridge loan when a section 7 loan is outstanding and for 30 days thereafter. 1098 As the Commodity Futures Trading Commission (CFTC) noted in a proposed rulemaking implementing an antievasion provision under title VII of the Dodd-Frank Act, “Structuring transactions and entities to evade the requirements of the Dodd-Frank Act could take any number of forms. As with the law of manipulation, the ‘methods and techniques’ of evasion are ‘limited only by the ingenuity of man.’” 76 FR 29818, 29866 (May 23, 2011) (quoting Cargill v. Hardin, 452 F.2d 1154, 1163 (8th Cir. 1971)). The Bureau’s approach to the anti-evasion clause in proposed § 1041.19 has been informed by this CFTC rulemaking, as discussed below. 1270 recently held that a loan agreement was unconscionable because, among other factors, the court found that the “purpose and effect” of the loan agreement was to evade the State’s payday lending law, which includes a cap on the total number of payday loans in a 12-month period and an anti-evasion provision.1099 States also have faced challenges in applying their laws to certain online lenders, including lenders claiming Tribal affiliation and offshore lenders. Furthermore, at the Federal level, lenders have been making loans that were narrowly structured to deliberately circumvent the scope of regulations to implement the Military Lending Act (MLA), which Congress enacted in 2006. For example, in response to the MLA’s implementing regulations that prohibited certain closed-end payday loans of 91 days or less in duration and vehicle title loans of 181 days or less in duration, lenders began offering payday loans greater than 91 days in duration and vehicle title loans greater than 181 days in duration, along with open-end products. The Department of Defense, which was responsible for drafting the MLA regulations, as well as numerous members of Congress, concluded that such practices were undermining the MLA’s consumer protections for service members and their families.1100 Given this historical background of a decade of widespread evasion of the protections supposedly conferred by the MLA, the Bureau determined that the anti-evasion provision in § 1041.19 was appropriate to include in the proposed rule. 1099 See James v. National Financial, LLC, 132 A.3d 799, 834 (Del. Ch. 2016). The lender structured a $200 loan as a 12-month installment loan with interest-only payments followed by a final balloon payment, with an APR of 838.45 percent. Id. at 803. The court also found a violation of TILA with regard to the disclosure of the APR in the loan contract. Id. at 838-39. This case and the Delaware payday law at issue are also discussed above in part II. 1100 The Department of Defense amended the MLA regulations in 2015 and the compliance date for the amendments is later this year. See 80 FR 43560 (Jul. 22, 2015) (final rule containing amendments). The preamble to the amendments included discussion of comments to the proposed rule from 40 U.S. Senators who wrote that the amendments were “essential to preventing future evasions” of the MLA regulations. Id. at 43561 (quoting letter from Jack Reed, et al., Nov. 25, 2014). 1271 In proposing § 1041.19 and its accompanying commentary, the Bureau relied on antievasion authority under section 1022(b)(1) of the Dodd-Frank Act, which provides that the Bureau’s director may prescribe rules “as may be necessary or appropriate to enable the Bureau to administer and carry out the purposes and objectives of the Federal consumer financial laws, and to prevent evasions thereof.”1101 Anti-evasion provisions are a feature of many Federal consumer financial laws and regulations.1102 In addition, anti-evasion provisions were included in a final rule issued in 2012 by the CFTC under title VII of the Dodd-Frank Act (the CFTC Anti-Evasion Rules).1103 One of the CFTC Anti-Evasion Rules provides that it is “unlawful to conduct activities outside the United States, including entering into agreements, contracts, and transactions and structuring entities, to willfully evade or attempt to evade any provision of” the Dodd-Frank Act title VII provisions or implementing CFTC regulations1104 and that the “[f]orm, label, and written 1101 The Bureau noted that Dodd-Frank Act section 1036(a) separately provides that it shall be unlawful for “any person to knowingly or recklessly provide substantial assistance to a covered person or service provider in violation of the provisions of section 1031, or any rule or order issued thereunder, and notwithstanding any provision of this title, the provider of such substantial assistance shall be deemed to be in violation of that section to the same extent as the person to whom such assistance is provided.” 12 U.S.C. 5536(a)(3). The Bureau did not rely on this authority for proposed § 1041.19, but noted that this statutory provision could be used in an enforcement action to address evasive conduct if a lender’s actions were taken with the substantial assistance of a non-covered person. 1102 See, e.g., Fair Credit Reporting Act, 15 U.S.C. 1681s(e)(1) (“The Bureau may prescribe regulations as may be necessary or appropriate to administer and carry out the purposes and objectives of this subchapter, and to prevent evasions thereof or to facilitate compliance therewith.”). 1103 See 77 FR 48208, 48297-48303 (Dec. 13, 2012) (Final Rule); 76 FR 29818, 29865-68 (May 23, 2011) (Proposed Rule). Section 721(c) of the Dodd-Frank Act required the CFTC to further define the terms “swap,” “swap dealer,” “major swap participant,” and “eligible contract participant” in order “[t]o include transactions and entities that have been structured to evade” subtitle A of title VII of the Dodd-Frank Act, and several other provisions of Dodd-Frank Act title VII reference the promulgation of anti-evasion rules. See 77 FR 48208, 48297 (Dec. 13, 2012). The CFTC Anti-Evasion Rules were promulgated as part of a larger rulemaking issued jointly by the CFTC and the Securities and Exchange Commission (SEC) under title VII of the Dodd-Frank Act, which established a comprehensive new regulatory framework for swaps and security-based swaps. Although the larger rule was issued jointly by the CFTC and the SEC, the anti-evasion provisions were adopted only by the CFTC. Id. at 48297-48302. The SEC declined to adopt any anti-evasion provisions under its Dodd-Frank Act discretionary anti-evasion authority. Id. at 48303. 1104 17 CFR 1.6(a). 1272 documentation of an agreement, contract, or transaction, or an entity, shall not be dispositive in determining whether the agreement, contract, or transaction, or entity, has been entered into or structured to willfully evade.”1105 Moreover, in the preamble for the final CFTC Anti-Evasion Rules, the CFTC provided interpretive guidance about the circumstances that may constitute evasion of the requirements of title VII of the Dodd-Frank Act. The CFTC differentiated between an action taken by a party solely for legitimate business purposes, which the CFTC stated would not constitute evasion, and an action taken by a party that based on a “consideration of all relevant facts and circumstances reveals the presence of a purpose that is not a legitimate business purpose,” which the CFTC stated could constitute evasion depending on the facts and circumstances.1106 The CFTC adopted a principles based approach because it found that adopting an alternative approach that provides a bright-line test of non-evasive conduct may provide potential wrong-doers with a roadmap for structuring evasive transactions. The Bureau believes that the CFTC Anti-Evasion Rules are an informative source of regulatory text and 1105 17 CFR 1.6(b). A separate anti-evasion provision deemed as a swap any agreement, contract, or transaction “that is willfully structured to evade any provision of” subtitle A of title VII. This provision contained similar language as 17 CFR 1.6(b) regarding the “form, label, and written documentation” of the transaction not being dispositive as to the determination of evasion. See 17 CFR 1.3(xxx)(6)(i), (iv). The CFTC defined willful conduct to include intentional acts or those taken with reckless disregard. 1106 See 77 FR at 48301-02; 76 FR at 29867. Among other sources for this distinction, the CFTC described Internal Revenue Service (IRS) guidance on the line between permissible tax avoidance and impermissible tax evasion. See 77 FR 48208, 48301-02; 76 FR 29818, 29867. The CFTC also addressed, in response to comments, whether avoidance of regulatory burdens is a legitimate business purpose. The CFTC wrote that the agency “fully expects that a person acting for legitimate business purposes within its respective industry will naturally weigh a multitude of costs and benefits associated with different types of financial transactions, entities, or instruments, including the applicable regulatory obligations.” 77 FR 48208, 48301. The CFTC further clarified that “a person’s specific consideration of regulatory burdens, including the avoidance thereof, is not dispositive that the person is acting without a legitimate business purpose in a particular case. The CFTC will view legitimate business purpose considerations on a case-by-case basis in conjunction with all other relevant facts and circumstances.” Id. 1273 interpretative guidance on agency use of anti-evasion authority granted under the Dodd-Frank Act.1107 Comments Received Several industry participants and trade associations raised questions about the Bureau’s reliance on the Dodd-Frank Act’s grant of authority to the CFPB’s director to promulgate rules to “prevent evasions” as the basis for its legal authority for the proposed rule’s anti-evasion provision. In particular, one commenter asserted that this legal authority should be construed narrowly to authorize only recordkeeping, reporting, and compliance requirements or to prohibit products and services where no reasonable expectation exists that consumers will use them in a lawful manner. Some commenters objected that exercising this authority would allow the Bureau to circumvent the constraints of the Administrative Procedure Act and impose restrictions without sufficient notice or specificity. Other industry commenters urged that the proposed anti-evasion clause should not be utilized because its purported breadth and ambiguity would lead to overreach that could adversely affect lenders that are responsible and committed to regulatory compliance. They noted that lenders are already obliged to comply with various State laws and with the Military Lending Act, and they contended that the anti-evasion clause is unnecessary in light of the Bureau’s existing authority to target and investigate unfair, deceptive, or abusive acts or practices. Many industry commenters urged that the rule either be made more specific— 1107 The Bureau emphasized that although the anti-evasion clause in proposed § 1041.19 and the accompanying commentary has been informed by the CFTC Anti-Evasion Rules, the Bureau was not formally adopting as the Bureau’s own position the interpretations drawn by the CFTC in the CFTC Anti-Evasion Rules’ preamble, nor did the Bureau endorse the reasoning and citations provided by the CFTC in the CFTC Anti-Evasion Rules’ preamble. 1274 without an anti-evasion clause—or that it be replaced instead with clear guidance to ensure compliance. They noted that the substantive and definitional provisions of the rule could be amended over time to address any loopholes that are found to harm consumers without including open-ended authority that they contend may create a trap for unwary lenders who believe, in good faith, that they are complying with the provisions of the rule. A group of chief legal officers echoed this advice by urging the Bureau to develop specific criteria to determine whether to bring enforcement actions because it would provide clear standards to lenders. Another industry commenter urged the Bureau to let the courts determine violations of law based on fact-specific circumstances and statutory interpretations rather than applying a broad antievasion clause. In contrast, consumer groups judged the anti-evasion clause to be an essential means of addressing evasive practices that would breach the intent of the rule while seeming to conform to its terms. They mentioned specific loopholes that exist under various State laws and described how those provisions are used to circumvent regulatory oversight in ways that are prevalent across the lending industry. One State Attorney General expressed support for a broad and flexible anti-evasion clause as necessary to prevent lenders from evading coverage by various means and to enable law enforcement to effectuate the purposes of the rule. Another commenter supported the clause but suggested supplementing it with additional bright-line rules to restrict certain fees and the bundling of covered loans with the sale of other goods and services. Many industry commenters and trade associations objected to the anti-evasion clause because of its alleged vagueness. They contended that, as a result, unfair effects could flow to lenders, including potential chilling effects on participation and innovation in the marketplace. 1275 In particular, they asserted that the proposed anti-evasion provision’s knowing or reckless standard for intent is too vague, open-ended, and indefinite and it exposes lenders to liability for non-compliance based on the Bureau’s own undefined notions of the spirit of the law, even where the lender is in technical compliance with the provisions of the rule. In addition, many industry commenters, while supportive of including an intent standard, thought it should be more specifically defined. They also objected to setting the threshold for intent at a “knowing or reckless” level because they thought it was too loose a standard for invoking such authority. They further contended that “intended evasions” should fall outside the scope of the rule, and an action should have to constitute an actual evasion to trigger a violation under the statute. A number of consumer and legal aid groups opposed the proposed “intent” provision, which they thought risked undermining the entire provision, as it would be potentially difficult for the Bureau to prove the lender’s state of mind. Others agreed and thought that the clause would set up time-consuming and costly legal battles that would actually facilitate evasions of the rule. They countered that the anti-evasion clause should be reworded simply to cover de facto evasions, without any importing of an intent standard into the clause. Several commenters further urged the Bureau not to prohibit acts or practices without lenders knowing what acts or practices were being proscribed. This objection was couched as a matter of elementary fairness and the legal requirement to provide sufficient notice before imposing liability. Commenters said that the anti-evasion clause is broad enough to permit the Bureau to label as a violation any action it perceives as politically distasteful, regardless of the specific provisions in the final rule. Some commenters focused on the Bureau’s second rationale for the proposal—that lenders of covered loans have a history of avoiding regulatory restrictions. 1276 They asserted that these examples of avoidance are really just evidence of lenders’ efforts to comply with those laws and regulations. One commenter objected that the anti-evasion clause would be likely to sow confusion in the complex system of modern interstate banking. Some industry commenters also were concerned that the breadth of the proposed antievasion clause would create a “chilling effect” that would disincentivize lenders from making loans, and could therefore cause some lenders to exit the market. By creating the potential to over-deter desirable conduct and punish undeserving actors, commenters warned that the clause was more likely to lead to significant litigation than to bolster regulatory effectiveness. At the same time, they contended that the open-ended nature of the clause would chill innovation and prevent market entry by lenders that would otherwise be willing to offer new products. The risks thus posed would tend to scare off investors and creditors, thereby increasing the cost of capital and discouraging more lending. Industry commenters also took issue with use of the phrase “solely for legitimate business purposes” in the commentary to the proposed rule. Specifically, the commentary stated that if the lender’s action is taken solely for legitimate business purposes, the lender’s action is not taken with the intent of evading the requirements. The commenters contended that the phrase was vague and not sufficiently defined in the proposal. One commenter asserted that this wording would allow the Bureau to reach as evasion any acts with a secondary purpose and instead the Bureau should be limited to reaching only acts that constitute a “disguised primary purpose,” as grounded in an evidentiary showing as a factual matter. Another commenter suggested exempting from the clause any change in practices that produces an economic benefit to the consumer. 1277 Consumer groups stated that an evasion should not be limited to a change in a lender’s practices, in order to capture new entrants to the markets with practices that would evade the rule. They also argued that the relevant time frame for gauging a pertinent shift in a lender’s practices should extend back to the issuance of the SBREFA framework of proposals, rather than the issuance of the final rule, which they deemed to be more consistent with an “all facts and circumstances” approach. One industry commenter asked the Bureau to clarify that compliance with the rule is itself a legitimate business reason to modify products and processes. Industry participants and trade associations objected to the Bureau’s statement in the proposal that anti-evasion provisions are a feature of many Federal consumer financial laws and regulations, which they claim is unfounded. They sought to distinguish on a variety of grounds the FCRA, the treatment in Regulation Z derived from the Home Ownership Equity and Protection Act (HOEPA), and the anti-evasion clause contained in the Dodd-Frank Act as administered by the CFTC. For example, one commenter noted that the FCRA has a statutory anti-evasion provision, while only Regulation Z contains limited anti-evasion clauses in its highcost mortgage provision, which was derived from HOEPA. Other commenters distinguished the CFTC’s anti-evasion clause from the proposal’s provision because it applies only to “willful” behavior; the parties to the regulated activity are generally more sophisticated than the consumer borrowers at issue here; and a person’s consideration of the regulatory burdens, including avoidance thereof, is not dispositive that the person is acting without a legitimate purpose. Several industry commenters concluded that the proposal’s anti-evasion provision was arbitrary and capricious, citing several of the issues identified above, including, among other things: the perceived lack of distinction in the proposal between proper and improper behavior; 1278 the Bureau’s reliance on the CFTC’s anti-evasion rule; the necessity of the provision in light of the Bureau’s other authority; and the perceived potential for a chilling of the markets. Many commenters also provided input into different aspects of the commentary set out in the proposal and how well it does or does not succeed in bolstering the proposed rule. In particular, some commenters criticized the commentary as exacerbating the concerns about vagueness with its list of “non-exhaustive” examples. One industry commenter noted that the limited examples do not guarantee that other regulators will take the same view, or that what is currently viewed as permissible under the proposed rule would remain so in the future, both of which raise liability concerns. On the other side, consumer groups also recommended revising and adding a number of examples to further their goal of strengthening the anti-evasion clause. A number of commenters also expressed differing views about the appropriate relationship or intersection between covered and non-covered loans for purposes of some of these provisions. Among other conduct, the first example in the proposal would pertain to a lender that routinely obtains a leveraged payment mechanism but does so more than 72 hours after origination. One attorney general observed that it was illustrative of the need for an anti-evasion clause. Several commenters noted, however, that this example should be strengthened to protect borrowers by removing the time limit altogether or covering loans any time a lender obtains a leveraged payment mechanism, regardless of when that occurs. An industry commenter stated that this example was too vague, because it did not specify how many borrowers were needed to meet the “routinely” standard. Another commented that an examiner at a later date should not be able to add further restrictions beyond the 72-hour period. One Tribal lender expressed its 1279 concern that the language used seemed like a warning that the Bureau will regularly find that the Tribal operations do not constitute legitimate business practices. Among other conduct, the second proposed example would pertain to a lender not conducting an ability-to-repay analysis and regularly charging a recurring late fee to borrowers to be paid biweekly while the loan is outstanding. Consumer groups offered suggestions about the second example in the proposal. They contended that the assumption that delinquency fees and re-borrowing fees are the same should be eliminated, and suggested that the Bureau should emphasize that the scenario could lack elements from the fact pattern and still constitute evasion. They further commented that the example did not provide very robust guidance about what constitutes evasion. They recommended modifying the definition of a loan sequence or covered loan to address the concerns underlying this example in a more effective manner. Consumer groups contended that the third proposed example which would involve, among other conduct, the lender charging a high penalty interest rate, was overly broad and advocated the use of a lower penalty rate to emphasize that not all of the elements in the example had to be present to constitute evasion. They also suggested that the rule should specify that the total cost of credit must include the penalty rate if the lender reasonably expects that a significant number of borrowers will trigger the penalty rate. Consumer groups also suggested that the reference point in the example for lenders’ past and current practices should be the SBREFA date. Regarding the fourth proposed example, which would include, among other conduct, the lender changing its practice such that its second presentment for a delinquent loan was for only 1280 $1, consumer groups recommended prohibiting the initiation of additional payment transfers after any failed attempt. The fifth proposed example would pertain to, among other conduct, a lender restructuring its loan product prior to the effective date of the final rule such that it is a covered loan subject to one of the conditional exceptions. The commentary suggests that the scenario offered is not indicative of evidence of a violation of the anti-evasion provision. An industry commenter stated that the fifth example suggests it might be an evasion to structure the loan product to be noncovered, but the example does not clarify how to avoid having such a loan product trigger the anti-evasion clause. Consumer groups also stated that the Bureau should adopt other examples for greater clarity about what constitutes an evasion. They suggested that if certain lenders unilaterally change the terms of an account after 72 hours to add a wage assignment, automatic transfers, or other leveraged payment mechanism, that should constitute an evasion. They also suggested that another example of evasion would be where the lender continues to use a leveraged payment mechanism without complying with the requirements of the payment provisions of the rule. Further, they suggested a list of more than a dozen ways lenders could evade the rule or certain of its requirements, which should be addressed to improve the proposal. One commenter, by contrast, asked the Bureau to adopt more examples of actions undertaken without intent to evade the rule, including the use of consumer notices, one-time ACH authorizations, and other mechanisms. A credit union trade association offered several ideas for how the anti-evasion clause could be clarified further, and asked the Bureau to clarify that the clause would not be used to create liability for credit unions that changed their lending programs to fall outside the 1281 scope of the rule. One set of academic commenters expressed concern that the definition of “annual percentage rate” could allow lenders to exclude late fees from the modified total cost of credit and structure rolled-over short-term loans to pass as long-term loans. Some commenters raised other miscellaneous suggestions. A trade association requested that if the Bureau keeps an anti-evasion clause, then it should extend a safe harbor for at least the first year after the effective date of the final rule. Another commenter urged that the Bureau should regularly examine records for data omissions and this provision should include specific language to address the consequences of any such data omissions. That commenter also sought language barring the practice of breaking up a payment request into smaller requests to avoid the reauthorization requirement. Consumer groups urged the Bureau to make clear that it will pay special attention to situations where lenders indicate they will attempt to expand or migrate to other industries and shift their unaffordable lending practices to those products. Finally, a trade association encouraged the Bureau to consult with prudential regulators about whether exempting depository institutions would incentivize certain entities in the payday lending market to convert to a bank status, which the commenter found to be implausible. And a set of chief legal officers urged the Bureau to consult with or defer to the States and incorporate some of their suggestions in the final rule, because the States have had more experience with these kinds of consumer loans. Final Rule Proposed § 1041.19 would have required that a lender must not take any action with the intent of evading the requirements of this part 1041. After considering the comments received, the Bureau concludes that the general anti-evasion provision as proposed is appropriate in the 1282 final rule to complement the specific, substantive requirements of the final rule by prohibiting a lender from taking action with the intent to evade those requirements. The only change from the proposed § 1041.19 to the final rule is technical in nature; its reference in the final rule is § 1041.13. In finalizing this provision, the Bureau is relying on its anti-evasion authority under section 1022(b)(1) of the Dodd-Frank Act, which provides that the Bureau’s director may prescribe rules “as may be necessary or appropriate to enable the Bureau to administer and carry out the purposes and objectives of the Federal consumer financial laws, and to prevent evasions thereof.” The Bureau is finalizing § 1041.13 for two primary reasons. First, the provision will address future lender conduct that is taken with the intent of evading the requirements of the rule but which the Bureau may not, or could not, have fully anticipated in developing the rule. The rule contains certain requirements that are specifically targeted at potential lender evasion and which rely on the Bureau’s authority to prevent evasion under section 1022(b)(1) of the DoddFrank Act. However, the Bureau cannot anticipate every possible way in which lenders could evade the requirements of the proposed rule. The Bureau concludes final § 1041.13 will provide flexibility to address future lender conduct that is taken with the intent of evading the proposed rule. By limiting avenues for potential evasion, § 1041.13 will enhance the effectiveness of the final rule’s specific, substantive requirements, and thereby preserve the consumer protections of the final rule. Second, the Bureau’s judgment is informed, in particular, by the history of evasive actions in this industry to circumvent restrictions in State laws and the coverage of the Military Lending Act, outlined above. 1283 In the commentary to the final rule, the Bureau modified the proposal’s commentary regarding the anti-evasion provision by removing the illustrative examples of lender actions that may have been taken with the intent of evading requirements of the rule outlined in proposed comment 19-2. Two illustrative examples can now be found in the commentary sections related to §§ 1041.5 and 1041.8 of the final rule. Specifically, the second example from proposed comment 19-2 is now found in the commentary for § 1041.5(e) of the final rule and the fourth example from proposed comment 19-2 is now found in the commentary for § 1041.8 of the final rule. Any modifications to those examples in the final rule are discussed above in the sectionby-section analysis of those provisions. In particular, the Bureau added to the final rule specific anti-evasion provisions about the ability-to-repay requirements and prohibited payment transfer attempts, and moved the illustrative examples from proposed § 1041.19 to those sections in the final rule to provide additional context for a violation of those specific anti-evasion provisions. Because of coverage changes and other considerations, including the comments it received, the Bureau deleted from the commentary for § 1041.13 of the final rule the remaining illustrative examples that were proposed in comment 19-2. In particular, the first example pertained to, among other conduct, a lender obtaining a leveraged payment mechanism 72 hours after the borrower received the loan proceeds. The proposed rule limited coverage of the abilityto-repay requirements for covered longer-term loans to loans for which the leveraged payment mechanism was taken within 72 hours of origination. However, under the final rule covered longer-term loans are subject only to the payment provisions, but not to the ability-to-repay underwriting provisions. Accordingly, in the final rule, the Bureau deleted the reference to the first example in the proposed rule’s commentary to avoid confusion. The Bureau deleted the 1284 third illustrative example in proposed comment 19-2 because it addressed evading the ability to repay requirements for longer-term loans, and in light of the changes to the coverage of the rule, it is of limited relevance. Likewise, the Bureau deleted the fifth illustrative example, in part, because of concerns raised about whether the counter-example of evidence not constituting a violation succeeded in providing adequate guidance. The comments the Bureau received about the inclusion of the illustrative examples were mixed, with some commenters seeking more examples to address certain situations and others finding the examples unhelpful and not sufficiently detailed. By relocating some of the examples and deleting others, the Bureau has attempted to balance the stated desire by commenters for clearer guidance about what conduct constitutes evasion and their suggestions that the anti-evasion provision should remain flexible. The Bureau has concluded that the specific anti-evasion provisions in the final rule and the related illustrative examples in the commentary will provide concrete guidance on specific types of evasions, while the general antievasion provision is necessary to allow the Bureau to prevent intentional evasions of the specific, substantive requirements of the final rule that it cannot yet anticipate at this time. In addition to deleting some of the proposal’s illustrative examples, the Bureau decided not to include any additional illustrative examples of evasion in the final rule, although many commenters suggested particular factual situations as possible examples and counter-examples of evasion. The Bureau reached this decision because of the comments it received highlighting concerns that undue weight may be placed on the specifics in any particular examples provided and hence they may be misconstrued as an exhaustive list of possible means of evasion that would be viewed as narrowing the concept that Congress explicitly incorporated into the Dodd-Frank Act. The 1285 Bureau thus disagrees with commenters that suggested a general anti-evasion provision is contrary to the statutory authority granted in section 1022(b)(1), which itself is expressly a general anti-evasion provision. Nothing in the Act suggests in any way that the Bureau’s authority to prevent evasions is limited, as some commenters have suggested. Nor does the Bureau agree that the Administrative Procedure Act is implicated if the Bureau exercises this direct statutory authority. In sum, the Bureau has decided to finalize, as it was proposed (and now renumbered), the general anti-evasion provision contained in § 1041.13 of the final rule. Although some commenters had questioned the Bureau’s references to anti-evasion features in other Federal consumer financial laws and regulations, the Bureau did not rely on those provisions in deciding upon its own authority to act in accordance with the express terms of the statute. Rather, the Bureau included references to other Federal consumer financial laws in the proposal merely because it found them to be informative. Because the CFTC’s source of authority for its Anti-Evasion Rules was the Dodd-Frank Act, the Bureau believed that provision to be of special interest regarding agency use of anti-evasion authority granted under the very same statute. The Bureau continues to find the CFTC Anti-Evasion Rules and other Federal consumer financial laws to be informative about the scope and nature of the Bureau’s antievasion provision, yet the Bureau does not formally adopt the CFTC’s interpretations as its own. As for the claim that an anti-evasion provision is unnecessary because of the Bureau’s UDAAP authority and lenders’ responsibilities to comply with other State and Federal laws, the Bureau does not find the claim persuasive. Instead, the Bureau concludes that an anti-evasion provision is necessary to ensure compliance with the substantive provisions of the final rule. Congress granted the Bureau to authority to promulgate rules to prevent evasions and thus, it is 1286 authorized to exercise its authority by finalizing a general anti-evasion provision. If Congress had intended that every evasion of the Bureau’s rules must also be an independent UDAAP, it would set out those requirements in the Dodd-Frank Act; however, it did not. In fact, it is wellestablished that violations of public policy—such as rules or other violations—do not in and of themselves constitute independent UDAAPs, in particular in the context of unfair acts or practices. Accordingly, the Bureau disagrees that its UDAAP authority negates the need for the anti-evasion provision because the Bureau may not be able to readily reach conduct that constitutes evasion using its existing UDAAP authority. In particular, the evasive conduct may be actionable without having to meet the stringent standards for UDAAP violations or with less expenditure of resources. Moreover, as described above, the historical background in this market indicates that lenders of covered loans have taken actions to circumvent and avoid compliance with various State and Federal regulatory restrictions designed to protect consumers, including the Military Lending Act. The Bureau places great weight on this recent historical experience and perceives it as considerable justification for being vigilant about similar conduct that may be engaged in to circumvent the provisions of this rule. The Bureau is not persuaded by the concerns raised about the purported breadth, ambiguity, and vagueness of a general anti-evasion provision. In particular, many commenters thought it would be important to identify much more specific conduct that would constitute evasion. Instead, the Bureau found compelling the arguments from commenters who urged that the anti-evasion provision should be maintained as a broad and flexible support for administering and enforcing the provisions of the rule. Almost by definition, the anti-evasion clause must be 1287 kept on a more general plane; if all the particulars could be specified in advance, they would all be written into the substantive provisions of the rule, even though that could prove cumbersome and add a good deal of complexity. As the CFTC noted in its anti-evasion rulemaking, providing bright-line tests of non-evasive conduct may provide potential wrong-doers with a roadmap for structuring evasive transactions. By contrast, however, the only real purpose to be served by an anti-evasion clause is to provide authority to address other situations that may arise but are not directly addressed by the specific provisions of the rule. Thus, the Bureau concludes that the anti-evasion clause is an important feature of this rule and that it must remain sufficiently flexible to prevent lenders from engaging in conduct designed to circumvent the rule in ways that could pose harms for consumers. Another point of contention is the intent requirement in the anti-evasion provision. Some commenters argued that it poses too low a standard and others argued that it is set too high. The Bureau has made the judgment that the requirement that a lender either knowingly or recklessly intends to evade the final rule is an important limitation on the Bureau’s exercise of its evasion authority. The intent requirement prevents the very outcome that some commenters fear— violations by unwary lenders acting in good faith. By its very terms, the intent requirement eliminates that possibility. The Bureau is thus finalizing § 1041.13 as proposed (and now renumbered), including its formulation of the intent standard as further explained in the related commentary. As the commentary, now finalized, sets out, a lender must act with knowing or reckless intent to evade the final rule in order to be liable under the anti-evasion provision. Intent is the state of mind accompanying an act. Ordinarily, state of mind cannot be directly proved but, 1288 instead must be inferred from the surrounding circumstances, as explained in the final rule commentary. As noted in the proposal, the intent standard in the final rule is consistent with the scienter standard in section 1036(a) of the Dodd-Frank Act for establishing that persons knowingly or recklessly provided substantial assistance to a covered person or service provider in violation of section 1031.1108 In the civil liability sphere, recklessness includes actions entailing an unjustifiably high risk of harm that is known or either so obvious it should be known.1109 Some commenters expressed concern that the intent standard would be a challenging threshold to meet. Yet the existence of such a standard is crucial to establishing that the lender has in fact engaged in the type of conduct that was intended to evade this rule, as opposed to being found liable for unintentional conduct. Because standards grounded in the intentions of the parties are well-established in the common law and are being developed in CFPB cases,1110 the Bureau is not persuaded that lenders would be confused or at a loss to know how to proceed or that the Bureau’s use of this authority would be unfettered and arbitrary. Accordingly, the Bureau has adopted this provision without altering the intent standard as originally proposed. Comment 13-1 of the final rule, which illustrates lender action taken with the intent of evading the requirements of the rule, is adopted in a form that remains unchanged from the proposal. Although several commenters raised concerns about this piece of the commentary, they appear to have misinterpreted it. In particular, it provides that “if the lender’s action is taken solely for legitimate business purposes, the lender’s action is not taken with the intent of 1108 The CFTC’s Anti-Evasion Rule’s scienter standard is willfulness which the CFTC interprets as including intentional or reckless acts. See Safeco Ins. Co. of America v. Burn, 551 US 47 (2007). 1109 See Safeco Ins. Co. of America v. Burn, 551 US 47 (2007). 1110 See CFPB v. Universal Debt and Payment Solutions, Civil Action No. 1:115-CF-00859 (D. Ga. September 2015). 1289 evading the requirements.”1111 It further provides that “if a consideration of all relevant facts and circumstances reveals the presence of a purpose that is not a legitimate business purpose, the lender’s action may have been taken with the intent of evading the requirements of” the final rule.1112 Both sentences must be read in conjunction. The existence of a non-legitimate business purpose does not mean that the lender necessarily intended to evade the rule’s requirements; it simply means that it may have done so. And commenters’ interpretation of the first sentence regarding “solely for legitimate business purposes” is misguided. As the commentary itself states, “the actual substance of the lender’s action as well as other relevant facts and circumstances will determine whether the lender’s action was taken with the intent of evading the requirements” of the rule. By its express terms, lenders who act solely from legitimate business purposes will not be subject to enforcement of this provision. Accordingly, a lender that modifies its practices to comply with the requirements of the final rule will not violate the anti-evasion provision unless it meets the threshold of acting with knowing or reckless intent to evade the requirements. Some commenters warned that this provision could create a “chilling effect” that would cause lenders not to make loans and to leave the market. To be sure, some lenders will likely change their practices in light of the final rule, including performing ability-to-repay underwriting of covered loans for the first time. However, it seems highly unlikely that the antievasion provision itself would be the cause of lenders changing their practices or exiting the market. In fact, the Bureau concludes that the intent requirement is a key element that undercuts 1111 1112 Comment 13-1 (emphasis added). Id. (emphasis added). 1290 arguments that the anti-evasion provision is unfair to lenders or will over-deter desirable conduct and punish undeserving actors. In terms of evaluating a lender’s practices under the anti-evasion provision, commenters made conflicting arguments that tend to underscore the need to maintain flexibility if this provision is to fulfill its intended purpose. Various limiting principles were suggested—such as that any changes in lender practices that produce an economic benefit for consumers should never be deemed to be evasions, or that conduct during one defined period or another should be established as a firm baseline—but none of them appears to be consistent with the general terms that Congress used to articulate and confer this authority. Nor was any sound justification offered for the suggestion that the Bureau should extend a safe harbor against its use of the antievasion provision for at least the first year after the effective date of the final rule. As stated in the commentary, the pertinent analysis instead is and should be the “actual substance of the lender’s action as well as other relevant facts and circumstances” and thus the Bureau made no changes to the commentary in this regard. Finally, in light of this discussion, the Bureau concludes that the final anti-evasion provision is not arbitrary and capricious. Lenders are on notice about the substantive provisions of the final rule and they are on notice that if they act with knowing or reckless intent to evade those provisions, they may be subject to the anti-evasion provision. Congress expressly authorized the Bureau to enact such a provision pursuant to the Dodd-Frank Act, and through this rulemaking process the Bureau has considered the relevant factors, including numerous public comments and its own analysis, to adopt this anti-evasion provision in § 1041.13 of the final rule. 1291 Section 1041.14 Severability Proposal Proposed § 1041.20 would have made the provisions of this rule separate and severable from one another. Comments Received Several commenters argued that the proposed rule should not include a severance provision because the various provisions of the proposal are interconnected and the proposal would create a whole new comprehensive regulatory framework. As such, if one provision is deemed invalid, they argued, the entire system should be deemed invalid. Commenters noted their impression that the proposal repeatedly emphasized that the provisions were designed to work in tandem, noting specifically the relationship between proposed § 1041.5 and § 1041.7. Final Rule The Bureau is finalizing proposed § 1041.20 as final § 1041.14, such that it now reads: “The provisions of this part are separate and severable from one another. If any provision is stayed or determined to be invalid, the remaining provisions shall continue in effect.” The final rule removes the phrase “it is the Bureau’s intention that” from the provision to clarify that the provision is not dependent on the Bureau’s intention. This is a standard severability clause of the kind that is included in most regulations and much legislation to clearly express agency intent about the course that is preferred if such events were to occur. The Bureau disagrees with commenters that the provisions are so interconnected that if one provision should fail, the others should, as well. The Bureau specifically designed the 1292 framework of the rule so that the fundamental protections will continue regardless of whether one or another provision is not effectuated. The rule anticipates certain contingencies. For example, lenders can still enter into loans made pursuant to final § 1041.5, regardless of whether there is a registered information system pursuant to § 1041.11. Lenders may not be able to do so under § 1041.6. In the absence of such protections, then under the terms of the rule itself, such lending is not available, and that framework should thus continue. Further, § 1041.6 is an exemption from § 1041.5, and thus, § 1041.5 alone should be more than sufficient to prevent the unfair and abusive practice identified in § 1041.4 if § 1041.6 should be overturned. Additionally, part B (§§ 1041.4 - .6) and part C (§§ 1041.7 - .9) are entirely separate, based on separate identified unfair and abusive practices, and thus, if either should fall, the other should remain intact and continue to operate. These examples are merely illustrative, and do not constitute a complete list of sections which are severable from each other, nor of reasons that sections can operate independently from each other. The Bureau designed each individual provision to operate independently and, thus the Bureau is finalizing the severability clause, as proposed. VI. Effective Date Proposed Rule The Bureau proposed that, in general, the final rule would take effect 15 months after publication in the Federal Register. The Bureau believed that 15 months struck the appropriate balance between providing consumers with necessary protections while giving covered persons adequate time to comply with all aspects of the final rule. In particular, the Bureau gave thought to the time necessary to implement the consumer reporting components of the proposal, in 1293 addition to the time that lenders would need to adjust their underwriting practices and prepare to provide new consumer disclosures. The Bureau proposed that proposed § 1041.17 (now final § 1041.11) would take effect 60 days after publication in the Federal Register with regard to registered information systems. The Bureau believed that this earlier effective date for § 1041.17 was appropriate to allow the standards and process for registration to be in place, which would be necessary for the information systems to be operational by the effective date of the other provisions of the final rule. Comments Received The Bureau received several comments suggesting that it should extend the effective date as to the general rule, with particular focus on 24 months after publication in the Federal Register as a proposed alternative. Commenters argued that 2 years would be necessary because they believed the rule would substantially change the core structure of the industry. One commenter cited the experience with the TILA-RESPA Integrated Disclosure Rule as evidence that complicated regulations require significant implementation time. That rule was initially published in the Federal Register on December 31, 2013, with an effective date of August 1, 2015,1113 but the effective date was extended to October 3, 2015, roughly 21 months after the initial rule was published.1114 Other commenters, more generally, suggested it would take more than 15 months, or “years,” to revise underwriting standards, develop new loan origination processes, train staff, upgrade systems to meet the new underwriting, disclosure, and recordkeeping requirements, and integrate their systems with the registered information systems. 1113 1114 78 FR 79730 (Dec. 31, 2013). 80 FR 43911 (July 24, 2015). 1294 Commenters also asked the Bureau more specifically to delay the date after which lenders will need to obtain a consumer report from a registered information system, citing concerns that lenders would be unable to make loans under the exemption in § 1041.6 if an information system is not registered sufficiently in advance of that data to allow lenders to rely on a consumer report from a registered information system as required under § 1041.6. Final Rule In light of comments received, and extended deadlines elsewhere in the rule, the Bureau is extending the effective date for the full rule by six months. The Bureau is finalizing an effective date for § 1041.11 of 60 days after publication in the Federal Register, and for the rest of the rule, 21 months after publication in the Federal Register. Accordingly, the standards and processes for registration as registered information systems will become operative 60 days after the final rule’s publication. However, it was persuaded that other time frames, based on the comments it received, should be extended. See the section-by-section analysis for §§ 1041.10 and 1041.11 for more details. The Bureau has extended deadlines for applying to be a registered information system found in § 1041.11(c)(3). It has also extended the amount of time an information system must be registered before a lender must furnish to it under § 1041.10(b). The combined amount of time extended for registration and preparation to furnish is 5 months. It is the Bureau’s intent to have information systems registered at least 180 days prior to the effective date of the rest of the rule such that lenders can furnish to and obtain reports from a registered information system, and make loans under § 1041.6, immediately upon that effective date. To help ensure that occurs, the 1295 Bureau needed to extend the effective date of the rule, in light of the extended deadlines in §§ 1041.10 and 1041.11, by at least 5 months. The timeline for implementation of the rule is as follows. Section 1041.11 goes into effect 60 days after publication of the rule in the Federal Register. The deadline to submit an application for preliminary approval to become a registered information system before the effective date of § 1041.10 is 90 days from the effective date of § 1041.11 (it was 30 days in the proposal). That means the deadline for applicants seeking preliminary approval is 150 days after publication in the Federal Register. Once the Bureau grants preliminary approval, the applicant will have an additional 120 days to submit an application to become a registered information system (it was 90 days in the proposal). Under § 1041.10(b), lenders will be required to furnish to a registered information system that has been registered for 180 days or more (it was 120 days or more in the proposal), or upon the effective date of § 1041.10, whichever is later. This will allow a period of at least 180 days for lenders to onboard to the registered information system and prepare to furnish. The Bureau believes an effective date of 21 months will accommodate these new periods and give the Bureau enough time to review applications. The Bureau also agrees that the industry may need additional time to implement the requirements of this rule. In setting an effective date for the final rule, the Bureau seeks to balance giving enough time for an orderly implementation period against the interest of enacting protections for consumers as soon as possible. The Bureau believes that by providing an additional 6 months, lenders should be able to reasonably adjust their practices to come into compliance with the rule. Of course, the Bureau will monitor the implementation period and make adjustments as appropriate. 1296 VII. Section 1022(b)(2) Analysis A. Overview In developing this final rule, the Bureau has considered the potential benefits, costs, and impacts as required by section 1022(b)(2) of the Dodd-Frank Act. Specifically, section 1022(b)(2) calls for the Bureau to consider the potential benefits and costs of a regulation to consumers and covered persons, including the potential reduction of access by consumers to consumer financial products or services, the impact on depository institutions and credit unions with $10 billion or less in total assets as described in section 1026 of the Dodd-Frank Act, and the impact on consumers in rural areas. In the proposal, the Bureau set forth a preliminary analysis of these effects and requested comments that could inform the Bureau’s analysis of the benefits, costs, and impacts of the proposal. In response, the Bureau received a number of comments on the topic. The Bureau has consulted with the prudential regulators and the Federal Trade Commission, including consultation regarding consistency with any prudential, market, or systemic objectives administered by such agencies. The Bureau specifically invited comment on all aspects of the data that it used to analyze the potential benefits, costs, and impacts of the proposed provisions. While some commenters provided additional empirical analyses and data, the Bureau notes that in some instances, the requisite data are not available or are quite limited. As a result, portions of this analysis rely, at least in part, on general economic principles, the Bureau’s experience and expertise in consumer financial markets, and qualitative evidence provided by commenters, while other portions rely on the data that the Bureau has collected and analyzed about millions of these loans. Many of the 1297 benefits, costs, and impacts of the final rule are presented in ranges, rather than as point estimates. The Bureau also discussed and requested comment on several potential alternatives, which it listed in the proposal’s Initial Regulatory Flexibility Analysis (IRFA) and also referenced in its Section 1022(b)(2) Analysis. A further detailed discussion of potential alternatives considered is provided in part VII.J and the Final Regulatory Flexibility Analysis (FRFA) in part VIII below. B. Major Provisions and Coverage In this analysis, the Bureau focuses on the benefits, costs, and impacts of the four major elements of the final rule: (1) the requirement to reasonably determine borrowers’ ability to repay covered short-term and longer-term balloon-payment loans according to their terms (along with the exemption allowing for a principal step-down approach to issuing a limited number of short-term loans); (2) certain limitations on attempts to initiate payment for covered loans; (3) the recordkeeping requirements associated with (1) and (2); and (4) the rule’s requirements concerning registered information systems. The discussion of impacts that follows is organized into these four main categories. Within each, the discussion is organized to facilitate a clear and complete consideration of the benefits, costs, and impacts of the major provisions of the rule. Impacts on depository institutions with $10 billion or less in total assets and on rural consumers are discussed separately below. There are two major classes of short-term lenders the Bureau expects to be affected by the ability-to-repay provisions of the rule: payday/unsecured short-term lenders, both storefront 1298 and online, and short-term vehicle title lenders. The Bureau also believes there is at least one bank that makes deposit advance product loans that are likely to be covered by these provisions. The Bureau recognizes that some community banks and credit unions occasionally make shortterm secured or unsecured loans, but the Bureau believes that those loans will generally fall within the exemption for alternative loans or the exemption for accommodation loans under § 1041.3(e) and (f). Similarly, the Bureau recognizes that some firms in the financial technology (fin tech) space are seeking to offer products designed to enable consumers to better cope with liquidity shortfalls, but the Bureau believes that those products, to a significant extent, will fall within the exclusion for wage advance programs under § 1041.3(d)(7) or the exclusion for nocost advances under § 1041.3(d)(8).1115 In addition to short-term lenders, lenders making longer-term balloon-payment loans (either vehicle title or unsecured) are also covered by the ATR requirements and the rule’s requirements concerning registered information systems. The Bureau believes there are many fewer such lenders, but notes that the following discussion applies to these lenders as well. The provisions relating to payment practices and related notices apply to any lender making a covered loan, either covered short-term loans, covered longer-term balloon-payment loans, or covered longer-term loans. However, payment withdrawals by lenders who also hold the consumer’s deposit account are exempt if they meet certain conditions. The payment provisions affect certain online lenders, who make loans with an APR above 36 percent and normally receive payments via ACH or other electronic means. In addition, storefront payday or 1115 The Bureau also believes many of the current “fintech” offerings fall outside of at least the ability-to-repay requirements of the rule, as they often focus on longer-term lending without balloon payments. 1299 payday installment lenders that receive payment via ACH or post-dated check, either for regular payments or when a borrower has failed to come to the store and make a cash payment in person, will be affected, as will some traditional finance companies if they make loans that meet the criteria for a covered longer-term loan. Lenders making vehicle title loans often do not obtain the same forms of account access, but those that do will also be affected. The provisions relating to recordkeeping requirements apply to any lender making covered loans, with additional requirements for lenders making covered short-term and longerterm balloon-payment loans. The provisions relating to the application process for entities seeking to become registered information systems govern any and all entities that apply to become such information systems.1116 The provisions relating to the requirements to operate as a provisionally registered or registered information system apply to any entity that becomes a provisionally registered or registered information system. The Bureau received many comments that seemed to mistakenly interpret the rule as a ban on payday and/or vehicle title loans. It should be noted that none of the above provisions, either on their own or in combination, constitutes a ban on covered lending. As such, the rule does not explicitly ban payday, vehicle title, longer-term balloon, or any other covered loans. While the Bureau estimates that there will be a substantial reduction in the volume of covered short-term payday loans made in response to the rule prior to any reforms that may occur in the market, the Bureau believes such loans will remain available to the vast majority of consumers facing a truly short-term need for credit (where permitted by State law). In fact, as described in 1116 In this section the Bureau’s references to registered information systems will generally include both provisionally registered information systems and registered information systems, as lenders will be required to report to both types of systems, and incur similar costs to do so. 1300 greater detail below, the Bureau’s simulations suggest that the rule will only restrict roughly 6 percent of borrowers from initiating a payday borrowing sequence they would have initiated absent the rule. In the case of short-term vehicle title loans, the Bureau acknowledges that a more substantial portion of lending will be curtailed.1117 C. Baseline for Consideration of Benefits, Costs, and Impacts In considering the potential benefits, costs, and impacts of the rule, the Bureau takes as the baseline for the analysis the regulatory regime that currently exists for the covered products and covered persons.1118 Given that the Bureau takes the status quo as the baseline, the analysis below focuses on providers that currently offer short-term loans and longer-term loans with balloon features, the potential entrants into the market for registered information systems required under this rule (although their participation is voluntary), and, to a lesser extent, providers of covered longer-term loans that face limits on their activities only through the intervention affecting payment practices. The baseline considers economic attributes of the relevant markets and the existing legal and regulatory structures applicable to providers. Most notably, the baseline recognizes the wide variation in State-level restrictions that currently exist. As described in greater detail in part II 1117 In this section the Bureau focuses most of its analysis on payday and vehicle title loans, rather than the longerterm balloon-payment loans that face similar coverage. The Bureau has observed that longer-term balloon-payment loans are currently less common, and have arisen mostly in response to regulatory regimes restricting or banning payday loans. As such, the Bureau has substantially less evidence about these loans. The Bureau does possess data for a single lender that made longer-term vehicle title loans with both balloon and amortizing payment schedules. These data show that loans with balloon payments defaulted at a substantially higher rate (see “CFPB Report on Supplemental Findings,” at 30), but do not provide much insight into the broader market for these loans. Still, the Bureau has concluded that they generally lead to similar harms due to their payment structures, and will experience similar effects from this rule. 1118 The Bureau has discretion in each rulemaking to choose the relevant provisions to discuss and to choose the most appropriate baseline for that particular rulemaking. 1301 above, there are now 35 States that either have created a carve-out from their general usury cap for payday loans or have no usury caps on consumer loans.1119 The remaining 15 States and the District of Columbia either ban payday loans or have fee or interest rate caps that payday lenders apparently find too low to sustain their business models. Further variation exists within States that allow payday loans, as States vary in their payday loan size limits and their rules related to rollovers (e.g., when rollovers are permitted and whether they are subject to certain limitations such as a numerical cap or requirements that the borrower must amortize the rollover by repaying part of the original loan amount with each payment made). Numerous cities and counties within these States have also passed local ordinances restricting the location, number, or product features of payday lenders.1120 Restrictions on vehicle title lending similarly vary across and within States, in a manner that often (but not always) overlaps with payday lending 1119 See Pew Charitable Trusts, “State Payday Loan Regulation and Usage Rates,” (Jan. 14, 2014), available at http://www.pewtrusts.org/en/multimedia/data-visualizations/2014/state-payday-loan-regulation-and-usage-rates (for a list of States). Other reports reach slightly different totals of payday authorizing States depending on their categorization methodology. See, e.g., Susanna Montezemolo, “The State of Lending in America & Its Impact on U.S. Households: Payday Lending Abuses and Predatory Practices,” at 32–33 (Ctr. for Responsible Lending 2013), available at http://www.responsiblelending.org/sites/default/files/uploads/10-payday-loans.pdf; Consumer Fed’n of Am., “Legal Status of Payday Loans by State,” available at http://www.paydayloaninfo.org/state-information (last visited Apr. 6, 2016) (lists 32 States as having authorized or allowed payday lending). Since publication of these reports, South Dakota enacted a 36 percent usury cap for consumer loans. Press Release, S.D. Dep’t of Labor and Reg., “Initiated Measure 21 Approved” (Nov. 10, 2016), available at http://dlr.sd.gov/news/releases16/nr111016_initiated_measure_21.pdf. Legislation in New Mexico prohibiting short-term payday and vehicle title loans will go into effect on January 1, 2018. Regulatory Alert, N.M. Reg. and Licensing Dep’t, “Small Loan Reforms,” available at http://www.rld.state.nm.us/uploads/files/HB%20347%20Alert%20Final.pdf. 1120 For a sample list of local payday ordinances and resolutions, see Consumer Fed’n of Am., “Controlling the Growth of Payday Lending Through Local Ordinances and Resolutions,” (Oct. 2012), available at www.consumerfed.org/pdfs/Resources.PDL.LocalOrdinanceManual11.13.12.pdf. 1302 restrictions. Overall, these restrictions leave fewer than half of States having vehicle title lenders.1121 Another notable feature of the baseline is the restriction in the Military Lending Act (MLA) to address concerns that servicemembers and their families were becoming over-indebted in high-cost forms of credit.1122 The MLA, as implemented by the Department of Defense’s regulation, requires, among other provisions, that the creditor may not impose a military annual percentage rate (MAPR) greater than 36 percent in connection with an extension of consumer credit to a covered borrower. In 2007, the Department of Defense issued its initial regulation under the MLA, limiting the Act’s application to closed-end loans with a term of 91 days or less in which the amount financed did not exceed $2,000; closed-end vehicle title loans with a term of 181 days or less; and closed-end tax refund anticipation loans.1123 This covered most shortterm and longer-term payday loans and vehicle title loans as well.1124 In considering the benefits, costs and impacts of the rule, the Bureau recognizes this baseline. More specifically, the Bureau notes that the rule will not have impacts, with some limited exceptions, for consumers in States that currently do not allow such lending. It is possible that consumers in these States do access such loans online, by crossing State lines, or through other means, and to the extent the rule limits such lending, they may be impacted. 1121 For a discussion of State vehicle title lending restrictions, see Consumer Fed’n of Am., Car Title Loan Regulation (Nov. 16, 2016), available at http://consumerfed.org/wp-content/uploads/2017/01/11-16-16-Car-TitleLoan-Regulation_Chart.pdf. 1122 The Military Lending Act, part of the John Warner National Defense Authorization Act for Fiscal Year 2007, was signed into law in October 2006. The interest rate cap took effect October 1, 2007. See 10 U.S.C. § 987. 1123 72 FR 50580 (Aug. 31, 2007). 1124 As noted earlier, effective October 2015 the Department of Defense expanded its definition of covered credit to include open-end credit and longer-term loans so that the MLA protections generally apply to all credit subject to the requirements of Regulation Z of the Truth in Lending Act, other than certain products excluded by statute. See 80 FR 43560 (July 22, 2015) (codified at 32 CFR part 232). 1303 Similarly, in States with more binding limits on payday lending, the rule will have fewer impacts on consumers and covered persons as the State laws may already be restricting lending. The overall effects of these more restrictive State laws were described earlier in part II. In the remaining States, which are those that allow lending covered by the rule without any binding limitations, the rule will have its most substantial impacts. Notably, the quantitative simulations discussed below reflect these variations in the baseline across States and across consumers with one exception. The data used inherently capture the nature of shocks to consumers’ income and payments that drive demand for covered loans. To the extent that these have not changed since the time periods covered by the data, they are captured in the simulations. The analysis also captures the statutory and regulatory environment at the time of the data. The implication is that to the extent that the environment has changed since 2011-2012, those changes are not reflected in the simulations. More specifically, the simulations will overstate the effect of the rule in those areas where regulatory changes since that time have limited lending, and will underestimate the effect of the rule in any areas where regulatory changes since that time have relaxed restrictions on lending. In general, the Bureau believes that the States have become more restrictive over the past five years so that the simulations here are more likely to overstate the effects of the rule. That said, the simulation results are generally consistent with the additional estimates, using other data and time periods, provided to the Bureau in comments. D. Description of the Market Failure The primary concern in this market, as described in Market Concerns—Underwriting and the section-by-section analysis of § 1041.4, is that many borrowers experience long and 1304 unanticipated durations of indebtedness. That is, the failures in the market do not necessarily impact the average borrower experience, but instead impact those borrowers who experience longer sequences of loans. If the likelihood of re-borrowing, and in particular re-borrowing that results in longer sequences is underestimated by customers when they take their initial loans, the existence of these sequences implies imperfect or incomplete information. This lack of information constitutes a potentially harmful market failure.1125 That the likelihood of these long sequences is underestimated or unanticipated is supported by empirical findings in the academic literature. The Bureau believes that Mann (2013) provides the most relevant data describing borrowers’ expected durations of indebtedness with payday loan products.1126 Many comments received in response to the proposal, including one from Professor Mann himself, suggest this is a widely held view. However, the Bureau’s consideration of the facts provided in Mann (2013) differs from the main points highlighted in the study, and reiterated in Professor Mann’s comment letter. This was discussed at length in Market Concerns—Underwriting and is addressed more completely, along with a discussion of the broader literature on the accuracy of borrowers’ expectations, in part VII.F.2. 1125 Note that the characterization of market failure here does not hinge only on the outcome of long sequences, but the unanticipated nature of that outcome. Also note that the typical customer anticipating his or her sequence length, or customers as a whole properly anticipating the average duration of indebtedness, is not a credible counterargument to this market failure. If few (or none) of the individuals who experience long sequences properly anticipated the likelihood that a sequence of this length might occur, that in and of itself would constitute a market failure. In assessing the costs and benefits of the rule, this section remains agnostic about the source of the information deficiency; however § 1041.4 describes the Bureau’s view about the nature and source of consumers’ inaccurate expectations. 1126 Ronald Mann, “Assessing the Optimism of Payday Loan Borrowers,” 21 Supreme Court Econ. Rev. 105, at 132 (2013). Also note that, while Mann’s approach is the most relevant for this rule, there are other studies that explore the accuracy of borrowers’ expectations about continued use of short-term loans. These studies are discussed in part VII.F.2 below. 1305 In summary, Mann asserts that borrowers are generally accurate in their predictions (citing the fact that 57 percent predict their time in debt within a 14-day window1127), that many anticipate re-borrowing (40 percent anticipated they would “continue their borrowing after its original due date”1128), and that borrowers were about as likely to overestimate their times in debt as they were to underestimate them. The Bureau did not contradict these findings in the proposal, nor does it attempt to do so now. However, the Bureau believes these data also provide strong evidence that those borrowers who experience long periods of indebtedness did not anticipate those experiences. For example, of the borrowers who remained in debt at least 140 days (10 biweekly loans), it appears that all (100 percent) underestimated their times in debt, with the average borrower in this group spending 119 more days in debt than anticipated (equivalent to 8.5 unanticipated rollovers). Of those borrowers who spent 90 or more days in debt (i.e., those most directly affected by the rule’s limits on re-borrowing under § 1041.6), it appears that more than 95 percent underestimated their time in debt, spending an average of 92 more days in debt than anticipated (equivalent to 6.5 unanticipated rollovers).1129 1127 Ronald Mann, “Assessing the Optimism of Payday Loan Borrowers,” 21 Supreme Court Econ. Rev. 105, at 123 (2013). Note that the reported value of 57 percent is out of respondents who answered the relevant question (approximately 80 percent of all survey respondents), meaning that only 46 percent of all survey respondents made predictions with this accuracy. 1128 See Ronald Mann, “Assessing the Optimism of Payday Loan Borrowers,” 21 Supreme Court Econ. Rev. 105, at 120 (2013). 1129 Theoretically, these findings can be reconciled with a rational expectations model, but only under very specific conditions. Specifically, one has to assume that borrowers have no or very little information on which to base their predictions of their length of indebtedness. In that case, the extreme outcomes are simply very rare realizations from some distribution of outcomes. To the extent that borrowers have information about their own financial circumstances (e.g., repeat borrowers know their past experience with payday loans), the above assumption cannot be plausibly maintained. And in fact, past experience is predictive of the future length of indebtedness: in a hazard model, the length of past loan sequences has an economically and statistically significant negative impact on the 1306 There is also evidence that even short-term borrowers do not fully expect the outcomes they realize. For example, only 40 percent of borrowers anticipated re-borrowing, but it appears that more than 70 percent of the customers Mann surveyed did in fact re-borrow. As such, even those borrowers who accurately predict their durations of indebtedness within a 14-day window are likely to have experienced unanticipated re-borrowing. Across all borrowers in the data, a line of “best fit” provided by Professor Mann describing the relationship between a borrower’s expected time in debt and the actual time in debt experienced by that borrower shows effectively zero slope (indicating no correlation between a borrower’s expectations and outcomes).1130 This shows that, regardless of whether borrowers experienced short or long durations of indebtedness, they did not systematically predict their outcomes with any sort of accuracy or precision. While many individuals appear to have anticipated short durations of use with reasonable accuracy (highlighted by Mann’s interpretation), borrowers’ individual predictions did not appear to be correlated with their actual outcomes, and virtually none accurately predicted long durations (which is the market failure described here).1131 E. Major Impacts of the Rule hazard of subsequent loan sequences ending, which implies that individuals with long sequences tend to have longer subsequent loan sequences. 1130 Again, technically these findings can be reconciled with a rational expectations model if one assumes that borrowers have no information on which to base their predictions of their length of indebtedness, but as argued in the preceding footnote, this assumption cannot be plausibly maintained. 1131 It should be noted that Professor Mann did not provide his data to the Bureau, either prior to the proposal, or in his comment in response to the proposal. In place of these data, the Bureau is relying on the charts and graphs he provided in his correspondence with and presentation to the Bureau. Among other things, these graphs depict the distribution of borrowers’ expectations and outcomes, but as they are scatterplots, counting the number of observations in areas of heavy mass (e.g., expecting no rollovers) is difficult. However, the scatterplot depicts only sequences up to approximately 170 days in length, while subsequent histograms of sequence length show a large portion of borrowers experiencing sequences of 200 or more days (approximately 13 percent). It appears these borrowers are not depicted on the scatterplots. As such, the analysis provided here may be somewhat imprecise. 1307 The primary impact of this rule, prior to any reforms it may prompt in market practices, will be a substantial reduction in the volume of short-term payday and vehicle title loans (measured in both number and total dollar value), and a corresponding decrease in the revenues that lenders realize from these loans. Simulations based on the Bureau’s data indicate that payday loan volumes will decrease by 62 percent to 68 percent, with a corresponding decrease in revenue.1132 Simulations of the impact on short-term vehicle title lending predict a decrease in loan volumes of 89 percent to 93 percent, with an approximately equivalent reduction in revenues. The specific details, assumptions, and structure of these simulations are described in detail below. The Bureau expects these declines will result in a sizable decrease in the number of storefronts, as was observed in States that experienced similar declines after adopting regulations of loan volumes (e.g., Washington). This decline may limit some physical access to credit for consumers, and this limit may be felt more acutely by consumers in rural areas. Additionally, the decrease in storefronts is likely to impact small lenders and lenders in rural areas more than larger lenders and those in areas of greater population density. However, borrowers in rural areas are expected to retain much of their access to these loans. In States with regulatory changes that led to decreases in storefronts, over 90 percent of borrowers had to travel an additional five miles or less in order to obtain such a loan. Additionally, the Bureau expects that online options will be available to the vast majority of current borrowers, including those in rural 1132 The Bureau ran a number of simulations based on different market structures that may result after the rule. The estimates cited here come from the specifications where lenders make loans under both the ATR and principal stepdown approaches. See part VII.F.1.c for descriptions of all the simulations conducted by the Bureau, and their results. 1308 areas.1133 Consumers may also substitute non-restricted borrowing options (e.g., longer-term loans not covered by the originations portion of the rule, credit cards, informal borrowing from family or friends, or other alternatives). As discussed further below, the welfare impacts of the decline in lending are expected to be positive for consumers, and negative for lenders. Decreased revenues (more precisely, decreased profits) in an industry with low concentration are expected to lead to exit by many current providers. Additionally, many of the restrictions imposed by the rule could have been voluntarily adopted by lenders absent the rule; that they were not implies the changes are likely to be at least weakly welfare-decreasing for lenders. As for the welfare impact on consumers, in an efficient market (one that is competitive, fully informed, and in which agents are rational and possess perfect foresight) a decrease in access to credit should decrease consumer welfare (though consumers would save an amount equal to the revenue lost by lenders). However, as discussed in Market Concerns—Underwriting, the section-by-section analysis for § 1041.4, and throughout this analysis, the payday and vehicle title lending markets exhibit characteristics consistent with a market failure. If some of the demand for these loans results from departures from rational expectations (or any other violation of neoclassical economic theory), reducing access may improve consumer welfare. To weigh these possible outcomes, the Bureau conducted a broad assessment of the literature pertaining to the welfare effects of short-term payday and vehicle title loans. A summary of this assessment is presented in part VII.F.2.c. 1133 This geographic impact on borrowers is discussed in the section on Reduced Geographic Availability of Covered Short-Term Loans in part VII.F.2.b.v below. 1309 The Bureau believes that the evidence on the impacts of the availability of payday loans on consumer welfare indeed varies. In general, the evidence to date suggests that access to payday loans appears to benefit consumers in circumstances where they use these loans for short periods to address an unforeseen and discrete need, such as when they experience a transitory and unexpected shock to their incomes or expenses. However, in more general circumstances, access to and intensive use of these loans appears to make consumers worse off. A more succinct summary is: access to payday loans may well be beneficial for those borrowers with discrete, short-term needs, but only if they are able to successfully avoid long sequences of loans. Short-term vehicle title borrowers are more likely to find that they are unable to obtain an initial loan because the principal step-down approach does not provide for vehicle title loans. Many of these consumers may choose to pursue a payday loan instead and seek to avail themselves of the principal step-down approach. However, as noted later, State restrictions and the financial condition of these borrowers may limit these options. As this rule will allow for continued access to the credit that appears most beneficial— that which assists consumers with discrete, short-term needs—the Bureau believes that much of the welfare benefit estimated in the literature will be preserved, despite the substantial reduction in availability of re-borrowing. Additionally, the rule limits the harm that may be realized by borrowers who experience long durations of indebtedness where the literature, albeit more limited, and the Bureau’s own analysis and study suggest the welfare impacts of prolonged re- 1310 borrowing are negative. Given this, the Bureau has concluded that the overall impacts of the decreased loan volumes resulting from the rule for consumers will be positive.1134 Relative to the considerations above, the remaining costs and benefits of this rule are much smaller. Most of these impacts manifest as administrative, compliance, or time costs; or as benefits from reductions in fraud or increased transparency. The Bureau expects most of these impacts to be fairly small on a per loan/customer/lender basis. These impacts include, inter alia, those applicable to the registered information systems envisioned by the rule’s requirements; those associated with furnishing requirements on lenders and consumers (e.g., cost to establish connection with registered information systems, benefit from reduced fraud); those associated with conducting an ATR assessment for loans that require such an assessment (e.g., cost to obtain a consumer report, benefit of decreased defaults); those associated with the increased requirements for record retention; those associated with disclosures regarding principal stepdown loans; those associated with the prescribed payment interventions (e.g., cost from additional disclosures, benefits from reduced NSF or overdraft fees); and the additional benefits associated with reduced loan volumes (e.g., changes in defaults or account closures). Each of these costs and benefits, broken down by market participant (lender, registered information system, consumer) is discussed in detail below. In addition, the Bureau has conducted a Final Regulatory Flexibility Analysis (FRFA), which describes the impact of the rule on small entities, responds to the significant issues raised 1134 Note that the Bureau has observed that longer-term balloon-payment loans are uncommon in the current market. As such, while the rule’s relative impact on these loans is expected to be similar to the impact on payday and vehicle title loans, the absolute magnitude of the impact on these loans is expected to be small. This is because the Bureau takes the current market as its baseline, and longer-term balloon-payment loans represent a small share of covered loans in this baseline. 1311 by the public comments and the Chief Counsel for Advocacy of the Small Business Administration regarding the proposal’s Initial Regulatory Flexibility Analysis, and describes changes made to the proposed rule in the final rule in response to these comments. The FRFA also provides an estimate of the number of small entities to which the final rule will apply; descriptions of the projected reporting, recordkeeping, and other compliance requirements of the rule; and a description of the steps the Bureau has taken to minimize the significant economic impact on small entities and a statement of the reasons for selecting the final rule over the other significant alternatives considered. The Bureau has also conducted a Paperwork Reduction Act (PRA) analysis to estimate the cost in burden hours and the dollar costs of the information collection requirements to the entities subject to the rule. The PRA separates these cost estimates into one-time and annual ongoing categories for total burden cost, labor burden hour cost, and labor burden dollar cost. Cost estimates are included for the requirements of the rule relating to disclosure, obtaining and furnishing consumer information, obtaining a consumer report, underwriting, registered information systems, prohibited payment transfers, and obtaining authorization for both small and large entities. F. Benefits and Costs of the Rule to Covered Persons and Consumers – Underwriting This section discusses the impacts of the provisions of the loan origination portions of the rule. Those provisions specifically relate to covered short-term loans and covered longer-term balloon-payment loans. The benefits and costs of these provisions may be affected by a shift to products not covered by the origination portions of the rule. For example, the potential for consumer substitution to longer-term installment and other loans may have implications for the 1312 effects of these provisions on those non-covered markets. The Bureau also acknowledges that some new products may develop in response to this rule, to cater to displaced demand for shortterm liquidity. In fact, many of the rule’s exclusions and exemptions are intended to encourage innovation in this market space. However, the potential evolution of substitutes in the market that may arise in response to this rule is beyond the scope of this analysis. Potential interactions with existing products are discussed as appropriate. The provisions discussed here include the requirements under § 1041.5 that lenders determine that applicants for short-term loans and longer-term balloon payment loans have the ability to repay the loan while still meeting their major financial obligations and paying for basic living expenses, as well as the alternative set of requirements for originating short-term loans discussed in § 1041.6. In this analysis, the practice of making loans after determining that the borrower has the ability to repay the loan will be referred to as the “ATR approach,” while the practice of making loans by complying with the alternative requirements under § 1041.6 will be referred to as the “principal step-down approach.” The procedural requirements for originations, and the associated restrictions on reborrowing, are likely to have a substantial impact on the markets for these products. In order to present a clear analysis of the benefits and costs of the rule, this section first describes the benefits and costs of the rule to covered persons and then discusses the implications of the rule for the overall markets for these products. The benefits and costs to consumers are then described. 1. Benefits and Costs to Covered Persons 1313 The rule imposes a number of procedural requirements on lenders making covered shortterm and longer-term balloon-payment loans, as well as imposing restrictions on the number of these loans that can be made. This section first discusses the benefits and costs of the procedural requirements for lenders using the ATR approach with regard to originating loans and furnishing certain related information to registered information systems over the life of the loan. This is followed by a discussion of the benefits and costs of the procedural requirements for lenders using the principal step-down approach. The final section discusses the potential impacts on loan volume and revenues of the underwriting and re-borrowing restrictions under both the ATR and the principal step-down approach. Most if not all of the provisions are activities that lenders could choose to engage in absent the rule. The benefits to lenders of those provisions are discussed here, but to the extent that lenders do not voluntarily choose to engage in the activities, it is likely the case that the benefits to lenders, in the lenders’ view, do not currently outweigh the costs to lenders.1135 The Bureau received many comments discussing the analysis of costs and benefits provided in the proposal. These comments came from industry, trade groups, consumer groups, customers, academic and other researchers, and others. Many of these comments offered general critiques of the assumptions made by the Bureau (e.g., with respect to time to process applications or cost to implement compliance systems), and others pointed out perceived deficiencies in the costs and benefits considered (e.g., should bolster discussion of the benefits from avoiding unaffordable payments, or should provide deeper consideration of the cost of 1135 It is possible that coordination problems limit the development of market improvements. This would be the case if such improvements are in the interest of each lender individually, but only if such improvements are undertaken by all lenders in the market. 1314 reduced access to credit). Relatively few comments offered data, evidence, or specific values for the costs or benefits likely to arise from the rule. Those comments that offered information of direct relevance to the analysis of costs and benefits have been considered—and where applicable, have been incorporated into—the analysis that follows. a. Procedural Requirements – ATR Approach Lenders making loans using the ATR approach need to comply with several procedural requirements when originating loans. Lenders need to consult their own records and the records of their affiliates to determine whether the borrower had taken out any prior short-term loans or longer-term balloon-payment loans that were still outstanding or were repaid within the prior 30 days. Lenders must obtain a consumer report from a registered information system (if available) in order to obtain information about the consumer’s borrowing history across lenders, and are required to furnish information regarding covered loans they originate to all registered information systems.1136 Lenders are also required to obtain and verify information about the amount of an applicant’s income (unless not reasonably available) and major financial obligations. Specifically, lenders must obtain a statement from applicants of their income and payments on major financial obligations, verification evidence where reasonably available regarding income, and a consumer report from a nationwide consumer reporting agency to verify major financial obligations. Lenders must assess that information and apply an estimate of the 1136 The Bureau received comments from a number of specialty consumer reporting agencies that indicated they believed themselves to be eligible to become registered information systems. Additionally, at least three of these companies have publically expressed interest in becoming registered information systems. As such, the Bureau believes there will be at least one registered information system when the market reaches steady-state. 1315 borrower’s basic living expenses in order to determine whether a consumer has the ability to repay the loan. Each of the procedural requirements entails costs that are likely to be incurred for loan applications, and not just for loans that are originated. Lenders will likely avoid incurring the full set of costs for each application by establishing procedures to reject applicants who fail a screen based on a review of partial information. For example, lenders are unlikely to collect any further information if their records show that a borrower is ineligible for a loan given the borrower’s prior borrowing history. The Bureau expects that lenders will organize their underwriting process so that the more costly steps of the process are only taken for borrowers who satisfy other requirements. Many lenders currently use other screens when making loans, such as screens meant to identify potentially fraudulent applications. If lenders employ these screens prior to collecting all of the required information from borrowers, that will eliminate the cost of collecting additional information on borrowers who fail those screens. But in most cases lenders will incur some of these costs evaluating loan applications that do not result in an originated loan, and in some cases lenders will incur all of these costs in evaluating loan applications that are eventually declined. Finally, lenders are required to develop procedures to comply with each of these requirements and train their staff in those procedures. The Bureau believes that many lenders use automated systems when originating loans and will modify those systems, or purchase upgrades to those systems, to incorporate many of the procedural requirements of the ATR approach. The costs of modifying or upgrading such a system and training staff are discussed 1316 below, in the discussion of the costs of developing procedures, upgrading systems, and training staff. i. Consulting Lender’s Own Records In order to consult its own records and those of any affiliates, a lender will need a system for recording loans that can be identified as being made to a particular consumer and a method of reliably accessing those records. The Bureau believes that lenders will most likely comply with this requirement by using computerized recordkeeping. A lender operating a single storefront will need a system of recording the loans made from that storefront and accessing those loans by consumer. A lender operating multiple storefronts or multiple affiliates will need a centralized set of records or a way of accessing the records of all of the storefronts or affiliates. A lender operating solely online will presumably maintain a single set of records; if it maintains multiple sets of records it will need a way to access each set of records. The Bureau believes that lenders must track their loans in order to service them. In addition, lenders need to track the borrowing and repayment behavior of individual consumers to reduce their credit risk, such as by avoiding lending to a consumer who has defaulted on a prior loan. And most States that allow payday lending have requirements that implicitly require lenders to have the ability to check their records for prior loans to a loan applicant, including limitations on renewals or rollovers, or cooling-off periods between loans. As such, existing business needs for recordkeeping ensure that most lenders already have the ability to comply with this provision, with the possible exception of lenders with affiliates that are run as separate operations. Still, there may be a small minority of lenders that currently do not have the capacity to comply with this requirement. 1317 Developing this capacity will enable these lenders to better service the loans they originate and to better manage their lending risk, such as by tracking the loan performance of their borrowers. Lenders that do not already have a records system in place will need to incur a one-time cost of developing such a system, which may require investment in information technology hardware and/or software. The Bureau estimates that purchasing necessary hardware and software will cost approximately $2,000, plus $1,000 for each additional storefront. The Bureau estimates that firms that already have standard personal computer hardware, but no electronic recordkeeping system, will need to incur a cost of approximately $500 per storefront. Lenders may instead contract with a vendor to supply part or all of the systems and training needs. For lenders that choose to access their records manually, rather than through an automated loan origination system, the Bureau estimates that doing so will take an average of nine minutes of an employee’s time. The Bureau received no comments from industry or trade groups asserting that a substantial number of lenders currently lack the ability to check their records for prior loans, or that implementing such a system would constitute an undue cost or burden. The Bureau believes this supports the benefit-cost framework laid out here. The Bureau did receive some comments noting that it had underestimated the costs associated with developing a system capable of allowing lender personnel to check the lender’s records, including by not accounting for training, maintenance, or furnishing costs. It was suggested by some commenters that these costs would be especially burdensome for small lenders. The Bureau addresses systems and training costs, and explicitly discusses the impacts on smaller lenders, in part VIII. The Bureau believes most lenders already have systems in place for which training must occur, and acknowledges that 1318 training for any new systems developed based on this rule would largely replace or be added to that training. ii. Obtaining a Consumer Report from a Registered Information System The Bureau believes that many lenders already obtain from third parties some of the information that will be included in the registered information system data. For example, in many States a private third party operates a database containing loan information on behalf of the State regulator, and many lenders utilize similar third parties for their own risk management purposes (e.g., fraud detection). However, the Bureau recognizes that there also is a sizable segment of lenders making short-term loans or longer-term balloon-payment loans that operate only in States without a State-mandated loan database, and who choose to make lending decisions without obtaining any data from a specialty consumer reporting agency. Lenders will receive benefits from being able to obtain timely information about an applicant’s borrowing history from a registered information system. This information will include reasonably comprehensive information about an applicant’s current outstanding covered loans, as well as his or her borrowing history with respect to such loans. Lenders that do not currently obtain consumer reports from specialty consumer reporting systems will benefit from reports from a registered information system through reduced risks of fraud and default. Additionally, the rule requires furnishing to registered information systems of all covered shortterm and longer-term balloon-payment loans, meaning that even lenders that already receive reports from specialty consumer reporting agencies will benefit by receiving more comprehensive and complete information. 1319 As noted above, the Bureau believes that many lenders use automated loan origination systems and will modify those systems or purchase upgrades to those systems such that they will automatically order a report from a registered information system during the lending process. For lenders that order reports manually, the Bureau estimates that it will take approximately nine minutes on average for a lender to request a report from a registered information system. For all lenders, the Bureau expects that access to a registered information system will be priced on a “per-hit” basis, where a hit is a report successfully returned in response to a request for information about a particular consumer at a particular point in time. The Bureau estimates that the cost per hit will be $0.50, based on pricing in existing relevant consumer reporting markets. The Bureau received comments from trade groups and lenders discussing the estimated “per hit” costs of the registered information system reports. The comments were approximately evenly split as to whether the estimated costs were substantially too low, slightly too low, or approximately accurate. A trade group representing mostly large depository institutions argued the cost is substantially too low, and cited its members’ average costs of $10.97 to purchase a credit report. Given the drastic difference between this cost and those stated by other commenters, the Bureau believes the credit reports referred to (e.g., tri-bureau credit reports) are not the type that would be purchased for this type of loan. This comparison did not seem relevant to the cost to obtain a report from a registered information system. A trade group representing small-dollar lenders also asserted the estimated cost was too low, citing its members’ average cost of $1 to obtain a credit report from a nationwide consumer reporting agency. Finally, a large small-dollar lender asserted the $0.50 estimate “appears to be right.” Given that registered information systems are likely to collect much less data than are collected 1320 by consumer reporting agencies operating in the market today, it follows that the cost of a report from a registered information system should be lower. Given that the comments received directly from lenders regarding the expected costs of a registered information system report argued the estimate is generally accurate, the Bureau continues to believe the cost per hit estimate of $0.50 is reasonable. Additionally, lenders will only need to pull a report from one registered information system. In the event that more than one registered information system enters the market, the Bureau believes that competition is likely to put downward pressure on the price of a report.1137 iii. Furnishing Information to Registered Information Systems Lenders making covered short-term and longer-term balloon-payment loans are required to furnish information about those loans to all information systems that have been registered with the Bureau for 180 days or more, have been provisionally registered with the Bureau for 180 days or more, or have subsequently become registered after being provisionally registered (generally referred to here as registered information systems). At loan consummation, the information furnished must include identifying information about the borrower, the type of loan, the loan consummation date, the principal amount borrowed or credit limit (for certain loans), and the payment due dates and amounts. While a loan is outstanding, lenders must furnish information about any update to information previously furnished pursuant to the rule within a reasonable period of time following the event prompting the update. And when a loan ceases to be an outstanding loan, lenders must furnish the date as of which the loan ceased to be 1137 As noted previously in this part, at least three specialty consumer reporting agencies have publicly expressed interest in becoming registered information systems. As such, the Bureau believes there will be at least one—and potentially multiple—registered information systems. 1321 outstanding and whether all amounts owed in connection with the loan were paid in full, including the amount financed, charges included in the cost of credit, and charges excluded from the cost of credit. Furnishing data to registered information systems will benefit all lenders by improving the coverage and quality of information available to lenders relative to the baseline. This will allow lenders to better identify borrowers who pose relatively high default risk, and the richer information and more complete market coverage will make fraud detection more effective relative to the baseline. Furnishing information to registered information systems also requires lenders to incur one-time and ongoing costs. One-time costs include those associated with establishing a relationship with each registered information system, and developing policies and procedures for furnishing the loan data and procedures for compliance with applicable laws.1138 Lenders using automated loan origination systems will likely modify those systems, or purchase upgrades to those systems, to incorporate the ability to furnish the required information to registered information systems.1139 The Bureau believes that large lenders rely on proprietary loan origination systems, and estimates the one-time programming cost for large respondents to update their systems to carry out the various functions to be 1,000 hours per entity.1140 The 1138 In the event that multiple registered information systems enter the market, the Bureau anticipates that some will choose to furnish information to the other registered information systems on behalf of the lender, as a way to compete for that lender’s business. Other third parties may also provide this service. 1139 Some software vendors that serve lenders that make payday and other loans have developed enhancements to enable these lenders to report loan information automatically to existing State reporting systems. 1140 In the PRA analysis prepared by the Bureau, the burden hours estimated to modify loan origination systems is 500. This is because only some of the system modifications are for functions related to information collections covered by the PRA. See Bureau of Consumer Fin. Prot., Paperwork Reduction Act Information Collection Request, Supporting Statement Part A, Payday, Vehicle Title and Certain High-Cost Installment Loans (12 CFR part 1322 Bureau believes small lenders that use automated loan origination systems rely on licensed software. Depending on the nature of the software license agreement, the Bureau estimates that the one-time cost to upgrade this software will be $10,000 for lenders licensing the software at the entity-level and $100 per “seat” (or user) for lenders licensing the software using a seatlicense contract. These systems are for furnishing information to, and receiving information from, registered information systems, obtaining consumer reports, and assessing ability to repay. Given the price differential between the entity-level licenses and the seat-license contracts, the Bureau believes that only small lenders with a significant number of stores will rely on the entity-level licenses. The ongoing costs will be the costs of accurately furnishing the data.1141 Lenders with automated loan origination and servicing systems with the capacity of furnishing the required data will have very low ongoing costs.1142 Lenders that report information manually will likely 1041) (posted Jul. 22), available at https://www.regulations.gov/document?D=CFPB-2016-0025-0002. The Bureau notes that these costs include the anticipated costs to establish connections to furnish to, and pull from, registered information systems. If more than one registered information system exists (as noted previously, multiple companies have publically expressed interest in becoming registered information systems), the programming costs may increase. The Bureau estimates this increase to be approximately 250 additional hours of programming per registered information system. 1141 The Bureau also received comments noting that lenders will have to incur additional costs associated with dispute resolution. One commenter specifically noted that consumers would dispute negative data contained on their reports which would require investigation along with company responses. The commenter cited a figure of $50,000 per year to handle these disputes and other costs of furnishing. The Bureau acknowledges there may be ancillary costs associated with such disputes, but believes that furnishing accurate data and compliance with the records management requirements should mitigate the costs associated with dispute resolutions (e.g. confirming the existence of the loan and any payments made). Additionally, many of the costs associated are expected to be borne by registered information systems, as the FCRA allows consumers to dispute information directly with the consumer reporting agency. As such, the $50,000 figure cited by the commenter seems inflated. Instead, the Bureau believes the costs associated with these activities are included in the ongoing costs associated with furnishing to registered information systems. 1142 The Bureau notes there could be modest per-loan furnishing costs (e.g., comparable to the costs of pulling from a registered information system). This will largely depend on the business model(s) adopted by registered information systems, and must be consistent with § 1041.11(b)(1), which requires registered information systems to 1323 do so through a web-based form, which the Bureau estimates will take three minutes to fill out for each loan at the time of consummation, when information is updated (as applicable), and when the loan ceases to be an outstanding loan. If multiple registered information systems exist and they do not share data, it may be necessary to incur this cost multiple times, unless there are services that report to all registered information systems on behalf of a lender. 1143 The Bureau notes that in States where a private third-party operates a database on behalf of a State regulator, some lenders are already required to provide information similar to that required under the rule, albeit to a single entity; such lenders thus have experience complying with this type of requirement. Where possible, the Bureau will also encourage the development of common data standards for registered information systems in order to reduce the costs of providing data to multiple information systems. iv. Obtaining Information and Verification Evidence about Income and Major Financial Obligations Lenders making loans under the ATR approach are required to collect information about the amount of income and major financial obligations from the consumer, make certain efforts to verify that information, and use that information to make an ability-to-repay determination. The impact on lenders with respect to applicants who a lender does not determine have the ability to repay, and are thus denied loans, is discussed separately. facilitate the timely and accurate transmission and processing of information in a manner that does not impose unreasonable costs or burdens on lenders. 1143 Should there be multiple registered information systems, the Bureau believes that one or more registered information systems or other third parties will offer to furnish information to all registered information systems on behalf of the lender. 1324 The Bureau believes that many lenders that make covered short-term and longer-term balloon-payment loans, such as storefront lenders making payday loans, already obtain some information on consumers’ income. Many of these lenders, however, only obtain income verification evidence the first time they make a loan to a consumer, or for the first loan following a substantial break in borrowing. Other lenders, such as some vehicle title lenders or some lenders operating online, may not currently obtain any income information, let alone income verification evidence, before issuing loans. In addition, many consumers likely have multiple income sources that are not all currently documented in the ordinary course of short-term lending. Under the rule, consumers and lenders might have incentives to provide and gather more income information than they do currently in order to establish the borrower’s ability to repay a given loan. The Bureau believes that most lenders that originate short-term and longerterm balloon-payment loans do not currently collect information on applicants’ major financial obligations, let alone attempt to verify such obligations, or determine consumers’ ability to repay a loan, as is required under the rule. As noted above, many lenders already use automated systems when originating loans. These lenders will likely modify those systems or purchase upgrades to those systems to automate many of the tasks that are required by the rule. Lenders are required to obtain a consumer report from a nationwide consumer reporting agency to verify applicants’ required payments under debt obligations unless, within the preceding 90 days, that lender has obtained a report that the lender retained and the consumer has not triggered a cooling-off period. See § 1041.5(c)(2)(ii)(D). As such, these consumer reports will usually be necessary to obtain only for the first loan in a new sequence of borrowing that 1325 begins more than 90 days since the last consumer report was obtained. This is in addition to the cost of obtaining a report from a registered information system, though the Bureau expects some registered information systems will provide consolidated reports.1144 Verification evidence for housing costs may be included on an applicant’s nationwide consumer report, if the applicant has a mortgage; otherwise the lender may reasonably rely on the consumer’s written statement as to housing expense. Based on industry outreach, the Bureau believes these reports will cost approximately $2.00 for small lenders and $0.55 for larger lenders. At least one trade group suggested this to be an accurate estimate, by noting its members pay around $1 per hit for such reports.1145 As with the ordering of reports from registered information systems, the Bureau believes that many lenders will modify or upgrade their loan origination system to allow the system to automatically order a national consumer report during the lending process at a stage in the process where the information is relevant, or to purchase combined reports from registered information systems that may offer them. For lenders that order reports manually, the Bureau estimates that it will take approximately nine minutes on average for a lender to request a report and incorporate it into the ATR determination. Lenders that do not currently collect income or verification evidence for income will need to do so. The Bureau estimates it will take roughly three to five minutes per application for lenders that use a manual process to gather and review information, for consumers who have straightforward documentation (e.g., pay stubs or bank statements). Some industry commenters 1144 The Bureau notes that, as discussed in the section-by-section analysis for § 1041.5(c)(2), lenders may order their information requests in a way that would minimize unnecessary impacts on consumers’ credit scores. Even with the consolidated reports envisioned here, lenders and the providers for the registered information systems could stagger the delivery of such reports so as to minimize the negative scoring impacts on consumers. 1145 Others suggested it would cost as high as $12 per hit, but the Bureau believes these estimates were unreasonably high. 1326 suggested this value was too low in the proposal, often citing cases where consumers may not have regular income from sources that provide documentation. The Bureau notes that many lenders already require such information prior to initiating loans. Additionally, the rule allows stated income to be used in appropriate cases where verification evidence is not reasonably available, reducing the average time cost associated with verification efforts. However, lenders will need to obtain a brief statement from consumers about their incomes and expenses prior to verification. As such, the Bureau believes the time estimates provided here to be reasonable. Some consumers may visit a lender’s storefront without the required income documentation and may have income for which verification evidence cannot be obtained. Lenders making loans online may face particular challenges obtaining verification evidence, especially for income. It may be feasible for online lenders to obtain scanned or photographed documents as attachments to an electronic submission; the Bureau understands that some online lenders are doing this today. And services that use other sources of information, such as checking account or payroll records, may mitigate the need for lenders to obtain verification evidence directly from consumers. Such services may be especially appealing to online lenders, to whom it might be more difficult to provide copies of physical pay stubs, bank statements, or other documentation of income. Additionally, for consumers with cash income that is not deposited into a deposit account, lenders will be allowed to rely on stated information, § 1041.5(c)(2)(ii), lowering the lenders’ costs relative to the proposal and the chance that a consumer is unable to complete an application. v. Making the Ability-to-Repay Determination 1327 Once information and verification evidence on income and major financial obligations has been obtained, the lender must use that information and evidence to make a reasonable determination that the consumer will have the ability to repay the contemplated loan. In addition to considering the information collected about income and major financial obligations, lenders must reasonably estimate an amount that the borrower needs for basic living expenses. They may do this in a number of ways, including, for example, collecting information directly from borrowers, using available estimates published by third parties, or basing estimates on their experience with similarly situated consumers. See comment 5(b)-2.i.C. The initial costs of developing methods and procedures for gathering information about major financial obligations and income, and estimating basic living expenses, are discussed further below. As noted above, the Bureau believes that many lenders use automated loan origination systems, and will modify these systems or purchase upgrades to these systems to make the ability-to-repay calculations. vi. Total Procedural Costs of the ATR Approach In total, the Bureau estimates that obtaining a statement from the consumer, taking reasonable steps to verify income, obtaining a report from a nationwide consumer reporting agency and a report from a registered information system, projecting the consumer’s residual income or debt-to-income ratio, estimating the consumer’s basic living expenses, and arriving at a reasonable ATR determination will take essentially no additional time for a fully automated electronic system and between 15 and 45 minutes for a fully manual system.1146 Numerous 1146 Note that times are increases above the baseline. That is, they represent additional time beyond that which is already taken to originate such loans, such as the time spent on income verification for payday loans. 1328 industry commenters suggested the estimate provided by the Bureau in the proposal (15 to 20 minutes) was too low. In response to these comments, the Bureau has increased its estimated time to manually underwrite these loans, but also notes that all major financial obligations should be obtainable either from a consumer report or consumer statement (in the example of rental expense). Further, total costs will depend on the existing utilization rates of, and wages paid to, staff that will spend time carrying out this work. To the extent that existing staff has excess capacity (that is, that a lender’s employees have time that is not fully utilized), the extra time to process applications for loans made via the ATR approach should not result in higher wage bills for the lender. Further, as the Bureau expects the majority of loans to be made via the principal step-down approach, the expected increase in staff hours necessary to comply with the new procedural requirements should be modest.1147 Still, to the extent that lenders must increase staff and/or hours to comply with the procedural requirements, they may experience increased costs from hiring, training, wages, and benefits. Dollar costs will include a consumer report from a nationwide consumer reporting agency costing between $0.55 and $2.00 and a report from a registered information system costing $0.50. Lenders relying on third-party services to gather verification information about income may face an additional small cost. vii. Developing Procedures, Upgrading Systems, and Training Staff 1147 In the Bureau’s simulations, the ratio of loans made via the principal step-down approach to those made via the ATR approach is approximately 14:1. 1329 Lenders need to develop policies and procedures to comply with the requirements of the ATR approach and train their staff in those procedures. Many of these requirements do not appear qualitatively different from many practices that most lenders already engage in, such as gathering information and documents from borrowers and ordering various types of consumer reports. Developing procedures to make a reasonable determination that a borrower has an ability to repay a loan without re-borrowing and while paying for major financial obligations and basic living expenses is likely to be a challenge for many lenders. The Bureau expects that vendors, law firms, and trade associations are likely to offer both products and guidance to lenders, potentially lowering the cost of developing procedures as service providers can realize economies of scale. Lenders must also develop a process for estimating borrowers’ basic living expenses if they choose not to make an individual determination for each customer. Some lenders may rely on vendors that provide services to determine ability to repay that include estimates of basic living expenses. Some methods for conducting an analysis to determine estimates of basic living expenses could be quite costly. There are a number of government data sources and online services, however, that lenders may be able to use to obtain living expense estimates. Additionally, lenders may rely on their experiences with similarly situated consumers in making this estimate, reducing the need to rely on individual measures or third parties. As noted above, the Bureau believes that many lenders use automated systems when originating loans and will incorporate many of the procedural requirements of the ATR approach into those systems. This will likely include an automated system to make the ability-to-repay determination; subtracting the component expense elements from income itself, or comparing the 1330 component expenses to income to develop a ratio, is quite straightforward and should not require substantial development costs. The costs of these systems are discussed above. One trade group commented that they believe the Bureau’s estimated systems costs to be too low, citing a survey of their members. However, the trade group’s members are not predominately involved in making loans that will be covered under the rule, so it is unclear how their estimates relate to the systems contemplated here. Additionally, the vast majority of the comments from more directly-related trade groups and lenders remained silent on these estimates, despite the invitation to provide feedback. As such, the Bureau has not changed these values from those put forth in the proposal. The Bureau estimates that lender personnel engaging in making loans will require approximately 5 hours per employee of initial training in carrying out the tasks described in this section and 2.5 hours per employee per year of periodic ongoing training.1148 b. Procedural Requirements – Principal Step-Down Approach The procedural requirements of the principal step-down approach will generally have less impact on lenders than the requirements of the ATR approach. Specifically, the rule does not mandate that lenders obtain information or verification evidence about income or major financial obligations, estimate basic living expenses, or complete an ability-to-repay determination prior to making loans that meet the requirements of the principal step-down approach.1149 1148 These training costs represent the total costs to comply with the rule, including training to conduct an underwriting assessment, pull a credit report, assess borrower history, and comply with disclosure requirements. The specific breakdown of these times can be found in part VIII. 1149 As discussed above, the Bureau believes that, in certain circumstances, lenders may choose to strengthen their internal screening processes in order to increase the probability that loans would be paid in full over a sequence of three principal step-down approach loans, since the rule would restrict further re-borrowing. 1331 Instead, lenders making loans under § 1041.6 must consult their internal records and those of affiliates, obtain reports from a registered information system, furnish information to all registered information systems, and make an assessment that certain loan requirements (such as principal limitations and restrictions on certain re-borrowing activity) are met. The requisite disclosures are discussed below. The requirement to consult the lender’s own records is slightly different than under the ATR approach, as the lender must check the records for the prior 12 months. This is unlikely to have different impacts on lenders, however, as any system that allows the lender to comply with the requirement to check its own records under the ATR approach should be sufficient for the principal step-down approach, and vice-versa. A lender will also have to develop procedures and train staff. i. Disclosure Requirement Lenders making short-term loans under the principal step-down approach are required to provide borrowers with disclosures, described in the section-by-section analysis of § 1041.6(e), with information about their loans and about the restrictions on future loans taken out using the principal step-down approach. One disclosure is required at the time of origination of a first principal step-down approach loan, where a borrower had not had a principal step-down approach loan within the prior 30 days. The other disclosure is required when originating a third principal step-down approach loan in a sequence, because the borrower would therefore be unable to take out another principal step-down approach loan within 30 days of repaying the loan being originated. The disclosures will need to be customized to reflect the specifics of the individual loan. 1332 By informing borrowers that they are not permitted to take out another covered loan for the full amount of their current loan within 30 days of repaying the current loan, the first disclosure may help lenders reduce defaults by borrowers who are unable to repay the loan, even in part, without re-borrowing. Lenders may have incentives to inform borrowers of this restriction to reduce their own risk, although it is unclear if they would choose to do so absent the requirement, if they believed that the restrictions on principal and re-borrowing were likely to discourage many borrowers who could repay from taking out loans made under the principal step-down approach. The Bureau believes that most, if not all, lenders have some disclosure system in place to comply with existing disclosure requirements. Lenders may enter data directly into the disclosure system, or the system may automatically collect data from the lenders’ loan origination system. For disclosures provided via mail, email, or text message, some disclosure systems forward the information necessary to prepare the disclosures to a vendor in electronic form, and the vendor then prepares and delivers the disclosures. For disclosures provided in person, disclosure systems produce a disclosure which the lender then provides to the borrower. Respondents will incur a one-time cost to upgrade their disclosure systems to comply with new disclosure requirements. The Bureau believes that large lenders rely on proprietary disclosure systems, and estimates the one-time programming cost for large respondents to update these systems to be 1,000 hours per lender. The Bureau believes small depositories and non-depositories rely on licensed disclosure system software. Depending on the nature of the software license agreement, the Bureau estimates that the cost to upgrade this software will be $10,000 for lenders licensing 1333 the software at the entity-level and $100 per seat for lenders licensing the software using a seatlicense contract. Given the price differential between the entity-level licenses and the seatlicense contracts, the Bureau believes that only small lenders with a significant number of stores will rely on entity-level licenses. In addition to the upgrades to the disclosure systems, the Bureau estimates that small storefront lenders will pay $200 to a vendor for a standard electronic origination disclosure form template. The Bureau estimates that providing disclosures in stores will take a store employee two minutes and cost $0.10. c. Effect on Loan Volumes and Revenue from Underwriting Requirements and Re-borrowing Limits The underwriting requirements under the ATR approach and the restrictions on certain re-borrowing under both the ATR approach and principal step-down approach will impact lenders’ loan volume in a way that the Bureau believes will likely be more substantial than the increase in compliance costs from implementing the requirements discussed above. The following section discusses these impacts by lender type since storefront and online payday lenders will have the option of using both the ATR approach and principal step-down approach, while vehicle title lenders are required to use the ATR approach. Any impacts on longer-term balloon-payment loans should be similar although, as noted, such loans are currently less common and the Bureau has substantially less data about these loans. The subsequent section discusses overall combined impacts on these markets from the reduction in lender revenue and the increased procedural costs. 1334 In order to simulate the effects of the rule, it is necessary to impose an analytic structure and make certain assumptions about the impacts of the rule, and apply these to the data. The Bureau conducted three simulations of the potential impacts of this rule on payday loan volumes. The first assumes all loans are issued using the ATR approach, and simulates the impacts from both the underwriting restriction (using assumed parameters informed by both Bureau and outside research) and the restrictions on re-borrowing. The second simulation assumes all loans are issued using the principal step-down approach. This approach simulates the impacts from the sequence limits and annual caps associated with these loans, and implicitly assumes no borrowers pass ATR after exhausting the loans made under the principal step-down approach. The final simulation assumes loans are issued via both the ATR and principal step-down approaches. For loans issued via the ATR approach, the Bureau simulates the effects of both the underwriting requirement and the restrictions on re-borrowing. Generally, this is the Bureau’s preferred simulation, as it most closely mirrors the market structure the Bureau expects in response to the rule.1150 In addition, the Bureau performed a single vehicle title simulation. As vehicle title loans are not eligible for the principal step-down approach, the simulation measures the impacts of the ATR approach. As with payday, the Bureau simulates the impacts from both the underwriting restriction and the restrictions on re-borrowing. The structure, assumptions, and data used by the Bureau are described below. i. Description of the Simulations of the Rule’s Impacts on Loan Volumes 1150 The Bureau also conducted a number of additional simulations as robustness checks. While not described here, their general results were consistent with those reported in this analysis. 1335 In general, the Bureau uses its data, described in part VII.F.1.c.ii, as the basis for the simulations. The simulations filter or constrain the observed data according to constraints imposed by the rule. In simulations where principal step-down approach loans are available, the Bureau always assumes principal step-down approach loans will be made to each consumer prior to any ATR approach loans as the Bureau believes that lenders will strictly favor issuing loans under the principal step-down approach over the ATR approach. Loans made under the principal step-down approach require substantially less underwriting (in effect just verifying the customer is eligible to borrow given his/her previous indebtedness). They are, therefore, faster and less costly to originate. Perhaps more importantly, the number and duration of ATR loans restrict lenders’ abilities to make subsequent loans to a consumer under the principal step-down approach. But there are no explicit caps on the number of loans or time in debt restricting the issuance of loans made under the ATR approach, beyond the sequence-level re-borrowing restriction. As such, lenders seeking to maximize loan volume, and borrowers seeking to maintain future borrowing options, would likely favor the principal step-down approach when available, even when customers are able to demonstrate the ability to repay.1151 For loans issued under the ATR approach, the Bureau assumes the loan amount will be unchanged from the amount observed in the data. This holds for both initial loans in a sequence and for all subsequent loans in that sequence. For loans issued under the principal step-down 1151 The Bureau does note that principal step-down approach loans do have potentially binding restrictions that may make them less desirable to a small subset of consumers (e.g., lower limits, forced principal step-down), and potentially a small set of lenders (those concerned with loan amount, rather than number of loans). However, the Bureau believes the speed and cost advantage of the principal step-down approach will largely outweigh these considerations. 1336 approach, the Bureau assumes that the amount borrowed in initial loans in a sequence will be the minimum of the observed loan amount in the data, or the maximum amount allowed by the rule (i.e., $500). Subsequent loans in a sequence will be the minimum of the observed loan amount in the data, or the maximum amount allowed by the rule for subsequent loans (i.e., two-thirds of the amount of the initial loan for a second loan and one-third of the amount of an initial loan for a third loan). With respect to the underwriting of loans, in those simulations where loans made via the principal step-down approach are available in the market, the Bureau assumes that all initial loans observed in the data are originated.1152 In contrast, simulations for payday loans under the ATR approach assume that only a fraction of consumers will qualify. To assess the impact of this reduction on loans and loan volumes, the fraction of borrowers assumed to qualify for ATR is applied to weight observations in the data that show revealed demand for ATR loans.1153 The Bureau’s analysis in the proposal attempted to calculate this fraction and comments received in response to the proposal provided additional information. Many of these comments note that modeling the ability to repay of borrowers is difficult without detailed information, though some comments attempt to provide evidence for the share of borrowers likely to pass an ATR 1152 The Bureau notes that the re-borrowing restrictions imposed by the rule may provide incentives for lenders to impose additional screens on borrowers. Under certain conditions, the limit to the revenue that can be realized via re-borrowing may drive lenders to attempt to screen out borrowers who are no longer profitable to lend to. The Bureau lacks evidence on if, how, and how frequently lenders would do this, and therefore the simulations do not attempt to model this possibility. But any such voluntary underwriting would further reduce the provision of credit. This implies that the simulation results may somewhat underestimate the overall reductions in loans and revenue if the price of and demand for these loans remains constant. 1153 As the specific loans that would pass ATR are unknown, the Bureau weights all potential loans by the ATR filter rate. If the loans that would pass an ATR assessment systematically vary in amount, propensity to re-borrow, or other such factors from the typical loans observed in the data, the simulations may overestimate or underestimate the impact of the ATR restriction (e.g., if a loan that would pass ATR is actually larger in amount, and rolled over more often than the typical loan, the estimated decreases in revenue by the simulations would be overstated). 1337 determination. The Bureau has reviewed these comments and, as appropriate, used their content to inform its assumptions. However, the Bureau continues to believe that determining the share of borrowers and particular loans likely to be impacted by an ATR assessment is necessarily imprecise. The details of the calculations are included below. The Bureau applies this underwriting filter to both payday and vehicle title loans. While the Bureau believes that the data and comments relating to the share of payday borrowers that could reasonably pass ATR are more informative than those relating to vehicle title borrowers, (e.g., no supporting evidence was provided to the Bureau in response to comments), the Bureau believes it is important to include an underwriting filter in its simulations of each market, and that the value of this filter may be similar across the affected products. In its ATR simulations, the Bureau assumes that each subsequent ATR loan would be subject to the same filter. That is, the probability of originating each subsequent loan is weighted by the value of the underwriting filter. It is true that any borrower who passes an ATR assessment on his or her initial loan will likely have the same residual income or DTI on each subsequent loan within a sequence (as the lender is not required to pull a new national consumer report if, within the preceding 90 days, that lender has obtained a report that the lender retained and the consumer has not triggered a cooling-off period, and a customer’s assessed ability to repay would only change if the information obtained about income or from a registered information system changed). However, the Bureau expects that the instances of re-borrowing should be less frequent for customers who pass an ATR assessment compared to customers who fail to satisfy an ATR determination. This is due to the fact that customers who are able to repay their loans according to the terms at origination are less likely to need to re-borrow compared to 1338 those customers who are expected to struggle to repay, and require a subsequent loan to repay the previous one. Additionally, lenders may reasonably interpret the borrower’s immediate return as an indicator that the borrower may lack the ability to repay the loan according to its terms, and decide not to extend an additional loan. The Bureau cannot identify from its data those specific customers who will demonstrate an ability to repay (and applies a weighting filter to account for the attrition induced by underwriting), let alone those near the margin of demonstrating an ability to repay (who are most likely to be voluntarily cut off by lenders). As such, assuming consistent attrition in subsequent loans is a way to account for the combined effects of ATR borrowers’ lower propensities to reborrow, coupled with lenders’ likely reassessments of those borrowers’ abilities to repay. Therefore, the Bureau assumes a constant decay of re-borrowing amongst those customers who originate an ATR loan. That is, for each new would-be ATR loan present in the data, the simulation accounts for the decline in loan volumes by weighting each loan by a value that represents the combined likelihood that a customer applies and is approved for that loan. Finally, with respect to re-borrowing restrictions, as stated previously, in simulations where loans made under the principal step-down approach are available, the Bureau assumes that all initial loans are taken out under the rule. Each subsequent loan observed in the data within 30 days of a prior loan (i.e., within a sequence) is also taken out, up to the limit imposed by the rule (e.g., three). For borrowers with sequences in excess of the limit and who have not reached any of the caps on loans under the principal step-down approach, the Bureau adopts one of two assumptions in each of its simulations: either the borrower returns immediately after the triggered cooling-off period (assumes need persists), or the borrower does not return after the 1339 cooling-off period (assumes need is obviated).1154 To the extent that long sequences reflect the difficulty that borrowers have paying off large single-payment loans, rather than borrowers repeatedly experiencing new income or expense shocks that lead to additional borrowing, it is more likely that borrowers will tend not to return to borrow once a loan sequence has ended and a 30-day period has expired. Regardless, the initial loan in each new distinct sequence for a borrower as observed in the data is always assumed to be initiated, until that borrower has reached his or her limit under the rule. When a borrower shows revealed demand for an ATR loan in the simulations (e.g., in simulations with only ATR loans or with both ATR and principal step-down approach loans where the borrower has exhausted his/her principal step-down approach loans), the Bureau applies an underwriting filter to the chance that the borrower takes the loan, as discussed above. As was the case under the principal step-down approach, for ATR borrowers with sequences in excess of the limit (and who pass the underwriting screen for each of the loans1155), the Bureau adopts one of two assumptions in each of its simulations: either the borrower returns immediately after the triggered cooling-off period (assumes need persists), or the borrower does not return after the cooling-off period (assumes need is obviated). As each new loan must pass 1154 Note that monthly borrowers are unlikely to be able to borrow loans via the principal step-down approach after the third loan in a 12-month period, as they will likely have reached the 90-day limit on indebtedness. 1155 Note again that the underwriting screens are taken to be independent. While it is likely that a borrower who is able to demonstrate ATR for an initial loan in a sequence will present with similar data for subsequent loans, the Bureau believes borrowers with a demonstrated ATR would be less likely to return to re-borrow. Additionally, lenders may take a borrower’s return as an indication they initially lacked the ability to repay, and may not originate subsequent loans barring a documented improvement in the borrower’s finances. As such, the underwriting filter can be viewed as a “combined probability of successfully re-borrowing” filter for second and third loans in a sequence. 1340 the ATR screen, there is a great deal of decay in the likelihood that a new sequence of ATR loans is initiated.1156 (a). Example: Payday Simulation In the simulation the Bureau estimates as most closely resembling anticipated market impacts, the Bureau assumes loans will be available under both ATR and principal step-down approaches. Consistent with the description above, the Bureau assumes all borrowers with revealed demand for six or more loans in a 12-month period will successfully take out loans under the principal step-down approach until the cap imposed by the rule, or until they reach a forced cooling-off period (after which, by assumption, they may or may not return). The Bureau also imposed an underwriting filter on the demand for and availability of all ATR loans (i.e., all loans in excess of the limit imposed by the principal step-down approach). Consumers are allowed to continue borrowing as permitted by the re-borrowing restriction and the underwriting filter. In practical terms, the re-borrowing rate for sequences of loans made via the ATR approach declines rapidly, as the underwriting filter compounds for each subsequent loan. The Bureau conducts this simulation under the assumption that borrowers with interrupted sequences return to attempt to borrow immediately after their cooling-off periods, and under the assumption that such borrowers do not attempt to borrow again until their next distinct sequence observed in the data. This provides upper and lower bounds for the estimated impacts under this simulation, though the range between these bounds is narrow, due to the low probability of both returning to re-borrow and being approved for a subsequent loan. 1156 In practice, this represents a small share of potential loans. For an ATR borrower to take a fourth loan, he or she would have had to pass four of the combined re-borrowing and ATR screens, making the probability of being eligible for such a loan p4, where p is the probability of passing the screen. 1341 (b). Example: Vehicle Title Simulation Vehicle title loans are only available under the ATR approach because principal stepdown loans cannot include vehicle security under § 1041.6(b)(3), limiting the assumptions required for simulations of this market. In the Bureau’s simulation for vehicle title loans, the Bureau imposes the same underwriting filter applied to payday loans. This means every loan observed in the data must pass the underwriting screen (and second loans must have passed the first screen, third loans must have passed the first and second screens, and so on). Consumers are allowed to continue borrowing as permitted by the re-borrowing restriction and underwriting filter, and trigger a 30-day cooling-off period if they reach a third loan. The Bureau conducts this simulation under the two different assumptions about borrowers that experience interrupted sequences: that borrowers with interrupted sequences return to attempt to borrow immediately after their cooling-off periods, and that such borrowers do not attempt to borrow again until their next distinct sequence observed in the data. This provides upper and lower bounds for the estimated impacts under this simulation. ii. Storefront Payday Lending: Impacts on Loan Volumes, Revenues, and Stores The Bureau has simulated the impacts of the lending restrictions on loan volumes assuming that lenders only make loans using the principal step-down approach relative to lending volumes today. The simulations measure the direct effect of the restrictions by starting with data on actual lending and then eliminating those loans that would not have been permitted if the regulation had been in effect. Possible responses by lenders or borrowers are not considered in the simulations, aside from the effect discussed above on borrowers who have loan sequences interrupted by the re-borrowing restrictions. Depending on the extent to which 1342 borrowers who have loan sequences cut off by the three-loan limit will return to borrow again after the 30-day period following the third loan, the estimated impact of the lending restrictions shows a decrease in the number of loans of 55 to 62 percent, and the estimated impact on total loan volume is a decrease of 71 to 76 percent. The simulated impact on revenue is greater than the impact on loan volume because of the loan-size restrictions of the principal step-down approach, with the “step down” in the allowable loan amounts for the second and third loans in a sequence having a greater impact than the $500 limit on initial loan size. The Bureau has also simulated the effects of imposing the ATR approach only (i.e., a market with no principal step-down approach loans). Under the ATR approach a new covered short-term loan cannot be made during the term of and for 30 days following a prior covered short-term loan made under the principal step-down approach. Additionally, new ATR loans can only be originated within 30 days of a previous ATR loan if such a loan would not constitute a fourth loan in a sequence. Using data and analysis provided in the proposal, and information received in comments responding to the proposal, the Bureau has estimated the share of borrowers who would be able to satisfy this requirement to be 33 percent of the would-be borrowers. The Bureau also uses this same value, 33 percent, for subsequent ATR loans to capture the dynamics explained above (i.e., the probability a borrower applies for, and is approved for, a subsequent loan). The Bureau views this, in the absence of specific evidence, as a very conservative assumption in that it generates a larger reduction in loans than would similarly justifiable assumptions (e.g., assuming a larger share of borrowers are able to pass the new, more streamlined ATR assessment; applying a single underwriting reduction at the 1343 sequence-level rather than the loan-level; etc.). However, the Bureau notes that the results are not particularly sensitive to using any similar fraction. Using the simulation approach described above and allowing only the ATR approach produces estimates of the reduction of loan volume and lender revenue of approximately 92 to 93 percent, relative to lending volume today. Again, these estimates vary depending on what is assumed about the behavior of borrowers after the end of the 30-day period following a loan, though these differences are small, as few borrowers will pass four ATR assessments in the simulations. The Bureau received some comments citing a study that criticizes the Bureau’s simulations, arguing they underestimate the reduction in loan volumes.1157 The study in question estimates that, under the principal step-down approach only, payday loan volumes would decrease by 79.6 percent, and under the ATR approach only payday loan volumes would decrease by 90.5 to 92.7 percent. The Bureau notes these differences are fairly small (less than four percentage points for the principal step-down approach only, and within two percentage points for the ATR approach only), and considers them broadly consistent with the Bureau’s findings. Further, the Bureau believes these differences are largely attributable to methodological differences in the identification of the loan sequences likely to be affected by the rule.1158 1157 For more details see nonPrime101 “Report 9 – Evaluating CFPB Simulations of the Impact of Proposed Rules on Storefront Payday Lending,” available at https://www.nonprime101.com/report-9-evaluating-cfpb-simulationsof-the-impact-of-proposed-rules-on-storefront-payday-lending/; “Update to Report 9 – ‘Being Precise About the Impact of ‘Principal Reduction’,” available at https://www.nonprime101.com/update-report-9/. 1158 Specifically, the nonPrime101 reports do not appear to account for the left-censoring of their data. Under the rule, these individuals would likely not be observed at this stage in their borrowing. The Bureau’s approach can be interpreted as the reduction in “steady-state” loan volumes (i.e., the level of reduced loans and revenues once the 1344 The Bureau received comments citing two additional and similar studies, which estimated the effects of the principal step-down approach (with no ATR approach loans) using data covering loans made by small lenders and loans made by large lenders. These studies estimate total revenue reductions of 82% and 83% respectively.1159 The Bureau again notes that these findings are broadly consistent with the Bureau’s findings, and that there are subtle but important methodological differences which may largely account for the differences in effect sizes.1160 The Bureau feels the methodology used in its simulations should generate the most accurate estimates of the steady-state effect on loans volumes in these markets. In the simulation the Bureau believes most closely mirrors the market likely to evolve in response to this rule, borrowers are assumed to be able to take out loans under the principal step-down approach, then continue re-borrowing subject to passing an ATR determination should they still have demand for such loans (again with a 33 percent chance of applying for and passing an ATR assessment for each new ATR sequence). This is the third simulation described above. This simulation produces estimates of the reduction of loan volume and lender revenue of approximately 51 to 52 percent, relative to lending volume in the data, with corresponding revenue decreases of 67 to 68 percent. Of note in this simulation is that approximately 40 percent of the reduction in revenue is market has adjusted to the rule). The Bureau has previously described its approach to dealing with the left-censoring (see, e.g., CFPB Data Point: Payday Lending, at 10), and does so again below. 1159 Arthur Baines et al., “Economic Impact on Small Lenders of the Payday Lending Rules Under Consideration by the CFPB,” Charles River Associates, (2015), available at http://www.crai.com/publication/economic-impact-smalllenders-payday-lending-rules-under-consideration-cfpb; Arthur Baines et al., “Economic Impact on Storefront Lenders of the Payday Lending Rules Proposed by the CFPB,” Charles River Associates (2016), available at http://www.crai.com/publication/economic-impact-storefront-lenders-payday-lending-rules-proposed-cfpb. 1160 In particular, a number of loans in their evaluation period are excluded for exceeding loan caps based on the number of loans taken in a pre-policy assessment period. However, the rule’s restrictions on allowed number of loans in a 12-month period will not encompass loans made in any period prior to the rule’s effective date. 1345 the result of limits on loan sizes (i.e., $500 max for principal step-down approach, and forced step-downs), with the remaining reduction attributable to re-borrowing restrictions. Estimating the share of payday loan borrowers for whom a lender could reasonably determine ability to repay the loan requires data on borrowers’ income, details about the prospective loans (especially the payments), and data on borrowers’ major financial obligations and estimated basic living expenses. In addition, lenders may satisfy the ATR requirements in a variety of ways (e.g., verification of income via pay stubs or bank statements vs. relying on stated income, or a residual income determination vs. a DTI assessment). It is also challenging to estimate the frequency with which borrowers will seek to initiate new loans sequences after a 30-day cooling-off period. All this necessarily complicates the estimation of the effects of the requirement. As already discussed, the Bureau has assumed 33 percent of would-be ATR borrowers will pass an initial ATR determination and that for each subsequent loan 33 percent of those borrowers would apply for and pass another ATR test. To the extent more applicants will apply for a loan and pass an ATR assessment, the ATR simulation estimates above will overstate the actual decline in lending; to the extent fewer applicants will apply for a loan and pass an ATR assessment, this simulation will understate the actual decline in lending. Given the importance of the assumption, the Bureau repeats here the analysis and discussion from the proposal of the share of borrowers who would be able to demonstrate an ability to repay a payday loan. Additional analyses using proprietary data were submitted to the 1346 Bureau in comments and these analyses are discussed immediately following.1161 The Bureau notes that the estimates provided by these analyses are all broadly consistent with one another. The data the Bureau uses include information on the income and loan amounts of payday borrowers. Data on major financial obligations and basic living expenses are only available at the household level, and only for certain obligations and expenses. In addition, only some of the obligation and expense data are available specifically for payday borrowers, and in no case is the obligation or expense data tied to specific loans. Given the limited information on major financial obligations and basic living expenses it is likely the case that estimates made using the available data will overstate the share of borrowers who would demonstrate an ability to repay a payday loan. In addition, lenders may adopt approaches to estimating basic living expenses that lead to fewer borrowers satisfying the lenders’ ATR evaluations. Also note that the data and discussion to follow focus on an assessment of residual income for determining ability to repay. While a debt to income (DTI) assessment is also permitted under § 1041.5(b), it is the Bureau’s expectation that the DTI approach will not lead to substantial differences compared to the residual income approach when assessing customers’ abilities to repay. Rather, the Bureau’s inclusion of DTI is intended to give lenders more flexibility in determining how to assess ATR. Data on payday loans and their associated individual borrower incomes were obtained under the Bureau’s supervisory authority.1162 These data cover a large number of payday loans 1161 The Bureau notes that the intent of these studies was to argue that an ability-to-repay assessment is not an effective means by which to reduce default. This Section 1022(b)(2) Analysis does not evaluate these claims or the analyses on which they are based, instead, it acknowledges the usefulness of their underlying data, and uses these data to inform assumptions about the share of borrowers who are likely to pass an ATR assessment. A discussion of the main conclusions of these studies is offered in the section-by-section for § 1041.5. 1347 originated by several lenders in over 30 States.1163 To ensure the sequences observed in the Bureau’s data are not impacted by left-censoring, the Bureau looks at borrowers who take their first loans in the second month of a lender’s data. The Bureau restricts the analysis to these sequences so that it can ensure it is able to observe the first loan in a sequence and thus accurately measure sequence duration.1164 In effect, this allows the Bureau to estimate the impact on lending volumes in the steady-state, as many of the loans observed in the first month’s data are deep into a sequence, and would not have been observed under the rule. Data on household expenditures comes from the 2010 BLS Consumer Expenditure Survey (CEX). These data contain information on some of the expenditures that make up major financial obligations, including housing obligations (rent or mortgage payments) and vehicle loan payments. The CEX also contains information on various categories of basic living expenses, including utilities, food, and transportation. These expense categories would need to be considered by lenders estimating basic living expenses. An important limitation of the data is that they do not contain information for all major financial obligations; in particular the data exclude such obligations as credit card payments, student loan payments, and payments on other small-dollar loans. 1162 These data have been used in prior Bureau publications including: CFPB Payday Loans and Deposit Advance Products White Paper; CFPB Data Point: Payday Lending; and CFPB Report on Supplemental Findings, and are discussed in more detail in those publications. 1163 Note that the Bureau’s data were collected from large payday lenders, and thus may not be representative of small lenders. However, the two Charles River Associates studies cited by commenters and discussed above estimated declines in loan volumes by lender size and found similar revenue impacts on small and large entities. See the Final Regulatory Flexibility Analysis for further discussion of these studies and the anticipated impacts on small lenders. 1164 See CFPB Data Point: Payday Lending, at 10 (for more details). 1348 As noted above, the CEX collects expenditure data at the household, rather than individual, level. Lenders are required to make the ATR determination for an individual borrower, which may include reasonable considerations of income from other persons to which the borrower can show access, contributions of other persons to major financial obligations and in certain cases to basic living expenses, see comments 5(a)(5)-3, 5(c)(1)-2, 5(b)-2.i.C.2. Given the lack of available information on individual expenditures, household level income and expenditures information is presented here, though the Bureau notes these may not be directly applicable to individual-level determinations of ATR. Because the data on payday loans collected under the Bureau’s supervisory authority contain information on borrowers’ individual incomes, the Bureau used a third source of data to map individual incomes to household incomes, with particular attention on this population. Data on both individual and household incomes come from the four waves of the FDIC National Survey of Unbanked and Underbanked Households that have been conducted as a special supplement to the Current Population Survey (CPS). This provides information on the distribution of household income for individuals with individual income in a certain range. The share of the population that takes one of these types of loans is fairly small, so income data on both payday and vehicle title borrowers is used to provide more robust information on the relationship between individual and household income for this population. The CPS collects information from 60,000 nationally representative respondents in each wave, of whom roughly two percent reported having taken out a payday and over one percent reported having taken out a 1349 vehicle title loan in the past 12 months in the most recent wave of the survey.1165 These data are the most extensive source of information on both individual and household income of such borrowers that the Bureau has been able to identify. Relative to the proposal, the Bureau has continued using data on household spending and income from the 2010 CEX, while including the latest wave of the 2015 FDIC Survey data. Compared to more recent CEX data, the data should better correspond to the borrower characteristics considered by lenders in the baseline loan origination data which are from 2011 and 2012. As noted below, the Bureau also continues to use the 2010 Survey of Consumer Finances for the same reason. Incorporating the additional wave of the FDIC survey data increases the small sample of observed payday and vehicle title borrowers, improving the estimated relationship between individual and household incomes. The differences in time periods should not introduce any bias as the four waves are centered roughly over the time periods of the loan and expense data, and the Bureau is only using the CPS data for the crosswalk between individual and household income. Table 1 shows the distribution of payday loan borrowers by their reported individual monthly income based on the loan data discussed above. As the table shows, roughly half of payday loans in the data were taken out by borrowers with monthly individual incomes below $2,000. 1165 Fed. Deposit Ins. Corp., “2015 FDIC National Survey of Unbanked and Underbanked Households,” at 34 (Oct. 20, 2016), available at https://www.fdic.gov/householdsurvey/2015/2015report.pdf (Calculations made using custom data tool.). The percentage of households reporting using payday loans varied from 3.5 percent to 1.7 percent over the four waves. The percentage of households reporting using auto-title loans ranged from 0.9 to 1.3 percent in the two waves where those data were collected. Fed. Deposit Ins. Corp., “2015 FDIC National Survey of Unbanked and Underbanked Households, Appendix Tables” (Oct. 20, 2016), available at https://www.fdic.gov/householdsurvey/2015/2015appendix.pdf. 1350 TABLE 1- Em mull": 13% ?nal?$999 14.4 mum?$1499 li'j 9-1. Him?$1999 n3 film-mm 14.11% tam-32m my E-?n Hum?Mm? 15% ?aw?Hm? 43% Hum?nun 53% ?aw?Hum Hum-Mum ?aw-Mum Sauce: analysis nflnan-m?alpafdaj' dam 1351 Table 2 provides the distribution of household monthly income among payday and vehicle title borrowers by their individual level of monthly income.1166 For instance, referring back to Table 1, 14 percent of payday loans in the loan data analyzed went to borrowers with individual incomes between $2,000 to $2,499 dollars per month (or $24,000 to $29,999 per year). As Table 2 shows, the median household income for a payday or vehicle title borrower with an individual monthly income in this range is $2,417 per month, with the mean household income slightly higher at $2,811 per month. 1166 Sarah Flood et al., “Integrated Public Use Microdata Series, Current Population Survey: Version 5.0” (dataset) (IPUMS CPS, Univ. of Minn., 2017). 1352 TABLE 2. I l' I Hun-?ght? 11? Mei:- m? I W52 415 2mg 2:11 2313 2?15 FDIC Natural ufUnhanl-zacl andUmiahankadesehul-?s dmim?mh? mm an? Drawn-Ii?elmin?ulastlimn?u. 1353 Table 3 shows the distribution of certain household expenditures by household monthly incomes. For instance, households with an income between $2,000 and $2,499 per month spend on average $756 on obligations which would fall within the category of major financial obligations, including rent or mortgage payments and vehicle loan payments. The same households spend an average of $763 on food, utilities, and transportation, which all are basic living expenses. As shown in the table, that leaves $689 to cover any other financial obligations, including payments on other forms of debt, other basic living expenses and payments on a new loan. 1354 TABLE 3- ?55 132 e: IUIIJ BLE- Em and 1355 Based on these data, it appears that payday borrowers need at least $1,500 in monthly household income to possibly have enough residual income to be able to repay a typical payday loan of $300-$400. However, this requires that the household have no other major financial obligations beyond housing and an auto loan, and does not factor into account all of the categories for basic living expenses defined in the rule. Table 4 provides more information about other typical major financial obligations of households that use payday loans. It shows the amount of outstanding debts and monthly payments for several categories of credit for households that used payday loans over a period of twelve months, as well as the share of those households that had each category of debt. This information comes from the 2010 Survey of Consumer Finances (SCF),1167 which has details on respondents’ assets, debts, and income, but the number of payday borrowers in the data is not large enough to allow estimating debts for borrowers in different income ranges.1168 1167 Relative to the proposal, the Bureau has continued to use the 2010 SCF data, as these better reflect contemporaneous debt obligations for borrowers observed in the baseline loan origination data. 1168 These estimates show a substantially lower share of borrowers with credit cards than was found in a study that matched payday loan data with credit report information. That study found that 59 percent of payday borrowers had an outstanding balance on at least one credit card, with an average outstanding balance of $2,900. 1356 TABLE 4- Fnrtin If . . g. . Barron: Datum?: Men 111 90* m. . In]. ?ulsl?uul?n? Ball-115 mm Cards $3,231 $210 51,3114: $1,131: 34% Clara: Ann-um? 53,151 31m Will Sim-D 5m 11mm; Imam: [liaising Farm-115? 5m $100 3:550 $1,390 lulu: oi (:er 519:5 520 ms $4435 45-6. [Car Lum? 5421 3204:: 33:54:: mu 35-2-1: tat-am Luau $114 151: ms mu 14% l?tlu Cum: Loans 52m 130 mm 55?: l??l-h Billing and Emil?Its an Credit Card and Charge Act-unis $3,551 $210 51,2114: $3.1m mm ?lm hunt-h: 591'? 3m $309 $1,710 939-1: hymn-ls Him [Honing and Car lain haunts ?'53 5?5? 339'? Salute: 21111] Federal Resume Board Burma].- nft'unimnar ijlmes year. aurora; Paganini: on lines ufuedit?n?m?ng Ila-m equity ?rm nfn?edi?. ?hdndespersnm?yunmdcars. mad-:5 1m vehicles- 1357 Table 4 shows that 34 percent of households with payday loans have outstanding credit card debt, with an average balance of nearly $3,300. An average credit card balance of approximately $3,300 requires a minimum monthly payment of around $100.1169 The table also shows that one-third of payday households have additional debts not associated with housing or vehicles, with average monthly payments of $263. Given these other major financial obligations, and the need to account for other basic living expenses, it seems likely that a household will need monthly income substantially higher than $1,500 to be able to demonstrate an ability to repay a typical payday loan. For example, households with at least $3,000 in monthly income seem to demonstrate an ability to repay a typical payday loan. Individuals in such households typically have roughly $2,500 in monthly income. And in the data the Bureau has analyzed, roughly onethird of payday borrowers have individual income above $2,500 per month. There is an additional caveat to this analysis: the CEX expenditure data are for all households in a given income range, not households of payday borrowers. If payday borrowers have unusually high expenses relative to their incomes, they will be less likely than the data suggest to be able to demonstrate an ability to repay a payday loan. Conversely, if payday borrowers have unusually low expenses relative to their incomes, they will be more likely to be able to borrower under the ATR approach. Given these borrowers’ needs for liquidity, it seems more likely that they have greater expenses relative to their income compared with households generally. This may be particularly true around the time that borrowers take out a payday loan, as this may be a time of unusually high expenses or low income. 1169 This assumes a 24 percent annual interest rate on the balance, with a minimum monthly payment calculated as all interest due plus one percent of the principal. 1358 As noted earlier, comments received in response to the proposal provided the Bureau with additional data that speak to payday borrowers’ residual incomes and the likely outcomes of an ability to repay assessment. The first of these data, shown in Table 5, were provided in a comment letter to the Bureau by an alternative credit bureau. Table 5 presents the percentages of current payday loans by the residual income level of the borrower. The residual incomes were calculated for a randomly sampled 1.65 million loan applicants in 2014. The calculation subtracted the following elements from a consumers’ stated monthly income: “Covered Loan” monthly debt obligation, traditional monthly debt obligation sourced from a national credit bureau, any applicable child or family support sourced from a national credit bureau, requested loan payment amount, monthly geo-aggregated estimate of housing costs (from Census data), and monthly estimate of utility and phone payments (from BLS data).1170 At least the basic living expenses comprised by this estimate of residual income are, as the commenter noted, incomplete, and thus the residual incomes in Table 5 are potentially higher than those that would result from an ability-to-repay assessment consistent with § 1041.5. 1170 See FactorTrust Inc. Comment Letter to the CFPB, dated October 6, 2016. The commenter did not provide more detail on the nature of the sample, which may include loans that are not covered under the rule (but were covered under the proposal). Also, the subtractions listed include “Covered Loan monthly debt obligation” and “requested loan payment amount.” The Bureau believes these refer to the same item. 1359 TABLE 5. Lammunuiul lh??h??unnt nsununm?qhu 11mm Pl'?llf LII-5 Amiinlh 511% 1.8% 1:415:11 7.9% 3.5% 415? 1: 410-1 3.2% 7.1% 410011 u??ul 14.156. 4501] to 1331-5. 111 II ligand:- 12.1% 12-1 aw". ?ll - 9.5% 9.1% nuln??m $1501? :25" 3.3% 7.2% 9.2% 9.3% Saute: Facm'l'mst Cambium ?ECFPEdatEdEktnh-H 6,2016, m. 19-21. 1360 As shown in Table 5, these data indicate that fewer than 50 percent of current payday loans are made to individuals with positive residual incomes, with slightly fewer first-time applicants having positive residual incomes (46.2 percent vs. 44.8 percent).1171 Setting aside the fact that as previously noted at least the subtractions for basic living expenses are incomplete, this still implies that the majority of payday loans would not pass an ability to repay determination. This finding is consistent with other studies that show that fewer than four in ten payday loan bookings passed a residual-income test.1172 Another report, submitted by the research arm of an alternative credit bureau, provided similar data.1173 Table 6 shows the percentage of storefront payday loan borrowers who would have had positive residual incomes after making a loan payment, and the percentage of all loans made to such borrowers. These percentages come from a sample of 90,000 storefront payday loans made in 2013, matched to debt obligations and two income measures (one each for the median observed income, and the most recently observed income). The residual-income measure subtracted from the borrower’s income debt service obligations and basic living expenses including shelter, food, transportation, communication, medical care, and dependent childcare (using BLS data to proxy where necessary). 1171 In this analysis, residual income refers to money left over after subtracting loan payments, financial obligations and some living expenses. Residual income in the rule is slightly different and refers to income minus major financial obligations and loan payments. Thus, whereas $0 residual income could indicate a borrower has ability to repay using the factor trust calculation, it would not under the rule’s calculation because funds would be needed to cover basic living expenses. 1172 FactorTrust “Underwriting Benchmarks: How Does Your Performance Stack Up?,” presentation to the 2017 CFSA Conference & Expo, at slide 20. 1173 nonPrime101, “Report 10: Is Consumer ‘Ability to Repay’ Predictive of Actual Repayment of Storefront Payday Loans?,” (2017), available at https://www.nonprime101.com/report-10-ability-to-pay/. 1361 TABLE I5- ?Bum-nu": Pen-at Inf LII-5- in Hus-1'! Plyan-Ilry WE?il'j' 2321-1?: 23.1111?: Median l'i'il'n': 13.711?: Imam} Mummy 14.111?: 13.911?: Total 33.1% 3&0? WE?il'j' 23-111?: 2511-1?: ?last Seen? Iii-Wad]? 15.611?: 13.611?: Imam 21.311?: 22.911?: Total 13.5% 31.1? Max:311: Hackelt. ?"3511an 10'. BCW?Ahilityta-Eepay?kadic?m Stumt'mm Payday :1 medium 1. 1362 The results in Table 6 show that between 28.5 and 33 percent of borrowers would have passed a residual-income test in these data. This appears somewhat lower than reported in Table 5, where 34 percent of borrowers had at least $500 in positive residual income (more than enough to cover the debt service on a payday loan). This difference could be due to the study’s sampling methodology, which may overstate loans in long sequences. Such loans may be suggestive of an inability to repay (see discussion of censoring above). As such, the Bureau considers these figures to be “conservative” (in that they may underestimate the share of borrowers who would pass an ATR assessment).1174 It is not known whether the applications that would fail to pass an ATR determination are more likely to be for one of a customer’s first six loans (which would not be subject to an ATR assessment if issued under the principal step-down approach). While first-time applicants do appear slightly more likely to have negative residual incomes, the residual income levels of applicants for a seventh (or greater) loan in a 12-month period may be higher or lower on average compared to the overall population of applications. As such, there is no strong evidence 1174 The differences also may reflect differences in the categories of expenses included as basic living expenses in the two analyses. The Bureau also received comments referencing other studies or analyses that provided less data, analytic rigor, or transparency; the Bureau placed less weight on the findings from such studies. For example, some commenters cited an analysis in an undated presentation by four industry representatives, including one of the specialty credit bureaus whose more detailed comment is noted above. This analysis claims that individuals earning less than $40,000 per year are “unlikely to qualify” for a $500 payday loan. However, this analysis is flawed. First, the presentation ignores the option for loans made under the principal step-down approach, which the Bureau expects to remain widely available. Second, the study’s sources for assumptions about the typical expenses faced by these households are not cited, and appear inflated relative to the levels shown in the available data (i.e., they assume $2,495 in “typical” monthly expenses, while Table 2 shows the median expense for individuals with this level of income is only $1,667). Third, this study applies a five percent “ATR buffer” that reduces the individual’s available income. This buffer was not part of the proposal (though it is similar to the considerations proposed by some consumer groups), and without this buffer, the individual in the presentation’s example actually would qualify for a $500 payday loan according to their calculations. In general, the Bureau considered carefully those analyses that provided or carefully cited reliable data, and discounted those that were less empirically grounded or had flaws similar to those noted here. 1363 that customers seeking their first loan under the ATR approach (which, as discussed previously, is likely to be their seventh loan in a 12-month period) would be more or less likely to pass an ATR assessment. As such, the evidence suggests that relatively few applicants would pass an ATR determination in order to continue borrowing beyond the limits imposed by the principal step-down approach. Based on these findings, the Bureau assumes for the purposes of its simulation that 33 percent of would-be borrowers can pass ATR. This number is near the lower end of the ranges identified by the Bureau’s analysis and in the first of the two comments described above and within the range of the second comment that independently attempted to measure the share of borrowers likely to pass an ATR assessment. While the 33 percent figure used here is a restrictive assumption (i.e., will result in a larger estimated decline in lending), the actual share of borrowers who will pass the ATR assessment in practice may differ from the value used here. To the extent that the value used in the Bureau’s simulations is too high (i.e., fewer borrowers would pass an ATR determination), the real decreases in loan volumes and revenues would be greater. To the extent that the value used in the Bureau’s simulations is, as suggested above, too low (i.e., more borrowers would pass an ATR determination), the real decreases in loan volumes and revenues would be smaller. However, given the magnitude of the decline in the ATR-only simulations, it appears that there is unlikely to be a substantial change to the estimates based on any reasonable assumption about the share of borrowers qualifying for ATR loans. The simulations of the re-borrowing restrictions and the ATR analysis presented thus far relate only to storefront loans. Online payday loans and vehicle title loans are considered next. iii. Online Payday Lending: Impacts 1364 The impact of the rule on the online payday market is more difficult to predict. There is no indication that online payday lenders will be more successful under the ATR approach than storefront lenders; in fact, it may be somewhat more difficult for them to satisfy the procedural requirements of that approach. The available information does not allow for reliably tracking sequences of online payday loans, as borrowers appear to change lenders much more often online and there is no comprehensive source of data on all online lenders. If very long sequences of loans are less common for online loans, however, the re-borrowing restrictions of both the ATR and principal step-down approaches will have a smaller impact on online lenders. There are additional relevant considerations for the impacts of the rule on online lenders relative to storefront lending. Unfortunately the direction and magnitudes of the impacts are not entirely clear. The decrease in online lending may be less relative to storefronts if the geographical contraction of storefronts leads more borrowers to seek loans from online lenders. Additionally, online lenders may have lower overhead costs and be able to better amortize onetime and per-location costs over broader potential borrowing populations. However, there could be negative selection into online lending (e.g., borrowers who are less likely to pass ATR assessments or are more likely to default) if storefront closings happen to displace less qualified customers. As such, the effects on online lenders are likely to be similar to those on storefront lenders, though the Bureau notes this actual impact on online lenders is much more difficult to predict. iv. Vehicle Title Lending: Impacts Vehicle title loans are not eligible for the principal step-down approach, and therefore lenders making only vehicle title loans will only be able to make such loans to borrowers who 1365 the lender is able to determine have the ability to repay the loan. Table 7 shows the distribution of individual incomes of single-payment vehicle title borrowers. 1366 TABLE [?sh-inlinn Inf-m nfSi'?E?Pi'flfl?nl Velide?lem Ininhalhlun?lhr Hmof Imam: Sal-5:99 1.9% 13.15 Slum-51:99 19.95: 515111-5139 Elli?- 3.9% ENE-SEES 15% 5.3% slum-$1999 41% SEE-SEES 191% Sam-56999 1.6% STEIN-WEI Sauna-$9999 [12% 5mm 1.2% Snunze: a na hm: of loan-level shale-payment vehicle title Iain dam- 1367 Table 7 shows that the incomes of vehicle title loan borrowers are slightly lower than those of payday loan borrowers. Vehicle title loans, however, are substantially larger than payday loans, with a median loan amount of nearly $700, twice that of payday loans.1175 Based on Tables 3 and 4, it appears that very few households with monthly income below $3,000 will be able to demonstrate an ability to repay a loan with a payment of $700, and even $3,000 will likely be insufficient. Based on the imputation of household numbers to individual borrowers, it appears that some individuals with monthly income between $1,500 and $2,000 will live in households with sufficient residual income to make a $700 payment, but that it is more likely that monthly individual income of $2,500 or more will be needed to have sufficient residual income to make such a payment. Table 7 shows that less than one third of vehicle title borrowers have monthly individual income above $2,500. Putting aside the difficulty of developing precise estimates of the share of borrowers who will be able to demonstrate an ability to repay a loan, it is likely that the share will be smaller for vehicle title borrowers than payday borrowers simply because vehicle title borrowers have slightly lower average incomes, and the average single-payment vehicle title loan is substantially larger than the average payday loan. However, the Bureau applied the same assumption as with payday loans about the share of borrowers who will pass an ATR assessment in the vehicle title simulations. Specifically, 33 percent of borrowers are assumed to pass the ATR screen. While it is likely that relatively fewer borrowers will pass an ATR determination for title loans, the 33 percent number was near the low end of the predicted ranges for borrowers passing ATR for payday. Additionally, the Bureau did not receive any comments with detailed analysis of the 1175 CFPB Payday Loans and Deposit Advance Products White Paper, at 15; CFPB Vehicle Title Report, at 6. 1368 share of borrowers likely to pass ATR for title loans. As such, while the Bureau has determined the 33 percent figure to be a reasonable assumption for the share of borrowers passing ATR assessments for both payday and title loans, it acknowledges that the figure is less precise for title loans. Vehicle title lenders also face the limitations of the ATR approach on making loans to borrowers during the term of, and for 30 days following, a prior covered short-term loan. The Bureau has run simulations of the share of single-payment vehicle title loans that are currently made that could still be made under the rule.1176 The simulations apply the same 33 percent ATR filter as was described for payday, and likewise assume that borrowers cannot take out a loan within 30 days of repaying a prior loan. Depending on whether borrowers who currently take out long sequences of loans will return to borrow again after a 30-day period following repayment of a loan, the Bureau estimates that the restrictions on short-term vehicle title lending will prevent between 89 and 93 percent of short-term vehicle title loans that are currently made, with an equivalent reduction in loan volume and revenue. Depending on the extent to which the underwriting restrictions on these lenders eliminate more loans (i.e., fewer than 33 percent of borrowers demonstrate ATR), the overall reduction in loans and revenue could be even greater. However, if more than 33 percent of borrowers can demonstrate ATR for each loan, the reduction in loans may be reduced. v. Overall Impacts on these Markets 1176 These are similar to the simulations described in CFPB Report on Supplemental Findings, at Chapter 6, though the results of the simulations presented there take account only of the re-borrowing restriction, while the results presented here add the underwriting filter. 1369 For the reasons discussed above, the Bureau believes that the rule will have a substantial impact on the markets for payday loans and single-payment vehicle title loans. The costs of the procedural requirements may have some impact on these markets, but the larger effects will come from the limitations on lending. Most of the costs associated with the procedural requirements of the rule are per-loan (or per-application) costs, what economists refer to as “marginal costs.” Standard economic theory predicts that marginal costs will be passed through to consumers, at least in part, in the form of higher prices. As discussed above in part II, however, many covered loans are being made at prices equal to caps that are set by State law or State regulation; lenders operating in States with binding price caps will not be able to recoup those costs through higher prices. The new procedural costs to lenders making loans using the principal step-down approach, however, will be quite small, primarily the costs of obtaining data from a registered information system and providing data to registered information systems. Lenders making vehicle title loans, which cannot be made under the principal step-down approach, will be required to incur the costs of using the ATR approach. If lenders make smaller loans to comply with the ATR requirements, however, the relative importance of procedural costs could increase. As described above, the limitations on lending included in the rule will have a substantial impact on the loan revenue of storefront payday and vehicle title lenders; the impact on online payday lenders is less clear, but is likely to be substantial as well. However, it is important to emphasize that these revenue projections do not account for lenders making changes to the terms of their loans to better fit the regulatory structure or offering other products, for instance by 1370 offering a longer-term vehicle title loan with a series of smaller periodic payments instead of offering a short-term vehicle title loan. The Bureau is not able to model these effects. A pattern of contractions in storefronts has played out in States that have imposed new laws or regulations that have had a similar impact on lending revenue, where revenue-per-store has generally remained fairly constant and the number of stores has declined in proportion to the decline in revenue.1177 To the extent that lenders cannot replace reductions in revenue by adapting their products and practices, Bureau research suggests that the ultimate net reduction in revenue will likely lead to contractions of storefronts of a similar magnitude, at least for stores that do not have substantial revenue from other lines of business, such as check cashing and selling money orders. With regard to evolution in product offerings, it is quite likely that lenders may respond to the requirements and restrictions in the rule by adjusting the costs and features of particular loans. They may also change the range of products that they offer. If lenders are able to make these changes, it will mitigate their revenue losses. On individual loans, a loan applicant may not demonstrate an ability to repay a loan of a certain size with a certain payment schedule. The lender may choose to offer the borrower a smaller loan or, if allowed in the State where the lender operates, a payment schedule with a comparable APR but a longer repayment period yielding smaller payments. Lenders may also make broader changes to the range of products 1177 CFPB Report on Supplemental Findings, at chapter 3; Wash. State Dep’t. of Fin. Insts., “2015 Payday Lending Report,”, at 5 (2015), available at http://www.dfi.wa.gov/sites/default/files/reports/2015-payday-lending-report.pdf; Adm’r of the Colo. Consumer Credit Unit, “Colorado Payday Lending—July Demographic and Statistical Information: July 2000 through December 2012,”; Adm’r of the Colo. Consumer Credit Unit, “Colorado Uniform Consumer Credit Code: Annual Report Composites,” available at https://coag.gov/uccc/info/ar. 1371 that they offer, shifting to longer-term, lower-payment installment loans, where these loans can be originated profitably within the limits permitted by State law.1178 Making changes to individual loans and to overall product offerings will impose costs on lenders even as it may serve to replace at least some lost revenues. Smaller individual loans generate less revenue for lenders. Shifting product offerings will likely have very little direct cost for lenders that already offer those products. These lenders will likely suffer some reduced profits, however, assuming that they found the previous mix of products to generate the greatest profits. Lenders who do not currently offer longer-term products but decide to expand their product range will incur a number of costs. These might include learning about or developing those products; developing the policies, procedures, and systems required to originate and to service the loans; training staff about the new products; and communicating the new product offerings to existing payday and single-payment vehicle title borrowers. 2. Benefits and Costs to Consumers a. Benefits to consumers The rule will benefit consumers by reducing the harm they suffer from the costs of extended sequences of payday loans and single-payment auto-title loans, from the costs of delinquency and default on these loans, from the costs of defaulting on other major financial obligations, and/or from being unable to cover basic living expenses in order to pay off covered 1178 An analysis by researchers affiliated with a specialty consumer reporting agency estimated that roughly half of storefront payday borrowers could demonstrate ability to repay a longer-term loan with similar size and APR to their payday loan, but noted that these loans would not be permitted in a number of States because of State lending laws and usury caps. nonPrime 101, “Report 8, Can Storefront Payday Borrowers Become Installment Loan Borrowers? Can Storefront Payday Lenders Become Installment Lenders?,” at 3 (Dec. 2, 2015) available at https://www.nonprime101.com/wp-content/uploads/2015/12/Report-8-Can-Storefront-Payday-Borrowers-BecomeInstallment-Loan-Borrowers-Web-61.pdf. 1372 short-term and longer-term balloon-payment loans.1179 Borrowers will also benefit from lenders adjusting their loan terms or their product mix, so that future loans are more predictable and ultimate repayment is more likely. i. Eliminating Extended Loan Sequences As discussed in detail above in Market Concerns—Underwriting, there is strong evidence that borrowers who take out storefront payday loans and single-payment vehicle title loans often end up taking out many loans in a row. This evidence comes from the Bureau’s own work, as well as analysis by independent researchers and analysts commissioned by industry. Each subsequent single-payment loan carries the same cost as the initial loan that the borrower took out, and there is evidence that many borrowers do not anticipate these long sequences of loans. Borrowers who do not intend or expect to have to roll over or re-borrow their loans, or expect only a short period of re-borrowing, incur borrowing costs that are several times higher than what they expected to pay. The limitations on making loans to borrowers who have recently had relevant covered loans will eliminate these long sequences of loans. The Bureau received many comments from industry, trade associations, and others arguing about consumers’ abilities to anticipate their borrowing patterns. The Bureau has addressed these comments previously in Market Concerns—Underwriting, the section-by-section analysis for § 1041.4 and part VII.D, and does so again here. There are several key findings that are raised by multiple sources, including analyses by the Bureau; by academic, industry, and other researchers; by State government agencies; in a 1179 As mentioned previously, the effects associated with longer-term balloon-payment loans are likely to be small relative to the effects associated with payday and vehicle title loans. This is because longer-term balloon-payment loans are uncommon in the baseline against which benefits are measured. 1373 report submitted by several of the SERs as part of the SBREFA process; and raised in comments. First, only a minority of new payday and single-payment vehicle title loans are repaid without reborrowing. With slight variation depending on the particular analysis, from approximately onein-three to one-in-four payday loans and approximately one-in-eight single-payment vehicle title loans is repaid without re-borrowing. In contrast, about half of loans lead to sequences at least four loans long, for both types of loans.1180 A significant percentage of borrowers have even longer sequences; about a third of either type of loan leads to sequences seven loans long, and about a quarter lead to sequences 10 loans long or longer. And, a small number of borrowers have extremely long sequences that go on for years. An analysis by an industry research group found that 30 percent of payday borrowers who took out a loan in a particular month also took out a loan in the same month four years later. For this group, the median time in debt over that period was over two years, and nine percent of the group had a loan in every pay period across the four years.1181 The Bureau believes the available empirical evidence demonstrates that borrowers who take out long sequences of payday loans and vehicle title loans do not anticipate those long sequences.1182 Aside from the Mann (2013) study, which is discussed further below, two 1180 See CFPB Data Point: Payday Lending, at 10–11; CFPB Vehicle Title Report, at 10–11; CFPB Report on Supplemental Findings, at Chapter 5; Arthur Baines et al., “Economic Impact on Small Lenders of the Payday Lending Rules Under Consideration by the CFPB,” Charles River Associates, (2015), available at http://www.crai.com/publication/economic-impact-small-lenders-payday-lending-rules-under-consideration-cfpb; Letter from Greg Gonzales, Comm’r,Tennessee Dep’t of Fin. Insts., to Hon. Bill Haslam, Governor and Hon. Members of the 109th General Assembly, at 8 (Apr. 12, 2016) (Report on the Title Pledge Industry), available at http://www.tennessee.gov/assets/entities/tdfi/attachments/Title_Pledge_Report_2016_Final_Draft_Apr_6_2016.pdf. 1181 nonPrime 101, “Report 7–C, A Balanced View of Storefront Payday Borrowing Patterns: Results from a Longitudinal Random Sample over 4.5 Years,” at tbl. A–7 (2016), available at https://www.nonprime101.com/datafindings/. 1182 The evidence described in this section is also discussed in Market Concerns—Underwriting. 1374 academic studies have asked payday and vehicle title borrowers about their expectations regarding how long it takes to repay payday loans, and not re-borrow shortly thereafter, and compared their responses with actual repayment behavior of the overall borrower population.1183 These studies did not compare borrowers’ predictions with their own borrowing experiences, but did show that borrowers appear, on average, somewhat optimistic about re-borrowing. Still, the average borrower experience may not be directly relevant to the impacts of this rule. Rather, as described in part VII.D, the more pertinent question in assessing the impacts of this rule’s restrictions is whether those borrowers who experience long sequences of re-borrowing properly anticipated these experiences. Two nearly identical surveys of payday borrowers commissioned by an industry trade group were conducted in 2013 and 2016, and asked borrowers who had recently repaid a loan and not re-borrowed if it had taken as long as the borrower had initially expected to repay the loan.1184 They found that the overwhelming majority of borrowers stated that it had not taken longer than they expected. This approach, however, may suffer from numerous problems, including recall bias (as borrowers were asked about what they expected in the past and whether their expectations were accurate) and “reverse” survivor bias (as only borrowers who successfully closed a sequence of loans are surveyed, and these borrowers are much less likely to have been in long borrowing sequences). It is also not clear from the wording of the survey if 1183 Kathryn Fritzdixon et al., “Dude, Where’s My Car Title?: The Law Behavior and Economics of Title Lending Markets,” 2014 U. IL L. Rev. 1013 (2014); Marianne Bertrand and Adair Morse, “Information Disclosure, Cognitive Biases and Payday Borrowing,” 66 Journal of Fin. 1865 (2011). 1184 Tarrance Group et al., “Borrower and Voter Views of Payday Loans,” Cmty. Fin. Servs. Ass’n of America (2016), available at http://www.tarrance.com/docs/CFSA-BorrowerandVoterSurvey-AnalysisF03.03.16.pdf; Harris Interactive, “Payday Loans and the Borrower Experience,” Cmty. Fin. Servs. Ass’n of America (2013), available at http://cfsaa.com/Portals/0/Harris_Interactive/CFSA_HarrisPoll_SurveyResults.pdf. 1375 borrowers are likely to have understood the question to refer to the actual loan they had recently repaid, or to the original loan they had taken out that led to the loan sequence. As discussed in the overview, Mann (2013) did ask borrowers about their expectations for re-borrowing and compared those with their actual borrowing experience, yielding insights more directly relevant for this rule.1185 As described in the proposal, the study found that borrowers who wound up with very long sequences of loans had rarely expected those long sequences; that only 40 percent of respondents expected to re-borrow at all (while more than 70 percent actually did re-borrow); and, that borrowers did not appear to become better at predicting their own borrowing, as those who had borrowed most heavily in the past were most likely to underestimate their future re-borrowing. This study was one of the most heavily cited by commenters, and the author himself provided a comment as well. Industry commenters and the author offered criticisms of the Bureau’s characterization of the study’s findings. However, the Bureau continues to believe the evidence suggests many borrowers did not anticipate their outcomes. Given the prevalence and intensity with which commenters cite this study, the Bureau offers a more detailed response here. Mann (2013) presents evidence that 51 percent of borrowers predict their outcomes within 7 days, 57 percent within 14 days, and 63 percent within 21 days,1186 and that borrower’s 1185 Ronald Mann, “Assessing the Optimism of Payday Loan Borrowers,” 21 Supreme Court Econ. Rev. 105 (2013), and correspondence between Prof. Mann and Bureau staff described in Market Concerns—Underwriting. 1186 Note that in performing these calculations, the paper ignores the 20 percent of respondents who did not respond to the questions (potentially because they were unable to offer a prediction of their time in debt). In terms of the share of all surveyed borrowers successfully predicting within a given window, these percentages in the paper translate to 41 percent within seven days, 46 percent within 14 days, and 51 percent within 21 days. 1376 errors were fairly symmetric around zero1187 (i.e., there was not evidence of systematic optimism or pessimism).1188 The Bureau appreciates Mann’s evidence and places significant weight on his findings, but does dispute his interpretation of those findings. The pertinent question for this rule, which limits long durations (but not discrete and short-term access), is: do the specific borrowers who will experience very long sequences anticipate these outcomes at the time they borrow? The answer to this question appears to be no. Mann did not include his data with his comment, which makes deeper exploration of his findings difficult.1189 However, using the paper and documents provided by the author to the Bureau, some useful findings can be discerned.1190 These include, inter alia: among borrowers taking 150+ days to clear a sequence, none (0 percent) predicted they would be in debt for even 100, and the average borrower spent 121 unanticipated days in debt (equivalent to more than 8.5 rollovers); among borrowers taking 90 or more days to clear their loans at least 95 percent 1187 Note that the paper does not offer the mean error, stating only that it is “close to zero.” It does divulge that the median error is three days, which is 10 percent of the predicted loan duration and over 20 percent of the initial loan term. This implies that even “average” borrowers may not be as precise in their predictions as the author implies. 1188 The Bureau notes this second point, but further notes that consumers who underestimate their ability to repay do not achieve additional benefit from the payday loan borrowing experience, though they do achieve better-thanexpected outcomes. Consumers who overestimate their ability to repay may suffer considerably over a long period of subsequent indebtedness. This asymmetry is what is addressed by the proposed rule, not the asymmetry in expected durations. 1189 As stated above in part VII.D, it should be noted that Professor Mann did not provide his data to the Bureau, either prior to the proposal, nor in his comment in response to the proposal. In place of these data, the Bureau is relying on the charts and graphs he provided in his correspondence with and presentation to the Bureau. As such the analysis provided here may be somewhat imprecise. 1190 Many of these findings were derived by analyzing the scatterplots depicting borrowers’ re-borrowing expectations and outcomes, provided in Attachment to Email from Ronald Mann, Professor, Columbia Law School, to Jialan Wang & Jesse Leary, Bureau of Consumer Fin. Prot. (Sept. 24, 2013, 1:32 EDT). The Bureau measured the distances of each discernable point on the plot to assess its coordinates. Because of the presence of line of best fit on the figure, some points near 28 days of expected indebtedness are obscured. This should not substantially impact the findings presented here, and would only serve to bias the results away from finding that borrowers with long sequences underestimate their durations of indebtedness. As previously noted, borrowers with exceptionally long sequences (including those displayed in subsequent slides of the author’s presentation) appear to be missing from this scatterplot. 1377 believed they would be in debt for shorter durations than they actually experienced, with the average borrower spending 92 unanticipated days in debt (equivalent to more than 6.5 rollovers); and among those borrowers taking 42 or more days to clear their loans (equivalent to the three loan sequence permitted under the rule) more than 90 percent underestimated their time in debt, with the average borrower experiencing 48 unanticipated days in debt (equivalent to more than three rollovers).1191 Additionally, a graph depicting the relationship between predicted and actual days in debt shows a regression line with no discernable slope. The Bureau believes this to be the clearest statistical evidence that there is no significant relationship between predicted and actual days in debt. If borrowers could have predicted precisely what would happen to them, the slope of the line would be equal to one. If borrowers’ predictions were generally (and positively) correlated with their actual outcomes, the slope of the line would be positive and non-trivial. If borrowers’ predictions were completely uncorrelated with their outcomes, the slope of the line would be zero. In the correspondence provided by the author, the slope of the line appears to be almost completely flat, and statistically indistinguishable from zero.1192 In other words: borrowers predictions had no discernable correlation with their outcomes, regardless of whether they experienced long periods of indebtedness. This finding of no discernable correlation between predictions and outcomes may seem inconsistent with the finding that many borrowers did accurately predict their durations within a 1191 Attachment to Email from Ronald Mann, Professor, Columbia Law School, to Jialan Wang & Jesse Leary, Bureau of Consumer Fin. Prot. (Sept. 24, 2013, 1:32 EDT). 1192 The Bureau estimates the actual slope of the line to be approximately 0.011, based on the Stata-generated graph provided to the Bureau by the author. See Attachment to E-mail from Ronald Mann, Professor, Columbia Law School, to Jialan Wang & Jesse Leary, Bureau of Consumer Fin. Prot. (Sept. 24, 2013, 1:32 EDT). And, again, the relationship is statistically insignificant. 1378 14-day window. Since so many borrowers expect short durations, and many borrowers experience these durations, it appears that they accurately predict their outcomes when, in fact, they are just as likely to have experienced longer durations. For example, in the Bureau’s data on payday loans, if all consumers predicted they would have no renewals, their actual sequence length would be within 14 days of the prediction 44 percent of the time. This is very similar to the 46 percent of borrowers in Mann’s data that are accurate in their predictions to within a 14day window (once those borrowers not reporting a prediction are included). Lastly, the paper itself presents direct evidence that a substantial minority of borrowers are unable to even offer a prediction of their outcomes. For example, approximately 20 percent of borrowers were unable to answer the question “…How long do you think it will be before you have saved enough money to go an entire pay period without borrowing from this lender? If you aren’t sure, please give your best estimate.”1193 In response to other questions in the survey, amongst borrowers who indicated they expected to roll the loan over, more than one-third did not (or could not) offer a prediction of how long they would continue borrowing.1194 Accounting for these non-responses means that the 57 percent of borrowers who Mann asserts predict their durations within a 14-day window actually represent less than half (46 percent) of all surveyed borrowers. Put another way, the paper’s findings are potentially instructive only for those borrowers who have enough confidence to make a prediction, and say little about the substantial 1193 Ronald Mann “Assessing the Optimism of Payday Loan Borrowers,” 21 Supreme Court Econ. Rev. 105, 121 (2013). 1194 Ronald Mann “Assessing the Optimism of Payday Loan Borrowers,” 21 Supreme Court Econ. Rev. 105, 121 (2013). 1379 fraction of borrowers who implicitly suggest or explicitly state they cannot predict their expected duration of indebtedness. In summary, the Bureau believes there are multiple implications of Mann’s findings. Specifically, it may be true that many borrowers accurately anticipate their debt durations, as Mann asserts in both his paper and comment. However, it is certainly true that most of those borrowers with long duration sequences did not accurately anticipate this outcome. Additionally, a large share of borrowers who anticipated no re-borrowing remain in debt for multiple loans, and many are unable to even offer a guess as to the duration of their indebtedness, let alone a precise prediction. Finally, there appears to be no discernable relationship between borrowers’ individual expectations, and their ultimate outcomes. Given the tenor of the comments received by the Bureau, the Bureau feels compelled to note that this rule does not ban payday or vehicle title lending. In fact, the Bureau expects the vast majority of borrowers to be permitted three-loan sequences under the principal step-down approach. It warrants mentioning that Mann (2013) shows that borrowers expect to be in debt an average of 36 days, and that more than 80 percent of borrowers expect clearance in 50 days or less, both of which fall within the approximate amount of time of indebtedness permitted under each sequence of loans under the rule.1195 As such, the evidence from Mann (2013) implies that the rule would not place a binding limit on the anticipated re-borrowing for the vast majority of his sample. 1195 See Ronald Mann, “Assessing the Optimism of Payday Loan Borrowers,” 21 Supreme Court Econ. Rev. 105, at at tbl. 3ii (2013). Note that a sequence of three biweekly loans covers approximately 42 days, which appears to be assigned to the same category as 50 days in the paper’s histogram. 1380 As mentioned, the Bureau received many comments suggesting that the cumulative available evidence shows borrowers anticipate their payday borrowing experiences. The Bureau believes the more thorough treatment of this literature offered here provides much in the way of support for the premise that those payday loan borrowers who experience long durations of debt failed to anticipate that this would occur. As such, the Bureau continues to believe the evidence strongly suggests there is a significant minority of borrowers who experience long durations of indebtedness that did not anticipate these outcomes, let alone the costly impacts thereof. It is less clear how large the benefits from the limitations on repeat borrowing will be for borrowers who take out online payday loans. As described above, available information does not allow for reliably tracking sequences of online payday loans, as borrowers appear to change lenders much more often online and there is no comprehensive source of data on all online lenders. If very long sequences of loans are less common for online loans, the costs of those sequences will be less and the benefits to consumers of preventing long sequences will be smaller. ii. Reduced Defaults and Delinquencies The Bureau estimates that borrowers taking out covered short-term and longer-term balloon-payment loans will experience substantially fewer defaults under the rule. As discussed in Market Concerns—Underwriting, the Bureau believes the consequences of defaults are harmful to consumers, and therefore reducing defaults provides a benefit to consumers. Consumers who default can become subject to harmful debt collection efforts. While delinquent, they may also seek to avoid default in ways that lead to a loss of control over budgeting for their other needs and expenses. In addition, 20 percent of single-payment vehicle title loan sequences 1381 end with borrowers losing their cars or trucks to repossession. Even borrowers who have not yet defaulted may incur penalty fees, late fees, or overdraft fees along the way and may find themselves struggling to pay other bills or meet their basic living expenses. There are at least three reasons generally to expect fewer defaults under the rule. First, borrowers who take out loans from lenders that use the ATR approach will go through a meaningful evaluation of their ability to make the payment or payments on the loan. The borrowers whom lenders determine have sufficient residual income or a low enough DTI ratio to cover each loan payment, make payments for major financial obligations, and meet basic living expenses over the term of the loan, and 30 days thereafter, will likely be better able to pay off their loans relative to the population of borrowers who currently take out these loans. Second, the reducing balances on loans made pursuant to the principal step-down approach should limit payment shocks to consumers. This step-down approach should lower the risk to lenders and borrowers of borrowers defaulting when a lender is unable to continue to lend to them (though some borrowers who would have re-borrowed the full amount of the initial loan may now default, if they are unable to successfully make the step-down payment). Third, lenders’ ability to make long sequences of loans to borrowers will be greatly curtailed, whether lenders use the ATR or principal step-down approach. Currently, borrowers who have difficulty repaying a loan in full usually have the option of paying just the finance charge and rolling the loan over, or repaying the loan and then quickly re-borrowing. The option to re-borrow may make borrowers willing to make a finance charge payment on a loan they know they cannot afford while still meeting their other obligations or expenditure needs. The option for continued re-borrowing allows borrowers to put off defaulting in the hopes they may 1382 ultimately be able to successfully repay the loan. If continued re-borrowing does not allow them to ultimately repay the loan, the lender will still have received multiple finance charges before the borrower defaults. To this point, Bureau research shows that nearly half of the consumers who experienced a default or a 30-day delinquency had fees over $60 in the month before their first default or 30-day delinquency.1196 Borrowers who are more likely to default are also more likely to have late payments; thus, reducing the rate of defaults will likely reduce the rate of late payments and the harm associated with those late payments. Late payments on payday loans, defined as a payment that is sufficiently late that the lender deposits the borrower’s check or attempts to collect using the ACH authorization, appear to range from seven1197 to over 10 percent.1198 At the borrower level, two different sources show that 39 to 50 percent of borrowers have a check deposited that bounces in their first year of payday borrowing.1199 These late payments are costly for borrowers. If a lender deposits a check or submits a payment request and it is returned for insufficient funds, the borrower’s bank or credit union will likely charge the borrower an NSF fee of approximately $35, and the lender may charge a returned-item fee. In addition, analysis the Bureau has conducted of payment requests from online lenders shows that a substantial 1196 Calculations using the Bureau’s payday loan dataset described above. “For the years ended December 31, 2011 and 2010, we deposited customer checks or presented an Automated Clearing House ("ACH") authorization for approximately 6.7 percent and 6.5 percent, respectively, of all the customer checks and ACHs we received and we were unable to collect approximately 63 percent and 64 percent, respectively, of these deposited customer checks or presented ACHs. Total charge-offs, net of recoveries, for the years ended December 31, 2011 and 2010 were approximately $106.8 million and $108 million, respectively.” Advance America, 2011 Annual Report (Form 10–K). 1198 Paige Marta Skiba and Jeremy Tobacman, “Payday Loans, Uncertainty, and Discounting: Explaining Patterns of Borrowing, Repayment, and Default,” (Vand. L. and Econ., Research Paper No. 08–33, 2008). 1199 Paige Marta Skiba and Jeremy Tobacman, “Payday Loans, Uncertainty, and Discounting: Explaining Patterns of Borrowing, Repayment, and Default,” (Vand. L. and Econ., Research Paper No. 08–33, 2008); Susanna Montezernollo and Sarah Wollf, “Payday Mayday: Visible and Invisible Payday Lending Defaults,” at 5 (Ctr for Responsible Lending 2015). 1197 1383 number of payments that are made are overdrafts.1200 Fees for overdrafts are generally equal to NSF fees at the same institution. Consumers will also benefit from mitigation of the harm from NSF and overdraft transactions by the limitations on payment practices and related notices described in the section-by-section analysis of §§ 1041.8 and 1041.9, and discussed later in this section. Default rates on individual payday loans are fairly low, 2 percent in the data the Bureau has analyzed.1201 But, as noted above, a substantial majority of borrowers takes out more than one loan in sequence before repaying the debt or defaulting. A more meaningful measure of default is therefore the share of loan sequences that end in default. The Bureau’s data show that, using a 30-day sequence definition, 20 percent of loan sequences end in default. Other researchers have found similar high levels of default at the borrower level. A study of payday borrowers in Texas found that 4.7 percent of loans were charged off but 30 percent of borrowers had a loan charged off in their first year of borrowing.1202 Less information is available on the delinquency and default rates for online payday loans. The available information is discussed in part II above, where the Bureau notes that one lender reports online single-payment loans have a charge-off rate substantially higher than that for storefront payday loans. In a 2014 analysis of its consumer account data, a major depository institution found that small-dollar lenders, which include lenders making a range of products 1200 The Bureau’s analysis shows that 6 percent of payment requests that were not preceded by a payment request that was returned for insufficient funds are returned for insufficient funds and 6 percent are paid as overdrafts. CFPB Online Payday Loan Payments. 1201 Default here is defined as a loan not being repaid as of the end of the period covered by the data or 30 days after the maturity date of the loan, whichever was later. 1202 Paige Marta Skiba and Jeremy Tobacman, “Payday Loans, Uncertainty, and Discounting: Explaining Patterns of Borrowing, Repayment, and Default,” at tbl. 2 (Vand. L. and Econ., Research Paper No. 08–33, 2008). 1384 including payday loans, had an overall return rate of 25 percent for ACH payments. The Bureau’s report on online payday loan payments practices presents rates of failed payments for online lenders exclusively.1203 It shows a lower rate of payment failure; six percent of payment attempts that were not preceded by a failed payment attempt themselves failed.1204 Default rates are more difficult to determine, but 42 percent of checking accounts with failed online loan payments are subsequently closed.1205 This provides a rough measure of default on these loans. Default rates on single-payment vehicle title loans are higher than those on payday loans. In the data analyzed by the Bureau, the default rate on all loans is nine percent, and the sequencelevel default rate is 31 percent.1206 In the data the Bureau has analyzed, five percent of all singlepayment vehicle title loans lead to repossession, and 18 percent of sequences of loans end with repossession. So, at the loan level and at the sequence level, slightly more than half of all defaults lead to repossession of the borrower’s vehicle. The range of potential impacts on a borrower of losing a vehicle to repossession depends on the transportation needs of the borrower’s household and the available transportation alternatives. According to two surveys of vehicle title loan borrowers, 15 percent of all borrowers report that they would have no way to get to work or school if they lost their vehicle 1203 CFPB Online Payday Loan Payments. CFPB Online Payday Loan Payments, at 13 tbl. 1. This analysis includes both online and storefront lenders. Storefront lenders normally collect payment in cash and only deposit checks or submit ACH requests for payment when a borrower has failed to pay in person. These check presentments and ACH payment requests, where the borrower has already failed to make the agreed-upon payment, have a higher rate of insufficient funds. 1205 CFPB Online Payday Loan Payments, at 24 tbl. 5. 1206 There is also evidence that the default rates on longer-term balloon-payment title loans are high. The Bureau has data for a single lender that made longer-term vehicle title loans with both balloon and amortizing payment schedules. Those loans with balloon payments defaulted at a substantially higher rate. See CFPB Report on Supplemental Findings, at 30. 1204 1385 to repossession.1207 Fully 35 percent of borrowers pledge the title to the only working vehicle in the household.1208 Even those with a second vehicle or the ability to get rides from friends or take public transportation would presumably experience significant inconvenience or even hardship from the loss of a vehicle. iii. Avoiding Harms from Making Unaffordable Payments Consumers will also benefit from a reduction in the other financial hardships that may arise because borrowers, having taken out a loan with unaffordable payments, feel compelled to take painful measures to avoid defaulting on the covered short-term and longer-term balloonpayment loans. If a lender has taken a security interest in the borrower’s vehicle, the borrower may decide not to pay other bills or forgo crucial expenditures because of the leverage that the threat of repossession gives to the lender. The repayment mechanisms for some covered shortterm loans and longer-term loans with balloon payments can also cause borrowers to lose control over their own finances. If a lender has the ability to withdraw payment directly from a borrower’s checking account, especially when the lender is able to time the withdrawal to the borrower’s payday, the borrower may lose control over the order in which payments are made and may be unable to choose to make essential expenditures before repaying the loan. iv. Changes to Loan Structure 1207 Kathryn Fritzdixon et al., “Dude, Where’s My Car Title?: The Law Behavior and Economics of Title Lending Markets,” 2014 U. IL L. Rev. 1013 (2014); Pew Charitable Trusts, “Auto Title Loans, Market Practices and Borrower Experiences,” at 1038 (2015), available at http://www.pewtrusts.org/~/media/assets/2015/03/autotitleloansreport.pdf. 1208 Kathryn Fritzdixon et al., “Dude, Where’s My Car Title?: The Law Behavior and Economics of Title Lending Markets,” 2014 U. IL L. Rev. 1013 (2014); Pew Charitable Trusts, “Auto Title Loans, Market Practices and Borrower Experiences,” at 1038 (2015), available at http://www.pewtrusts.org/~/media/assets/2015/03/autotitleloansreport.pdf. 1386 Consumers may benefit if lenders respond to the rule by modifying the terms of individual loans or if lenders adjust the range of products they offer. Borrowers offered smaller loans may benefit if this enables them to repay the loan, when they would otherwise be unable to repay. This will mitigate a borrower’s exposure to the costs associated with re-borrowing, default, or the costs of being unable to pay for other financial obligations or living expenses. If lenders shift from payday loans or single-payment vehicle title loans to longer-term loans, consumers may benefit from lower payments that make it more feasible for the borrowers to repay. Given the high rate of unanticipated re-borrowing of short-term loans, the financing costs of longer-term loans, provided they disclose their terms clearly and do not utilize balloon or leveraged payments, may be easier for borrowers to predict, and therefore borrowers may be less likely to end up in a loan that is substantially more expensive than they anticipated. b. Costs to Consumers and Access to Credit The procedural requirements of the rule will make the process of obtaining a loan more time consuming and complex for some borrowers. The restrictions on lending included in the rule will reduce the availability of storefront payday loans, online payday loans, single-payment vehicle title loans, longer-term balloon-payment loans, and other loans covered by the rule. Borrowers may experience reduced access to new loans (i.e., loans that are not part of an existing loan sequence). Some borrowers will also be prevented from rolling loans over or re-borrowing shortly after repaying a prior loan. And, some borrowers may still be able to borrow, but for smaller amounts or with different loan structures, and find this less preferable than the terms they would have received absent the rule. 1387 The Bureau received many comments suggesting that the consideration of costs to consumers was incomplete. Notably, comments suggested that the speed of obtaining funds would be reduced, leading to consumer harm; that the welfare implications of reducing the access to covered loans needed to be more adequately considered; that the Bureau should more explicitly consider the costs of moving to “inferior” alternatives due to the reduction in covered loans; and that the Bureau declined to provide monetary estimates of harm. The Bureau attempts to address each of these (as well as additional comments) in the subsections below. However, one general response is that the estimated restriction on consumer access to credit is not as severe as implied by these comments. The rule does not impose a ban on payday lending, and the Bureau expects the vast majority of consumers will experience minimal, if any, reduction in access to credit. The Bureau’s simulations (discussed above) show that the restrictions on re-borrowing and underwriting imply that only 5.9 to 6.2 percent of borrowers will be prohibited from initiating a sequence of loans they would have initiated absent the rule.1209 That is, since most consumers take out six or fewer loans each year, and are not engaged in long sequences of borrowing, most will not find their preferred borrowing patterns interrupted by the rule’s requirements and prohibitions. As will be discussed below, if borrowers derive greater benefits from their initial loans compared to subsequent loans, the impacts of these restrictions will have limited (and potentially positive) impacts on consumer welfare. i. Impacts of Procedural Requirements 1209 As previously mentioned, the Bureau does not attempt to predict the impact of any voluntary underwriting activities that would be undertaken by lenders providing loans under the principal step-down approach (e.g., to screen out likely defaulters who would have been profitable under a regime with unlimited rollovers). Any reduction in lending that might result from such a strategic response to this rule would further reduce the provision of credit compared to the estimates provided here. 1388 The procedural requirements for lenders will make the process of obtaining a loan more time consuming for some borrowers. This will depend on whether lenders use the ATR approach or the principal step-down approach, and the extent to which lenders automate their lending processes. In particular, borrowers taking out payday loans originated under the principal step-down approach from lenders that automate the process of checking their records and obtaining a report from a registered information system will see little, if any, increase in the time to obtain a loan. Notably, this should not substantially reduce the speed at which customers can take out a first loan (or a first loan after 30 or more days without a covered short-term loan or longer-term loan with a balloon payment). As such, those consumers who experience discrete, unanticipated, and infrequent shocks are unlikely to be negatively impacted by the rule’s procedural requirements.1210 Borrowers taking out loans from lenders using the ATR approach are more likely to experience additional complexity. Online payday borrowers and vehicle title borrowers are required to provide documentation of the amount of their income, which currently is often not required. Both storefront and online borrowers will be asked to fill out a form listing the amount of their income and payments on major financial obligations. Even when additional documentation is not required and a customer statement of income or expenses is sufficient, the process by which a lender may obtain these values is likely to take additional time, and lead to additional scrutiny, than was the case prior to the rule. As such, customers seeking loans under 1210 Some commenters suggested that the procedural requirements would reduce both speed and access to short-term credit, leading to consumer harm for those consumers who experience unanticipated shocks to their finances (e.g., car repair or hospital bill). As the principal step-down approach is likely to be the primary means through which customers get infrequent loans to deal with shocks of this nature, the procedural requirements are unlikely to bind on customers dealing with these events. 1389 the ATR approach will likely experience reductions in the speed they receive funds and/or access to credit. While the Bureau expects many lenders to automate much of the ATR determination, there may still be lenders that rely, partially or completely, on manual underwriting processes.1211 Estimates of the time required to manually process an application for a loan made via the ATR approach vary substantially. In the proposal, the Bureau assumed manual calculations of ATR would take less than 20 minutes. A large lender noted in its comment that manually processing applications in the U.K. takes one to four hours, and a trade group representing mostly large depository institutions suggested that three hours was a viable estimate. Comments received from a trade group representing covered title lenders and based on information provided by Small Entity Representatives shows that the increased time to process a manual ATR determination is 15-45 minutes. The last of these seems to be based on the most applicable information (e.g., covered lenders in the U.S.), and thus informs the Bureau’s estimates. Thus, if a lender orders consumer reports manually and performs the calculations by hand necessary to determine that the borrower has the ability to repay the loan, the Bureau estimates this could add 15-45 minutes to the borrowing process. And if a borrower is unaware that it is necessary to provide certain documentation required by the lender, this may require a second trip to the lender, increasing the costs borne by the borrower. Finally, borrowers taking out loans online 1211 It is likely that those stores able to determine ATR more rapidly and at a lower cost (e.g., via an automated process) will have a competitive advantage. Given the reduction in stores anticipated in this section, in steady-state the Bureau has concluded that relatively few lenders will employ a manual process, and those that do will be the ones who are able to streamline their assessments. 1390 may need to upload verification evidence, such as by taking a photograph of a pay stub, or facilitate lender access to other information sources. ii. Reduced Access to Initial Loans Initial covered short-term loans—i.e., those taken out by borrowers who have not recently had a covered short-term loan—are presumably taken out because of a need for credit that is not the result of prior borrowing of covered short-term loans. Borrowers may be unable to take out new loans (those originated more than 30 days after their last loan) for at least two reasons: they may only have access to loans made under the ATR approach and be unable to demonstrate an ability to repay the loan under the rule, or they may be unable to satisfy any underwriting requirements adopted by lenders. Payday borrowers are not likely to be required to satisfy the ATR requirement unless and until they have exhausted the limits on loans available to them under the principal step-down approach, or unless the borrower is seeking a loan in excess of $500. However, to obtain loans under the principal step-down approach, borrowers may be required to satisfy more exacting underwriting requirements than are applied today. Moreover, after exhausting the limits on principal step-down approach loans, borrowers are required to satisfy the ATR requirement in order to obtain a new loan. The direct effects of the principal step-down approach on borrowers’ ability to take out loans will be quite limited, provided the borrowers did not have an active loan within the past 30 days. The Bureau estimates that only about five percent of initial payday loans (those that are not part of an existing sequence) will be prevented by the annual limits, and roughly six percent of borrowers will be prohibited from initiating a new sequence of loans they would have started 1391 absent the rule. That is, only about five percent of the loans that are most likely to reflect a new need for credit will be affected by these annual limits on borrowing. These affected borrowers will then have to satisfy the ATR test in order to obtain a new loan. Vehicle title borrowers are more likely to find that they are unable to obtain an initial loan because the principal step-down approach does not provide for vehicle title loans and thus these borrowers must satisfy the ATR requirement. Many of these consumers could choose to pursue a payday loan instead and seek to avail themselves of the principal step-down approach. However, there are two States that permit vehicle title loans but not payday loans, and 15 percent of vehicle title borrowers do not have a checking account, and thus may not be eligible for a payday loan under the lender’s own rules (as borrowers without a checking account are allowed to obtain a loan under this rule).1212 In addition, many States limit the size of payday loans but not the size of vehicle title loans, so some borrowers may prefer a vehicle title loan. For all of these borrowers, their ability to obtain an initial loan will depend upon their ability to demonstrate an ability to repay and satisfy any other underwriting requirements the lender may impose. Consumers who are unable to obtain a new loan because they cannot satisfy the ATR requirement and have exhausted or cannot qualify for a loan under the principal step-down approach will bear some costs from reduced access to credit. They may be forced to forgo certain purchases,1213 incur high costs from delayed payment of existing obligations, incur high 1212 The 2015 FDIC National Survey of Unbanked and Underbanked Households finds that 12.4 percent of consumers reporting having used an auto title loan in the prior 12 months are unbanked. 1213 Specifically, consumers may react to reduced access to short-term loans by decreasing their short-run consumption. However, to the extent they avoid long sequences of loans, and the fees associated with them, their 1392 costs and other negative impacts by simply defaulting on bills, or they may choose to borrow from sources that are more expensive or otherwise less desirable. Some borrowers may overdraft their checking account; depending on the amount borrowed, an overdraft on a checking account may be more expensive than taking out a payday or single-payment vehicle title loan. Similarly, “borrowing” by paying a bill late may lead to late fees or other negative consequences like the loss of utility service. Other consumers may turn to friends or family when they would rather borrow from a lender. Still others may seek other types of credit, like longer-term loans not covered by the origination portions of this rule, credit cards, or other alternatives. And, some consumers may take out online loans from lenders that do not comply with this regulation.1214 Survey evidence provides some information about what borrowers are likely to do if they do not have access to these loans. Using the data from the CPS Unbanked/Underbanked supplement, researchers found that the share of households using pawn loans increased in States that banned payday loans, to a level that suggested a large share of households that would otherwise have taken out payday loans took out pawn loans, instead.1215 A 2012 survey of payday loan borrowers found that a majority indicated that if payday loans were unavailable they would reduce expenses, delay bill payment, borrow from family or friends, and pawn personal items. Some did indicate, however, that they would get a bank or credit union loan or use a longer-term consumption may increase. One study of consumption responses to payday loan access shows that overall consumption increases as payday loan use declines. See Brian Baugh, “What Happens When Payday Borrowers Are Cut Off from Payday Lending? A Natural Experiment,” Fisher College of Bus., Ohio State U. 2015). 1214 It has been suggested that some borrowers might turn to in-person illegal lenders, or “loan sharks.” The Bureau is unaware of any data on the current prevalence of illegal lending in the United States by individuals. Nor is the Bureau aware of any data suggesting that such illegal lending is more prevalent in States in which payday lending is not permitted than in States which permit payday lending or of any evidence that the amount of such lending has increased in States which adopted a prohibition on payday lending. 1215 Neil Bhutta et al., “Consumer Borrowing after Payday Loan Bans.” 59 J. of L. and Econ. 225 (2016). 1393 credit card to cover expenses.1216 Finally, data collected by the Bureau from banks that ceased offering deposit advance products (“DAP loans”), showed that there was no evidence that reduced access to these products led to greater rates of overdraft or account closure.1217 In many comments received by the Bureau it was suggested that more consideration be given to the alternatives that displaced borrowers may turn to absent available payday or title loans. Overdraft fees, “illegal loan sharks,” and pawn loans were specifically mentioned as inferior forms of credit that borrowers denied a payday or title loan may utilize. The Bureau agrees that these are indeed valid potential costs, and considered them in the proposal. The Bureau notes that its summary and analysis of the related literature and empirical evidence suggests that intensive payday borrowers experienced increase welfare from reduced use of these loans. This outcome reflects the net effects of any substitution patterns or reductions in borrowing. iii. Limits on Loan Size Lenders making loans using the principal step-down approach could not make loans larger than $500. This will limit the availability of credit to borrowers who would otherwise seek a larger loan, and either do not have access to loans under the ATR approach or cannot demonstrate their ability to repay the larger loan. In the data analyzed by the Bureau, however, the median payday loan is only $350, and some States impose a $500 maximum loan size, so 1216 Pew Charitable Trusts, “Payday Lending in America: Who Borrows, Where They Borrow, and Why,” at 16 (Report 1, 2012), available at http://www.pewtrusts.org/~/media/legacy/uploadedfiles/pcs_assets/2012/pewpaydaylendingreportpdf.pdf (reporting $375 as the average). 1217 CFPB Report on Supplemental Findings, at 35-39. The Bureau notes, however, that if demand for short-term liquidity is inelastic and outside options are limited, a decrease in access to one option will necessarily increase the demand for its substitutes. 1394 most existing payday loans would fall at or below the $500 maximum.1218 Any borrowers that would have preferred a vehicle title loan but instead obtain a payday loan originated under the principal step-down approach because of the rule may be more affected by the loan size limit, as the median single-payment vehicle title loans is for nearly $700.1219 There are additional restrictions on loan sizes made via the principal step-down approach that apply to the second and third loans in a sequence. That is, each subsequent loan in a sequence made using the principal step-down approach must decrease by at least one-third the amount of the original loan. For example, a $450 initial loan would mean borrowers are restricted to no more than $300 for a second loan, and no more than $150 for a third loan. In the Bureau’s preferred simulation, described in part VII.F.1.c, around 40 percent of the reduction in loan revenues were the result of $500 cap on initial loans and the principal stepdown, with the remaining reduction attributable to re-borrowing restrictions. Put another way, the reduction in revenues (which correspond to total amounts borrowed) predicted by the Bureau’s simulations are partially, though not primarily, attributed to changes in maximum loans amounts.1220 iv. Limits on Re-borrowing For storefront payday borrowers, most of the reduction in the availability of credit will likely be due to borrowers who have recently taken out loans being unable to roll their loans over 1218 CFPB Payday Loans and Deposit Advance Products White Paper, at 15. CFPB Vehicle Title Report, at 7 tbl. 1. 1220 Note that the Bureau’s simulations do not consider the possible strategic responses to the amortization features of loans made via the principal step-down approach. For example, some lenders may encourage borrowers to take out larger initial loans to ensure increased access to credit on the second and third loans in a sequence. To the extent this increases initial loan sizes, the Bureau’s estimates may overstate the expected decreases in lender revenues and borrowers’ access to credit. 1219 1395 or borrow again within a short period of time. As discussed above, the Bureau believes that most storefront payday lenders will employ the principal step-down approach to making loans. If lenders only make loans under the principal step-down approach, each successive loan in a sequence will have to reduce the amount borrowed by one-third of the original principal amount, with a maximum of three loans per sequence, and borrowers will only be able to take out six covered short-term loans in a 12-month period or be in debt on such loans for at most 90 days over the course of any 12-month period.1221 This restriction could limit borrowers paid monthly to as few as three loans per year, depending on when they take out their loans relative to when they are paid. If lenders make both ATR and principal step-down approach loans, borrowers who can demonstrate an ability to repay a loan will be able to take out ATR approach loans after they have reached the cap on loans issued via the principal step-down approach. As described above, consumers will benefit from not having long sequences of loans and the associated higher than anticipated borrowing costs. Some borrowers, however, may experience costs from not being able to continue to re-borrow. For example, consider a borrower who has a loan due and is unable to repay one-third of the original principal amount (plus finance charges and fees), but who anticipates an upcoming influx of income. This borrower may experience additional costs if unable to re-borrow the full amount due because of the restrictions imposed by the rule. These costs could include the costs of being delinquent on the loan and having a check deposited or ACH payment request submitted, either of which may lead 1221 Prior loans made using the ATR approach would count towards the maximum number of loans and maximum time-in-debt limits of the principal step-down approach. 1396 to an NSF fee. Borrowers in this situation may reasonably expect to eventually repay the loan, given the upcoming influx, but may simply default if they are not permitted to re-borrow. The Bureau does not believe, however, that the restrictions on lending will necessarily lead to increases in borrowers defaulting on payday loans, in part because the step-down provisions of the principal step-down approach are designed to help consumers reduce their debt over subsequent loans. This step-down approach should reduce the risk of payment shock and lower the risk to lenders and borrowers of borrowers defaulting when a lender is unable to continue to lend to them (though some borrowers who would have re-borrowed the full amount of the initial loan may now default, if they are unable to successfully make the step-down payment). Additionally, the Bureau’s simulations indicate that the majority of reduced access to credit will result from the re-borrowing restrictions, rather than initial loan size cap and forced step-down features of loans made via the principal step-down approach. It is also possible that some borrowers or lenders will strategically respond to the step-down provisions by taking out larger initial loans to ensure that subsequent loans in a sequence are sufficient to cover anticipated expenses. Finally, borrowers anticipating an influx of more than three pay periods in the future may find it more appropriate to pursue a longer-term loan (where permitted), meaning they should be less prevalent in the market for short-term loans. Borrowers taking out single-payment vehicle title loans will also be much less likely to be able to roll their loans over or borrow again within a short period than they are today. They will potentially suffer the same costs as by payday borrowers taking out loans under the ATR approach who would prefer to roll over or re-borrow rather than repay their loan without reborrowing. 1397 v. Reduced Geographic Availability of Covered Short-Term Loans Consumers will also have somewhat reduced physical access to payday storefront locations. Bureau research on States that have enacted laws or regulations that substantially impacted the revenue from storefront lending indicates that the number of stores has declined roughly in proportion to the decline in revenue.1222 Because of the way payday stores locate, however, this has had much less impact on the geographic availability of payday loans. Nationwide, the median distance between a payday store and the next closest payday store is only 0.3 miles. When a payday store closes in response to laws that reduce revenue, there is usually a store nearby that remains open. For example, across several States with regulatory changes, between 93 and 95 percent of payday borrowers had to travel less than five additional miles to find a store that remained open. This is roughly equivalent to the median travel distance for payday borrowers nationwide. Using the loan volume impacts previously calculated above for storefront lenders exclusively using the principal step-down approach (which were about 7176 percent without accounting for additional ATR lending or for changes in product terms or mixes1223) the Bureau forecasts that a large number of storefronts will close under the rule, but 1222 CFPB Report on Supplemental Findings, at Chapter 3. This is consistent with theoretical research showing that State price caps should lead to fewer stores and more borrowers per store. See Mark Flannery & Katherine Samolyk, “Payday Lending: Do the Costs Justify the Price?,” (FDIC Ctr. for Fin. Res., Working Paper No. 2005–09, 2005), available at https://www.fdic.gov/bank/analytical/cfr/2005/wp2005/cfrwp_2005-09_flannery_samolyk.pdf; Mark Flannery & Katherine Samolyk, “Scale Economies at Payday Loan Stores,” at 233–259 (Proceedings of the Federal Reserve Bank of Chicago’s 43rd Annual Conference on Bank Structure and Competition 2007). It is also consistent with empirical analysis showing a correlation between State price caps and the number of stores per State resident. Pew Charitable Trusts, Fact Sheet, “How State Rate Limits Affect Payday Loan Prices” (Apr. 2014), available at http://www.pewtrusts.org/~/media/legacy/uploadedfiles/pcs/contentlevel_pages/fact_sheets/stateratelimitsfactsheet.pdf. 1223 It is important to note that the estimates for the reduction in lending above may underestimate impacts in some ways and overestimate them in others. For example, store closures may cause total lending to fall further. A small share of potential borrowers will lose easy access to stores. In addition, the reduced physical presence and therefore visibility of stores, even in areas where as store is fairly close by, may lead to some consumers not taking out loans, 1398 that consumers’ geographic access to stores will not be substantially affected in most areas. c. Evidence on the Benefits and Costs to Consumers of Access to Payday and other Covered Loans Most studies of the effects of payday loans on consumer welfare have relied on Statelevel variation in laws governing payday lending.1224 Most of these studies rely on an “intent to treat” identification strategy, where access to payday loans is used as a proxy for actual use. While certainly instructive, the Bureau believes findings from such studies are generally less compelling than those based on individual-level data that are able to identify actual payday borrowers and use. A third class of studies addressing questions around payday focuses on experiments, either in the field or in laboratory settings. Within this literature, most studies have examined storefront payday loans; the literature studying online loans and vehicle title lending is much smaller; and there is even less direct evidence on longer-term balloon-payment loans. The Bureau notes that all of the studies vary in their empirical rigor and the connection of their causal inference to their documented findings. As such, the Bureau, based on its experience and expertise, finds some studies to be more compelling than others. The Bureau discussed many of these studies in the proposal; additional studies are discussed here in light of comments received on the proposal. or borrowing less, because they are not reminded as frequently of the availability of payday loans. Some lenders, however, may successfully adapt to the regulation by, for example, broadening the range of products they offer. The ability to do this will vary across States and across individual lenders. 1224 This section focuses on the benefits and costs to consumers from payday lending. The literature on consumers’ understanding and expectations regarding payday lending, notably Mann (2013), is discussed earlier in this section and above in Market Concerns—Underwriting. Other strands of the literature related to payday and small-dollar lending (e.g., those addressing the populations of borrowers, endogenous market entry by lenders, changes in behavior or outcomes not related to regulatory changes, and academic studies of the business models or market structure) were also reviewed by the Bureau, but are not discussed here. 1399 As noted above, the rule does not ban payday or other covered short-term loans or longerterm balloon-payment loans. In fact, the Bureau believes that covered short-terms loans will still be available to consumers facing a truly short-term need for credit in States that allow them. In contrast, most research has focused almost exclusively on the question of what happens when all access to a given form of credit is eliminated, as opposed to restricted. This is often referred to as the extensive margin (access), rather than the intensive margin (use, once accessed). As noted above, the available evidence from States that have imposed strong restrictions on lending, but not outright or de facto bans, suggests that, even after large contractions in this industry, loans remain widely available, and access to physical locations is not unduly limited. To the extent that ability to repay and/or shorter loan sequences are associated with beneficial borrowing, this should not unduly restrict the positive welfare for consumers associated with borrowing to cover discrete needs. That said, if the benefits from borrowing are realized from later loans in a 12-month period, and are concentrated predominately in the segment of borrowers who would not pass an ATR assessment, the rule will more substantially reduce the benefits realized by borrowers. As noted at the end of this section however, the Bureau believes that the literature implies the greatest benefits consumers receive from access to credit are realized early in a borrowing sequence. i. Intent-to-Treat Studies As mentioned previously, intent-to-treat studies focus on the availability of credit to larger populations of individuals, rather than focusing on the actual usage of that credit. Many of these studies focus on the changes resulting after States institute bans on payday lending. For example, Morgan and Strain (2008) study a number of State law changes over a ten-year period, 1400 and find that payday bans were associated with higher rates of bounced checks.1225 They also found that bans were associated with higher rates of complaints about debt collectors to the FTC, but lower rates of Chapter 13 bankruptcy filings. In an update to that paper, Morgan et al. (2012) expand the time frame, analyze more State-level payday bans, and consider the impacts of enabling payday lending as well.1226 They again find evidence that bounced checks and complaints about debt collectors to the FTC increase, and Chapter 13 bankruptcy filings decrease in response to limits on payday lending. They also find that the service fees received on deposit accounts by banks operating in a single State tend to increase with limits on payday lending, and interpret this as an indication that payday loans help to avoid overdraft fees. In contrast, Campbell, et al. (2008) found that Georgia’s payday ban appeared to improve consumer’s outcomes, as consumers living in counties further from bordering States that allowed payday lending had lower rates of involuntary checking account closures.1227 Bhutta et al. (2016), using data from the Current Population Survey, show some evidence of increased use of alternative forms of high-interest credit (e.g., pawn loans) when access to payday loans was restricted. 1228 Additionally, they present weak evidence of an increase in involuntary account closings after the imposition of State bans of payday loans, but this effect did not persist. In data collected by the Bureau from banks that ceased offering deposit advance products (“DAP 1225 Donald P. Morgan and Michael R. Strain, “Payday Holiday: How Households Fare after Payday Credit Bans” (Fed. Reserve of N.Y. Staff Reports No. 309, 2008). 1226 Donald P. Morgan and Ihab Seblani, “How Payday Credit Access Affects Overdrafts and Other Outcomes,” 44 J. of Money, Credit, and Banking 519 (2012). 1227 Dennis, F. Campbell et al., “Bouncing Out of the Banking System: An Empirical Analysis of Involuntary Bank Account Closures,” 36 J. of Banking and Fin. 1224 (2012). 1228 Neil Bhutta et al., “Consumer Borrowing after Payday Loan Bans.” 59 J. of L. and Econ. 225 (2016). 1401 loans”), there was no evidence that reduced access to these products led to greater rates of overdraft or account closure.1229 Melzer (2011) measured access to payday loans of people in States that do not allow payday lending using distance to the border of States that permit payday lending.1230 He measured the effects of access on the payment of mortgages, rent and utilities, and found that greater access causes greater difficulty in paying these basic expenses, as well as delays in needed medical care. In a follow-up study, Melzer (2016), found higher Supplemental Nutritional Assistance Program (food stamp) usage and lower child-support payments with greater payday availability. 1231 Two additional studies exploit State-level variation in access to estimate the impacts of payday loans by looking at similarly situated counties. Desai & Elliehausen (2017) compare counties in States that ban payday lending (Georgia, North Carolina, and Oregon) with adjacent States that allow such lending.1232 While the authors cannot observe whether or to what extent payday borrowing is actually occurring in these counties, it appears that legislation in the States curbing payday lending had very small, mostly positive, effects on delinquencies. Edminson (2011) uses a similar identification approach (county-level analysis with varying payday restrictions), but does not limit the analysis to counties in adjacent States.1233 This study 1229 CFPB Report on Supplemental Findings, at 39. Brian T. Melzer, “The Real Costs of Credit Access: Evidence from the Payday Lending Market,” 126 Quarterly J. of Econ. 517 (2011). 1231 Brian T. Melzer, “Spillovers from Costly Credit.” Review of Fin. Studies (forthcoming NW Univ., Kellogg Sch. of Management, Dep’t of Finance, 2013). 1232 Chintal A. Desai and Gregory Elliehausen, “The Effect of State Bans of Payday Lending on Consumer Credit Delinquencies,” 64 Quarterly Review of Econ. and Fin. 94 (2017). 1233 Kelly D. Edminston, “Could Restrictions on Payday Lending Hurt Consumers?” at 37-38 (Fed. Reserve Bank of K.C. Econ. Review 31, 2011). 1230 1402 concludes that restrictive payday regimes are associated with lower average credit scores, even when income is accounted for. Zinman (2010) conducted a survey of payday loan users in Oregon and Washington both before and after a new law took effect in Oregon that limited the size of payday loans and reduced overall availability of these loans.1234 He showed that the law appeared to increase consumer hardship, measured by unemployment and qualitative self-assessments of current and expected future financial conditions, over the subsequent five months. An alternative to the State-level variation in extensive access to payday loans is to look at the intensive concentration of lenders in a geographical area as a proxy for payday loan availability. For example, Morse (2011) looked at zip code-level data to assess the impact of the availability of payday loans in particular circumstances, natural disasters.1235 Using information about the concentration of payday lenders by zip code and linking it to data on natural disasters, she found that greater access to payday lending in times of disaster—which may generalize to unexpected personal emergencies—reduces home foreclosures and small property crime. Dobridge (2014) found that, in normal times, access to payday loans reduced consumer wellbeing, as measured by purchases of consumer durable goods.1236 But, similar to Morse (2011), Dobridge found that in times of severe weather, access to payday loans allowed consumers to smooth consumption and avoid declines in food spending or missed mortgage payments. Carrell and Zinman (2014) also developed a measure of payday loan access similar to that used by 1234 Jonathan Zinman, “Restricting Consumer Credit Access: Household Survey Evidence on Effects Around the Oregon Rate Cap,” 34 J. of Banking and Fin. 546 (2010). 1235 Adair Morse, “Payday Lenders: Heroes or Villains?,” 102 J. of Fin. Econ. 28 (2011). 1236 Christine L. Dobridge, “Heterogeneous Effects of Household Credit: The Payday Lending Case” (Wharton Sch., Univ. of Penn., Working Paper, 2014). Note that this paper relies on a State-level approach (similar to Melzer, 2011), as opposed to the more intensive measures used by Morse (2011). 1403 Morse (2011) and linked it to the job performance of Air Force personnel, showing that greater access to payday lending leads to worse job performance to such an extent that fewer are eligible for reenlistment.1237 Carter and Skimmyhorn (2016) used an alternative identification strategy, utilizing the differential access to payday loans associated with the implementation of the Military Lending Act (MLA). The MLA effectively banned payday loans to military personnel, allowing the authors to measure the impact of payday loans on financial well-being and labor market outcomes of soldiers in the Army.1238 Unlike Carrell and Zinman who also focused on military personnel, Carter and Skimmyhorn found no effects. They speculated that some of the difference in the outcomes of the two preceding studies could reflect the fact that re-enlisting in the Army was easier than re-enlisting in the Air Force during the periods covered by the respective studies. Another study also used the implementation of the MLA to measure the effects of payday loans on the ability of consumers to smooth their consumption between paydays, and found that access to payday loans did appear to make purchasing patterns less concentrated around paydays (Zaki, 2013).1239 This study also found some evidence that access to payday loans increased what the author referred to as “temptation purchases,” specifically alcohol and consumer electronics. 1237 Scott E. Carrell and Jonathan Zinman, “In Harm’s Way? Payday Loan Access and Military Personnel Performance,” 27 Rev. of Fin. Studies 2805 (2014). 1238 Susan Payne Carter and William Skimmyhorn “Much Ado About Nothing? New Evidence on the Effects of Payday Lending on Military Members,” (forthcoming Rev. of Econ. and Stats, 2016). 1239 Mary Zaki, “Access to Short-term Credit and Consumption Smoothing within the Paycycle” (FEEM. Working Paper No. 007.2016, 2016), available at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2741001. 1404 Among these intent-to-treat studies, industry comments most often cited Morgan and Strain (2008), Zinman (2010), Morse (2011), and Morgan et al. (2012), along with a related study that is no longer available.1240 Many of these commenters argued that these studies suggest strong, positive welfare impacts of access to payday lending. However, Morgan and Strain (2008) relies on a methodology that severely undermines their conclusions. Specifically, Morgan and Strain’s (2008) assertion that checks are returned more frequently from the nonauthorizing payday States of Georgia and North Carolina relies on data that intermingles those States’ data with that of numerous authorizing States (e.g., Louisiana, Alabama, and Tennessee).1241 Additionally, the complaints data they cite are limited by the fact that the FTC is unlikely to receive complaints about payday lending (at the time, State regulators were more likely to receive such complaints). As such, the complaints measure the authors employ may not indicate the actual rate of credit-related complaints, let alone overall consumer satisfaction. While Morgan et al. (2012) expands on the previous studies by including more States (contributing to the policy variation needed for identification), and additional outcome measures (e.g., bank fee income), they fail to adequately address the shortcomings of their previous studies. For example, this study once again employs the measure of complaints received by the FTC. It also relies on data sources that comingle returned checks from States with payday bans with those from States that permit payday, which their difference-in-difference identification 1240 Donald P. Morgan, “Defining and Detecting Predatory Lending” (Fed. Reserve Bank of N.Y. Staff Report No. 273, 2007). FRBNY webpage indicates report was “removed at the request of the author.” 1241 Donald P. Morgan and Michael R. Strain, “Payday Holiday: How Households Fare after Payday Credit Bans,” (Fed. Reserve of N.Y. Staff Report No. 309, 2008), available at https://www.newyorkfed.org/research/staff_reports/sr309.html (similarly mischaracterizes authorizing and nonauthorizing States, e.g., asserting North Carolina to be a non-authorizing State despite having 500+ payday lenders during the period analyzed.). 1405 approach may not adequately address. For example, the Atlanta check processing center (CPC) is coded as “banned” even after States that allow payday (e.g., Alabama and Louisiana) are absorbed; the Oregon payday ban is never coded into their data since the CPC for Oregon is in Seattle (and Washington allows payday); etc.1242 The biggest addition to the paper relative to Morgan and Strain (2008) is that Morgan et al. (2012) analyze a new outcome to support the notion that payday limits are associated with an increase in overdrafts by looking at bank revenues realized through fees. However, their proxy for overdraft fees includes all service fees on deposit accounts at a time when the prevalence of overdraft was changing, and they limit their sample of banks to only those operating in a single State, limiting both the accuracy and generalizability of their finding. Finally, most of the findings in Morgan et al. (2012) are not robust but rather highly sensitive to the choice of specification. For example, the point estimates and significance levels change a great deal in response to the inclusion or exclusion of State-specific time trends; the service fee findings are dependent on using a log fees per capita measure, rather than the more natural fees per capita or log fees; and their findings for the impacts of State-level bans on returned checks become insignificant when questionable demographic variables are excluded from the regressions.1243 1242 These findings were obtained from a brief analysis of the data used by Morgan et al. (2012), see Donald P. Morgan and Ihab Seblani, “How Payday Credit Access Affects Overdrafts and Other Outcomes,” 44 J. of Money, Credit, and Banking 519 (2012). 1243 The authors note their coefficients of interest “were insignificant in regressions using (unlogged) levels of fee income and income per capita.” Donald P. Morgan and Ihab Seblani, “How Payday Credit Access Affects Overdrafts and Other Outcomes,” 44 J. of Money, Credit, and Banking 519, at n.16 (2012). The findings about the sensitivity of the returned checks estimates were achieved by analyzing the Morgan et al. (2012) data available at id. It should also be noted that the Bureau finds other weaknesses in the analytic approach employed in this study. Specifically, the difference-in-difference approach for returned checks relies on observations at the check processing 1406 Zinman (2010) was also frequently cited by industry comments. Those comments point to the qualitative findings that survey respondents indicate greater levels of “financial hardships” after a payday ban. However, the quantitative findings show indications that the welfare effects of the ban may have been positive (e.g., lower rates of phone disconnections, greater rates of on time bill payment).1244 Additionally, the findings rely on a small survey conducted across only two States where idiosyncratic effects may drive many of the results. As such, the Bureau believes the actual welfare implications from this study are hard to generalize. Priestly (2014), another paper frequently mentioned in industry comments, is more clear on the welfare implications of payday, and specifically re-borrowing. The author’s results indicate, for example, that each rollover in 2008-2009 was associated with a .109-point increase in a customer’s VantageScore (a credit score similar to FICO). The Bureau believes these benefits are quite small, as Priestly’s findings suggest that the average consumer in her sample would need to roll a payday loan over more than nine times (at a cost of approximately $135 per $100 borrowed) in order to increase his or her VantageScore by one point. For the average customer in Priestly’s sample, this would represent an increase from 587 to 588, deep enough into the subprime range that such a change would be unlikely to have any practical value. center (CPC) level, yet a single CPC may process checks from many States, some of which ban payday, some of which allow it, and some of which have no explicit allowance or ban. The authors attempt to control for this using a very large number of dummy variables to capture CPC mergers, but this results in estimates that are highly sensitive to specification assumptions. Additionally, the study appears to code in “sharp” policies where the policy is actually “fuzzy,” which would cause identification problems (e.g., they code a payday ban for P.A. in 2007, when the last payday lender exited the market, even though there had been a longer decline since 2006 when the legislation was passed). There are additional econometric issues with this study’s approach, but the Bureau believes those cited here are sufficient to cast doubt on the strength of the reported findings. 1244 Phone disconnections were explored in greater detail in the working paper version. See Jonathan Zinman, “Restricting Consumer Credit Access: Household Survey Evidence on Effects Around the Oregon Rate Cap,” (Dartmouth College, 2008), available at http://www.dartmouth.edu/~jzinman/Papers/Zinman_RestrictingAccess_oct08.pdf. 1407 The Morse (2011) study differs from the other intent-to-treat studies most cited by commenters, as it focuses on a source of variation more relevant to this rule (endogenous concentrations of lenders, rather than restrictions on locations), and its welfare implications are more nuanced. Specifically, Morse finds that borrowers appear “better off” in the face of unexpected shocks (i.e., those that lead to discrete needs) with access to payday loans. While the outcome measures used in the study (e.g., home foreclosures) limit the generalizability of the findings (as homeowners may not be representative of the typical payday borrower), the Bureau believes this study is methodologically sound and the findings are large and significant enough to warrant deep consideration. However, the Bureau has found little in this study to imply that a limit on continued use of payday loans (rather than a limit on the availability of short-term credit for discrete needs) would necessarily decrease borrowers’ welfare. ii. Individual-level studies Other studies, rather than using differences across States in the availability of payday loans, have used data on the actual borrowers who apply for loans and are either offered loans or are rejected. These individual-level studies offer more direct insight into the effects of payday loans, rather than the effect of access measured by the intent-to-treat studies. Skiba and Tobacman (2009) used this approach to find that taking out a payday loan increases the likelihood that the borrower will file for Chapter 13 bankruptcy.1245 They found that initial approval for a payday loan essentially doubled the bankruptcy rate of borrowers. Bhutta, et al., (2015) used a similar approach to measure the causal effects of storefront borrowing on 1245 Paige Marta Skiba and Jeremy Tobacman. “Do Payday Loans Cause Bankruptcy?,” (Vand. U. Sch. of L., L. and Econ., Working Paper No. 11-13, 2011), available at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=1266215. 1408 borrowers’ credit scores.1246 They found that obtaining a loan had no impact on how the consumers’ credit scores evolved over the following months. The authors noted, however, that applicants generally had very poor credit scores both prior to and after borrowing (or being rejected for) a payday loan. In each of these studies, the authors were unable to determine whether borrowers that were rejected by the lender from which they had data were able to take out a loan from another lender. Two other studies have used data on payday borrowing and repayment behavior to compare changes over time in credit scores for different groups of borrowers. Priestley (2014), discussed above, measured changes over time in credit scores for borrowers who re-borrowed different numbers of times, and found that in some cases it appeared that borrowers who reborrowed more times had slightly more positive changes in their credit scores.1247 These differences were not economically meaningful, however, implying borrowers would need to rollover a loan more than nine times (at an average total cost of $135 per $100 borrowed) to see a one-point increase in their VantageScores.1248 Mann (2014) compared the changes in credit scores of borrowers who defaulted on their loans with borrowers who did not, and also found no difference.1249 Similar to the Bhutta, et al. (2015) study, neither the Priestly nor Mann studies found a meaningful effect of payday loan borrowing behavior on credit scores. Unlike Bhutta, et 1246 Neil Bhutta et al., “Payday Loan Choices and Consequences,” 47 J. of Money, Credit and Banking 223 (2015). Jennifer Priestly, “Payday Loan Rollovers and Consumer Welfare” (Kennesaw State U., Dep’t of Stats. and Analytical Sciences 2014). 1248 The Priestley study also compared changes over time in credit scores of payday borrowers in different States, and attributed those differences to differences in the States’ payday regulations. This ignores differences in who chooses to take out payday loans in different States, and ignores the different changes over time in the broader economic conditions in different States. 1249 Ronald Mann, “Do Defaults on Payday Loans Matter?,” (Colum. L. and Econ., Working Paper No. 509, 2015), available at https://papers.ssrn.com/sol3/Papers.cfm?abstract_id=2560005. 1247 1409 al. (2015), however, if either had measured an effect it would have simply been a finding of correlation, as neither had a way of identifying an effect as causal. Gathergood, et al. (2016),1250 used an approach similar to that used by Skiba and Tobacman (2014) and Bhutta, et al., (2015) to study the effects of taking out payday loans on United Kingdom borrowers’ future overdrafting, rates of delinquency on other loan products, subjective well-being, and feelings of regret about borrowing. The products studied are similar to payday loans in the United States, primarily single-payment loans due in roughly 30 days. While the UK market includes storefront lenders, it is dominated by online lenders. The authors found that online payday loans led to higher rates of bank overdraft and delinquencies on other loans. While it had no effect on subjective measures of well-being, borrowers did report regretting the decision to take out the payday loan. Baugh (2015) used the closure of dozens of online payday lenders, which cut off borrowers’ access to such loans and other high-cost online credit, to measure the effects of these loans on consumers’ consumption, measured via expenditures on debit and credit cards, and on overdrafts and insufficient funds transactions.1251 He found that losing access to these loans, especially for consumers who had been heavy users of these loans, led to increased consumption and fewer overdrafts or NSF transactions. iii. Experimental Studies 1250 John Gathergood et al., “Comments on: How do Payday Loans Affect Consumers?” (NBER Summer Institute– L. and Econ. 2015). 1251 Brian Baugh, “What Happens When Payday Borrowers Are Cut Off From Payday Lending? A Natural Experiment,) (Ph.D. dissertation, Ohio State Univ., 2015), available at http://fisher.osu.edu/supplements/10/16174/Baugh.pdf. 1410 There have also been at least three studies of the impacts of payday loans that rely on experimental approaches. Bertrand and Morse (2011) run an experiment providing three types of information disclosures about the costs and re-borrowing rates of payday loans at the time borrowers receive their loans from a storefront payday lender.1252 The disclosures are found to reduce the incidence of re-borrowing by 6-11 percent and the average amount borrowed by 1223 percent relative to the control group, with stronger results for borrowers self-reporting higher degrees of self-control. Fusaro and Cirillo (2011) conduct an experiment in which some borrowers are given nofee loans and their re-borrowing rates are compared to borrowers who are given loans with normal fees.1253 They find that re-borrowing rates are not different between the two groups. This could lead to at least two possible and compatible conclusions: that the cost does not drive a cycle of debt, and/or that the single-payment structure is a key factor that drives unaffordability, not merely the fee. Commenters also referenced a third experimental study, Wilson et al. (2010).1254 In this study the authors conducted a laboratory experiment designed to test whether access to payday loans improves or worsens the likelihood of “financial survival” or financial health in the face of expense shocks. The authors found that the students engaged in the game were more likely to successfully manage financial shocks if they had access to payday loans. However, when they explore the intensity of usage, they find that participants who utilize 10 or more loans over the 30 1252 Marianne Bertrand, and Adair Morse, “Information, Disclosure, Cognitive Bias, and Payday Borrowing,” 66 J. of Fin. and Econ. 1865 (2011). 1253 Marc A. Fusaro & Patricia J. Cirillo, “Do Payday Loans Trap Consumers in a Cycle of Debt?,” (2011), available at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=1960776. 1254 Bart J. Wilson et al., “An Experimental Analysis of the Demand for Payday Loans,” 10 B.E. J. of Econ. Analysis & Policy (2010). 1411 experimental months find themselves at greater risk than they would under a regime that bans payday loans. iv. Discussion of Literature The Bureau received numerous comments selectively citing the studies listed above, and making reference to particular results of interest to the commenters. Generally, industry and trade group commenters favored studies that imply access improves consumer outcomes (e.g., Priestly (2014), Zinman (2010)); consumer groups favored studies that imply access harms consumers (e.g., Skiba and Tobacman (2015), Baugh (2015)); and academic researchers referenced numerous studies highlighting the ambiguity or uncertainty illustrated by the literature. The Bureau has considered the comments carefully and gives weight to the studies in proportion to their applicability to the rule, generalizability, and methodological soundness.1255 Additionally, and as much as possible, the Bureau has endeavored to rely on the descriptive (positive) findings of the studies, and not the authors’ interpretations (often normative) of those findings. In reviewing the existing literature, the Bureau notes that the evidence on the impacts of the availability of payday loans on consumer welfare indeed varies. In general, the evidence to 1255 The Bureau received numerous comments calling into question the objectivity of some studies funded by industry. These issues have also been noted in the press. See, e.g., Ben Walsh and Ryan Grim, “Emails Show ProPayday Loan Study Was Edited by the Payday Loan Industry,” Huffington Post, Nov. 2, 2015, available at http://www.huffingtonpost.com/entry/payday-loan-study_us_5633d933e4b00aa54a4e4273; Christopher Werth, “Tracking the Payday-Loan Industry’s Ties to Academic Research,” Freakonomics, Apr. 6, 2014, available at http://freakonomics.com/podcast/industry_ties_to_academic_research/. At least one of these studies appears to have given editorial and content control to an industry lobbyist. Others failed to reference the financial and other support received from the group in any of their acknowledgements, as is the best practice in such research. Still others mention the support received, but assert the group had no influence on the study or its findings (a similar assertion was made in the study where influence was documented). Such comments are to be expected in any contentious policy debate. Overall, the Bureau attempted to judge each study on its merits. As such, findings from these industry studies are generally weighted by their methodological soundness (in terms of data collection and analysis). 1412 date suggests that access to payday loans appears to benefit consumers in circumstances where they use these loans for short periods to address an unforeseen and discrete need, such as when they experience a transitory and unexpected shock to their incomes or expenses. However, in more general circumstances, access to and intensive use of these loans appears to make consumers worse off. A more succinct summary is: access to payday loans may well be beneficial for those borrowers with discrete, short-term needs, but only if they can succeed in avoiding long sequences of loans. There is also some limited evidence about the welfare effects of “intensive” users of payday. It should be noted, however, that there are no studies the Bureau is aware of that directly evaluate the welfare impacts of the seventh and later loans taken by a borrower in a 12month span.1256 There are also no studies on the welfare effects of payday loans made specifically to borrowers who would have failed an ATR assessment. Since the rule’s restrictions should only bind for individuals who demand a seventh loan in a 12-month period and cannot demonstrate an ability to repay, there are no studies that speak directly to the likely impacts of the regulation. As this rule will allow for continued access to the credit that appears to benefit consumers with discrete needs, the Bureau believes that the rule limits the potential harm other borrowers may experience while maintaining much of the welfare gains consumers realize from access to these loans. G. Benefits and Costs of the Rule to Covered Persons and Consumers – Payments and Notices 1256 Bart J. Wilson et al., “An experimental analysis of the demand for payday loans,” 10 B.E. J. of Econ. Analysis & Policy (2010) (This analysis does show that once a participant takes 10 or more loans in a 30-month span, the loans appear to be more harmful than helpful to financial survival.) 1413 The rule limits how lenders initiate payments on a covered loan from a borrower’s account and imposes two notice requirements relating to such payments. Specifically, if two consecutive prior attempts to withdraw payment through any channel from a borrower’s account have failed due to insufficient funds, lenders are prohibited from continuing to attempt to withdraw payment from a borrower’s account, unless the lender obtains a new and specific authorization to make further withdrawals from the consumer’s account. The rule also requires lenders of covered loans to provide a notice to a borrower before the initial withdrawal attempt and before initiating an unusual withdrawal attempt. A special notice is also required to be sent to the borrower if the lender can no longer continue to initiate payment directly from a borrower’s account because two consecutive prior attempts had failed due to insufficient funds. The impacts of these proposals are discussed here for all covered loans. Note that the Bureau expects that unsuccessful payment withdrawal attempts will be less frequent under the rule. This is because of the notice of irregular withdrawals; and it is also true because the ability-to-repay provisions or the requirements of the conditional exemption loans will reduce the frequency with which borrowers receive loans that they do not have the ability to repay. This should in turn lessen the impacts of the limitation on payment withdrawal attempts and the number of instances where a lender is required to notify consumers that the lender is no longer permitted to attempt to withdraw payments from a borrower’s account. Most if not all of the requirements in this portion of the rule are activities that lenders could have chosen to engage in absent the rule. As such, the Bureau believes that, while there are potential benefits to lenders, the restrictions are expected to impose some costs on these 1414 covered persons.1257 That said, the Bureau is aware that many lenders have practices of not continuing to attempt to withdraw payments from a borrower’s account after one or more failed attempts, and that some depository institutions do not assess additional fees to customers when continued attempts to withdraw from their accounts are made. In addition, some lenders provide upcoming-payment notices to borrowers in some form. 1. Limitation on Payment Withdrawal Attempts The rule prevents lenders from attempting to withdraw payment from a consumer’s account if two consecutive prior payment attempts made through any channel are returned for nonsufficient funds. The lender can resume initiating payment if the lender obtains from the consumer a new and specific authorization to collect payment from the consumer’s account. a. Benefits and Costs to Covered Persons The rule will impose costs on lenders by limiting their use of payment methods that allow them to withdraw funds directly from borrowers’ accounts, and by imposing the cost of obtaining a renewed authorization from the consumer or using some other method of collecting payment. There may be some benefits to lenders of reduced attempts to withdraw funds following repeated failures, as other methods of collecting may be more successful. The impact of this restriction depends on how often a lender previously attempted to collect from a consumers’ account after more than two consecutive failed transactions, and how often the lender was successful in doing so. Based on industry outreach, the Bureau understands that some lenders had already established a practice of not continuing to attempt to collect using 1257 This is simply a revealed preference argument that to the extent that lenders did not voluntarily choose to engage in the activities, it is likely the case that the benefits to lenders do not outweigh the costs to lenders (at least in the lenders’ views). 1415 these means after one or two failed attempts. These lenders would not incur costs from the restriction. Additionally, some depository institutions have disallowed repeated attempts to collect using these means; lenders attempting to collect from such depositories would also not incur costs from this restriction. The Bureau has analyzed the ACH payment request behavior of lenders making payday or payday installment loans online. The Bureau found that about half the time that an ACH payment request fails, the lender makes at least two additional ACH payment requests.1258 The likelihood of a successful payment request after a request that was returned for insufficient funds is quite low. Only 30 percent of requests that follow a failed request succeed, only 27 percent of third requests succeed, and after that the success rate is below 20 percent.1259 The Bureau found that only 7 to 10 percent of the payments attempted through the ACH system came after two failed payments requests, equivalent to $55 to $219 per borrower from whom a payment was collected after the two failed attempts.1260 These payments would have been prevented if the rule had been in place at the time. The Bureau notes that under the restriction, lenders can still seek payment from borrowers by engaging in other lawful collection practices. As such, the preceding are high-end estimates of the impact this restriction would have had on the collection 1258 CFPB Online Payday Loan Payments, at 14 tbl. 2. Lenders make at least one additional request after a failed payment request 74 percent of the time. Two-thirds of these are followed by a third request, if the second also fails. These calculations exclude multiple requests made on the same day, as those requests are unlikely to be intentional re-presentments of failed attempts because the lender is unlikely to know that a payment failed on the same day it was submitted and be able to re-present the request on the same day. The data used in the Bureau’s analysis were for 18 months in 2011 and 2012. Changes to the rules governing the ACH system in the fall of 2015 may have reduced the frequency with which lenders continue to make payment requests after one or more payment attempts have failed. 1259 CFPB Online Payday Loan Payments, at 13 tbl. 1. 1260 CFPB Report on Supplemental Findings, at 150. These impacts may be lower now than they were at the time covered by the data analyzed by the Bureau, due to changes in industry practices and to changes in the rules governing the ACH system referred to in note CFPB Online Payday Loan Payments, at 14 tbl. 2. 1416 efforts of these lenders. These other forms of lawful collection practices, however, may be more costly for lenders than attempting to collect directly from a borrower’s account. After the limitation is triggered by two consecutive failed attempts, lenders are required to send a notice to consumers. To seek a new and specific authorization to collect payment from a consumer’s account, the lender can send a request with the notice and may need to initiate additional follow-up contact with the consumer. The Bureau believes that this will most often be done in conjunction with general collections efforts and will impose little additional cost on lenders, other than the costs associated with the disclosures, discussed below. To the extent that lenders assess returned item fees when an attempt to collect a payment fails and are subsequently able to collect on those fees, this rule may reduce lenders’ revenues. Lenders will also need the capability of identifying when two consecutive payment requests have failed. The Bureau believes that the systems lenders use to identify when a payment is due, when a payment has succeeded or failed, and whether to request another payment will have the capacity to identify when two consecutive payments have failed, and therefore this requirement will not impose a significant new cost. b. Benefits and Costs to Consumers Consumers will benefit from the restriction because it will reduce the fees they are charged by the lender and the fees they are charged by their depository institution. Many lenders charge a returned item fee when a payment is returned for insufficient funds. Borrowers will benefit if the reduced number of failed ACH payment requests also results in reductions in the number of these fees, to the extent that they are eventually paid. Borrowers may also benefit from a reduction in the frequency of checking account closure, to be discussed below. 1417 Each time an ACH transaction is returned for insufficient funds, the borrower is likely to be charged an NSF fee by her financial institution. In addition, each time a payment is paid by the borrower’s financial institution when the borrower does not have sufficient funds in the account to cover the full amount of the payment, the borrower is likely to be charged an overdraft fee. Overdraft and NSF fees each average $34 per transaction.1261 As noted above, most re-presentments1262 of failed payment requests themselves fail, leading to additional NSF fees. In addition, about a third of all re-presentments that succeed only succeed because the borrower’s financial institution paid it as an overdraft, likely leading to an overdraft fee. The Bureau’s analysis of online lender payment practices shows that borrowers who have two payment withdrawal attempts fail are charged additional fees on subsequent payment attempts of $64 to $87. These costs would be prevented by the rule.1263 The restriction on repeated attempts to withdraw payments from a borrower’s checking account may also reduce the rate of account closure, as account closures appear to be associated with failed withdrawal attempts. This benefits borrowers by allowing them to maintain their existing account so as to better manage their overall finances. It also allows them to avoid the possibility of a negative record in the specialty consumer reporting agencies that track involuntary account closures, which can make it difficult to open a new account and effectively cut the consumer off from access to the banking system and its associated benefits. In the data 1261 CFPB Online Payday Loan Payments, at 2. For the purposes of its analysis, the Bureau referred to any payment request following a failed payment request as a “re-presentment.” The only exception was when multiple payment requests were submitted on the same day; if two or more failed, only the first failed payment request was considered a re-presentment. 1263 The Bureau notes that at least one depository institution limits the fees charged to consumers from multiple attempts to drawn on an account by payday lenders. To the extent that this type of policy is being voluntarily adopted, the net benefits of this limitation might decrease (due to an increase in the benefits present in the baseline). 1262 1418 studied by the Bureau, account holders who took out online payday loans were more likely to have their accounts closed by their financial institution than were other account holders, and this difference was substantially higher for borrowers who had NSF online loan transactions.1264 Borrowers with two consecutive failures by the same lender are significantly more likely to experience an involuntary account closure by the end of the sample period than accountholders generally (43 percent versus 3 percent, respectively).1265 While there is the potential for a number of confounding factors, transactions that were NSFs could contribute to account closure in at least two ways. First, the fees from repeated payment attempts add to the negative balance on the deposit account, making it more difficult for a borrower to bring the account balance positive and maintain a positive balance. And, if a lender is repeatedly attempting to extract money from an account, the borrower may feel that the only way to regain control of her finances is to cease depositing money into the account and effectively abandon it. The reduced ability to collect by repeatedly attempting to withdraw payments from a borrower’s account may increase lenders’ credit losses, which may, in turn reduce the availability or raise the cost of credit. As discussed in the consideration of the costs to lenders, this reduction in collections is likely to be quite small. And, as noted above in the discussion of the impacts of the ATR requirements, many lenders already charge the maximum price allowed by State law. 2. Required Notice Prior to Attempt to Collect Directly from a Borrower’s Account 1264 1265 CFPB Online Payday Loan Payments, at 24. CFPB Report on Supplemental Findings, at 151 n. 177. 1419 The rule also requires lenders to provide consumers with a notice prior to the first lenderinitiated attempt to withdraw payment from consumers’ accounts, including ACH entries, postdated signature checks, remotely created checks, remotely created payment orders, and payments run through the debit networks. The notice is required to include the date the lender will initiate the payment request; the payment channel; the amount of the payment; the breakdown of that amount to principal, interest, and fees; the loan balance remaining if the payment succeeds; the check number if the payment request is a signature check or RCC; and contact information for the consumer to reach the lender. There are also separate notices required prior to unusual payments. a. Benefits and Costs to Covered Persons These notices may reduce delinquencies and related collections activities if consumers take steps to ensure that they have funds available to cover loan payments, such as delaying or forgoing other expenditures, making deposits into their accounts, or contacting the lender to make alternative arrangements. Costs to lenders of providing these notices will depend heavily on when the lender provides the notice and, should they provide a notice after origination, whether they are able to provide the notice via email, text messages, or on paper at origination or have to send notices through paper mail. In practice, the Bureau expects most lenders to provide the notice of initial payment withdrawal at origination, minimizing the transmission costs. This can either be done via a written disclosure (at a storefront), or as a PDF attachment, or webpage sent along with an electronic short notice sent via an email or text (for either storefront or online lenders). The variation in costs of notices provided after origination (either regular notices, or notices in 1420 advance of unusual payments) is due in part to differences in transmission costs between different channels. Most borrowers are likely to have Internet access and/or a mobile phone capable of receiving text messages, and during the SBREFA process multiple SERS reported that most borrowers, when given the opportunity, opt in to receiving notifications via text message. The Bureau has intentionally structured the rule to encourage transmission by email or text message because it believes those channels are the most effective for consumers, as well as less burdensome for lenders. However, should the lender choose to send paper notifications via regular mail, they would incur higher costs of transmission, as well as administrative costs associated with providing the notification early enough to ensure sufficient time for it to be received by the consumer. The Bureau believes that all lenders affected by the new disclosure requirements have some system in place to comply with existing disclosure requirements, such as those imposed under Regulation Z, 12 CFR part 1026, and Regulation E, 12 CFR part 1005. Lenders enter data directly into the disclosure system, or the system automatically collects data from the lenders’ loan origination system. For disclosures provided via mail, email, text message, or immediately at the time of origination, the disclosure system often forwards the information necessary to prepare the disclosures to a vendor in electronic form, and the vendor then prepares and delivers the disclosures. Lenders will incur a one-time burden to upgrade their disclosure systems to comply with new disclosure requirements. Lenders will need to update their disclosure systems to compile the necessary loan information to send to the vendors that will produce and deliver the disclosures relating to payments. The Bureau believes that large lenders rely on proprietary disclosure systems, and 1421 estimates the one-time programming cost for large respondents to update these systems to be 1,000 labor hours per entity. The Bureau believes small lenders rely on licensed disclosure system software. Depending on the nature of the software license agreement, the Bureau estimates that the cost to upgrade this software will be $10,000 for lenders licensing the software at the entity-level and $100 per seat for lenders licensing the software using a seat-license contract. For lenders using seat license software, the Bureau estimates that each location for small lenders has on average three seats licensed. Given the price differential between the entitylevel licenses and the seat-license contracts, the Bureau believes that only small lenders with a significant number of stores will rely on the entity-level licenses. Lenders with disclosure systems that do not automatically pull information from the lenders’ loan origination or servicing system will need to enter payment information into the disclosure system manually, so that the disclosure system can generate payment disclosures. The Bureau estimates that this will require two minutes per loan in addition to the two minutes to provide the disclosures. Lenders would need to update this information if the scheduled payments were to change. For disclosures delivered through the mail, the Bureau estimates that vendors will charge two different rates, one for high volume mailings and another for low volume mailings. For the high volume mailings, the Bureau estimates vendors will charge $0.53 per disclosure. However, the Bureau expects high volume mailings to be infrequent, as follow-up disclosures are only necessary for unusual payments and reauthorizations. For the low-volume mailings, the Bureau estimates vendors will charge $1.00 per disclosure. For disclosures delivered through email, the Bureau estimates vendors will charge $0.01 to create and deliver each email such that it complies 1422 with the requirements of the rule. For disclosures delivered through text message, the Bureau estimates vendors will charge $0.08 to create and deliver each text message such that it complies with the requirements of the rule. The vendor will also need to provide either a PDF attachment of the full disclosure or a web page where the full disclosure linked to in the text message is provided. The cost of providing this PDF attachment or web disclosure is included in the cost estimate of providing the text message. Finally, for disclosures delivered on paper at origination, the Bureau estimates costs will be $0.10 per disclosures. Again, the Bureau believes that virtually all notifications will be provided at the time of origination (for regular notices), or electronically via text or email (for notifications of unusual payments). As such, the mailing costs discussed here are expected to be almost completely avoided. In addition to the costs associated with providing notices, this requirement may impact the frequency with which lenders initiate withdrawal attempts and lenders’ revenue. On timing, lenders are likely to disclose all regular payment schedules at origination, and must provide notices on unusual payments in advance of their initiation. This lag time could affect lenders’ decisions as to the timing and frequency of withdrawal attempts. With regard to revenue, the impacts are uncertain: payment revenue will be reduced if the notices lead to consumers taking steps to avoid having payments debited from their accounts, including placing stop-payment orders or paying other expenses or obligations prior to the posting of the payment request. Alternatively, if the notices help borrowers to ensure that funds are available to cover the payment request, this will reduce lenders’ losses from non-payment, although it will also lower lenders’ returned-item fee revenue. 1423 b. Benefits and Costs to Consumers Receiving notices prior to an upcoming unusual payment will benefit consumers by allowing them to take those payments into account when managing the funds in their accounts. This will allow them to reduce the likelihood that they will run short of funds to cover either the upcoming payment or other obligations. The notice will also help borrowers who have written a post-dated check or authorized an ACH withdrawal, or remotely created check or remotely created payment order, to avoid incurring NSF fees. These fees can impose a significant cost on consumers. In data the Bureau has analyzed, for example, borrowers who took out loans from certain online lenders paid an average of $92 over an 18 month period in overdraft or NSF fees on the payments to, or payment requests from, those lenders.1266 The information in the notices may also benefit borrowers who need to address errors or unauthorized payments, by making it easier for the borrower to resolve errors with the lender or obtain assistance through their financial institution prior to the payment withdrawal being initiated. Some consumers may incur costs for notices sent by text. Consumers can avoid these costs by choosing email;1267 the Bureau requires that lenders must provide an email delivery option whenever they are providing a text or other electronic delivery option. As some commenters noted, costs associated with the disclosures might be passed on the consumers. However, the Bureau believes the costs associated with the disclosures will be 1266 CFPB Online Payday Payments, at 3. It is possible that some consumers may only have access to email via data-limited plans (e.g., smartphones), and thus receiving emails could impose costs as well. However, there are numerous ways to avoid the cost of accessing email (e.g., public libraries or facilities that offer free WiFi). As such, the Bureau considers the cost of receiving an email to be negligible. 1267 1424 limited, as noted above. Specifically the costs will be much lower than under the proposed rule, which would have required a disclosure before each payment withdrawal attempt. Ultimately, the Bureau believes these costs to consumers will be small in relation to the overall cost of the loan. 3. Required Notice When Lender Can No Longer Collect Directly from a Borrower’s Account The rule requires a lender to provide a borrower with a notice of consumer rights within three days of a second consecutive unsuccessful attempt to collect payment from a borrower’s account. This notice will identify the loan, explain that the lender is no longer able to attempt to collect payment directly from the borrower’s account, and provide the consumer a record of the two failed attempts to collect funds. a. Benefits and Costs to Covered Persons This provision may benefit lenders if it leads to consumers contacting the lender to provide a new authorization to withdraw payments from the borrower’s account or make other payment arrangements. However, lenders would likely have attempted to make contact with borrowers to obtain payment even in the absence of this requirement. The requirement will impose on lenders the cost of providing the notice. Lenders already need to track whether they can still attempt to collect payments directly from a borrower’s account, so identifying which borrowers should receive the notice should not impose any additional cost on lenders. The Bureau also expects that lenders normally attempt to contact borrowers in these circumstances in an attempt to identify other means of obtaining payment. If they are contacting the consumer via mail, the lender will be able to include the required notice in that mailing. 1425 The Bureau expects that lenders will incorporate the ability to provide this notice into their payment notification process. The Bureau estimates that vendors will charge $0.53 per notice sent via paper mail for lenders that send a large number of mailings and $1.00 per notice for lenders that send a small volume of mailing. For disclosures delivered through email, the Bureau estimates vendors will charge $0.01 to create and deliver each email such that it complies with the requirements of the rule. For disclosures delivered through text message, the Bureau estimates vendors will charge $0.08 to create and deliver each text message. The vendor will also need to provide either a PDF attachment of the full disclosure or a web page where the full disclosure linked to in the text message would be provided. The cost of providing this PDF attachment or web disclosure is included in the cost estimate of providing the text message. b. Benefits and Costs to Consumers Consumers will benefit from the notice because it will inform them that the lender cannot continue to collect payment directly from their account without their express permission. Absent this notice, borrowers may believe that they are obligated to re-authorize a lender to begin collecting directly from their account, when in many cases the borrower has the option to repay the loan through some other means that carries less risk of fees and provides the borrower with greater control over the timing and prioritization of their expenditures. Conversely, absent some communication from the lender, the borrower may not realize that payment can no longer be withdrawn and, as a result, fail to make payments on a loan. Some consumers may incur costs for notices sent by text. Consumers can avoid these costs by choosing email or paper delivery of the notices. The Bureau does not believe the required disclosures will impose any other costs on consumers. 1426 H. Benefits and Costs of the Rule to Covered Persons and Consumers – Recordkeeping The rule requires lenders to maintain sufficient records to demonstrate compliance with the rule. This includes, among other records, loan records; materials collected during the process of originating loans, including the information used to determine whether a borrower had the ability to repay the loan, if applicable; records of reporting loan information to a registered information system, as required; and, records of attempts to withdraw payments from borrowers accounts, and the outcomes of those attempts. 1. Benefits and Costs to Covered Persons The Bureau believes that some of the records that lenders are required to maintain would have already been maintained in the ordinary course of business. Given the very low cost of electronic storage, however, the Bureau did not believe that these new requirements would impose a meaningful new burden on lenders. However, a number of trade groups provided comments suggesting there are indeed costs associated with retaining these records. These comments note that lenders may incur some costs in developing a document retention policy, obtaining additional computer storage space to maintain the documents, programming the computer system to keep the documents for 36 months, training employees to comply with the recordkeeping requirements, and monitoring the implementation of these new procedures modify systems. The Bureau acknowledges these costs but believes them to be small. The development of retention policy should be straightforward, as the requirements are not opaque. Computer storage is inexpensive and even the largest lenders should not require more than one terabyte of additional storage to manage the retention of their files enterprise-wide (and that assumes their 1427 computer systems are already storage-constrained). As such, the Bureau estimates this cost to be less than $50 per lender if they wish to purchase additional storage themselves (e.g., a portable hard drive), or $10 per month if they wish to lease storage (e.g., from one of the many online cloud storage vendors). There may be a need to develop procedures and train staff to retain materials that they would not normally retain in the ordinary course of business, as well as design systems to generate and retain required records; those costs are included in earlier estimates of the costs of developing procedures, upgrading systems, and training staff. The Bureau also finds that maintaining the records will facilitate lenders’ ability to comply, and document their compliance, with other aspects of the rule. 2. Benefits and Costs to Consumers Consumers will benefit from the requirement to maintain records sufficient to demonstrate compliance because this will make compliance by lenders more likely, and facilitates enforcement of the rule, ensuring that consumers receive the benefits of the rule. I. Benefits and Costs of the Rule to Covered Persons and Consumers – Registered Information Systems As discussed above, the rule will generally require lenders to report covered loans to registered information systems in close to real time. Entities wishing to become registered information systems must apply to the Bureau to become registered. The process for becoming a registered information system prior to the effective date of § 1041.11 requires an entity to submit an application for preliminary approval with information sufficient to determine that the entity would be reasonably likely to satisfy the conditions to become a registered information system. These conditions include, among other things, that the entity possesses the technical capabilities 1428 to carry out the functions of a registered information system; that the entity has developed, implemented, and maintains a program reasonably designed to ensure compliance with all applicable Federal consumer financial laws; and that the entity has developed, implemented, and maintains a comprehensive information security program. If an entity obtains preliminary approval to become a registered information system from the Bureau, it will need to submit an application to be a registered information system that includes certain written third-party assessments contemplated by the rule. The rule also permits the Bureau to require an entity to submit to the Bureau additional information and documentation to facilitate determination of whether the entity satisfies the eligibility criteria to become a registered information system, or otherwise to assess whether registration of the entity will pose an unreasonable risk to consumers. On or after the effective date of § 1041.11, the rule contemplates a slightly different twostage process. Specifically, an entity can become provisionally registered by submitting an application that contains information and documentation sufficient to determine that the entity satisfies the conditions to become a registered information system, including the written thirdparty assessments contemplated by the rule. Lenders will be required to furnish information to a provisionally registered system, but a consumer report from such a system will not satisfy the lenders’ obligations under the rule to check borrowing history until a 240-day period from the date of provisional registration has expired, after which time the system will be deemed a fully registered information system. Once an entity is a registered information system under either process, the rule requires the entity to submit biennial assessments of its information security program. 1429 The Bureau expects that applicants to become registered information systems will be primarily, or exclusively, existing consumer reporting agencies. These entities have the technical capacity to receive data on consumer loans from a large number of entities and, in turn, deliver that data to a large number of entities. Depending on their current operations, some firms that wish to apply to become registered information systems may need to develop additional capabilities to satisfy the requirements of the rule. These requirements include that an entity possess the technical capability to receive specific information from lenders immediately upon furnishing, using reasonable data standards that facilitate the timely and accurate transmission and processing of information in a manner that does not impose unreasonable costs or burdens on lenders, as well as the technical capability to generate a consumer report containing all required information substantially simultaneous to receiving the information from a lender. Because firms currently operating as consumer reporting agencies must comply with applicable existing laws and regulations, including Federal consumer financial laws and the Standards for Safeguarding Customer Information, the Bureau also expects that they should already have programs in place to ensure such compliance. 1. Benefits and Costs to Covered Persons The rule will benefit firms that apply to become registered information systems by requiring lenders to furnish information regarding most covered loans to all registered information systems and to obtain a consumer report from a registered information system before originating most covered loans. The requirement to furnish information will provide registered information systems with data on borrowing of covered short-term and longer-term balloon payment loans. The requirement to obtain a consumer report before originating covered short1430 term and longer-term balloon-payment loans will ensure that there will be a market for these reports, which will provide a source of revenue for registered information systems. Registered systems will also be well-positioned to offer lenders supplemental services, for instance in providing assistance with determining consumers’ ability to repay. Any firm wishing to become a registered information system will need to incur the costs of applying to the Bureau. For some firms these costs may consist solely of compiling information about the firms’ practices, capabilities, and policies and procedures, all of which should be readily available, and obtaining the required third-party written assessments. Some firms may choose to invest in additional technological or compliance capabilities so as to be able to satisfy the requirements for registered information systems. Firms currently operating as consumer reporting agencies must comply with applicable existing laws and regulations, including Federal consumer financial laws and the Standards for Safeguarding Customer Information. As such, it is the Bureau’s expectation that these firms have programs in place to ensure such compliance. However, the independent assessments of these programs outlined in the rule may impose additional costs for some firms. Once approved, a registered information system will be required to submit biennial assessments of its information security program. Firms that already obtain independent assessments of their information security programs at least biennially, similar to those contemplated in the rule, will incur very limited additional costs. Firms that do not obtain biennial independent assessments similar to those contemplated in the rule will need to incur the cost of doing so, which may be substantial. 2. Benefits and Costs to Consumers 1431 The requirement that registered information systems have certain technical capabilities will ensure that the consumer reports that lenders obtain from these systems are sufficiently timely and accurate to achieve the consumer protections that are the goal of this part. This will benefit borrowers by facilitating compliance with the rule’s ability to repay requirements and the conditional exemption in § 1041.6 to the ability to repay requirements. J. Alternatives Considered In preparing the rule, the Bureau has considered a number of alternatives to the provisions. The alternatives discussed here are: ● Limits on re-borrowing covered short-term loans without an ability-to-repay requirement; ● An ability-to-repay requirement for short-term loans with no principal step-down approach; ● Disclosures as an alternative to the ability-to-repay requirement; and ● Limitations on withdrawing payments from borrowers’ accounts without such disclosures. In this section, the major alternatives are briefly described and their potential impacts relative to each provision are discussed. 1. Limits on Re-borrowing of Covered Short-Term Loans without an Ability-to-Repay Requirement The Bureau considered not imposing a requirement that lenders making covered shortterm and longer-term balloon-payment loans determine the ability of borrowers to repay the loans, and instead proposing solely to limit the number of times that a lender could make a 1432 covered short-term loan to a borrower. Such a restriction could take the form of either a limit on the number of loans that could be made in sequence or a limit on the number of loans that could be made in a certain period of time. The impacts of such an approach would depend on the specific limitation adopted. One approach the Bureau considered would have been to prevent a lender from making a covered short-term loan to a borrower if that loan would be the fourth covered short-term loan to the borrower in a sequence. A loan would be considered part of the same sequence as a prior loan if it were taken out within 30 days of when the prior loan were repaid or otherwise ceased to be outstanding. A limit on repeated lending of this type would have procedural costs similar to the principal step-down approach, and therefore lower than the ATR approach to making short-term loans. The Bureau simulated the effects of a “principal step-down approach only” policy. More specifically, the simulation assumed one possible implementation of this type of policy: a threeloan sequence cap, a six-loan annual cap, and a principal step-down requirement within each sequence. In this simulation, loan volumes and revenues decreased by 71-76 percent. Without an annual cap on loans, the impacts of this alternative on payday or vehicle title lender revenue would likely be less than the current rule. The ATR approach and the repeated lending limit both place a three-loan cap on loan sequences, but the ATR approach imposes the requirement that a lender not make a first loan without determining the borrower has the ability to repay the loan. The repeated lending limit without an annual cap on loans would likely also have less impact on payday lender revenue than would the principal step-down approach. The principal 1433 step-down approach limits loan sequences to no more than three loans, but, in addition, imposes loan size limitations and limits borrowers to no more than six loans in a year and no more than 90 days in debt per year on a covered short-term loan. While payday lenders could make loans using the ATR approach to borrowers who had reached the annual limits for loans issued via the principal step-down approach, the ATR approach will likely limit the total loans available to many consumers. The Bureau believes that limiting repeated lending should create stronger incentives to underwrite borrowers for ability to repay than exist in the current market. This is due to the reduction in expected revenue from loan sequences that would be cut off after the threshold is reached, rather than being able to continue for as long as the consumer is able to sustain rollover payments. However, a rule that relied solely on limiting repeat lending would increase the risk that borrowers take out loans that they would not have the ability to repay relative to the rule. This alternative would also lack the protections of the principal step-down approach, which include mandatory reductions in loan size across a sequence of loans. The Bureau believes that this step-down system will make it more likely that borrowers will successfully repay a loan or short loan sequence than would a limit on repeated lending, which might produce more defaults at the point that further re-borrowing would be prohibited. And, without the principal step-down approach’s limits on the number of loans per year and the limit on the time in debt, some borrowers might effectively continue their cycle of re-borrowing by returning as soon the 30-day period has ended. 2. An Ability-to-Repay Requirement for Short-Term Loans with No Principal Step-Down Approach 1434 The Bureau also considered the ATR approach without the principal step-down approach for covered short-term loans. Many consumer groups suggested this alternative. Without the principal step-down approach, lenders would be required to incur the expenses of the ATR approach for all payday loans. This effect, together with the impact of the ATR requirements, would have a larger impact on the total volume of payday loans that could be originated than would the rule. The Bureau simulated the effects of an “ATR approach only” policy, applying the same assumption that 33 percent of borrowers would qualify for an initial ATR loan (see part VII.F.1.c for more details on the Bureau’s simulations); and, as described in part VII.F.1.c, using various assumptions about how borrowers behave when the loan sequences are cut off. In this simulation, loan volumes and revenues decreased by 92 to 93 percent. Borrowers who could not demonstrate an ability to repay the loan would be unable to take out a payday loan. 3. Disclosures as an Alternative to the Ability-to-Repay Requirement The Bureau considered whether to require disclosures to borrowers warning of the risk of re-borrowing or default, rather than the ATR approach and the principal step-down approach, and the Bureau received a number of comments asserting that this approach would be sufficient or more advantageous, as discussed in the section-by-section analysis above. The Bureau believes that a disclosure-only approach would have lower procedural costs for lenders than would the ATR approach or the principal step-down approach. Requiring lenders to prepare disclosures that were customized to a particular loan would impose some additional cost over current practices. If lenders could simply provide standardized disclosures, that would impose almost no additional cost on lenders. 1435 A disclosure-only approach would also have substantially less impact on the volume of covered short-term lending. Evidence from a field trial of several disclosures designed specifically to warn of the risks of re-borrowing and the costs of re-borrowing showed that these disclosures had a marginal effect on the total volume of payday borrowing.1268 Analysis by the Bureau of similar disclosures implemented by the State of Texas showed a reduction in loan volume of 13 percent, consistent with the limited magnitude of the impacts from the field trial.1269 The Bureau believes that a disclosure-only approach would also have substantially less impact on the harms consumers experience from long sequences of payday and single-payment vehicle title loans. Given that loans in very long sequences make up well over half of all payday and single-payment vehicle title loans, a reduction of 13 percent in total lending has only a marginal impact on those harms. In addition, analysis by the Bureau of the impacts of the disclosures in Texas shows that the probability of re-borrowing on a payday loan declined by approximately 2 percent once the disclosure was put in place, indicating that high levels of reborrowing and long sequences of payday loans remain a significant source of consumer harm. A disclosure-only approach would also not change the lender’s incentives to encourage borrowers to take out long sequences of covered short-term loans. Given the evidence of unanticipated re-borrowing discussed above in Market Concerns— Underwriting, borrowers are likely to dismiss warnings of possible negative outcomes as not applying to them, and to not focus on disclosures of the possible harms associated with a 1268 Marianne Bertrand and Adair Morse, “Information Disclosure, Cognitive Biases and Payday Borrowing,” 66 J. of Fin. 1865 (2011). 1269 See CFPB Supplemental Findings, section 3. 1436 negative outcome that they do not anticipate experiencing. To the extent the borrowers have thought about the likelihood that they themselves will default on a loan, a general warning about how often people default is unlikely to cause them to revise their own expectations about the chances they themselves will default. Additionally, there is evidence that borrowers are generally aware of the average durations of sequences, but in spite of this are not good at predicting whether or not they themselves will experience a long duration.1270 As such, warnings about the potential for long durations are also unlikely to elicit changes in these borrowers’ behaviors. The Bureau received comments suggesting that the potential for disclosures to impact behavior in this market was not fully considered. They pointed to the research of Bertrand and Morse (cited above), to the Texas disclosure law (described and analyzed above), and for the finding that disclosures alerting borrowers to the availability of payment plans in certain States increase participation in said payment plans. While the Bureau believes disclosures can be effective in certain applications—especially when there is a market failure resulting in a lack of information about a more immediate and certain outcome—the available evidence suggests that a disclosure-only intervention in this market would yield substantially lower benefits to consumers than the ATR with principal step-down approach in the rule. The Bureau discusses this topic in the section-by-section analysis in part V as well. 4. Limitations on Withdrawing Payments from Borrowers’ Accounts without Disclosures 1270 See the discussion in Market Concerns—Underwriting and above in this section of Ronald Mann, “Assessing the Optimism of Payday Loan Borrowers,” 21 Sup. Ct. Econ. Rev. 105 (2013). 1437 The Bureau considered including the limitation on lenders continuing to attempt to withdraw payment from borrowers’ accounts after two sequential failed attempts to do so, but not including the required initial disclosure of usual payments or the additional disclosure in the event of unusual payments, or the notice that would be sent when a lender could no longer continue to attempt to collect payments from a borrower account. The impacts of excluding the upcoming payment notices would simply be to not cause lenders and borrowers to experience the benefits and costs that are described in the discussion of the impacts of those provisions. With regard to the notice that a lender could no longer attempt to withdraw payment from a borrower’s account, the primary effect would be analogous, and the benefits and costs are described in the discussion of the impacts of the provision that would require that notice. However, there may also have been a particular interaction if lenders had been prevented from continuing to attempt to withdraw payment from a borrower’s account, but the borrower did not receive a notice explaining that. Absent some communication from the lender, the borrower may not realize that payment would no longer be withdrawn and, as a result, fail to make payments on a loan. Lenders would presumably reach out to borrowers to avoid this eventuality. In addition, absent the notice, borrowers may have been more likely to believe that they are required to provide lenders with a new authorization to continue to withdraw payments directly from their accounts, when they may have been better off using some alternative method of payment. K. Potential Impact on Depository Creditors with $10 Billion or Less in Total Assets The Bureau believes that depository institutions and credit unions with less than 10 billion dollars in assets rarely originate loans that are covered by this rule. To the extent depository institutions do make loans in this market, many of those loans would be exempted 1438 under § 1041.3(e) or (f) as alternative or accommodation loans. L. Impact on Consumers in Rural Areas Consumers in rural areas will have a greater reduction in the availability of covered shortterm and longer-term balloon-payment loans originated through storefronts relative to consumers living in non-rural areas. As described in part VII.F.1.c, the Bureau estimates that the restrictions on making these loans will likely lead to a substantial contraction in the markets for storefront payday loans and storefront single-payment vehicle title loans.1271 The Bureau has analyzed how State laws in Colorado, Virginia, and Washington that led to significant contraction in the number of payday stores in those States affected the geographic availability of storefront payday loans in those States.1272 In those States, nearly all borrowers living in nonrural areas (defined as Metropolitan Statistical Areas or “MSA”) still had physical access to a payday store.1273 A substantial minority of borrowers living outside of MSAs, however, no longer had a payday store readily available following the contraction in the industry. In Colorado, Virginia, and Washington, 37 percent, 13 percent, and 30 percent of borrowers, respectively, would need to travel at least five additional miles to reach a store that remained open.1274 In Virginia, almost all borrowers had a store that remained open within 20 miles of their previous store.1275 And, in Washington 9 percent of borrowers would have to travel at least 20 additional miles.1276 While many borrowers who live outside of MSAs do travel that far to 1271 The Bureau reiterates that, given their limited prevalence, data on longer-term balloon-payment loans is scant. The effects on these types of loans are extrapolations from the empirical findings on short-term loans. 1272 CFPB Supplemental Findings. 1273 CFPB Supplemental Findings at 95 tbl. 17. 1274 CFPB Supplemental Findings. 1275 CFPB Supplemental Findings. 1276 CFPB Supplemental Findings. 1439 take out a payday loan, many do not,1277 and the additional travel distance resulting from closures of rural storefronts will impose a cost on these borrowers and may make borrowing from storefront lenders impractical or otherwise cause them to choose not to borrow from such lenders. Rural borrowers for whom visiting a storefront payday lender becomes impracticable retain the option to seek covered loans from online lenders, subject to the restrictions of State and local law. The Bureau has not been able to study a similar contraction in the single-payment vehicle title market, but expects that the relative impacts on rural and non-rural consumers will be similar to what has occurred in the payday market. That is, rural consumers are likely to experience a greater reduction in the physical availability of single-payment vehicle title loans made through storefronts than borrowers living in non-rural areas. The Bureau received numerous comments suggesting that the proposal’s consideration of rural borrowers was incomplete. However, the specific shortcoming cited was almost universally that rural borrowers displaced by the contraction in storefront lenders may not retain access via online lenders if they do not have access to the Internet. In assessing this, the Bureau notes that rural populations are less likely to have access to high-speed broadband compared to the overall population (39 percent vs 10 percent).1278 However, the bandwidth and speed required to access an online payday lender is minimal; even if high-speed access is currently beneficial to seeking an online loan, lenders can scale down the bandwidth requirements if the 1277 CFPB Supplemental Findings at 93 tbl. 15. Darrell M. West and Jack Karsten, “Rural and Urban America Divided by Broadband Access,” Brookings Institution, Techtank, July 18, 2016, available at https://www.brookings.edu/blog/techtank/2016/07/18/rural-andurban-america-divided-by-broadband-access/. 1278 1440 latent demand for loans amongst rural borrowers is sufficient to justify doing so. Additionally, the Bureau believes most potential borrowers in rural communities will likely be able to access the Internet by some means (e.g., dial up, or access at the public library or school). While the ease of access and quality of experience for bandwidth-limited rural customers may be lower than for non-rural customers, the Bureau believes that there will still be reasonable access for rural customers in need of loans. Additionally, mobile broadband access is growing rapidly in rural areas, with 67 percent of adults in these areas reporting they own a smartphone.1279 Additional commenters noted that some online payday lenders operate in rural areas, and that some comprise large shares of their local economies. If these lenders are amongst the number the Bureau expects to contract, this could impose a cost on these rural communities that would be avoided by more densely populated areas experiencing similar labor market shocks. However, if the cost advantages realized by lenders in rural areas (e.g., lower overhead, lower wages afforded by lower costs of living) give them a competitive advantage over online lenders in more densely populated areas, they may be less likely to contract. However, the Bureau acknowledges that at least some rural lenders will be substantially impacted by the rule. Given the available evidence, the Bureau believes that, other than the greater reduction in the physical availability of covered short-term loans made through storefronts, a potentially small relative reduction in access to any covered short-term loans, and the risk of negative labor market shocks to some rural areas in which online lenders comprise a significant share of employment, consumers living in rural areas will not experience substantially different effects of the regulation than other consumers. 1279 Pew Research Center, “Digital gap between rural and nonrural America persists.” May 19, 2017. 1441 VIII. Regulatory Flexibility Analysis The Regulatory Flexibility Act (RFA) generally requires an agency to conduct an Initial Regulatory Flexibility Analysis (IRFA) and a Final Regulatory Flexibility Analysis (FRFA) of any rule subject to notice-and-comment rulemaking requirements.1280 These analyses must “describe the impact of the proposed rule on small entities.”1281 An IRFA or FRFA is not required if the agency certifies that the proposal will not have a significant economic impact on a substantial number of small entities.1282 The Bureau also is subject to certain additional procedures under the RFA involving the convening of a panel to consult with small entity representatives prior to proposing a rule for which the IRFA is required.1283 A. Overview of the Bureau’s Approach In the proposal the Bureau did not certify that the proposal would not have a significant impact on a substantial number of small entities within the meaning of the RFA. Accordingly, the Bureau convened and chaired a Small Business Review Panel under the Small Business Regulatory Enforcement Fairness Act (SBREFA) to consider the impact of the rule on small entities that would be subject to the rule and to obtain feedback from representatives of such small entities. The Small Business Review Panel for the proposal is discussed in the SBREFA Report. The proposal also contained an IRFA pursuant to section 603 of the RFA, which among other things estimated the number of small entities that would be subject to the proposal. In this IRFA, the Bureau described the impact of the proposal on those entities, drawing on the 1280 5 U.S.C. 601, et seq. 5 U.S.C. 603(a). 1282 5 U.S.C. 605(b). 1283 5 U.S.C. 609. 1281 1442 proposal’s Section 1022(b)(2) Analysis. The Bureau also solicited comments on any costs, recordkeeping requirements, compliance requirements, or changes in operating procedures arising from the application of the proposal to small businesses; comments regarding any Federal rules that would duplicate, overlap, or conflict with the proposal; and comments on alternative means of compliance for small entities. Comments that addressed the impact on small entities are discussed below. Many of these comments implicated individual provisions of the final rule or the Bureau’s Section 1022(b)(2) Analysis and are also addressed in those parts. Similar to its approach in the proposal, the Bureau is not certifying that the final rule will not have a significant economic impact on a substantial number of small entities. Instead, the Bureau has completed a FRFA as detailed below. Section 604(a) of the RFA sets forth the required elements of the FRFA. Section 604(a)(1) requires the FRFA to contain a statement of the need for, and objectives of, the rule.1284 Section 604(a)(2) requires a statement of the significant issues raised by the public comments in response to the IRFA, a statement of the assessment of the Bureau of such issues, and a statement of any changes made in the proposed rules as a result of such comments.1285 Section 604(a)(3) requires the response of the Bureau to any comments filed by the Chief Counsel for Advocacy of the Small Business Administration in response to the proposed rule, and a detailed statement of any change made to the proposed rule in the final rule as a result of the comments.1286 The FRFA further must contain a description of and, where feasible, provide 1284 5 U.S.C. 604(a)(1). 5 U.S.C. 604(a)(2). 1286 5 U.S.C. 604(a)(3). 1285 1443 an estimate of the number of small entities to which the final rule will apply.1287 Section 604(a)(5) requires a description of the projected reporting, recordkeeping, and other compliance requirements of the rule, including an estimate of the classes of small entities that will be subject to the requirement and the types of professional skills necessary for the preparation of the report or record.1288 Finally, section 604(a)(6) requires a description of the steps the Bureau has taken to minimize the significant economic impact on small entities consistent with the stated objectives of applicable statutes, including a statement of the factual, policy, and legal reasons for selecting the alternative adopted in the final rule and why each one of the other significant alternatives to the rule considered by the agency which affect the impact on small entities was rejected; and a description of the steps the agency has taken to minimize any additional cost of credit for small entities.1289 B. Rationale and Objectives of the Final Rule As discussed in Market Concerns—Underwriting and Market Concerns—Payments above, the Bureau is concerned that practices in the market for payday, vehicle title, longer-term balloon-payment loans, and certain other longer-term loans utilizing leveraged payment mechanisms pose significant risk of harm to consumers. In particular, the Bureau is concerned about the harmful impacts on consumers of the practice of making these loans without making a reasonable determination that the consumer has the ability to repay the loan while paying for major financial obligations and basic living expenses. In addition, the Bureau is concerned that lenders in this market are using their ability to initiate payment withdrawals from consumers’ 1287 5 U.S.C. 604(a)(4). 5 U.S.C. 604(a)(5). 1289 5 U.S.C. 604(a)(6). 1288 1444 accounts in ways that harm consumers. To address these concerns, the Bureau is issuing the final rule pursuant to its authority under the Dodd-Frank Act in order to identify certain unfair and abusive acts or practices in connection with certain consumer credit transactions, to set forth requirements for preventing such acts or practices, to exempt loans meeting certain conditions from those requirements, to prescribe requirements to ensure that the features of those consumer credit transactions are fully, accurately, and effectively disclosed to consumers, and to prescribe processes and criteria for registration of information systems. The legal basis for the rule is discussed in detail in the legal authority analysis in part IV and in the section-by-section analysis in part V. 1. Public Comments on the IRFA and the Bureau’s Views and Treatment of Those Comments In accordance with section 603(a) of the RFA, the Bureau prepared an IRFA. In the IRFA, the Bureau estimated the possible costs for small entities with respect to the reporting, recordkeeping, and compliance requirements of the proposed rule against a pre-statute baseline. The Bureau requested comment on the IRFA. A number of comments specifically addressed the IRFA or raised concerns regarding the burden of compliance with the rule for small entities. These comments are discussed first. Those comments that repeated the same issues raised by the Office of Advocacy of the U.S. Small Business Administration are addressed in the next section of the FRFA, below. While many additional comments referred to economic impacts affecting all entities, this FRFA discussion focuses on comments addressing impacts that are particular to or differential for small entities, supplementing the discussion in the section-by-section analysis in part V, and the consideration of the broader impacts in the Section 1022(b)(2) Analysis in part VII. 1445 The significant comments addressing the IRFA or compliance burdens for small entities raised specific concerns falling into one of the following general categories: anticipated direct costs to small entities unaccounted for or unquantified in the IRFA; direct costs to small entities accounted for but underestimated; the lack of estimates for revenue losses specific to small entities; indirect effects on costs or prices faced by small entities not addressed; alternatives to the proposed rule which were not addressed or not appropriately considered; conflicts with existing laws and regulations not addressed; and categories of small entities not included in the analysis. a. Comments Asserting Anticipated Direct Costs to Small Entities Not Accounted for in the IRFA Commenters raised concerns about costs arising from several requirements of the rule which, they asserted, were unaccounted for or unquantified in the IRFA. First, commenters raised concerns that although the IRFA states that small entities may contract with attorneys, consultants, and vendors for assistance in complying with the ability-to-repay, disclosure, and reporting requirements of the rule, these costs were not made explicit. Related comments expressed concern that the need for small entities to contract with attorneys and vendors was in conflict with the Bureau’s statement that professional skills beyond those of existing employees would be required in only rare circumstances. The Bureau acknowledges that the need to contract with attorneys, consultants, and vendors may entail new costs for some small entities. For those small lenders which already maintain compliance processes for existing rules or regulations, the Bureau believes that the marginal added cost will be limited. In addition, some changes to the final rule which simplify 1446 the ability-to-repay verification and calculation requirements may lessen the need for these services. For those small entities that do not have relationships with these types of service providers under their current business process, the one-time costs may be larger. Second, commenters expressed concern that the costs associated with the 36 month recordkeeping requirement of the rule would be more substantial than the discussion in the IRFA implied. In the case of recordkeeping, Regulation Z, implementing TILA, has a general record retention rule that lenders “shall retain evidence of compliance” for “two years after the date disclosures are required to be made or action is required to be taken.”1290 In addition, as discussed in greater detail in the Background section, a number of States (including Colorado, Texas, Virginia, and Washington) have record retention requirements specific to payday loans, and numerous others have payday lending requirements which implicitly require some form of recordkeeping for compliance. Thus, the Bureau believes the 36 month recordkeeping requirement constitutes only an adjustment or extension of existing processes, with limited costs. Still, commenters noted that lenders may incur some costs in developing a document retention policy, obtaining additional computer storage space to maintain the documents, programming the computer system to keep the documents for 36 months (and then delete them), training employees to comply with the recordkeeping requirements, and monitoring the implementation of these new procedures. The Bureau acknowledges these costs but believes them to be small. The development of retention policies should be straightforward, as the requirements are not opaque. Computer storage is inexpensive and even the largest lenders should not require more than one terabyte of additional storage to manage the retention of their 1290 Regulation Z, 12 CFR 1026.25(a). 1447 files enterprise-wide (and that assumes their computer systems are already storage-constrained). As such, the Bureau estimates this cost to be less than $50 per lender if they wish to purchase additional storage themselves (e.g., a portable hard drive) with any associated operations and maintenance costs, or $10 per month if they wish to lease storage (e.g., from one of the many online cloud storage vendors). There may be a need to develop procedures and train staff to retain materials that they would not normally retain in the ordinary course of business, as well as design systems to generate and retain the required records; those costs are included in earlier estimates of the costs of developing procedures, upgrading systems, and training staff. The Bureau also believes that maintaining the records will facilitate lenders’ ability to comply with, and to document their compliance with, other aspects of the rule. Third, commenters stated that tracking failed payment withdrawals would require new systems and procedures to be developed, at a cost not specified in the IRFA. While the Bureau acknowledges that some entities may face costs in modifying existing systems to comply with the recordkeeping and payment processing requirements of the rule, these requirements largely build on processes required by existing laws or necessitated by standard business practice. b. Comments Asserting That Direct Costs to Small Entities Were Underestimated Commenters raised concerns that, among the costs to small entities quantified in the IRFA, some of the Bureau’s estimates of required time and financial costs were too low. Comments stated that compliance with the ability-to-repay requirements would be more costly and take employees longer than the Bureau had estimated. In particular, comments from industry trade associations and others asserted that the complexity of the proposed rule meant 1448 that verification and documentation of evidence for the ability-to-repay calculations would take longer than the Bureau’s estimate of three to five minutes. Similarly, the commenters raised concerns that making the ability-to-repay determination would take longer than 15 to 20 minutes for manual decisions, and that the Bureau’s statement that automated decisions would take essentially no time neglected to account for the time required for employees to monitor and maintain the automated decision-making system. Based on a survey of community banks, one industry trade association stated that respondents anticipate three hours of processing time on average to complete ability-to-repay verification and determination. As discussed in the sectionby-section analysis for § 1041.5, part VII, and part VIII.C, in response to these concerns the Bureau has lessened the documentation requirements and simplified the calculations for the ability-to-repay determination in the following respects. First, if verification evidence for income is not reasonably available, lenders may reasonably rely on stated amounts for income. Second, if the verification evidence for major financial obligations (e.g., the borrower’s credit report) does not include a particular obligation, lenders reasonably may rely on the stated amount of such obligation. Third, lenders will not be required to perform a credit check if they have already done so in the past 90 days and the consumer has not recently triggered a cooling-off period following a three-loan sequence. Fourth, lenders can use either a residual income or debt-to-income ratio when making the abilityto-repay determination, and the income and expenses can be based on a snapshot of the relevant calendar month rather than a time period which depends on the length of the loan. Fifth, lenders are not required to track the timing of income receipts or payments on major financial obligations. Finally, the Bureau has eliminated the presumptions of unaffordability attached to 1449 the second and third loan in a sequence made under the ability-to-repay requirements, likely reducing the underwriting costs for these loans and increasing the number of consumers determined to have the ability to repay such a loan. While these changes should reduce small entities’ time costs for compliance with the ability-to-repay requirements, the Bureau has increased its estimate of the total time to conduct a manual ability-to-repay determination to 15-45 minutes. This estimate is consistent with comments received from a trade group representing covered lenders and information provided by Small Entity Representatives. Commenters also raised concerns that the Bureau’s time estimates for initial and periodic ongoing training estimates were too low. The Bureau has reviewed its assessment, and the broader set of comments, and has concluded that the training estimates laid out were reasonable. The Bureau has clarified that the training estimates are per employee engaged in the relevant business process. Across a number of business processes, commenters raised concerns that the Bureau’s estimates for the one-time costs to update policies, systems, and materials were underestimated. Regarding the disclosure requirements of the proposed rule, commenters stated that the time and costs to develop and ensure disclosures are accurate was underestimated. Similarly, commenters also stated that the estimated one-time costs to update credit reporting systems were too low. Finally, commenters stated that the Bureau’s estimates of the costs to upgrade general computer systems -- separate from licensed underwriting, credit reporting, and disclosure systems -- were underestimated. The Bureau appreciates these comments, but believes its estimates, and the cost 1450 framework used throughout the rule, are accurate. Throughout the rule, the Bureau has updated its estimates when appropriate, as in the case of possible setup costs for furnishing to multiple registered information systems, and believes these changes and the corresponding discussions in part VII where the Section 1022(b)(2) Analysis address these concerns. c. Comments Asserting That the IRFA Did Not Estimate Lost Revenue for Smaller Entities In the proposed rule, the Bureau estimated the loss of revenue from the proposal (see for example the section in the proposed Section 1022(b)(2) Analysis on “Effect on Loan Volumes and Revenue From Underwriting Requirements and Restrictions on Certain Re-borrowing”). These costs, while not specifically estimated for small entities, were also referenced in the IRFA. Even assuming uniform compliance with the rule across large and small entities, the Bureau believes that the revenue impacts could differ between large and small entities. As noted below in more detail in the next section of this FRFA, the Bureau does not have data, and commenters provided only minimal evidence, that allow for the separate estimation of revenue impacts for small lenders. This issue is also discussed in part VII.F.1.c. d. Comments Asserting Additional Indirect Effects on Costs and Prices Commenters raised concerns regarding indirect costs and impacts on small entities resulting from the responses of lenders or other market participants to the rule. Several commenters stated that lenders themselves may face higher costs of obtaining credit due to the rule’s impact on their profitability. Commenters also noted that lenders would face adjustment costs if they were to shift their portfolio of products away from covered loans. Related comments stated that if lenders were to forgo leveraged payment mechanisms on new originations in response to the rule, loan defaults were likely to increase. One commenter raised 1451 the concern that a reduction in the total size of the market could require vendors and consultants for small entities to raise prices charged for services provided. Commenters raised concern over possible increased litigation risk for lenders. The Bureau appreciates these comments, and acknowledges that small lenders may face higher costs of credit, and that business practice adjustments would likely impact both the costs and revenues of these firms. Litigation risks and the pricing of vendor or consulting services could also change in response to the rule. While the exact form of these indirect costs is uncertain and the Bureau does not have the data available to estimate them, small lenders may face a relatively higher burden than larger lenders, given their smaller scale over which to spread fixed investments, and their potentially more limited access to financing options. These impacts are likely to be larger for small lenders that are highly specialized in short-term loans, or longerterm balloon-payment loans, or vehicle title loans not eligible for the exemption in § 1041.6, and smaller for those with more diversified product portfolios. e. Comments Asserting That Certain Alternatives Were Not Addressed or Appropriately Considered Regarding the IRFA, commenters expressed concern that the Bureau failed to provide a meaningful explanation for why it declined to pursue significant alternatives to the proposed rule. The IRFA included discussions of four significant alternatives to the proposed rule, which referred to more detailed analyses in the section-by-section discussions and the Section 1022(b)(2) Analysis. The Bureau believes its discussion of the alternatives provided in the IRFA, along with the alternatives considered in the proposal’s Section 1022(b)(2) Analysis, provided sufficient explanation for the choice of regulatory approach. However, in order to 1452 provide improved detail and clarity, part VIII.D below includes additional discussion in response to comments. The Bureau received a number of comments requesting exemptions for small entities. The Bureau is finalizing an exemption for accommodation loans, which are loans made by lenders that make fewer than 2,500 covered short-term loans and covered longer-term balloonpayment loans a year, and for which covered short-term loans and covered longer-term balloonpayment loans make up less than 10 percent of annual receipts. Additionally, the Bureau has adjusted its exemption for alternative loans to ensure that all PAL loans, and loans made by nonFederal credit unions which match the characteristic of a PAL loan, are exempt. This exemption should significantly reduce burden for smaller credit unions and other companies. Further, in response to comments the Bureau has substantially adjusted the rule in order to lessen the burdens of compliance, and also to reduce the degree to which the rule will impact total loan volumes, as noted above and in the section-by-section analysis for §§ 1041.5 and 1041.8. Even with these changes, there will still be a significant impact on small entities. The Bureau declines to completely exempt small entities because it believes many smaller entities, especially payday and vehicle title lenders, are engaging in the unfair and abusive practices identified in §§ 1041.4 and 1041.7. These practices cause substantial harm to consumers, and an exemption for small entities that would allow the practices to continue, albeit only at smaller companies, would substantially undermine the goals of this rule and permit a significant amount of consumer harm to continue. f. Comments Asserting That Conflicts with Existing Law Were Not Considered The IRFA requires identification, to the extent practicable, of all relevant Federal rules 1453 which may duplicate, overlap, or conflict with the proposed rule. Several trade association commenters raised concerns that the Bureau had not identified E-SIGN and ECOA/Regulation B as duplicate or overlapping rules. One comment stated that the proposed rule conflicts with E-SIGN and Regulation E because it adopts a different and new definition for consumer consent to receive electronic disclosures. The Bureau believes there is no conflict with E-SIGN because E-SIGN is not implicated by the consent process laid out in the rule. The Bureau decided not to use the E-Sign framework because of concerns raised in the SBREFA process about the burden of E-SIGN and the policy consideration of using an electronic disclosure consent process that is tailored to the small-dollar origination process and the situation the consumer is providing consent for. The Bureau also believes that the framework for obtaining consent for electronic notifications is more appropriate for the specific purposes of the notices in this rule. Another comment raised concerns about conflicts with EFTA, Regulation E, and Regulation CC. EFTA and Regulation E were discussed in the Market Concerns—Payments and section-by-section analysis for §§ 1041.7 and 1041.8. There are no provisions in EFTA, Regulation E, and Regulation CC that require or limit re-presentments of payments; those regulations do not conflict, duplicate or overlap with the limit on re-presentments. There are longstanding private network rules regarding repeat presentments that similarly do not raise conflicts. One comment stated that the proposed rule conflicts with ECOA because it does not permit lenders to consider household income or expenses in making an ability-to-repay determination. Similarly, another comment expressed concern that considerations in ECOA and Regulation B for co-habitation arrangements, including “spouses, cosigners, roommates, parents 1454 and adult children residing together, adult-children and elderly parents residing together,” do not fit neatly into the proposal’s documentation requirements for income, obligations, and living expenses. It also noted that “the consumer reporting and registered information systems do not address how such information is reported under those varying arrangements.” In the section-bysection analysis of § 1041.5, the Bureau discusses changes made to the ability-to-repay requirements of the final rule which now permits lenders to consider third party income to which a consumer has a reasonable expectation of access, to consider whether other persons are contributing towards the consumer’s payment of major financial obligations, and to consider whether other persons are contributing towards the consumer’s payment of basic living expenses when a lender chooses to itemize basic living expenses. As noted in the section-by-section analysis of § 1041.5 above, the Bureau believes that the requirements of the rule do not conflict with ECOA or Regulation B.1291 The Bureau also received comments suggesting that it had failed to consider the overlap between the proposal’s provisions relating to registered information systems and to lenders’ obligation to furnish to registered information systems, on the one hand, and the Fair Credit Reporting Act, Regulation V, the Gramm Leach Bliley Act, Regulation P or the Privacy Rule, and the Safeguards Rule, on the other hand. The commenter claimed that the Bureau had opened the door to numerous Regulation V issues relating to proper compliance with the duties of users and furnishers of information in registered information systems, and that the Bureau had not 1291 Under the RFA, rules are duplicative or overlapping if they are based on the same or similar reasons for the regulation, the same or similar regulatory goals, and if they regulate the same classes of industry. Rules are conflicting when they impose two conflicting regulatory requirements on the same classes of industry. The Bureau does not believe these standards are met in this case. 1455 considered legal issues around the privacy and data security of said data. Yet these laws do not conflict with the rule in any way. To the contrary they would all have the same effect as they are applicable, and they would operate to address the issues raised by the commenter here in the same manner that they do in other areas of the economy. g. Comments Asserting That Categories of Entities Were Not Included A small number of commenters raised concerns regarding the impacts of the proposed rule on Indian tribes, which the IRFA did not separately address. The Bureau did not specifically analyze effects on Indian tribes, as it does not consider them to be small entities under the RFA, consistent with the interpretation provided by the Small Business Administration’s comment. However, as many Tribal lenders may be small lenders, and many exist in rural areas, there is the potential for a more acute impact of the rule on Tribal lenders. This coincides with the impact on small and rural entities, and is therefore considered within the discussion of the impacts on those lenders. 2. Response to the Small Business Administration Chief Counsel for Advocacy The SBA Office of Advocacy (Advocacy) provided a formal comment letter to the Bureau in response to the proposed rule. Among other things, this letter expressed concern about the following issues: the burden of complying with the ability-to-repay requirements; the lack of estimates for the impact of the ability-to-repay requirements on lender revenues; the length of the cooling-off period; the lack of an exception for loans to address an emergency; the interaction of the rule with State laws; the impact of the rule on credit unions, small communities, and Indian tribes; the lack of clarity of the business loan exemption; the effect of the rule on lender’s own cost of credit; and the implementation date of the final rule. 1456 Advocacy expressed concern that the ability-to-repay requirements in the proposed rule would be burdensome. The proposed rule would have required lenders to verify a consumer’s net income, debt obligations, and housing expenses; project basic living expenses, net income, and obligations for a time period based on the term of the loan; and use this information to calculate the consumer’s ability to repay the loan. Advocacy expressed concern that these requirements were complicated and extensive, turning an uncomplicated product into a complex product. Advocacy also expressed concern that many customers may not qualify for loans under the ability to repay requirements, particularly in small rural communities where lenders contend that lending is relationship based. Advocacy encouraged the Bureau to eliminate some of the ability-to-repay requirements, and suggested eliminating the credit check requirement as one possibility. In response to comments from Advocacy and the public, the Bureau has made changes to the ability-to-repay requirements to reduce compliance costs for small entities of both obtaining evidence and making the ability-to-repay determination. For example, if verification evidence for income is not reasonably available, lenders may reasonably rely on stated amounts for income. Additionally, verification evidence is no longer required for rental housing expenses. The Bureau estimates that these changes will reduce the time and expense of obtaining the information required to make an ability to repay determination, particularly for lenders serving customers with income or expenses that are difficult to document. And while the Bureau believes that the credit check requirement is necessary to properly project a consumer’s debt obligations, lenders will not be required to perform a credit check if they have already done so in the past 90 days and the consumer has not recently triggered a cooling-off period following a 1457 three-loan sequence. This change maintains the integrity of the ability to repay requirements, while eliminating some marginal costs that both Advocacy and the Bureau suggest are higher for small lenders compared to larger lenders. Additional changes were made to final rule to reduce the burden of making the ability-torepay determination. Lenders can use either residual income or debt-to-income ratio when making the ability-to-repay determination, and the income and expenses can be based on a snapshot of the relevant calendar month rather than a time period which depends on the length of the loan. The Bureau expects these changes to ease implementation of the ability-to-repay requirement, particularly for smaller lenders who have less scale over which to recoup their fixed investment in compliance requirements. Finally, the Bureau has eliminated the presumptions of unaffordability attached to the second and third loan in a sequence made under the ability-torepay requirements, likely reducing the underwriting costs for these loans and increasing the number of consumers determined to have the ability to repay such a loan. In addition to compliance burdens, Advocacy expressed concern that the IRFA did not provide separate estimates of the impact of the ability-to-repay requirements, or the proposed rule as a whole, on revenue for small entities. The Bureau does not have data that allow for the separate estimation of revenue impacts for small lenders. However, even assuming uniform compliance with the rule across large and small entities, the Bureau believes that the revenue impacts could differ between large and small entities. This possibility is discussed in part VII.F.1.c. However, that discussion is based on economic theory and reasoning, as the Bureau lacks the data required to differentiate the potential impacts on small and large lenders. 1458 In contrast, two studies of loan-level data cited by commenters suggest the impacts on revenue may be similar for small and large entities.1292 The studies separately simulated the effects of the proposed rule on a dataset of loans made by small lenders and on a dataset of loans made by large lenders, estimating total revenue reductions of 82% and 83% respectively. As described earlier, the Bureau’s updated estimates in the Section 1022(b)(2) Analysis in part VII.F.1.c indicate smaller reductions in revenue from the final rule relative to the proposed rule; however, the Bureau is not able to differentiate the impacts for smaller entities. As a result, the Bureau has no evidence to suggest the revenue impacts on small entities will exceed those on larger entities, but remains sympathetic to that possibility. While not directly addressing revenue impacts, data on market concentration before and after payday lending laws were implemented in Colorado suggest that overall impacts were larger for small lenders. Colorado implemented its payday lending laws in 2010, and the share of storefront locations operated by the ten largest companies increased from 64% to 78% between 2009 and 2011.1293 Note that the provisions and market context of the Colorado law differ from those in this rule. Beyond the ability-to-repay requirements, Advocacy stated that the 30-day cooling-off period for re-borrowing will harm small businesses. As a result of the SBREFA panel, the 1292 Arthur Baines et al., “Economic Impact on Small Lenders of the Payday Lending Rules Under Consideration by the CFPB,” Charles River Associates (2015), available at http://www.crai.com/publication/economic-impact-smalllenders-payday-lending-rules-under-consideration-cfpb; Arthur Baines et al., “Economic Impact on Storefront Lenders of the Payday Lending Rules Proposed by the CFPB,” Charles River Associates (2016), available at http://www.crai.com/publication/economic-impact-storefront-lenders-payday-lending-rules-proposed-cfpb. Note that these estimates assume lenders use the principal step-down approach, rather than ability-to-repay, due to data limitations. 1293 See Adm’r of the Colo. Consumer Credit Unit, “Colorado Payday Lending—July Demographic and Statistical Information: July 2000 through December 2009,”; Adm’r of the Colo. Consumer Credit Unit, “Colorado Payday Lending—July Demographic and Statistical Information: July 2000 through December 2011,”; Adm’r of the Colo. Consumer Credit Unit, “Colorado Uniform Consumer Credit Code: Annual Report Composites,” available at https://coag.gov/uccc/info/ar. 1459 Bureau reduced the cooling-off period from 60 to 30 days, for which Advocacy expressed appreciation. However, Advocacy asserted that the size of the revenue reductions estimated by the Bureau may be detrimental to small entities, and encouraged the Bureau to consider a shorter cooling-off period. Additionally, Advocacy noted that consumers may have bills due more frequently than monthly, in which case the 30-day cooling-off period may prevent the consumer from obtaining funds to meet these needs. While the Bureau considered a range of cooling-off periods in the rulemaking process, the 30-day period was chosen, consistent with the re-borrowing period described in the sectionby-section analysis above, so that borrowers must go a full billing cycle across all their liabilities before being permitted to take out another loan. This aligns the rule with the idea that short-term loans are intended to cover unexpected and temporary financial shocks, rather than persistent income deficits relative to expenses. See the section-by-section analysis for §§ 1041.4 and 1041.5 for more details. Advocacy encouraged the Bureau to provide an exemption for consumers who have experienced and unexpected emergency, and to provide clear guidance on what qualifies as an emergency. The Bureau has not created an exception for consumers who have experienced an emergency, as defining an emergency in such a way that does not allow broader evasion of the rule’s requirements was not feasible. The Bureau believes that the alternatives to the ability-torepay requirements present in the rule will make credit available to these consumers enduring unusual and nonrecurring expenses or drops in income. Specifically, the Bureau expects a consumer will be able to obtain no less than six loans in a 12-month period, without needing to 1460 satisfy any ability to repay requirements. The Bureau further expects this will be sufficient to address the vast majority of discrete events, such as emergencies and/or unexpected shocks to a consumer’s income or expenses. This issue was discussed in greater depth above in Market Concerns—Underwriting. Advocacy noted that many States have addressed the issue of payday loans through their own lawmaking. Small entities in States with existing payday lending laws have already made changes to their practices to comply with these laws. Advocacy encouraged the Bureau to recognize the States’ ability to make the appropriate choices for their citizens and exempt from the rule small businesses that operate in States that currently have payday lending laws. The Bureau has considered how this rule will interact with the existing State payday lending laws, which are discussed in greater detail in part II and part VII.C. Given the varying stringency of State payday lending laws, the Bureau has found evidence of harm to consumers even in States with these laws, as discussed earlier. As such, the Bureau believes that State exemptions would be inconsistent with the objectives of the rule. As noted earlier, for those lenders in States with stricter limits on lending, lenders will experience relatively low compliance costs and smaller impacts from the rule, as the rule will be relatively less binding on them. Advocacy raised concerns that the Bureau had underestimated the rule’s impact on small credit unions. In particular, Advocacy expressed concerns over the minimum length required for loans made by credit unions, under the PAL program administered by the NCUA. The proposed rule required loans made under the alternative PAL approach to be at least 46 days in length, while NCUA requires a minimum length of only 30 days. Advocacy also raised concerns that 1461 the all-in APR calculation required by the proposal may require credit unions to perform additional calculations to populate new forms, disclosures, compliance training, and other resources. Advocacy encouraged the Bureau to recognize the NCUA’s expertise in the area of credit unions and exempt small credit unions from the proposed rule. While the Bureau believes that exempting small credit unions entirely would be inconsistent with the objectives of the rule, several changes have been made to the final rule to address the concerns and burden for small credit unions. First, the Bureau has lowered the minimum length of a loan made under the PAL Approach to 30 days, bringing the requirements into alignment with those of NCUA. In addition, the Bureau has added a safe harbor to any loans made by Federal credit unions in compliance with the PAL program as set forth by NCUA. Finally, the Bureau has added an exemption for entities offering loans on an accommodation basis that would otherwise be covered loans, as evidenced by the volume of such loans that an entity makes in absolute terms and relative to its overall business. The Bureau believes that most small credit unions will fall within this exemption. Thus the compliance costs of the rule will be significantly reduced for small credit unions, as well as other small entities, which make loans that follow the PAL Approach. Advocacy expressed concern about the impact of the rule on small rural communities and Tribal businesses and communities. Consumers in rural communities may have fewer options for accessing credit than consumers in more populated areas. Advocacy also stated that consolidation of lenders will be more difficult in these areas, and the resulting long distances between lenders may further reduce credit access. Advocacy relayed the concerns of Tribal representatives regarding the impact of the rule on their communities, many of which are 1462 economically disadvantaged. Advocacy encouraged the Bureau to consider the detrimental effects that the proposed rule may have on small rural communities, and to work with federally recognized Indian tribes to resolve the issue of Tribal consultation and Tribal sovereignty. The Bureau acknowledges that the effects of the rule may be felt differentially in communities depending on their population density, density of lenders, income, and wealth. Specifically, the Bureau considered the impact of consolidation by estimating the additional distance a rural customer may have to travel after this rule in part VII.F.2.b.v and part VII.L. Regarding the specific effects on small lenders, the Bureau believes that the changes made in the final rule described above will mitigate some of the burden associated with compliance in rural or Tribal areas. Advocacy thanked the Bureau for clarifying that the proposed rule would not apply to business loans, and encouraged the Bureau to provide clear guidance on what qualifies as a small business loan. Advocacy stated that some small businesses do use payday loan products to finance their businesses, and this source of financing is important to their operations. Advocacy raised concerns that even with clear guidance, sources of credit for small businesses may be reduced if a large percentage of payday lenders cease operating due to the rule. In addition, Advocacy noted that if the rule affects the revenue stream of payday lenders, those lenders themselves may face higher costs of credit. Advocacy encouraged the Bureau to perform a full analysis of the impact that this rulemaking may have on the cost of credit for small entities as required by the RFA. The Bureau’s rule is not intended to effect business loans, and the definitions of covered loans reflect this fact. Only loans extended to a consumer primarily for personal, family, or 1463 household purposes are covered by the rule. The Bureau appreciates the concern for a possible reduction in business loan availability due to lender exit, and acknowledges that those business relying on products offered by payday lenders may have to travel further to obtain credit, or seek credit from alternative sources. (e.g., online lenders). Regarding the potentially higher cost of credit to payday lenders themselves, Advocacy’s point is well taken. The Bureau’s analysis has focused on estimating the direct effects of the rule, as the indirect effects rely heavily on lender’s responses to the rule, and the Bureau does not have data which could be used to quantify these effects. Finally, Advocacy encouraged the Bureau to allow at least 24 months for small entities to comply with the rule, in part because small entities have undergone a number of other regulatory changes, including due to the implementation of State lending laws and the Military Lending Act. The Bureau appreciates the concern regarding the required adjustments to small entities operations, and has increased the effective date to 21 months. The Bureau believes this is a sufficient period for compliance with the final rule. C. Effect of the Rule on Small Entities 1. Description and Estimate of the Number of Small Entities to Which the Final Rule Will Apply As discussed in the Small Business Review Panel Report, for purposes of assessing the impacts of the rule on small entities, “small entities” is defined in the RFA to include small businesses, small nonprofit organizations, and small government jurisdictions.1294 A “small business” is determined by application of SBA regulations and reference to the North American 1294 5 U.S.C. 601(6). 1464 Industry Classification System (NAICS) classifications and size standards.1295 Under such standards, banks and other depository institutions are considered “small” if they have $550 million or less in assets, and for most other financial businesses, the threshold is average annual receipts (i.e., annual revenues) that do not exceed $38.5 million.1296 During the SBREFA process, the Bureau identified four categories of small entities that may be subject to the proposed rule for purposes of the RFA. The categories and the SBA small entity thresholds for those categories are: (1) commercial banks, savings associations, and credit unions with up to $550 million in assets; (2) nondepository institutions engaged in consumer lending or credit intermediation activities with up to $38.5 million in annual revenue; (3) nondepository institutions engaged in other activities related to credit intermediation activities with up to $20.5 million in annual revenue; and (4) mortgage and non-mortgage loan brokers with up to $7.5 million in annual revenue. The following table provides the Bureau’s revised estimates of the number and types of entities that may be affected by the rule:1297 1295 5 U.S.C. 601(3). The current SBA size standards are found on SBA’s website at http://www.sba.gov/content/table-small-business-size-standards. 1296 5 U.S.C. 601(3). 1297 In the Small Business Review Panel Report at Chapter 9.1, a preliminary estimate of affected entities and small entities was included in a similar format (a chart with clarifying notes). See Small Business Review Panel Report, at 26 tbl. 9.1.1, 27 tbl. 9.1.2. 1465 Table 1: Estimated Number and Types of Affected Entities and Small Entities by NAICS lE'ode Estimated Estimated Industry Code Small Entity Number of Number of Threshold Total Entities Small Entities Commercial Banks, Savings Institutions, and Credit 522110; $550 million in - a 522120; assets 12,256 10,44? mm? 522130 Institutions Engaged in Conmmier $33.5 million in Lending or Credit annual revenues I I 1' Ion Aetieitiesh 522290 5,523 5,403 Institutions Engaged in Other Aetivities $20.5 million in Related to lCredit annual revenues I I 1' Ion Aetis'itiesh 522390 4,201 4,549 2.5 million in Mortgage and?jlon- Mortgage ual revenues Loan 522310 2,00? 45,01? $30.5 million in name 2 I en dingh annual revenues 522291 3,206 3,130 a Total number of entities and small entities 1I:II.ras estimated based ontiie 2'01]r lBall Report. ofU.S. Businesses for 2012- Total meher ofentities and small entities was estimated based on to Census Bmeau?s Statistics 1466 As discussed in the Small Business Review Panel Report, the NAICS categories are likely to include firms that do not extend credit that will be covered by the rule. In addition, some of these firms may qualify for exemptions under the rule. The following Table 2 provides the Bureau’s estimates, not accounting for exemptions, of the numbers and types of small entities within particular segments of primary industries that may be affected by the rule: 1467 Table 1: Estimated Number and Types nf Affected Small Entities by Industry Categnry Estinute? Number Hams Industry Hares small Entitv ramall Entities rlu-aallulu a $2115 millien in annual $th Payday m? 5 52255111 revemre 2,21s slulaauul I-?ayday lenders Uperating $2-5 millien in annual - as Hrskersa 5223?] revemre 229' I: $33.5 million in annual $th mum Lend? 5222511 revemre 15?? . . :1 $535 millien in annual Stme?'ent 1ifeh1ele Tltle Lenders 522293 rev 312 $2115 millien er 5s_5 - :1 5222113; millien in annual lenders - 52255111 revemre 124 Credit ?Dime 5221311 $5511 millienin assets 145115 . 5221 111; Banks Ihn?s 5221211 $5511 millienin assets usual The mnnlier at small stnre?'ent payday lenders is estimated using licensee ?'um State regulators, ?rm revenue infmmatien ?'em public ?lings and nun?public snurees? and, for a mnnlier nf States, industry market research relying untelepliene directery listings. State supplementedhy leeatinn infmmatien prepared by Steven Graves and Christopher Petersen vailable at Flap'a'i?llww. urns. tnfUE _pdl_addr.1rl5. Based on mese stunner:1 are 2,256 stme?'ent payday lenders in the United States- Based en mepnliliely? vailable revemre information, at least 33 ef the ?lnn have revenue alieve me small entity est {time remaining ?rms eperate a very small Therefore, while slime ef ?rms nlitlmut publicly available inferma?en may have revenue abnve the small entity tlireslielui, the interest efheing inclusive all are tn be small entities. 1468 everall erefreut installment lenders. Jehu Hecht; ?Altemalive Financial Serviees: lnnevalingte Meet emerNeedsinanEvelving Regulaterv Franhewe?t,? {2014} (Stephens, Inc-, slidepresentatien} en A. report estimated te sterefrentinstallment lender lecaliens. ased eupuhliclv?availahle infernlatien, SE efthe largest ?nns have revenue aheve efpeten?aleetiinates sterel'reut instalhnentlen?ers eriginste ecvered leans. tate ?nancialregulators ?lings andfremueu?puhlic scanees. PredaterPraclice-s: 'I'heState (Ctr. urRespensihleLending??lE} milu?issr Based cesthereareappmsimatelyE-i sterefreutvehiclehtle lendersintheUnitedStates- Baseden revenue infennatilen1 at least 3D cf the ?nns have revenue aheve the mall entity ?lreeheli Thueferewhilesemeef ?nnswi?icanpuhlicly available nenPIiinel?l, Report 1: Pro?ling Intemet Small-Della Lending? Basic Demographics and Lean Characteristiee1 at 3 ellsr-lending-Final.pdf. 'I'heestiniatesfer assumedeuet u?gimteceveredleans 1469 2. Projected Reporting, Recordkeeping, and other Compliance Requirements of the Rule The rule imposes new reporting, recordkeeping, and compliance requirements on certain small entities. These requirements and the costs associated with them are discussed below. a. Reporting Requirements and Their Costs for Small Entities The rule imposes new reporting requirements to ensure that lenders making covered short-term and longer-term balloon-payment loans under the rule have access to timely and reasonably comprehensive information about a consumer’s current and recent borrowing history with other lenders, as discussed in the section-by-section analysis for § 1041.10. This section discusses these reporting requirements and their associated costs on small entities. Lenders making covered short-term or longer-term balloon-payment loans are required to furnish information about those loans to all information systems that have been registered with the Bureau for 180 days or more, have been provisionally registered with the Bureau for 180 days or more, or have subsequently become registered after being provisionally registered (generally referred to here as registered information systems). At loan consummation, the information furnished needs to include identifying information about the borrower, the type of loan, the loan consummation date, the principal amount borrowed or credit limit (for certain loans), and the payment due dates and amounts. While a loan is outstanding, lenders need to furnish any update to information previously furnished pursuant to the rule within a reasonable period of time following the event prompting the update. And when a loan ceases to be an outstanding loan, lenders must furnish the date as of which the loan ceased to be outstanding and whether all amounts owed in connection with the loan were paid in full, including the amount financed, charges included in the cost of credit, and charges excluded from the cost of credit. Furnishing information to registered information systems will require small entities to 1470 incur one-time and ongoing costs. One-time costs include those associated with establishing a relationship with each registered information system and developing policies and procedures for furnishing the loan data.1298 Lenders using automated loan origination systems will likely modify those systems, or purchase upgrades to those systems, to incorporate the ability to furnish the required information to registered information systems.1299 The ongoing costs will be those of accurately furnishing the data.1300 Lenders with automated loan origination and servicing systems with the capacity to furnish the required data will have very low ongoing costs. Lenders that furnish information manually will likely do so through a web-based form, which the Bureau estimates will take three minutes to fill out for each loan at the time of consummation, when information is updated (as applicable), and when the loan ceases to be an outstanding loan. If multiple registered information systems exist, it may be necessary to incur this cost multiple times, unless there are services that furnish to all registered information systems on behalf of a lender.1301 The Bureau notes that some lenders in States where a private third-party operates a 1298 If multiple registered information systems exist, lenders may be able to contract with a third party to furnish to all registered information systems on their behalf. This third party may be one of the registered information systems, as they may provide this service to make them a more attractive option to lenders. 1299 Some software vendors that serve lenders that make payday and other loans have developed enhancements to enable these lenders to report loan information automatically to existing State reporting systems. 1300 The Bureau also received comments noting that lenders will have to incur additional costs associated with dispute resolution. One commenter specifically noted that consumers would dispute negative data contained on their reports which would require investigation along with company responses. The commenter cited a figure of $50,000 per year to handle these disputes and other costs of furnishing. The Bureau acknowledges there may be ancillary costs associated with such disputes, but believes that furnishing accurate data and compliance with the records management requirements should mitigate the costs associated with dispute resolutions (e.g. confirming the existence of the loan and any payments made). Additionally, many of the costs associated are expected to be borne by registered information systems, as the FCRA allows consumers to dispute information directly with the consumer reporting agency. As such, the $50,000 figure cited by the commenter seems inflated. Instead, the Bureau believes the costs associated with these activities are included in the ongoing costs associated with furnishing to registered information systems. 1301 Should there be multiple registered information systems, the Bureau expects that one or more registered information systems or other third parties will offer to furnish information to all registered information systems on behalf of the lender. 1471 database on behalf of State regulators are already required to provide information similar to that required under the rule, albeit to a single entity; such lenders thus have experience complying with this type of requirement. Where possible, the Bureau will also encourage the development of common data standards for registered information systems in order to reduce the costs of providing data to multiple information systems. In addition to the costs of developing procedures for furnishing the specified information to registered information systems, lenders will also need to train their staff in those procedures. The Bureau estimates that lender personnel engaging in furnishing information will require approximately half an hour of initial training in carrying out the tasks described in this section and 15 minutes of periodic ongoing training per year. b. Recordkeeping Requirements and Their Costs for Small Entities The rule imposes new data retention requirements for the requirements to assess borrowers’ ability to repay and alternatives to the requirement to assess borrowers’ ability to repay for both short-term and longer-term balloon-payment loans by requiring lenders to maintain evidence of compliance in electronic tabular format for certain records. The retention period is 36 months, as discussed above in the section-by-section analysis for § 1041.12. The data retention requirement in the rule may result in costs to small entities. The Bureau believes that not all small lenders currently maintain data in an electronic tabular format. To comply with the record retention provisions, therefore, lenders originating short-term or longer-term balloon-payment loans may be required to reconfigure existing document production and retention systems. For small entities that maintain their own compliance systems and software, the Bureau does not believe that adding the capacity to maintain data in an electronic tabular format will impose a substantial burden. The Bureau believes that the primary cost will 1472 be one-time systems changes that could be accomplished at the same time that systems changes are carried out to comply with the provisions of §§ 1041.5 and 1041.6 of the rule. Similarly, small entities that rely on vendors will likely rely on vendor software and systems to comply in part with the data retention requirements. In addition to the costs described above, lenders will also need to train their staff in record retention procedures. The Bureau estimates that lender personnel engaging in recordkeeping will require approximately half an hour of initial training in carrying out the tasks described in this section and 15 minutes of periodic ongoing training per year. c. Compliance Requirements and Their Costs for Small Entities The analysis below discusses the costs of compliance for small entities of the following major provisions: (i) ability-to-repay requirements for covered short-term and longer-term balloon-payment loans, including the requirement to obtain a consumer report from a registered information system; and a conditional exemption providing an alternative to those specific underwriting criteria for short-term loans, including notices to consumers taking out loans originated under this alternative; and (ii) provisions relating to payment practices that limit continuing attempts to withdraw money from borrowers’ accounts after two consecutive failed attempts; and payment notice requirements. The discussions of the impacts are organized into the two main categories of provisions listed above—those relating to underwriting and those related to payments. Within each category, the discussion is organized to facilitate a clear and complete consideration of the impacts of these major provisions of the rule on small entities. In considering the potential impacts of the rule, the Bureau takes as the baseline for the analysis the regulatory regime that currently exists for the covered products and covered 1473 persons.1302 These include State laws and regulations; Federal laws, such as the MLA, FCRA, FDCPA, TILA, EFTA, ECOA, E-SIGN, and the regulations promulgated under those laws; and, with regard to depository institutions that make covered loans, the guidance and policy statements of those institutions’ prudential regulators.1303 The rule includes several exemptions, and in places it is useful to discuss their benefits, costs, and impacts relative to those of the core provisions of the proposed regulation. The baseline for evaluating the full potential benefits, costs, and impacts of the proposal, however, is the current regulatory regime as of the issuance of the proposal. The discussion here is confined to the direct costs to small entities of complying with the requirements of the rule. Other impacts, such as the impacts of limitations on loans that could be made under the rule, are discussed at length above. The Bureau believes that, except where otherwise noted, the impacts discussed there would apply to small entities. i. Underwriting for Covered Short-Term and Longer-Term Balloon-Payment Loans (a). Requirement to Assess Borrowers’ Ability To Repay The rule will require that lenders determine that applicants for short-term and longer-term balloon-payment loans have the ability to repay the loan while still meeting their major financial obligations and paying basic living expenses. For purposes of this discussion, the practice of making loans after determining that the borrower has the ability to repay the loan will be referred 1302 The Bureau has discretion in each rulemaking to choose the relevant provisions to discuss and to choose the most appropriate baseline for that particular rulemaking. 1303 See, e.g., FDIC, Fin. Institution Letter FIL–14–2005, “Payday Lending Programs: Revised Examination Guidance,” (Revisd 2015), available at https://www.fdic.gov/news/news/financial/2005/fil1405.pdf; OCC, Guidance on Supervisory Concerns and Expectations Regarding Deposit Advance Product, 78 FR 70624 (Nov. 26, 2013); Guidance on Supervisory Concerns and Expectations Regarding Deposit Advance Products, 78 FR 70552 (Nov. 26, 2013). 1474 to as the “ATR approach.” Lenders making loans using the ATR approach will need to comply with several procedural requirements when originating loans. The Bureau’s assessment of the benefits, costs, and other relevant impacts on small entities of these procedural requirements are discussed below. The Bureau believes that many lenders use automated systems when underwriting loans and will modify those systems, or purchase upgrades to those systems, to incorporate many of the procedural requirements of the ATR approach. The costs of modifying such a system or purchasing an upgrade are discussed below, in the discussion of the costs of developing procedures, upgrading systems, and training staff. (1). Consulting Lender’s Own Records and Costs to Small Entities Under the rule, lenders will need to consult their own records and the records of their affiliates to determine whether the borrower had taken out any prior short-term loans or longerterm balloon-payment loans that were still outstanding or were repaid within the prior 30 days. To do so, a lender will need a system for recording loans that can be identified as being made to a particular consumer and a method of reliably accessing those records. The Bureau has concluded that lenders will most likely comply with this requirement by using computerized recordkeeping. A lender operating a single storefront will need a system of recording the loans made from that storefront and accessing those loans by consumer. A lender operating multiple storefronts or multiple affiliates will need a centralized set of records or a way of accessing the records of all of the storefronts or affiliates. A lender operating solely online will presumably maintain a single set of records; if it maintained multiple sets of records, it will need a way to access each set of records. The Bureau believes that most small entities already have the ability to comply with this 1475 provision, with the possible exception of those with affiliates that are run as separate operations. Lenders’ own business needs likely lead them to have this capacity. Lenders need to be able to track loans in order to service the loans. In addition, lenders need to track the borrowing and repayment behavior of individual consumers to reduce their credit risk, such as by avoiding lending to a consumer who has defaulted on a prior loan. And most States that allow payday lending have requirements that implicitly require lenders to have the ability to check their records for prior loans to a loan applicant, including limitations on renewals or rollovers or cooling-off periods between loans. Despite these various considerations, however, there may be some lenders that currently do not have the capacity to comply with this requirement. Small entities that do not already have a records system in place will need to incur a onetime cost of developing such a system, which may require investment in information technology hardware and/or software. Lenders may instead contract with a vendor to supply part or all of the systems and training needs. As noted above, the Bureau believes that many lenders use automated loan origination systems and will modify those systems or purchase upgrades to those systems such that they would automatically access the lender’s own records. For lenders that access their records manually, rather than through an automated origination system, the Bureau estimates that accessing and utilizing these records in the ATR determination will take an average of nine minutes of an employee’s time. The Bureau received no comments from industry or trade groups asserting that a substantial number of lenders currently lack the ability to check their record for prior loans, or that implementing such a system would constitute an undue cost or burden. The Bureau believes this supports the cost framework laid out here. 1476 (2). Obtaining a Consumer Report from a Registered Information System; Costs to Small Entities Under the rule, small entities will have to obtain a consumer report from a registered information system containing timely information about an applicant’s borrowing history, if one or more such systems were available. The Bureau believes that many lenders likely already obtain from third parties some of the information that will be included in the registered information system data, such as in States where a private third-party operates a database containing loan information on behalf of the State regulator or for their own risk management purposes, such as fraud detection. However, the Bureau recognizes that there also is a sizable segment of lenders making short-term loans that operate only in States without a State-mandated loan database and that make lending decisions without obtaining any data from a specialty consumer reporting agency. As noted above, the Bureau believes that many small entities use automated loan origination systems and will modify those systems or purchase upgrades to those systems such that they will automatically order a report from a registered information system during the lending process. For lenders that order reports manually, the Bureau estimates that it will take approximately nine minutes on average for a lender to request a report from a registered information system and utilize the report in the ATR determination. For all lenders, the Bureau expects that access to a registered information system will be priced on a “per-hit” basis, where a hit is a report successfully returned in response to a request for information about a particular consumer at a particular point in time. Based on industry outreach, the Bureau estimates that the cost to small entities would be $0.50 per hit, based on pricing in existing relevant consumer reporting markets. 1477 The Bureau received comments from trade groups and lenders discussing the estimated “per hit” costs of the registered information system reports. The comments were approximately evenly split as to whether the estimated costs were substantially too low, slightly too low, or approximately accurate. A trade group representing mostly large depository institutions argued the cost is substantially too low, and cited its members’ average costs of $10.97 to purchase a credit report. Given the drastic difference between this cost and those stated by other commenters, the Bureau believes the credit reports referred to (e.g., tri-bureau credit reports) are not the type that would be purchased for this type of loan. This comparison did not seem relevant to the cost to obtain a report from a registered information system. A trade group representing small-dollar lenders also asserted the estimated cost was too low, citing its members’ average cost of $1 to obtain a consumer report from a nationwide consumer reporting reporting agency. Finally, a large small-dollar lender asserted the $0.50 estimate “appears to be right.” Given that registered information systems are likely to collect much less data than are collected by consumer reporting agencies operating in the market today, it follows that the cost of a report from a registered information system should be lower. Given that the comments received directly from lenders regarding the expected costs of a registered information system report argued the estimate is generally accurate, the Bureau continues to believe the cost per hit estimate of $0.50 is reasonable. (3). Assessing Ability to Repay Based on Information and Verification Evidence about Income and Major Financial Obligations; Costs to Small Entities Lenders making loans under the ATR approach are required to collect information about the amount of income and major financial obligations, make reasonable efforts to verify that information, and use that information to make an ability-to-repay determination. 1478 The Bureau believes that many small entities that make short-term loans, such as small storefront lenders making payday loans, already obtain some information on consumers’ income. Many of these lenders, however, only obtain income verification evidence the first time they make a loan to a consumer or for the first loan following a substantial break in borrowing. Other lenders, such as some vehicle title lenders or some lenders operating online, may not currently obtain income information at all, let alone verification evidence for that information, before issuing loans. In addition, many consumers likely have multiple income sources that are not all currently documented in the ordinary course of short-term lending. Under the rule, consumers and lenders might have incentives to provide and gather more income information than they do currently in order to establish the borrower’s ability to repay a given loan. The Bureau believes that most lenders that originate short-term loans and longer-term loans with balloon payments do not currently collect information on applicants’ major financial obligations, let alone attempt to verify obligations, nor do they determine consumers’ ability to repay a loan, as will be required under the rule. There are two types of costs entailed in making an ATR determination: the cost of obtaining and verifying evidence where possible and the cost of making an ATR determination consistent with that evidence. As noted above, many lenders already use automated systems when originating loans. These lenders will likely modify those systems or purchase upgrades to those systems to automate many of the tasks that would be required by the rule. Under the rule, small lenders will be required to obtain a consumer report from a nationwide consumer reporting agency to verify the amount of payments for debt obligations, unless that lender has obtained a report in the preceding 90 days or the consumer has triggered a 1479 cooling-off period at the end of a three-loan sequence. As such, these consumer reports will typically only be necessary to obtain for the first loan in a new sequence of borrowing that begins more than 90 days since the last consumer reports was obtained. This will be in addition to the cost of obtaining a report from a registered information system, though the Bureau expects some registered information systems will provide consolidated reports. Based on industry outreach, the Bureau believes these reports will cost approximately $2.00 for small entities. As with the ordering of reports from registered information systems, the Bureau believes that many small entities will modify their loan origination system or purchase an upgrade to that system to allow the system to automatically order a consumer report from a nationwide consumer reporting agency during the lending process at a stage in the process where the information is relevant. For lenders that order reports manually, the Bureau estimates that it would take approximately nine minutes on average for a lender to request a report and utilize it in the ATR determination. Small entities that do not currently collect income or verification evidence for income will need to do so. The Bureau estimates it will take roughly three to five minutes per application for lenders that use a manual process to gather and review information a for consumers who have straightforward documentation (e.g., pay stubs), and incorporate the information into the ATR determination. Some industry commenters suggested this value was too low in the proposal, often citing cases where consumers may not have regular income from sources that provide documentation. The Bureau notes that many lenders already require such information prior to initiating loans. Additionally, the rule now allows stated income to be used in appropriate cases, mitigating the time costs associated with more rigorous verification efforts. As such, the Bureau believes the time estimates provided here to be reasonable. Some consumers may visit a lender’s storefront without the required documentation and 1480 may have income for which verification evidence cannot be obtained electronically. Small entities making loans online may face particular challenges obtaining verification evidence, especially for income. It may be feasible for online lenders to obtain scanned or photographed documents as attachments to an electronic submission; the Bureau understands that some online lenders are doing this today with success. And services that use other sources of information, such as checking account or payroll records, may mitigate the need for lenders to obtain verification evidence directly from consumers. Additionally, for consumers with cash income that is not deposited into a depository account, lenders will be allowed to rely on stated information, minimizing the lenders’ costs and the chance that a consumer is unable to complete an application. Once information and verification evidence on income and major financial obligations has been obtained, the lender must use that information and evidence to make a reasonable determination that the consumer will have the ability to repay the contemplated loan. In the process of considering the information collected about income and major financial obligations, lenders will need to estimate an amount that the borrower needs for basic living expenses. They may do this in a number of ways, including, for example, collecting information directly from borrowers, using available estimates published by third parties, or basing it on their experience with similarly situated consumers. In total, the Bureau estimates that obtaining a statement from the consumer and taking reasonable steps to verify income and required payments for major financial obligations, projecting the consumer’s residual income, estimating the consumer’s basic living expenses, and arriving at a reasonable ATR determination will take essentially no additional time for a fully automated electronic system and between 15 and 45 minutes for a fully manual system. 1481 Numerous industry commenters suggested the estimate provided by the Bureau in the proposal (15 to 2010 minutes) was too low. In response to these comments, the Bureau has increased its estimated time to manually underwrite these loans, but also notes that all major financial obligations should be obtainable either from a consumer report or consumer statement (in the example of rental expense). Further total costs will depend on the existing utilization rates of and wages paid to staff that will spend time carrying out this work. To the extent that existing staff has excess capacity (that is, that a lender’s employees have time that is not fully utilized), the extra time to process applications for loans made via the ATR approach should not result in higher wage bills for the lender. Further, as the Bureau expects the majority of loans to be made via the principal stepdown approach, the expected increase in staff hours necessary to comply with the new procedural requirements should be modest. Still, to the extent that lenders must increase staff and/or hours to comply with the procedural requirements, they may experience increased costs from hiring, training, wages, and benefits. Dollar costs include a report from a registered information system costing $.50 and a consumer report from a nationwide consumer reporting agency containing housing costs estimates costing $2.00. Lenders relying on electronic services to gather verification information about income would face an additional small cost. (4). Developing Procedures, Upgrading Systems, and Training Staff; Costs to Small Entities Small entities will need to develop procedures to comply with the requirements of the ATR approach and train their staff in those procedures. Many of these requirements do not appear qualitatively different from many practices that most lenders already engage in, such as gathering information and documents from borrowers and ordering various types of consumer 1482 reports. Developing procedures to make a reasonable determination that a borrower has an ability to repay a loan without re-borrowing and while paying for major financial obligations and living expenses is likely to be a challenge for many small entities. The Bureau expects that vendors, law firms, and trade associations are likely to offer both products and guidance to lenders, potentially lowering the cost of developing procedures as service providers can realize economies of scale. Lenders must also develop a process for estimating borrowers’ basic living expenses if they choose not to make an individual determination for each customer. Some lenders may rely on vendors that provide services to determine ability to repay that include estimates of basic living expenses. Some methods of conducting an analysis to determine estimates of basic living expenses could be quite costly. There are a number of government data sources and online services, however, that lenders may be able to use to obtain living expense estimates. Additionally, lenders may rely on their experiences with similarly situated consumers in making this estimate, reducing the need to rely on individual measures or third parties. As noted above, the Bureau believes that many lenders use automated systems when originating loans and would incorporate many of the procedural requirements of the ATR approach into those systems. This will likely include an automated system to make the abilityto-repay determination; subtracting the component expense elements from income itself is quite straightforward and should not require substantial development costs. The Bureau believes small lenders that use automated loan origination systems rely on licensed software. Depending on the nature of the software license agreement, the Bureau estimates that the one-time cost to upgrade this software will be $10,000 for lenders licensing the software at the entity-level and $100 per seat for lenders licensing the software using a seat-license contract. Given the price differential 1483 between the entity-level licenses and the seat-license contracts, the Bureau believes that only small entities with a significant number of stores will rely on the entity-level licenses. One trade group commented that they believe this to be too low an estimate of the associated costs, citing a survey of their members. However, the trade group’s members are not predominately involved in making loans that will be covered under the rule, so it is unclear how their estimates relate to the systems contemplated here. Additionally, the vast majority of the comments from more directly-related trade groups, lenders, etc. remained silent on these estimates, despite the invitation to provide feedback. As such, the Bureau has not changed these values from those put forth in the proposal. The Bureau estimates that lender personnel engaging in making loans would require approximately 4 hours per employee of initial training in carrying out the tasks described in this section and 2 hours per employee of periodic ongoing training per year.1304 (b). Principal step-down approach as an alternative to the underwriting criteria used to assess the borrower’s ability to repay; costs to small entities The rule includes an alternative approach, as opposed to using the underwriting criteria specified in § 1041.5, for originating certain short-term loans as in § 1041.6. In this section, the practice of making loans by complying with the alternative requirements under § 1041.6 will be referred to as the “principal step-down approach.” The procedural requirements of the principal step-down approach will generally have less impact on small lenders than the requirements of the ATR approach. Lenders that make shortterm loans under the principal step-down approach will not have to obtain information or 1304 Note that the Bureau expects that this training would be in addition to the training relating to furnishing loan information as discussed in part VIII.C.2.a and recordkeeping as discussed in part VIII.C.2b. 1484 verification evidence about income or major financial obligations, estimate basic living expenses, or complete an ability-to-repay determination prior to making loans. The rule will instead require only that lenders making loans under § 1041.6 consult their internal records and those of affiliates, access reports from a registered information system, furnish information to all registered information systems, and make an assessment as part of the origination process that certain loan requirements (such as principal limitations and restrictions on certain re-borrowing activity) are met. The requirement to consult the lender’s own records is slightly different than under the ATR Approach, as the lender must check the records for the prior 12 months. This is unlikely to have different impacts on small lenders, however, as any system that allows the lender to comply with the requirement to check its own records under the ATR approach should be sufficient for the principal step-down approach and vice-versa. A lender will also have to develop procedures and train staff. Small entities making short-term loans under the principal step-down approach will be required to provide borrowers with a disclosure, described in the section-by-section analysis of § 1041.6(e), with information about their loans and about the restrictions on future loans taken out using the principal step-down approach. One disclosure will be required at the time of origination of a first principal step-down approach loan, where a borrower had not had a principal step-down approach loan within the prior 30 days. The other disclosure will be required when originating a third principal step-down approach loan in a sequence because the borrower will therefore be unable to take out another principal step-down approach loan within 30 days of repaying the loan being originated. The disclosures will need to be customized to reflect the specifics of the individual loan. The Bureau believes that all small entities have some disclosure system in place to 1485 comply with existing disclosure requirements. Lenders may enter data directly into the disclosure system, or the system may automatically collect data from the lenders’ loan origination system. For disclosures provided via mail, email, or text message, some disclosure systems forward the information necessary to prepare the disclosures to a vendor in electronic form, and the vendor then prepares and delivers the disclosures. For disclosures provided in person, disclosure systems produce a disclosure that the lender then provides to the borrower. Respondents will incur a one-time cost to upgrade their disclosure systems to comply with new disclosure requirements. The Bureau believes that small lenders generally rely on licensed disclosure system software. Depending on the nature of the software license agreement, the Bureau estimates that the cost to upgrade this software will be $10,000 for lenders licensing the software at the entitylevel and $100 per seat for lenders licensing the software using a seat- license contract. Given the price differential between the entity-level licenses and the seat- license contracts, the Bureau believes that only small lenders with a significant number of stores will rely on entity-level licenses. In addition to the upgrades to the disclosure systems, the Bureau estimates that small storefront lenders will pay $200 to a vendor for a standard electronic origination disclosure form template. The Bureau estimates that providing disclosures in stores will take a store employee two minutes and cost $0.10. ii. Payment Practices and Related Notices for Certain Covered Loans; Costs to Small Entities The rule limits how lenders initiate payments on a covered loan from a borrower’s account and imposes two notice requirements relating to such payments. The impacts of these 1486 provisions are discussed here for all covered loans. Note that the Bureau believes that the requirement to assess ATR before making a shortterm or longer-term balloon-payment loan, or to comply with one of the conditional exemptions, will reduce the frequency with which borrowers receive loans that they do not have the ability to repay. This should make unsuccessful payment withdrawal attempts less frequent, and lessen the impacts of the limitation on payment withdrawal attempts and the requirement to notify consumers when a lender is no longer permitted to attempt to withdraw payments from a borrower’s account. (a). Limitation on Payment Withdrawal Attempts; Costs to Small Entities The rule prevents lenders from attempting to withdraw payment from a consumer’s account if two consecutive prior attempts to withdraw payment made through any channel are returned for nonsufficient funds. The lender can resume initiating payment if the lender obtains from the consumer a new and specific authorization to collect payment from the consumer’s account. The impact of this restriction depends on how often the lender attempts to collect from a consumers’ account after more than two consecutive failed transactions and how often they succeed in doing so. Based on industry outreach, the Bureau understands that some small entities already have a practice of not continuing to attempt to collect using these means after one or two failed attempts. These lenders will not incur costs from the restriction. Additionally, some depository institutions disallowed repeated attempts to collect using these means; lenders attempting to collect from such depositories would also not incur costs from this restriction. While not specific to small lenders, the Section 1022(b)(2) Analysis discusses the Bureau’s analysis of ACH payment request behavior of online lenders making payday or payday 1487 installment loans. The Bureau found that only 7 to 10 percent of the payments attempted through the ACH system came after two failed payments requests.1305 Under the restriction, lenders can still seek payment from their borrowers by engaging in other lawful collection practices. As such, the preceding are high-end estimates of the impact of this restriction on the collection efforts of these lenders. These other forms of lawful collection practices, however, may be more costly for lenders than attempting to collect directly from a borrower’s account. After the limitation is triggered by two consecutive failed attempts, lenders are required to send a notice to consumers. To seek a new and specific authorization to collect payment from a consumer’s account, the lender can send a request with the notice and may need to initiate additional follow-up contact with the consumer. The Bureau believes that this will most often be done in conjunction with general collections efforts and will impose little additional cost on lenders. To the extent that lenders assess returned item fees when an attempt to collect a payment fails and lenders are subsequently able to collect on those fees, this rule may reduce lenders’ revenues. Small entities will also need the capability of identifying when two consecutive payment requests have failed. The Bureau believes that the systems small entities use to identify when a payment is due, when a payment has succeeded or failed, and whether to request another payment will have the capacity to identify when two consecutive payments have failed, and therefore this requirement will not impose a significant new cost. The Bureau received comments stating that tracking failed payment withdrawals would 1305 CFPB Report on Supplemental Findings, at 150 tbl. 32. These impacts may be lower now than they were at the time covered by the data analyzed by the Bureau, due to changes in industry practices and to changes in the rules governing the ACH system. 1488 require new systems and procedures to be developed, at a cost not specified in the IRFA. While the Bureau acknowledges that some small entities may face costs in modifying existing systems to comply with the recordkeeping and payment processing requirements of the rule, these requirements largely build on processes required by existing laws or necessitated by standard business practice. (b). Required Notice to Collect Directly from a Borrower’s Account; Costs to Small Entities The rule will require lenders to provide consumers with a notice prior to the first lenderinitiated attempt to withdraw payment from consumers’ accounts, including ACH entries, postdated signature checks, remotely created checks, remotely created payment orders, and payments run through the debit networks. The notice will be required to include the date the lender will initiate the payment request, the payment channel, the amount of the payment, the breakdown of that amount to principal, interest, and fees, the loan balance remaining if the payment succeeds, the check number if the payment request is a signature check or RCC, and contact information for the consumer to reach the lender. There are separate notices required prior to unusual payments. This provision will not apply to small lenders making loans under the PAL approach or making accommodation loans. The costs to small entities of providing these notices will depend heavily on whether they are able to provide the notice via email, text messages, or on paper at origination or will have to send notices through regular mail. In practice, the Bureau expects most small lenders to provide the notice of initial payment withdrawal at origination, minimizing the transmission costs. This can either be done via a written disclosure (at a storefront), or as a PDF attachment or webpage sent via an email or text (for either storefront or online lenders). The variation in costs of notices 1489 provided after origination (either regular notices, or notices in advance of unusual payments) is due in part to differences in transmission costs between different channels. Most borrowers are likely to have Internet access or a mobile phone capable of receiving text messages, and during the SBREFA process multiple SERs reported that most borrowers, when given the opportunity, opt in to receiving notifications via text message. The Bureau has intentionally structured the rule to encourage transmission by email or text message because it believes those channels are the most effective for consumers, as well as less burdensome for lenders. However, should the lender choose to send paper notifications via regular mail, they would incur higher costs of transmission, as well as administrative costs associated with providing the notification early enough to ensure sufficient time for it to be received by the consumer. The Bureau believes that small entities that will be affected by the new disclosure requirements have some disclosure system in place to comply with existing disclosure requirements, such as those imposed under Regulation Z, 12 CFR part 1026 and Regulation E, 12 CFR part 1005. Lenders enter data directly into the disclosure system or the system automatically collects data from the lenders’ loan origination system. For disclosures provided via mail, email, text message, or immediately at the time of origination, the disclosure system often forwards to a vendor, in electronic form, the information necessary to prepare the disclosures, and the vendor then prepares and delivers the disclosures. Lenders will incur a onetime burden to upgrade their disclosure systems to comply with new disclosure requirements. Small lenders will need to update their disclosure systems to compile necessary loan information to send to the vendors that would produce and deliver the disclosures relating to payments. The Bureau believes small lenders rely on licensed disclosure system software. Depending on the nature of the software license agreement, the Bureau estimates that the cost to 1490 upgrade this software would be $10,000 for lenders licensing the software at the entity-level and $100 per seat for lenders licensing the software using a seat-license contract. For lenders using seat license software, the Bureau estimates that each location for small lenders has on average three seats licensed. Given the price differential between the entity-level licenses and the seatlicense contracts, the Bureau believes that only small entities with a significant number of stores will rely on the entity-level licenses. Small entities with disclosure systems that do not automatically pull information from the lenders’ loan origination or servicing system will need to enter payment information into the disclosure system manually so that the disclosure system can generate payment disclosures. The Bureau estimates that this will require two minutes per loan in addition to the two minutes to provide the disclosures. Lenders will need to update this information if the scheduled payments were to change. For disclosures delivered through the mail, the Bureau estimates that vendors would charge two different rates, one for high volume mailings and another for low volume mailings. The Bureau understands that small entities will likely generate a low volume of mailings and estimates vendors will charge such lenders $1.00 per disclosure. For disclosures delivered through e-mail, the Bureau estimates vendors will charge $0.01 to create and deliver each e- mail such that it complies with the requirements of the rule. For disclosures delivered through text message, the Bureau estimates vendors will charge $0.08 to create and deliver each text message such that it complies with the requirements of the rule. The vendor would also need to provide either a PDF attachment of the full disclosure or a web page where the full disclosure linked to in the text message is provided. The cost of providing this PDF attachment or web disclosure is included in the cost estimate of providing the text message. Finally, for disclosures delivered on 1491 paper at origination, the Bureau estimates costs will be $0.10 per disclosures. Again, the Bureau believes that virtually all notifications will be provided at the time of origination (for regular notices), or electronically via text or email (for notifications of unusual payments). As such, the mailing costs discussed here are expected to be almost completely avoided. (c). Required Notice When Lender Can No Longer Collect Directly from a Borrower’s Account; Costs to Small Entities The rule will require a lender that has made two consecutive unsuccessful attempts to collect payment through any channel from a borrower’s account to provide a borrower, within three business days of learning of the second unsuccessful attempt, with a consumer rights notice explaining that the lender is no longer able to attempt to collect payment directly from the borrower’s account, along with information identifying the loan and a record of the two failed attempts to collect funds. The requirement will impose on small entities the cost of providing the notice. Lenders already need to track whether they can still attempt to collect payments directly from a borrower’s account, so identifying which borrowers should receive the notice should not impose any additional cost on lenders. The Bureau also expects that lenders normally attempt to contact borrowers in these circumstances to identify other means of obtaining payment. If they are contacting the consumer via mail, the lender will be able to include the required notice in that mailing. The Bureau expects that small entities will incorporate the ability to provide this notice into their payment notification process. The Bureau estimates that vendors will charge $1.00 per notice for small entities that send a small volume of mailing. For disclosures delivered through 1492 email, the Bureau estimates vendors will charge $0.01 to create and deliver each e-mail such that it complies with the requirements of the proposed rule. For disclosures delivered through text message, the Bureau estimates vendors will charge $0.08 to create and deliver each text message. The vendor would also need to provide either a PDF attachment of the full disclosure or a web page where the full disclosure linked to in the text message would be provided. The cost of providing this PDF attachment or web disclosure is included in the cost estimate of providing the text message. (d). Estimate of Small Entities Subject to the Rule and Costs for Preparing Reports and Records Section 604(a)(5) of the RFA also requires an estimate of the type of professional skills necessary for the preparation of the reports or records. The Bureau does not anticipate that, except in certain rare circumstances, any professional skills will be required for recordkeeping and other compliance requirements of this rule that are not otherwise required in the ordinary course of business of the small entities affected by the proposed rule. Parts VIII.C.2.b and VIII.C.2.c summarize the recordkeeping and compliance requirements of the rule that will affect small entities. As discussed above, the Bureau believes that vendors will update their software and provide small creditors with the ability to retain the required data. The one situation in which a small entity would require professional skills that are not otherwise required in the ordinary course of business will be if a small creditor does not use computerized systems to store information relating to originated loans and therefore will either need to hire staff with the ability to implement a machine-readable data retention system or contract with one of the vendors that provides this service. The Bureau believes that the small entities will otherwise have the professional skills necessary to comply with the proposed rule. 1493 The Bureau believes efforts to train small entity staff on the updated software and compliance systems will be reinforcing existing professional skills sets above those needed in the ordinary course of business. In addition, although the Bureau acknowledges the possibility that certain small entities may have to hire additional staff as a result of certain aspects of the rule, the Bureau has no evidence that such additional staff will have to possess a qualitatively different set of professional skills than small entity staff employed currently. The Bureau presumes that additional staff that small entities may need to hire will generally be of the same professional skill set as current staff. Several commenters raised concerns that the initial implementation of the rule’s requirements may require legal or consulting skills beyond those of employees at typical small lenders. The Bureau acknowledges this concern, and believes these costs are accounted for in earlier estimates of the one-time costs of developing procedures, upgrading systems, and training staff. D. The Bureau’s Efforts to Minimize the Economic Impact on Small Entities Section 604(a)(6) of the RFA requires the Bureau to describe in the FRFA the steps taken to minimize the significant economic impact on small entities consistent with the stated objectives of applicable statutes. The Bureau has taken numerous steps to minimize the significant economic impact on small entities consistent with the stated objectives of applicable statues. These include simplification of the ability-to-repay requirements, expanded exclusions from the rule, expanded exemptions for alternative loans and accommodation loans, increased flexibility and reduced number of required payment disclosures, and a later effective date as described in the Bureau’s responses to public comments and the SBA Office for Advocacy. 1. Consideration of Alternatives to the Final Rule and Their Impact on Small Entities 1494 In the IRFA, four significant alternatives to the proposed rule were considered, but the Bureau decided that none of them would accomplish the stated objectives of Title X of the DoddFrank Act while minimizing the impact of the rule on small entities.1306 In this section, the Bureau presents its considerations in that regard. Four significant alternatives are briefly described and their impacts on small entities relative to the adopted provisions are discussed below. The discussion of each alternative includes a statement of the factual, policy, and legal reasons for selecting the adopted provisions and rejecting the significant alternatives. The alternatives discussed here are: ● Limits on re-borrowing of short-term loans without an ability-to-repay requirement; ● An ATR requirement for short-term loans with no principal step-down approach; ● Disclosures as an alternative to the ability-to-repay requirement; and ● Limitations on withdrawing payments from borrowers’ accounts without disclosures. In addition to the significant alternatives outlined above, the Bureau has considered comments on alternatives to specific provisions of the rule, discussed in the section-by-section analysis of each corresponding section. a. Limits on Re-borrowing Short-Term Loans without an Ability-to-Repay Requirement As an alternative to the ability-to-repay requirements in § 1041.5 for short-term loans, the Bureau considered a limitation on the overall number of short-term loans that a consumer could take in a loan sequence or within a short period of time. This alternative would limit consumer injury from extended periods of re-borrowing on short-term loans. However, as discussed further in part VII.J.1, the Bureau has concluded that a limitation on re-borrowing without a 1306 5 U.S.C. 603(c). 1495 requirement to determine the consumer’s ability to repay the loan will not provide sufficient protection against consumer injury from making a short-term loan without reasonably determining that the consumer will have the ability to repay the loan. Accordingly, the Bureau finds that a limitation on repeat borrowing alone will not be consistent with the stated objectives of Title X to identify and prevent unfair, deceptive, or abusive acts or practices. However, the Bureau has made changes to the ability-to-pay requirements to reduce the burden of compliance for small entities, as described in the Bureau’s responses to the SBA Office for Advocacy. b. An ATR Requirement for Short-Term Loans with No Principal Step-down Approach The Bureau considered adopting the ability-to-repay requirements in § 1041.5 for shortterm loans without adopting the alternative approach for originating certain short-term loans as described in § 1041.6. In the absence of the principal step-down approach, lenders would be required to make a reasonable determination that a consumer has the ability to repay a loan and to therefore incur the costs associated with the ability-to-repay requirements for every short-term application that they process. However, the Bureau has determined that the principal step-down approach will provide sufficient structural consumer protections while reducing the compliance burdens associated with the ATR approach on lenders and permitting access to less risky credit for borrowers for whom it may be difficult for lenders to make a reasonable determination that the borrower has the ability to repay a loan, but who may nonetheless have sufficient income to repay the loan and also meet other financial obligations and basic living expenses. Comments from small entities expressed particular concern that the ability-to-repay requirements would be burdensome given their smaller scale over which to spread fixed cost investments. In addition, comments suggested that because small lenders base some lending decisions on their personal relationship with customers, the full ability-to-repay assessment was not 1496 necessary for all loan originations. Accordingly, the Bureau has concluded that providing the principal step-down approach as described in § 1041.6 will help minimize the economic impact of the proposed rule on small entities without undermining consumer protections in accordance with the stated objectives of Title X to identify and prevent unfair, deceptive, or abusive acts or practices. c. Disclosures as an Alternative to the Ability-to-Repay Requirement As an alternative to substantive regulation of the consumer credit transactions that will be covered by the rule, the Bureau considered whether enhanced disclosure requirements would prevent the consumer injury that is the focus of the rule and minimize the impact of the proposal on small entities. In particular, the Bureau considered whether the disclosures required by some States would accomplish the stated objectives of Title X of the Dodd-Frank Act. The Bureau is adopting, in §§ 1041.6 and 1041.9 requirements that lenders make specific disclosures in connection with certain aspects of a transaction. Analysis by the Bureau indicates that a disclosure-only approach would have substantially less impact on the volume of short-term lending, but also would have substantially less impact on the harms consumers experience from long sequences of payday and singlepayment vehicle title loans, as discussed further in part VII.J.3. Because the Bureau has concluded that disclosures alone would be ineffective in warning borrowers of those risks and preventing the harms that the Bureau seeks to address with the proposal, the Bureau is not adopting disclosure as an alternative to the ability-to-repay and other requirements of the rule. d. Limitations on Withdrawing Payments from Borrowers’ Accounts without Disclosures The Bureau considered including the prohibition on lenders attempting to collect payment from a consumer’s accounts when two consecutive attempts have been returned due to a 1497 lack of sufficient funds in § 1041.8 unless the lender obtains a new and specific authorization, but not including the required disclosures of upcoming payment withdrawals (both the first and unusual payments) or the notice by lenders to consumers alerting them to the fact that two consecutive withdrawal attempts to their account have failed and the lender can therefore no longer continue to attempt to collect payments from a borrower account. This alternative would reduce lenders’ one-time costs of upgrading their disclosure systems as well as the incremental burden of providing each disclosure. The Bureau finds, however, that in the absence of the disclosures, consumers face an increased risk of injury in situations in which lenders intend to initiate a withdrawal in a way that deviates from the loan agreement or prior course of conduct between the parties. In addition, consumers would face an increased risk of believing that they are required to provide lenders with a new authorization to continue to withdraw payments directly from their accounts when they may be better off using some alternative method of payment. To reduce the burden for small entities and other lenders, after the first payment, any payment withdrawals for usual payments do not require a disclosure under the final rule. Relative to the proposed rule, this change will decrease compliance costs for small entities while still accomplishing the stated objectives of the rule. Some commenters expressed concern that the Bureau’s position on disclosures – that they are an insufficient alternative to the ability-to-repay requirements but beneficial for payment withdrawals, is inconsistent. Yet the mandated disclosures in these situations address different harms. The primary harm from re-borrowing is unlikely to be resolved by disclosures that long 1498 sequences may occur, as borrowers seem to understand the average duration of sequences,1307 but cannot accurately predict their own durations.1308 For re-borrowing, providing evidence about the average would therefore not address the market failure. However, disclosures about payments are different, as they are more immediate and inform the borrower of more certain events. Therefore, the Bureau has determined that they are an appropriate intervention here. 2. The Bureau’s Efforts to Minimize Any Additional Cost of Credit for Small Entities Section 603(d) of the RFA requires the Bureau to consult with small entities about the potential impact of the proposed rule on the cost of credit for small entities and related matters. In the FRFA, the Bureau is required to provide a description of the steps taken to minimize any additional cost of credit for small entities.1309 To satisfy these statutory requirements, the Bureau provided notification to the Chief Counsel that the Bureau would collect the advice and recommendations of the same small entity representatives identified in consultation with the Chief Counsel through the SBREFA process concerning any projected impact of the proposed rule on the cost of credit for small entities.1310 The Bureau sought to collect the advice and recommendations of the small entity representatives during the Small Business Review Panel Outreach Meeting regarding the potential impact on the cost of business credit because, as small financial service providers, the SERs could provide valuable input on any such impact related to the proposed rule.1311 1307 See, e.g., Marianne Bertrand and Adair Morse, “Information Disclosure, Cognitive Biases and Payday Borrowing,” 66 J. of Fin. 1865 (2011). 1308 Ronald Mann, “Assessing the Optimism of Payday Loan Borrowers,” 21 Sup. Ct. Econ. Rev. 105 (2013). 1309 5 U.S.C. 604(a)(6). 1310 See 5 U.S.C. 603(d)(2)(A). The Bureau provided this notification as part of the notification and other information provided to the Chief Counsel with respect to the SBREFA process pursuant to section 609(b)(1) of the RFA. 1311 See 5 U.S.C. 603(d)(2)(B). 1499 At the Small Business Review Panel Outreach Meeting, the Bureau asked the SERs a series of questions regarding issues about the cost of business credit.1312 The questions were focused on two areas. First, the SERs were asked whether, and how often, they extend to their customers covered loans to be used primarily for personal, family, or household purposes but that are used secondarily to finance a small business, and whether the proposals then under consideration would result in an increase in their customers’ cost of credit. Second, the Bureau inquired as to whether the proposals under consideration would increase the SERs’ cost of credit. In general, some of the SERs expressed concern that the proposals under consideration would have a substantial impact on the cost of business credit, both by reducing access to credit for their customers that are using loans to fund small business operations and by making their businesses less creditworthy. As discussed in the Small Business Review Panel Report, the Panel recommended that the Bureau cover only loans extended primarily for personal, family, or household purposes.1313 The Bureau agreed with that recommendation, and so in § 1041.3(b), the rule does in fact specify that it will apply only to loans that are extended to consumers primarily for personal, family, or household purposes. Loans that are made primarily for a business, commercial, or agricultural purpose will not be subject to this part. Nonetheless, the Bureau recognizes that some covered loans may nonetheless be used in part or in whole to finance small businesses, both with or without the knowledge of the lender. The Bureau also recognizes that the rules will impact the ability of some small entities to access business credit themselves. As discussed more fully part VII.J and just above in this section, in developing the rule, the Bureau has considered a number of alternative approaches, 1312 1313 See Small Business Review Panel Report, at 25. See id. at 33. 1500 yet for the reasons stated it has concluded that none of them would achieve the statutory objectives while minimizing the cost of credit for small entities. IX. Paperwork Reduction Act Under the Paperwork Reduction Act of 1995 (PRA),1314 Federal agencies are generally required to seek approval from the Office of Management and Budget (OMB) for information collection requirements prior to implementation. Under the PRA, the Bureau may not conduct or sponsor and, notwithstanding any other provision of law, a person is not required to respond to an information collection unless the information collection displays a valid control number assigned by OMB. OMB has tentatively assigned control #3170-0064 to these collections of information, however this control number is not yet active. This final rule contains information collection requirements that have not yet been approved by the OMB and, therefore, are not effective until OMB approval is obtained. The unapproved information collection requirements are listed below. A complete description of the information collection requirements, including the burden estimate methods, is provided in the information collection request (ICR) that the Bureau has submitted to OMB under the requirements of the PRA. The Bureau believes the following aspects of the rule would be information collection requirements under the PRA: (1) development, implementation, and continued use of notices for covered short-term loans made under § 1041.6, upcoming payment notices (including unusual payment notices), and consumer rights notices; (2) obtaining a consumer report from a registered information system; (3) furnishing information about consumers’ borrowing behavior to each 1314 44 U.S.C. 3501 et seq. 1501 registered information system; (4) retrieval of borrowers’ national consumer report information; (5) collection of consumers’ income and major financial obligations during the underwriting process; (6) obtaining a new and specific authorization to withdraw payment from a borrower’s deposit account after two consecutive failed payment transfer attempts; (7) application to be a registered information system; (8) biennial assessment of the information security programs for registered information systems; (9) retention of loan agreement and documentation obtained when making a covered loan, and electronic records of origination calculations and determination, records for a consumer who qualifies for an exception to or overcomes a presumption of unaffordability, loan type and term, and payment history and loan performance. The Bureau received a fairly significant number of comments pertaining to the expected burden of the proposal, including burdens accounted for in the PRA. Some of those comments specifically noted the PRA, and argued that the proposed collections of information did not fill a legitimate regulatory purpose. Specifically, they claimed that the paperwork burden, in particular the collection and verification of income and debt information, did not serve a legitimate purpose and would not advance the goal of ensuring that loans would be made based on a reasonable assessment of the borrower’s ability to repay. As explained in detail in the section-by-section analysis, especially the section-by-section analysis for § 1041.5, as well as the Section 1022(b)(2) Analysis in part VII, the Bureau has significantly reduced the burden associated with the rule’s requirements in response to comments it received which stated concerns that the proposed requirements would be too onerous. As finalized, and as described above, the Bureau is confident that each of the collections of information is worth the burden and serves an important purpose. Specific to the verification of income and debt requirements, the Bureau believes that these requirements are not overly 1502 burdensome. In many cases, covered lenders already verify income. Verification of debt will be achievable through obtaining consumer reports, an approach that would not burden consumers, and is consistent with industry practices in most other credit markets. These requirements advance the stated goal of assessing ability to repay because they ensure that lenders verify essential variables for a reasonable ability-to-repay determination, and they combat significant risks associated with lenders’ potential evasion of the rule. Pursuant to 44 U.S.C. 3507, the Bureau will publish a separate notice in the Federal Register announcing the submission of these information collection requirements to OMB as well as OMB’s action on these submissions, including the OMB control number and expiration date. The Bureau has a continuing interest in the public’s opinion of its collections of information. At any time, comments regarding the burden estimate, or any other aspect of the information collection, including suggestions for reducing the burden, may be sent to the Consumer Financial Protection Bureau (Attention: PRA Office), 1700 G Street NW, Washington, D.C. 20552, or by email to CFPB_Public_PRA@cfpb.gov. Title of Collection: Payday, Vehicle Title, and Certain High-Cost Installment Loans. OMB Control Number: 3170-0064 Type of Review: New collection (Request for a new OMB control number). Affected Public: Private Sector. Estimated Number of Respondents: 9,900. Estimated Total Annual Burden Hours: 8,075,337. 1503 List of Subjects in 12 CFR Part 1041 Banks, Banking, Consumer protection, Credit, Credit Unions, National banks, Registration, Reporting and recordkeeping requirements, Savings associations, Trade practices. Authority and Issuance For the reasons set forth above, the Bureau adds 12 CFR part 1041 to Chapter X in Title 12 of the Code of Federal Regulations, as set forth below: PART 1041—PAYDAY, VEHICLE TITLE, AND CERTAIN HIGH-COST INSTALLMENT LOANS Subpart A—General Sec. 1041.1 Authority and purpose. 1041.2 Definitions. 1041.3 Scope of coverage; exclusions; exemptions. Subpart B—Underwriting 1041.4 Identification of unfair and abusive practice. 1041.5 Ability-to-repay determination required. 1041.6 Conditional exemption for certain covered short-term loans. Subpart C—Payments 1041.7 Identification of unfair and abusive practice. 1041.8 Prohibited payment transfer attempts. 1041.9 Disclosure of payment transfer attempts. Subpart D—Information Furnishing, Recordkeeping, Anti-Evasion, and Severability 1041.10 Information furnishing requirements. 1041.11 Registered information systems. 1041.12 Compliance program and record retention. 1041.13 Prohibition against evasion. 1041.14 Severability. Appendix A to Part 1041—MODEL FORMS Supplement I to Part 1041—Official Interpretations Authority: 12 U.S.C. 5511, 5512, 5514(b), 5531(b), (c), and (d), 5532. 1504 Subpart A—General § 1041.1 Authority and purpose. (a) Authority. The regulation in this part is issued by the Bureau of Consumer Financial Protection (Bureau) pursuant to Title X of the Dodd-Frank Wall Street Reform and Consumer Protection Act (12 U.S.C. 5481, et seq.). (b) Purpose. The purpose of this part is to identify certain unfair and abusive acts or practices in connection with certain consumer credit transactions and to set forth requirements for preventing such acts or practices. This part also prescribes requirements to ensure that the features of those consumer credit transactions are fully, accurately, and effectively disclosed to consumers. This part also prescribes processes and criteria for registration of information systems. § 1041.2 Definitions. (a) Definitions. For the purposes of this part, the following definitions apply: (1) Account has the same meaning as in Regulation E, 12 CFR 1005.2(b). (2) Affiliate has the same meaning as in 12 U.S.C. 5481(1). (3) Closed-end credit means an extension of credit to a consumer that is not open-end credit under § 1041.2(a)(16). (4) Consumer has the same meaning as in 12 U.S.C. 5481(4). (5) Consummation means the time that a consumer becomes contractually obligated on a new loan or a modification that increases the amount of an existing loan. (6) Cost of credit means the cost of consumer credit as expressed as a per annum rate and is determined as follows: 1505 (i) Charges included in the cost of credit. The cost of credit includes all finance charges as set forth by Regulation Z, 12 CFR 1026.4, but without regard to whether the credit is consumer credit, as that term is defined in 12 CFR 1026.2(a)(12), or is extended to a consumer, as that term is defined in 12 CFR 1026.2(a)(11). (ii) Calculation of the cost of credit—(A) Closed-end credit. For closed-end credit, the cost of credit must be calculated according to the requirements of Regulation Z, 12 CFR 1026.22. (B) Open-end credit. For open-end credit, the cost of credit must be calculated according to the rules for calculating the effective annual percentage rate for a billing cycle as set forth in Regulation Z, 12 CFR 1026.14(c) and (d). (7) Covered longer-term balloon-payment loan means a loan described in § 1041.3(b)(2). (8) Covered longer-term loan means a loan described in § 1041.3(b)(3). (9) Covered person has the same meaning as in the Dodd-Frank Wall Street Reform and Consumer Protection Act, 12 U.S.C. 5481(6). (10) Covered short-term loan means a loan described in § 1041.3(b)(1). (11) Credit has the same meaning as in Regulation Z, 12 CFR 1026.2(a)(14). (12) Electronic fund transfer has the same meaning as in Regulation E, 12 CFR 1005.3(b). (13) Lender means a person who regularly extends credit to a consumer primarily for personal, family, or household purposes. (14) Loan sequence or sequence means a series of consecutive or concurrent covered short-term loans or covered longer-term balloon-payment loans, or a combination thereof, in which each of the loans (other than the first loan) is made during the period in which the 1506 consumer has a covered short-term loan or covered longer-term balloon-payment loan outstanding and for 30 days thereafter. For the purpose of determining where a loan is located within a loan sequence: (i) A covered short-term loan or covered longer-term balloon-payment loan is the first loan in a sequence if the loan is extended to a consumer who had no covered short-term loan or covered longer-term balloon-payment loan outstanding within the immediately preceding 30 days; (ii) A covered short-term or covered longer-term balloon-payment loan is the second loan in the sequence if the consumer has a currently outstanding covered short-term loan or covered longer-term balloon-payment loan that is the first loan in a sequence, or if the consummation date of the second loan is within 30 days following the last day on which the consumer’s first loan in the sequence was outstanding; (iii) A covered short-term or covered longer-term balloon-payment loan is the third loan in the sequence if the consumer has a currently outstanding covered short-term loan or covered longer-term balloon-payment loan that is the second loan in the sequence, or if the consummation date of the third loan is within 30 days following the last day on which the consumer’s second loan in the sequence was outstanding; and (iv) A covered short-term or covered longer-term balloon-payment loan would be the fourth loan in the sequence if the consumer has a currently outstanding covered short-term loan or covered longer-term balloon-payment loan that is the third loan in the sequence, or if the consummation date of the fourth loan would be within 30 days following the last day on which the consumer’s third loan in the sequence was outstanding. 1507 (15) Motor vehicle means any self-propelled vehicle primarily used for on-road transportation. The term does not include motor homes, recreational vehicles, golf carts, and motor scooters. (16) Open-end credit means an extension of credit to a consumer that is an open-end credit plan as defined in Regulation Z, 12 CFR 1026.2(a)(20), but without regard to whether the credit is consumer credit, as defined in 12 CFR 1026.2(a)(12), is extended by a creditor, as defined in 12 CFR 1026.2(a)(17), is extended to a consumer, as defined in 12 CFR 1026.2(a)(11), or permits a finance charge to be imposed from time to time on an outstanding balance as defined in 12 CFR 1026.4. (17) Outstanding loan means a loan that the consumer is legally obligated to repay, regardless of whether the loan is delinquent or is subject to a repayment plan or other workout arrangement, except that a loan ceases to be an outstanding loan if the consumer has not made at least one payment on the loan within the previous 180 days. (18) Service provider has the same meaning as in the Dodd-Frank Wall Street Reform and Consumer Protection Act, 12 U.S.C. 5481(26). (19) Vehicle security means an interest in a consumer’s motor vehicle obtained by the lender or service provider as a condition of the credit, regardless of how the transaction is characterized by State law, including: (i) Any security interest in the motor vehicle, motor vehicle title, or motor vehicle registration whether or not the security interest is perfected or recorded; or (ii) A pawn transaction in which the consumer’s motor vehicle is the pledged good and the consumer retains use of the motor vehicle during the period of the pawn agreement. 1508 (b) Rule of construction. For purposes of this part, where definitions are incorporated from other statutes or regulations, the terms have the meaning and incorporate the embedded definitions, appendices, and commentary from those other laws except to the extent that this part provides a different definition for a parallel term. § 1041.3 Scope of coverage; exclusions; exemptions. (a) General. This part applies to a lender that extends credit by making covered loans. (b) Covered loan. Covered loan means closed-end or open-end credit that is extended to a consumer primarily for personal, family, or household purposes that is not excluded under paragraph (d) of this section or conditionally exempted under paragraphs (e) or (f) of this section; and: (1) For closed-end credit that does not provide for multiple advances to consumers, the consumer is required to repay substantially the entire amount of the loan within 45 days of consummation, or for all other loans, the consumer is required to repay substantially the entire amount of any advance within 45 days of the advance; (2) For loans not otherwise covered by (1): (i) For closed-end credit that does not provide for multiple advances to consumers, the consumer is required to repay substantially the entire balance of the loan in a single payment more than 45 days after consummation or to repay such loan through at least one payment that is more than twice as large as any other payment(s). (ii) For all other loans, either: (A) The consumer is required to repay substantially the entire amount of an advance in a single payment more than 45 days after the advance is made or is required to make at least one payment on the advance that is more than twice as large as any other payment(s); or 1509 (B) A loan with multiple advances is structured such that paying the required minimum payments may not fully amortize the outstanding balance by a specified date or time, and the amount of the final payment to repay the outstanding balance at such time could be more than twice the amount of other minimum payments under the plan; or (3) For loans not otherwise covered by (1) or (2), if both of the following conditions are satisfied: (i) The cost of credit for the loan exceeds 36 percent per annum, as measured: (A) At the time of consummation for closed-end credit; or (B) At the time of consummation and, if the cost of credit at consummation is not more than 36 percent per annum, again at the end of each billing cycle for open-end credit, except that: (1) Open-end credit meets the condition set forth in this paragraph (B) in any billing cycle in which a lender imposes a finance charge, and the principal balance is $0; and (2) Once open-end credit meets the condition set forth in this paragraph (B), it meets the condition set forth in paragraph (B) for the duration of the plan. (ii) The lender or service provider obtains a leveraged payment mechanism as defined in paragraph (c) of this section. (c) Leveraged payment mechanism. For purposes of paragraph (b) of this section, a lender or service provider obtains a leveraged payment mechanism if it has the right to initiate a transfer of money, through any means, from a consumer’s account to satisfy an obligation on a loan, except that the lender or service provider does not obtain a leveraged payment mechanism by initiating a single immediate payment transfer at the consumer’s request. (d) Exclusions for certain types of credit. This part does not apply to the following: 1510 (1) Certain purchase money security interest loans. Credit extended for the sole and express purpose of financing a consumer’s initial purchase of a good when the credit is secured by the property being purchased, whether or not the security interest is perfected or recorded. (2) Real estate secured credit. Credit that is secured by any real property, or by personal property used or expected to be used as a dwelling, and the lender records or otherwise perfects the security interest within the term of the loan. (3) Credit cards. Any credit card account under an open-end (not home-secured) consumer credit plan as defined in Regulation Z, 12 CFR 1026.2(a)(15)(ii). (4) Student loans. Credit made, insured, or guaranteed pursuant to a program authorized by subchapter IV of the Higher Education Act of 1965, 20 U.S.C. 1070 through 1099d, or a private education loan as defined in Regulation Z, 12 CFR 1026.46(b)(5). (5) Non-recourse pawn loans. Credit in which the lender has sole physical possession and use of the property securing the credit for the entire term of the loan and for which the lender’s sole recourse if the consumer does not elect to redeem the pawned item and repay the loan is the retention of the property securing the credit. (6) Overdraft services and lines of credit. Overdraft services as defined in 12 CFR 1005.17(a), and overdraft lines of credit otherwise excluded from the definition of overdraft services under 12 CFR 1005.17(a)(1). (7) Wage advance programs. Advances of wages that constitute credit if made by an employer, as defined in the Fair Labor Standards Act, 29 U.S.C. 203(d), or by the employer’s business partner, to the employer’s employees, provided that: (i) The advance is made only against the accrued cash value of any wages the employee has earned up to the date of the advance; and 1511 (ii) Before any amount is advanced, the entity advancing the funds warrants to the consumer as part of the contract between the parties on behalf of itself and any business partners, that it or they, as applicable: (A) Will not require the consumer to pay any charges or fees in connection with the advance, other than a charge for participating in the wage advance program; (B) Has no legal or contractual claim or remedy against the consumer based on the consumer’s failure to repay in the event the amount advanced is not repaid in full; and (C) With respect to the amount advanced to the consumer, will not engage in any debt collection activities if the advance is not deducted directly from wages or otherwise repaid on the scheduled date, place the amount advanced as a debt with or sell it to a third party, or report to a consumer reporting agency concerning the amount advanced. (8) No-cost advances. Advances of funds that constitute credit if the consumer is not required to pay any charge or fee to be eligible to receive or in return for receiving the advance, provided that before any amount is advanced, the entity advancing the funds warrants to the consumer as part of the contract between the parties: (i) That it has no legal or contractual claim or remedy against the consumer based on the consumer’s failure to repay in the event the amount advanced is not repaid in full; and (ii) That, with respect to the amount advanced to the consumer, such entity will not engage in any debt collection activities if the advance is not repaid on the scheduled date, place the amount advanced as a debt with or sell it to a third party, or report to a consumer reporting agency concerning the amount advanced. (e) Alternative loan. Alternative loans are conditionally exempt from § 1041. Alternative loan means a covered loan that satisfies the following conditions and requirements: 1512 (1) Loan term conditions. An alternative loan must satisfy the following conditions: (i) The loan is not structured as open-end credit, as defined in § 1041.2(a)(16); (ii) The loan has a term of not less than one month and not more than six months; (iii) The principal of the loan is not less than $200 and not more than $1,000; (iv) The loan is repayable in two or more payments, all of which payments are substantially equal in amount and fall due in substantially equal intervals, and the loan amortizes completely during the term of the loan; and (v) The lender does not impose any charges other than the rate and application fees permissible for Federal credit unions under regulations issued by the National Credit Union Administration at 12 CFR 701.21(c)(7)(iii). (2) Borrowing history condition. Prior to making an alternative loan under § 1041.3(e), the lender must determine from its records that the loan would not result in the consumer being indebted on more than three outstanding loans made under this section from the lender within a period of 180 days. The lender must also make no more than one alternative loan under § 1041.3(e) at a time to a consumer. (3) Income documentation condition. In making an alternative loan under § 1041.3(e), the lender must maintain and comply with policies and procedures for documenting proof of recurring income. (4) Safe harbor. Loans made by Federal credit unions in compliance with the conditions set forth by the National Credit Union Administration at 12 CFR 701.21(c)(7)(iii) for a Payday Alternative Loan are deemed to be in compliance with the requirements and conditions of paragraphs (e)(1), (e)(2), and (e)(3) of this section. 1513 (f) Accommodation loans. Accommodation loans are conditionally exempt from § 1041. An accommodation loan is a covered loan if at the time that the loan is consummated: (1) The lender and its affiliates collectively have made 2,500 or fewer covered loans in the current calendar year, and made 2,500 or fewer such covered loans in the preceding calendar year; and (2)(i) During the most recent completed tax year in which the lender was in operation, if applicable, the lender and any affiliates that were in operation and used the same tax year derived no more than 10 percent of their receipts from covered loans; or (ii) If the lender was not in operation in a prior tax year, the lender reasonably anticipates that the lender and any of its affiliates that use the same tax year will derive no more than 10 percent of their receipts from covered loans during the current tax year. (3) Provided, however, that covered longer-term loans for which all transfers meet the conditions in § 1041.8(a)(1)(ii), and receipts from such loans, are not included for the purpose of determining whether the conditions of paragraphs (f)(1) and (f)(2) of this section have been satisfied. (g) Receipts. For purposes of paragraph (f) of this section, receipts means “total income” (or in the case of a sole proprietorship “gross income”) plus “cost of goods sold” as these terms are defined and reported on Internal Revenue Service (IRS) tax return forms (such as Form 1120 for corporations; Form 1120S and Schedule K for S corporations; Form 1120, Form 1065 or Form 1040 for LLCs; Form 1065 and Schedule K for partnerships; and Form 1040, Schedule C for sole proprietorships). Receipts do not include net capital gains or losses; taxes collected for and remitted to a taxing authority if included in gross or total income, such as sales or other taxes collected from customers but excluding taxes levied on the entity or its employees; or amounts 1514 collected for another (but fees earned in connection with such collections are receipts). Items such as subcontractor costs, reimbursements for purchases a contractor makes at a customer’s request, and employee-based costs such as payroll taxes are included in receipts. (h) Tax year. For purposes of paragraph (f) of this section, “tax year” has the meaning attributed to it by the IRS as set forth in IRS Publication 538, which provides that a “tax year” is an annual accounting period for keeping records and reporting income and expenses. Subpart B—Underwriting § 1041.4 Identification of unfair and abusive practice. It is an unfair and abusive practice for a lender to make covered short-term loans or covered longer-term balloon-payment loans without reasonably determining that the consumers will have the ability to repay the loans according to their terms. § 1041.5 Ability-to-repay determination required. (a) Definitions. For purposes of this section: (1) Basic living expenses means expenditures, other than payments for major financial obligations, that a consumer makes for goods and services that are necessary to maintain the consumer’s health, welfare, and ability to produce income, and the health and welfare of the members of the consumer’s household who are financially dependent on the consumer. (2) Debt-to-income ratio means the ratio, expressed as a percentage, of the sum of the amounts that the lender projects will be payable by the consumer for major financial obligations during the relevant monthly period and the payments under the covered short-term loan or covered longer-term balloon-payment loan during the relevant monthly period, to the net income that the lender projects the consumer will receive during the relevant monthly period, all of which projected amounts are determined in accordance with paragraph (c). 1515 (3) Major financial obligations means a consumer’s housing expense, required payments under debt obligations (including, without limitation, outstanding covered loans), child support obligations, and alimony obligations. (4) National consumer report means a consumer report, as defined in section 603(d) of the Fair Credit Reporting Act, 15 U.S.C. 1681a(d), obtained from a consumer reporting agency that compiles and maintains files on consumers on a nationwide basis, as defined in section 603(p) of the Fair Credit Reporting Act, 15 U.S.C. 1681a(p). (5) Net income means the total amount that a consumer receives after the payer deducts amounts for taxes, other obligations, and voluntary contributions (but before deductions of any amounts for payments under a prospective covered short-term loan or covered longer-term balloon-payment loan or for any major financial obligation); provided that, the lender may include in the consumer’s net income the amount of any income of another person to which the consumer has a reasonable expectation of access. (6) Payment under the covered short-term loan or covered longer-term balloon-payment loan: (i) Means the combined dollar amount payable by the consumer at a particular time following consummation in connection with the covered short-term loan or covered longer-term balloon-payment loan, assuming that the consumer has made preceding required payments and in the absence of any affirmative act by the consumer to extend or restructure the repayment schedule or to suspend, cancel, or delay payment for any product, service, or membership provided in connection with the loan; (ii) Includes all principal, interest, charges, and fees; and (iii) For a line of credit is calculated assuming that: 1516 (A) The consumer will utilize the full amount of credit under the covered short-term loan or covered longer-term balloon-payment loan as soon as the credit is available to the consumer; and (B) The consumer will make only minimum required payments under the covered shortterm loan or covered longer-term balloon-payment loan for as long as permitted under the loan agreement. (7) Relevant monthly period means the calendar month in which the highest sum of payments is due under the covered short-term or covered longer-term balloon-payment loan. (8) Residual income means the sum of net income that the lender projects the consumer will receive during the relevant monthly period, minus the sum of the amounts that the lender projects will be payable by the consumer for major financial obligations during the relevant monthly period and payments under the covered short-term loan or covered longer-term balloonpayment loan during the relevant monthly period, all of which projected amounts are determined in accordance with paragraph (c). (b) Reasonable determination required. (1)(i) Except as provided in § 1041.6, a lender must not make a covered short-term loan or covered longer-term balloon-payment loan or increase the credit available under a covered short-term loan or covered longer-term balloonpayment loan, unless the lender first makes a reasonable determination that the consumer will have the ability to repay the loan according to its terms. (ii) For a covered short-term loan or covered longer-term balloon-payment loan that is a line of credit, a lender must not permit a consumer to obtain an advance under the line of credit more than 90 days after the date of a required determination under this paragraph (b), unless the 1517 lender first makes a new determination that the consumer will have the ability to repay the covered short-term loan or covered longer-term balloon-payment loan according to its terms. (2) A lender’s determination of a consumer’s ability to repay a covered short-term loan or covered longer-term balloon-payment loan is reasonable only if either: (i) Based on the calculation of the consumer’s debt-to-income ratio for the relevant monthly period and the estimates of the consumer’s basic living expenses for the relevant monthly period, the lender reasonably concludes that: (A) For a covered short-term loan, the consumer can make payments for major financial obligations, make all payments under the loan, and meet basic living expenses during the shorter of the term of the loan or the period ending 45 days after consummation of the loan, and for 30 days after having made the highest payment under the loan; and (B) For a covered longer-term balloon-payment loan, the consumer can make payments for major financial obligations, make all payments under the loan, and meet basic living expenses during the relevant monthly period, and for 30 days after having made the highest payment under the loan; or (ii) Based on the calculation of the consumer’s residual income for the relevant monthly period and the estimates of the consumer’s basic living expenses for the relevant monthly period, the lender reasonably concludes that: (A) For a covered short-term loan, the consumer can make payments for major financial obligations, make all payments under the loan, and meet basic living expenses during the shorter of the term of the loan or the period ending 45 days after consummation of the loan, and for 30 days after having made the highest payment under the loan; and 1518 (B) For a covered longer-term balloon-payment loan, the consumer can make payments for major financial obligations, make all payments under the loan, and meet basic living expenses during the relevant monthly period, and for 30 days after having made the highest payment under the loan. (c) Projecting consumer net income and payments for major financial obligations—(1) General. To make a reasonable determination required under paragraph (b) of this section, a lender must obtain the consumer’s written statement in accordance with paragraph (c)(2)(i), obtain verification evidence to the extent required by paragraph (c)(2)(ii), assess information about rental housing expense as required by paragraph (c)(2)(iii), and use those sources of information to make a reasonable projection of the amount of a consumer’s net income and payments for major financial obligations during the relevant monthly period. The lender must consider major financial obligations that are listed in a consumer’s written statement described in paragraph (c)(2)(i)(B) of this section even if they cannot be verified by the sources listed in paragraph (c)(2)(ii)(B) of this section. To be reasonable, a projection of the amount of net income or payments for major financial obligations may be based on a consumer’s written statement of amounts under paragraph (c)(2)(i) of this section only as specifically permitted by paragraphs (c)(2)(ii) or (iii) or to the extent the stated amounts are consistent with the verification evidence that is obtained in accordance with paragraph (c)(2)(ii) of this section. In determining whether the stated amounts are consistent with the verification evidence, the lender may reasonably consider other reliable evidence the lender obtains from or about the consumer, including any explanations the lender obtains from the consumer. (2) Evidence of net income and payments for major financial obligations—(i) Consumer statements. A lender must obtain a consumer’s written statement of: 1519 (A) The amount of the consumer’s net income, which may include the amount of any income of another person to which the consumer has a reasonable expectation of access; and (B) The amount of payments required for the consumer’s major financial obligations. (ii) Verification evidence. A lender must obtain verification evidence for the amounts of the consumer’s net income and payments for major financial obligations other than rental housing expense, as follows: (A) For the consumer’s net income: (1) The lender must obtain a reliable record (or records) of an income payment (or payments) directly to the consumer covering sufficient history to support the lender’s projection under paragraph (c)(1) if a reliable record (or records) is reasonably available. If a lender determines that a reliable record (or records) of some or all of the consumer’s net income is not reasonably available, then, the lender may reasonably rely on the consumer’s written statement described in paragraph (c)(2)(i)(A) of this section for that portion of the consumer’s net income. (2) If the lender elects to include in the consumer’s net income for the relevant monthly period any income of another person to which the consumer has a reasonable expectation of access, the lender must obtain verification evidence to support the lender’s projection under paragraph (c)(1) of this section. (B) For the consumer’s required payments under debt obligations, the lender must obtain a national consumer report, the records of the lender and its affiliates, and a consumer report obtained from an information system that has been registered for 180 days or more pursuant to § 1041.11(c)(2) or is registered pursuant to § 1041.11(d)(2), if available. If the reports and records do not include a debt obligation listed in the consumer’s written statement described in 1520 paragraph (c)(2)(i)(B) of this section, the lender may reasonably rely on the written statement in determining the amount of the required payment. (C) For a consumer’s required payments under child support obligations or alimony obligations, the lender must obtain a national consumer report. If the report does not include a child support or alimony obligation listed in the consumer’s written statement described in paragraph (c)(2)(i)(B) of this section, the lender may reasonably rely on the written statement in determining the amount of the required payment. (D) Notwithstanding paragraphs (c)(2)(ii)(B) and (c)(2)(ii)(C) of this section, the lender is not required to obtain a national consumer report as verification evidence for the consumer’s debt obligations, alimony obligations, and child support obligations if during the preceding 90 days: (1) The lender or an affiliate obtained a national consumer report for the consumer, retained the report under § 1041.12(b)(1)(ii), and checked it again in connection with the new loan; and (2) The consumer did not complete a loan sequence of three loans made under this section and trigger the prohibition under paragraph (d)(2) of this section since the previous report was obtained. (iii) Rental housing expense. For a consumer’s housing expense other than a payment for a debt obligation that appears on a national consumer report obtained pursuant to paragraph (c)(2)(ii)(B) of this section, the lender may reasonably rely on the consumer’s written statement described in paragraph (c)(2)(i)(B) of this section. (d) Additional limitations on lending—covered short-term loans and covered longer-term balloon-payment loans—(1) Borrowing history review. Prior to making a covered short-term 1521 loan or covered longer-term balloon-payment loan under this section, in order to determine whether any of the prohibitions in this subsection are applicable, a lender must obtain and review information about the consumer’s borrowing history from the records of the lender and its affiliates, and from a consumer report obtained from an information system that has been registered for 180 days or more pursuant to § 1041.11(c)(2) or is registered with the Bureau pursuant to § 1041.11(d)(2), if available. (2) Prohibition on loan sequences of more than three covered short-term loans or covered longer-term balloon-payment loans made under § 1041.5. A lender must not make a covered short-term loan or covered longer-term balloon-payment loan under § 1041.5 during the period in which the consumer has a covered short-term loan or covered longer-term balloonpayment loan made under § 1041.5 outstanding and for 30 days thereafter if the new covered short-term loan or covered longer-term balloon-payment loan would be the fourth loan in a sequence of covered short-term loans, covered longer-term balloon-payment loans, or a combination of covered short-term loans and covered longer-term balloon-payment loans made under § 1041.5. (3) Prohibition on making a covered short-term loan or covered longer-term balloonpayment loan under § 1041.5 following a covered short-term loan made under § 1041.6. A lender must not make a covered short-term loan or covered longer-term balloon-payment loan under § 1041.5 during the period in which the consumer has a covered short-term loan made under § 1041.6 outstanding and for 30 days thereafter. (e) Prohibition against evasion. A lender must not take any action with the intent of evading the requirements of this section. 1522 § 1041.6 Conditional exemption for certain covered short-term loans. (a) Conditional exemption for certain covered short-term loans. Sections 1041.4 and 1041.5 do not apply to a covered short-term loan that satisfies the requirements set forth in paragraphs (b) through (e) of this section. Prior to making a covered short-term loan under this section, a lender must review the consumer’s borrowing history in its own records, the records of the lender’s affiliates, and a consumer report from an information system that has been registered for 180 days or more pursuant to § 1041.11(c)(2) or is registered with the Bureau pursuant to § 1041.11(d)(2). The lender must use this borrowing history information to determine a potential loan’s compliance with the requirements in paragraphs (b) and (c) of this section. (b) Loan term requirements. A covered short-term loan that is made under this section must satisfy the following requirements: (1) The loan satisfies the following principal amount limitations, as applicable: (i) For the first loan in a loan sequence of covered short-term loans made under this section, the principal amount is no greater than $500. (ii) For the second loan in a loan sequence of covered short-term loans made under this section, the principal amount is no greater than two-thirds of the principal amount of the first loan in the loan sequence. (iii) For the third loan in a loan sequence of covered short-term loans made under this section, the principal amount is no greater than one-third of the principal amount of the first loan in the loan sequence. (2) The loan amortizes completely during the term of the loan and the payment schedule provides for the lender allocating a consumer’s payments to the outstanding principal and interest and fees as they accrue only by applying a fixed periodic rate of interest to the 1523 outstanding balance of the unpaid loan principal during every scheduled repayment period for the term of the loan. (3) The lender and any service provider do not take vehicle security as a condition of the loan, as defined in § 1041.2(a)(19). (4) The loan is not structured as open-end credit, as defined in § 1041.2(a)(16). (c) Borrowing history requirements. Prior to making a covered short-term loan under this section, the lender must determine that the following requirements are satisfied: (1) The consumer has not had in the past 30 days an outstanding covered short-term loan under § 1041.5 or covered longer-term balloon-payment loan under § 1041.5; (2) The loan would not result in the consumer having a loan sequence of more than three covered short-term loans under this section; and (3) The loan would not result in the consumer having during any consecutive 12-month period: (i) More than six covered short-term loans outstanding; or (ii) Covered short-term loans outstanding for an aggregate period of more than 90 days. (d) Restrictions on making certain covered loans and non-covered loans following a covered short-term loan made under the conditional exemption. If a lender makes a covered short-term loan under this section to a consumer, the lender or its affiliate must not subsequently make a covered loan, except a covered short-term loan made in accordance with the requirements in § 1041.6, or a non-covered loan to the consumer while the covered short-term loan made under this section is outstanding and for 30 days thereafter. (e) Disclosures. (1) General form of disclosures. 1524 (i) Clear and conspicuous. Disclosures required by this paragraph (e) must be clear and conspicuous. Disclosures required by this section may contain commonly accepted or readily understandable abbreviations. (ii) In writing or electronic delivery. Disclosures required by this paragraph (e) must be provided in writing or through electronic delivery. The disclosures must be provided in a form that can be viewed on paper or a screen, as applicable. This provision is not satisfied by a disclosure provided orally or through a recorded message. (iii) Retainable. Disclosures required by this paragraph (e) must be provided in a retainable form. (iv) Segregation requirements for notices. Notices required by this paragraph (e) must be segregated from all other written or provided materials and contain only the information required by this section, other than information necessary for product identification, branding, and navigation. Segregated additional content that is not required by this paragraph (e) must not be displayed above, below, or around the required content. (v) Machine readable text in notices provided through electronic delivery. If provided through electronic delivery, the notices required by paragraphs (e)(2)(i) and (ii) of this section must use machine readable text that is accessible via both web browsers and screen readers. (vi) Model Forms—(A) First loan notice. The content, order, and format of the notice required by paragraph (e)(2)(i) of this section must be substantially similar to Model Form A-1 in appendix A to this part. (B) Third loan notice. The content, order, and format of the notice required by paragraph (e)(2)(ii) of this section must be substantially similar to Model Form A-2 in appendix A to this part. 1525 (vii) Foreign language disclosures. Disclosures required under this paragraph may be made in a language other than English, provided that the disclosures are made available in English upon the consumer’s request. (2) Notice requirements—(i) First loan notice. A lender that makes a first loan in a sequence of loans made under this section must provide to a consumer a notice that includes, as applicable, the following information and statements, using language substantially similar to the language set forth in Model Form A-1 in appendix A to this part: (A) Identifying statement. The statement “Notice of restrictions on future loans,” using that phrase. (B) Warning for loan made under this section. (1) Possible inability to repay. A statement that warns the consumer not to take out the loan if the consumer is unsure of being able to repay the total amount of principal and finance charges on the loan by the contractual due date. (2) Contractual due date. Contractual due date of the loan made under this section. (3) Total amount due. Total amount due on the contractual due date. (C) Restriction on a subsequent loan required by Federal law. A statement that informs a consumer that Federal law requires a similar loan taken out within the next 30 days to be smaller. (D) Borrowing limits. In a tabular form: (1) Maximum principal amount on loan 1 in a sequence of loans made under this section. (2) Maximum principal amount on loan 2 in a sequence of loans made under this section. (3) Maximum principal amount on loan 3 in a sequence of loans made under this section. (4) Loan 4 in a sequence of loans made under this section is not allowed. 1526 (E) Lender name and contact information. Name of the lender and a telephone number for the lender and, if applicable, a URL of the website for the lender. (ii) Third loan notice. A lender that makes a third loan in a sequence of loans made under this section must provide to a consumer a notice that includes the following information and statements, using language substantially similar to the language set forth in Model Form A-2 in appendix A to this part: (A) Identifying statement. The statement “Notice of borrowing limits on this loan and future loans,” using that phrase. (B) Two similar loans without 30-day break. A statement that informs a consumer that the lender’s records show that the consumer has had two similar loans without taking at least a 30-day break between them. (C) Restriction on loan amount required by Federal law. A statement that informs a consumer that Federal law requires the third loan to be smaller than previous loans in the loan sequence. (D) Prohibition on subsequent loan. A statement that informs a consumer that the consumer cannot take out a similar loan for at least 30 days after repaying the loan. (E) Lender name and contact information. Name of the lender and a telephone number for the lender and, if applicable, a URL of the website for the lender. (3) Timing. A lender must provide the notices required in paragraphs (e)(2)(i) and (e)(2)(ii) of this section to the consumer before the applicable loan under § 1041.6 is consummated. 1527 Subpart C—Payments § 1041.7 Identification of unfair and abusive practice. It is an unfair and abusive practice for a lender to make attempts to withdraw payment from consumers’ accounts in connection with a covered loan after the lender’s second consecutive attempts to withdraw payments from the accounts from which the prior attempts were made have failed due to a lack of sufficient funds, unless the lender obtains the consumers’ new and specific authorization to make further withdrawals from the accounts. § 1041.8 Prohibited payment transfer attempts. (a) Definitions. For purposes of this section and § 1041.9: (1) Payment transfer means any lender-initiated debit or withdrawal of funds from a consumer’s account for the purpose of collecting any amount due or purported to be due in connection with a covered loan. (i) Means of transfer. A debit or withdrawal meeting the description in paragraph (a)(1) of this section is a payment transfer regardless of the means through which the lender initiates it, including but not limited to a debit or withdrawal initiated through any of the following means: (A) Electronic fund transfer, including a preauthorized electronic fund transfer as defined in Regulation E, 12 CFR 1005.2(k). (B) Signature check, regardless of whether the transaction is processed through the check network or another network, such as the automated clearing house (ACH) network. (C) Remotely created check as defined in Regulation CC, 12 CFR 229.2(fff). (D) Remotely created payment order as defined in 16 CFR 310.2(cc). 1528 (E) When the lender is also the account-holder, an account-holding institution’s transfer of funds from a consumer’s account held at the same institution, other than such a transfer meeting the description in paragraph (a)(1)(ii) of this section. (ii) Conditional exclusion for certain transfers by account-holding institutions. When the lender is also the account-holder, an account-holding institution’s transfer of funds from a consumer’s account held at the same institution is not a payment transfer if all of the conditions in this paragraph (a)(1)(ii) are met, notwithstanding that the transfer otherwise meets the description in paragraph (a)(1) of this section. (A) The lender, pursuant to the terms of the loan agreement or account agreement, does not charge the consumer any fee, other than a late fee under the loan agreement, in the event that the lender initiates a transfer of funds from the consumer’s account in connection with the covered loan for an amount that the account lacks sufficient funds to cover. (B) The lender, pursuant to the terms of the loan agreement or account agreement, does not close the consumer’s account in response to a negative balance that results from a transfer of funds initiated in connection with the covered loan. (2) Single immediate payment transfer at the consumer’s request means: (i) A payment transfer initiated by a one-time electronic fund transfer within one business day after the lender obtains the consumer’s authorization for the one-time electronic fund transfer. (ii) A payment transfer initiated by means of processing the consumer’s signature check through the check system or through the ACH system within one business day after the consumer provides the check to the lender. 1529 (b) Prohibition on initiating payment transfers from a consumer’s account after two consecutive failed payment transfers—(1) General. A lender must not initiate a payment transfer from a consumer’s account in connection with any covered loan that the consumer has with the lender after the lender has attempted to initiate two consecutive failed payment transfers from that account in connection with any covered loan that the consumer has with the lender. For purposes of this paragraph (b), a payment transfer is deemed to have failed when it results in a return indicating that the consumer’s account lacks sufficient funds or, if the lender is the consumer’s account-holding institution, it is for an amount that the account lacks sufficient funds to cover. (2) Consecutive failed payment transfers. For purposes of the prohibition in this paragraph (b): (i) First failed payment transfer. A failed payment transfer is the first failed payment transfer from the consumer’s account if it meets any of the following conditions: (A) The lender has initiated no other payment transfer from the account in connection with the covered loan or any other covered loan that the consumer has with the lender. (B) The immediately preceding payment transfer was successful, regardless of whether the lender has previously initiated a first failed payment transfer. (C) The payment transfer is the first payment transfer to fail after the lender obtains the consumer’s authorization for additional payment transfers pursuant to paragraph (c) of this section. (ii) Second consecutive failed payment transfer. A failed payment transfer is the second consecutive failed payment transfer from the consumer’s account if the immediately preceding payment transfer was a first failed payment transfer. For purposes of this paragraph (b)(2)(ii), a 1530 previous payment transfer includes a payment transfer initiated at the same time or on the same day as the failed payment transfer. (iii) Different payment channel. A failed payment transfer meeting the conditions in paragraph (b)(2)(ii) of this section is the second consecutive failed payment transfer regardless of whether the first failed payment transfer was initiated through a different payment channel. (c) Exception for additional payment transfers authorized by the consumer—(1) General. Notwithstanding the prohibition in paragraph (b) of this section, a lender may initiate additional payment transfers from a consumer’s account after two consecutive failed payment transfers if the additional payment transfers are authorized by the consumer in accordance with the requirements and conditions in this paragraph (c) or if the lender executes a single immediate payment transfer at the consumer’s request in accordance with paragraph (d) of this section. (2) General authorization requirements and conditions—(i) Required payment transfer terms. For purposes of this paragraph (c), the specific date, amount, and payment channel of each additional payment transfer must be authorized by the consumer, except as provided in paragraph (c)(2)(ii) or (iii) of this section. (ii) Application of specific date requirement to re-initiating a returned payment transfer. If a payment transfer authorized by the consumer pursuant to this paragraph (c) is returned for nonsufficient funds, the lender may re-initiate the payment transfer, such as by re-presenting it once through the ACH system, on or after the date authorized by the consumer, provided that the returned payment transfer has not triggered the prohibition in paragraph (b) of this section. (iii) Special authorization requirements and conditions for payment transfers to collect a late fee or returned item fee. A lender may initiate a payment transfer pursuant to this paragraph (c) solely to collect a late fee or returned item fee without obtaining the consumer’s authorization 1531 for the specific date and amount of the payment transfer only if the consumer has authorized the lender to initiate such payment transfers in advance of the withdrawal attempt. For purposes of this paragraph (c)(2)(iii), the consumer authorizes such payment transfers only if the consumer’s authorization obtained under paragraph (c)(3)(iii) of this section includes a statement, in terms that are clear and readily understandable to the consumer, that payment transfers may be initiated solely to collect a late fee or returned item fee and specifies the highest amount for such fees that may be charged and the payment channel to be used. (3) Requirements and conditions for obtaining the consumer’s authorization—(i) General. For purposes of this paragraph (c), the lender must request and obtain the consumer’s authorization for additional payment transfers in accordance with the requirements and conditions in this paragraph (c)(3). (ii) Provision of payment transfer terms to the consumer. The lender may request the consumer’s authorization for additional payment transfers no earlier than the date on which the lender provides to the consumer the consumer rights notice required by § 1041.9(c). The request must include the payment transfer terms required under paragraph (c)(2)(i) of this section and, if applicable, the statements required by paragraph (c)(2)(iii) of this section. The lender may provide the terms and statements to the consumer by any one of the following means: (A) In writing, by mail or in person, or in a retainable form by email if the consumer has consented to receive electronic disclosures in this manner under § 1041.9(a)(4) or agrees to receive the terms and statements by email in the course of a communication initiated by the consumer in response to the consumer rights notice required by § 1041.9(c). (B) By oral telephone communication, if the consumer affirmatively contacts the lender in that manner in response to the consumer rights notice required by § 1041.9(c) and agrees to 1532 receive the terms and statements in that manner in the course of, and as part of, the same communication. (iii) Signed authorization required—(A) General. For an authorization to be valid under this paragraph (c), it must be signed or otherwise agreed to by the consumer in writing or electronically and in a retainable format that memorializes the payment transfer terms required under paragraph (c)(2)(i) of this section and, if applicable, the statements required by paragraph (c)(2)(iii) of this section. The signed authorization must be obtained from the consumer no earlier than when the consumer receives the consumer rights notice required by § 1041.9(c) in person or electronically, or the date on which the consumer receives the notice by mail. For purposes of this paragraph (c)(3)(iii)(A), the consumer is considered to have received the notice at the time it is provided to the consumer in person or electronically, or, if the notice is provided by mail, the earlier of the third business day after mailing or the date on which the consumer affirmatively responds to the mailed notice. (B) Special requirements for authorization obtained by oral telephone communication. If the authorization is granted in the course of an oral telephone communication, the lender must record the call and retain the recording. (C) Memorialization required. If the authorization is granted in the course of a recorded telephonic conversation or is otherwise not immediately retainable by the consumer at the time of signature, the lender must provide a memorialization in a retainable form to the consumer by no later than the date on which the first payment transfer authorized by the consumer is initiated. A memorialization may be provided to the consumer by email in accordance with the requirements and conditions in paragraph (c)(3)(ii)(A) of this section. 1533 (4) Expiration of authorization. An authorization obtained from a consumer pursuant to this paragraph (c) becomes null and void for purposes of the exception in this paragraph (c) if: (i) The lender subsequently obtains a new authorization from the consumer pursuant to this paragraph (c); or (ii) Two consecutive payment transfers initiated pursuant to the consumer’s authorization fail, as specified in paragraph (b) of this section. (d) Exception for initiating a single immediate payment transfer at the consumer’s request. After a lender’s second consecutive payment transfer has failed as specified in paragraph (b) of this section, the lender may initiate a payment transfer from the consumer’s account without obtaining the consumer’s authorization for additional payment transfers pursuant to paragraph (c) of this section if: (1) The payment transfer is a single immediate payment transfer at the consumer’s request as defined in paragraph (a)(2) of this section; and (2) The consumer authorizes the underlying one-time electronic fund transfer or provides the underlying signature check to the lender, as applicable, no earlier than the date on which the lender provides to the consumer the consumer rights notice required by § 1041.9(c) or on the date that the consumer affirmatively contacts the lender to discuss repayment options, whichever date is earlier. (e) Prohibition against evasion. A lender must not take any action with the intent of evading the requirements of this section. 1534 § 1041.9 Disclosure of payment transfer attempts. (a) General form of disclosures—(1) Clear and conspicuous. Disclosures required by this section must be clear and conspicuous. Disclosures required by this section may contain commonly accepted or readily understandable abbreviations. (2) In writing or electronic delivery. Disclosures required by this section must be provided in writing or, so long as the requirements of paragraph (a)(4) of this section are satisfied, through electronic delivery. The disclosures must be provided in a form that can be viewed on paper or a screen, as applicable. This provision is not satisfied by a disclosure provided orally or through a recorded message. (3) Retainable. Disclosures required by this section must be provided in a retainable form, except for electronic short notices delivered by mobile application or text message under paragraph (b) or (c) of this section. (4) Electronic delivery. Disclosures required by this section may be provided through electronic delivery if the following consent requirements are satisfied: (i) Consumer consent—(A) General. Disclosures required by this section may be provided through electronic delivery if the consumer affirmatively consents in writing or electronically to the particular electronic delivery method. (B) Email option required. To obtain valid consumer consent to electronic delivery under this paragraph, a lender must provide the consumer with the option to select email as the method of electronic delivery, separate and apart from any other electronic delivery methods such as mobile application or text message. 1535 (ii) Subsequent loss of consent. Notwithstanding paragraph (a)(4)(i) of this section, a lender must not provide disclosures required by this section through a method of electronic delivery if: (A) The consumer revokes consent to receive disclosures through that delivery method; or (B) The lender receives notification that the consumer is unable to receive disclosures through that delivery method at the address or number used. (5) Segregation requirements for notices. All notices required by this section must be segregated from all other written or provided materials and contain only the information required by this section, other than information necessary for product identification, branding, and navigation. Segregated additional content that is not required by this section must not be displayed above, below, or around the required content. (6) Machine readable text in notices provided through electronic delivery. If provided through electronic delivery, the payment notice required by paragraph (b) and the consumer rights notice required by paragraph (c) of this section must use machine readable text that is accessible via both web browsers and screen readers. (7) Model Forms—(i) Payment notice. The content, order, and format of the payment notice required by paragraph (b) of this section must be substantially similar to Model Forms A3 through A-4 in appendix A to this part. (ii) Consumer rights notice. The content, order, and format of the consumer rights notice required by paragraph (c) of this section must be substantially similar to Model Form A-5 in appendix A to this part. 1536 (iii) Electronic short notice. The content, order, and format of the electronic short notice required by paragraph (b) of this section must be substantially similar to Model Clauses A-6 and A-7 in appendix A to this part. The content, order, and format of the electronic short notice required by paragraph (c) of this section must be substantially similar to Model Clause A-8 in appendix A to this part. (8) Foreign language disclosures. Disclosures required under this section may be made in a language other than English, provided that the disclosures are made available in English upon the consumer's request. (b) Payment notice—(1) General. Prior to initiating the first payment withdrawal or an unusual withdrawal from a consumer’s account, a lender must provide to the consumer a payment notice in accordance with the requirements in this paragraph (b) as applicable. (i) First payment withdrawal means the first payment transfer scheduled to be initiated by a lender for a particular covered loan, not including a single immediate payment transfer initiated at the consumer’s request as defined in § 1041.8(a)(2). (ii) Unusual withdrawal means a payment transfer that meets one or more of the conditions described in paragraph (b)(3)(ii)(C) of this section. (iii) Exceptions. The payment notice need not be provided when the lender initiates: (A) The initial payment transfer from a consumer’s account after obtaining consumer authorization pursuant to § 1041.8(c), regardless of whether any of the conditions in paragraph (b)(3)(ii)(C) of this section apply; or (B) A single immediate payment transfer initiated at the consumer’s request in accordance with § 1041.8(a)(2). 1537 (2) First payment withdrawal notice—(i) Timing. (A) Mail. If the lender provides the first payment withdrawal notice by mail, the lender must mail the notice no earlier than when the lender obtains payment authorization and no later than six business days prior to initiating the transfer. (B) Electronic delivery. (1) If the lender provides the first payment withdrawal notice through electronic delivery, the lender must send the notice no earlier than when the lender obtains payment authorization and no later than three business days prior to initiating the transfer. (2) If, after providing the first payment withdrawal notice through electronic delivery pursuant to the timing requirements in paragraph (b)(2)(i) of this section, the lender loses the consumer’s consent to receive the notice through a particular electronic delivery method according to paragraph (a)(4)(ii) of this section, the lender must provide notice of any future unusual withdrawal, if applicable, through alternate means. (C) In person. If the lender provides the first payment withdrawal notice in person, the lender must provide the notice no earlier than when the lender obtains payment authorization and no later than three business days prior to initiating the transfer. (ii) Content requirements. The notice must contain the following information and statements, as applicable, using language substantially similar to the language set forth in Model Form A-3 in appendix A to this part: (A) Identifying statement. The statement, “Upcoming Withdrawal Notice,” using that phrase, and, in the same statement, the name of the lender providing the notice. (B) Transfer terms—(1) Date. Date that the lender will initiate the transfer. (2) Amount. Dollar amount of the transfer. 1538 (3) Consumer account. Sufficient information to permit the consumer to identify the account from which the funds will be transferred. The lender must not provide the complete account number of the consumer, but may use a truncated version similar to Model Form A-5 in appendix A to this part. (4) Loan identification information. Sufficient information to permit the consumer to identify the covered loan associated with the transfer. (5) Payment channel. Payment channel of the transfer. (6) Check number. If the transfer will be initiated by a signature or paper check, remotely created check (as defined in Regulation CC, 12 CFR 229.2(fff)), or remotely created payment order (as defined in 16 CFR 310.2 (cc)), the check number associated with the transfer. (C) Payment breakdown. In a tabular form: (1) Payment breakdown heading. A heading with the statement “Payment Breakdown,” using that phrase. (2) Principal. The amount of the payment that will be applied to principal. (3) Interest. The amount of the payment that will be applied to accrued interest on the loan. (4) Fees. If applicable, the amount of the payment that will be applied to fees. (5) Other charges. If applicable, the amount of the payment that will be applied to other charges. (6) Amount. The statement “Total Payment Amount,” using that phrase, and the total dollar amount of the payment as provided in paragraph (b)(2)(ii)(B)(2) of this section. (7) Explanation of interest-only or negatively amortizing payment. If applicable, a statement explaining that the payment will not reduce principal, using the applicable phrase 1539 “When you make this payment, your principal balance will stay the same and you will not be closer to paying off your loan” or “When you make this payment, your principal balance will increase and you will not be closer to paying off your loan.” (D) Lender name and contact information. Name of the lender, the name under which the transfer will be initiated (if different from the consumer-facing name of the lender), and two different forms of lender contact information that may be used by the consumer to obtain information about the consumer’s loan. (3) Unusual withdrawal notice—(i) Timing. (A) Mail. If the lender provides the unusual withdrawal notice by mail, the lender must mail the notice no earlier than 10 business days and no later than six business days prior to initiating the transfer. (B) Electronic delivery. (1) If the lender provides the unusual withdrawal notice through electronic delivery, the lender must send the notice no earlier than seven business days and no later than three business days prior to initiating the transfer. (2) If, after providing the unusual withdrawal notice through electronic delivery pursuant to the timing requirements in paragraph (b)(3)(i)(B) of this section, the lender loses the consumer’s consent to receive the notice through a particular electronic delivery method according to paragraph (a)(4)(ii) of this section, the lender must provide notice of any future unusual withdrawal attempt, if applicable, through alternate means. (C) In person. If the lender provides the unusual withdrawal notice in person, the lender must provide the notice no earlier than seven business days and no later than three business days prior to initiating the transfer. (D) Exception for open-end credit plans. If the unusual withdrawal notice is for an openend credit plan as defined in § 1041.2(a)(16), the lender may provide the unusual withdrawal 1540 notice in conjunction with the periodic statement required under Regulation Z, 12 CFR 1026.7(b), in accordance with the timing requirements of that section. (ii) Content requirements. The unusual withdrawal notice must contain the following information and statements, as applicable, using language substantially similar to the language set forth in Model Form A-4 in appendix A to this part: (A) Identifying statement. The statement, “Alert: Unusual Withdrawal,” using that phrase, and, in the same statement, the name of the lender that is providing the notice. (B) Basic payment information. The content required for the first withdrawal notice under paragraphs (b)(2)(ii)(B) through (D) of this section. (C) Description of unusual withdrawal. The following content, as applicable, in a form substantially similar to the form in Model Form A-4 in appendix A to this part: (1) Varying amount—(i) General. If the amount of a transfer will vary in amount from the regularly scheduled payment amount, a statement that the transfer will be for a larger or smaller amount than the regularly scheduled payment amount, as applicable. (ii) Open-end credit plans. If the payment transfer is for an open-end credit plan as defined in § 1041.2(a)(16), the varying amount content is required only if the amount deviates from the scheduled minimum payment due as disclosed in the periodic statement required under Regulation Z, 12 CFR 1026.7(b). (2) Date other than date of regularly scheduled payment. If the payment transfer date is not a date on which a regularly scheduled payment is due under the terms of the loan agreement, a statement that the transfer will be initiated on a date other than the date of a regularly scheduled payment. 1541 (3) Different payment channel. If the payment channel will differ from the payment channel of the transfer directly preceding it, a statement that the transfer will be initiated through a different payment channel and a statement of the payment channel used for the prior transfer. (4) For purpose of re-initiating returned transfer. If the transfer is for the purpose of reinitiating a returned transfer, a statement that the lender is re-initiating a returned transfer, a statement of the date and amount of the previous unsuccessful attempt, and a statement of the reason for the return. (4) Electronic delivery—(i) General. When the consumer has consented to receive disclosures through electronic delivery, the lender may provide the applicable payment notice required by paragraph (b)(1) of this section through electronic delivery only if it also provides an electronic short notice, except for email delivery as provided in paragraph (b)(4)(iii) of this section. (ii) Electronic short notice—(A) General content. The electronic short notice required by this paragraph (b) must contain the following information and statements, as applicable, in a form substantially similar to Model Clause A-6 in appendix A to this part: (1) Identifying statement, as required under paragraphs (b)(2)(ii)(A) and (b)(3)(ii)(A) of this section; (2) Transfer terms. (i) Date, as required under paragraphs (b)(2)(ii)(B)(1) and (b)(3)(ii)(B) of this section; (ii) Amount, as required under paragraphs (b)(2)(ii)(B)(2) and (b)(3)(ii)(B) of this section; (iii) Consumer account, as required and limited under paragraphs (b)(2)(ii)(B)(3) and (b)(3)(ii)(B) of this section; 1542 (3) Website URL. When the full notice is being provided through a linked URL rather than as a pdf attachment, the unique URL of a website that the consumer may use to access the full payment notice required by paragraph (b) of this section. (B) Additional content requirements. If the transfer meets any of the conditions for unusual attempts described in paragraph (b)(3)(ii)(C) of this section, the electronic short notice must also contain the following information and statements, as applicable, using language substantially similar to the language in Model Clause A-7 in appendix A to this part: (1) Varying amount, as defined under paragraph (b)(3)(ii)(C)(1) of this section; (2) Date other than due date of regularly scheduled payment, as defined under paragraph (b)(3)(ii)(C)(2) of this section; and (3) Different payment channel, as defined under paragraph (b)(3)(ii)(C)(3) of this section. (iii) Email delivery. When the consumer has consented to receive disclosures through electronic delivery, and the method of electronic delivery is email, the lender may either deliver the full notice required by paragraph (b)(1) of this section in the body of the email or deliver the full notice as a linked URL webpage or PDF attachment along with the electronic short notice as provided in paragraph (b)(4)(ii) of this section. (c) Consumer rights notice—(1) General. After a lender initiates two consecutive failed payment transfers from a consumer’s account as described in § 1041.8(b), the lender must provide to the consumer a consumer rights notice in accordance with the following requirements: (2) Timing. The lender must send the notice no later than three business days after it receives information that the second consecutive attempt has failed. 1543 (3) Content requirements. The notice must contain the following information and statements, using language substantially similar to the language set forth in Model Form A-5 in appendix A to this part: (i) Identifying statement. A statement that the lender, identified by name, is no longer permitted to withdraw loan payments from the consumer’s account. (ii) Last two attempts were returned. A statement that the lender’s last two attempts to withdraw payment from the consumer’s account were returned due to non-sufficient funds, or, if applicable to payments initiated by the consumer's account-holding institution, caused the account to go into overdraft status. (iii) Consumer Account. Sufficient information to permit the consumer to identify the account from which the unsuccessful payment attempts were made. The lender must not provide the complete account number of the consumer, but may use a truncated version similar to Model Form A-5 in appendix A to this part. (iv) Loan identification information. Sufficient information to permit the consumer to identify any covered loans associated with the unsuccessful payment attempts. (v) Statement of Federal law prohibition. A statement, using that phrase, that in order to protect the consumer’s account, Federal law prohibits the lender from initiating further payment transfers without the consumer’s permission. (vi) Contact about choices. A statement that the lender may be in contact with the consumer about payment choices going forward. (vii) Previous unsuccessful payment attempts. In a tabular form: (A) Previous payment attempts heading. A heading with the statement “previous payment attempts.” 1544 (B) Payment due date. The scheduled due date of each previous unsuccessful payment transfer attempted by the lender. (C) Date of attempt. The date of each previous unsuccessful payment transfer initiated by the lender. (D) Amount. The amount of each previous unsuccessful payment transfer initiated by the lender. (E) Fees. The fees charged by the lender for each unsuccessful payment attempt, if applicable, with an indication that these fees were charged by the lender. (viii) CFPB information. A statement, using that phrase, that the Consumer Financial Protection Bureau created this notice, a statement that the CFPB is a Federal government agency, and the URL to www.consumerfinance.gov/payday-rule]. This statement must be the last piece of information provided in the notice. (4) Electronic delivery—(i) General. When the consumer has consented to receive disclosures through electronic delivery, the lender may provide the consumer rights notice required by paragraph (c) through electronic delivery only if it also provides an electronic short notice, except for email delivery as provided in paragraph (c)(4)(iii) of this section: (ii) Electronic short notice—(A) Content. The notice must contain the following information and statements, as applicable, using language substantially similar to the language set forth in Model Clause A-8 in appendix A to this part: (1) Identifying statement, as required under paragraph (c)(3)(i) of this section; (2) Last two attempts were returned, as required under paragraph (c)(3)(ii) of this section; (3) Consumer account, as required and limited under paragraph (c)(3)(iii) of this section; 1545 (4) Statement of Federal law prohibition, as required under paragraph (c)(3)(v) of this section; and (5) Website URL. When the full notice is being provided through a linked URL rather than as a PDF attachment, the unique URL of a website that the consumer may use to access the full consumer rights notice required by paragraph (c) of this section. (iii) Email delivery. When the consumer has consented to receive disclosures through electronic delivery, and the method of electronic delivery is email, the lender may either deliver the full notice required by paragraph (c)(1) of this section in the body of the email or deliver the full notice as a linked URL webpage or PDF attachment along with the electronic short notice as provided in paragraph (c)(4)(ii) of this section. Subpart D—Information Furnishing, Recordkeeping, Anti-Evasion, and Severability § 1041.10 Information furnishing requirements. (a) Loans subject to furnishing requirement. For each covered short-term loan and covered longer-term balloon-payment loan a lender makes, the lender must furnish the loan information described in paragraph (c) of this section to each information system described in paragraph (b)(1) of this section. (b) Information systems to which information must be furnished. (1) A lender must furnish information as required in paragraphs (a) and (c) of this section to each information system that, as of the date the loan is consummated: (i) Has been registered with the Bureau pursuant to § 1041.11(c)(2) for 180 days or more; or 1546 (ii) Has been provisionally registered with the Bureau pursuant to § 1041.11(d)(1) for 180 days or more or subsequently has become registered with the Bureau pursuant to § 1041.11(d)(2). (2) The Bureau will publish on its website and in the Federal Register notice of the provisional registration of an information system pursuant to § 1041.11(d)(1), registration of an information system pursuant to § 1041.11(c)(2) or (d)(2), and suspension or revocation of the provisional registration or registration of an information system pursuant to § 1041.11(h). For purposes of paragraph (b)(1) of this section, an information system is provisionally registered or registered, and its provisional registration or registration is suspended or revoked, on the date that the Bureau publishes notice of such provisional registration, registration, suspension, or revocation on its website. The Bureau will maintain on the Bureau’s website a current list of information systems provisionally registered pursuant to § 1041.11(d)(1) and registered pursuant to § 1041.11(c)(2) and (d)(2). In the event that a provisional registration or registration of an information system is suspended, the Bureau will provide instructions on its website concerning the scope and terms of the suspension or revocation. (c) Information to be furnished. A lender must furnish the information described in this paragraph, at the times described in this paragraph, concerning each covered loan as required in paragraphs (a) and (b) of this section. A lender must furnish the information in a format acceptable to each information system to which it must furnish information. (1) Information to be furnished at loan consummation. A lender must furnish the following information no later than the date on which the loan is consummated or as close in time as feasible to the date the loan is consummated: (i) Information necessary to uniquely identify the loan; 1547 (ii) Information necessary to allow the information system to identify the specific consumer(s) responsible for the loan; (iii) Whether the loan is a covered short-term loan or a covered longer-term balloonpayment loan; (iv) Whether the loan is made under § 1041.5 or § 1041.6, as applicable; (v) The loan consummation date; (vi) For a loan made under § 1041.6, the principal amount borrowed; (vii) For a loan that is closed-end credit: (A) The fact that the loan is closed-end credit; (B) The date that each payment on the loan is due; and (C) The amount due on each payment date. (viii) For a loan that is open-end credit: (A) The fact that the loan is open-end credit; (B) The credit limit on the loan; (C) The date that each payment on the loan is due; and (D) The minimum amount due on each payment date. (2) Information to be furnished while loan is an outstanding loan. During the period that the loan is an outstanding loan, a lender must furnish any update to information previously furnished pursuant to this section within a reasonable period of the event that causes the information previously furnished to be out of date. (3) Information to be furnished when loan ceases to be an outstanding loan. A lender must furnish the following information no later than the date the loan ceases to be an outstanding loan or as close in time as feasible to the date the loan ceases to be an outstanding loan: 1548 (i) The date as of which the loan ceased to be an outstanding loan; and (ii) Whether all amounts owed in connection with the loan were paid in full, including the amount financed, charges included in the cost of credit, and charges excluded from the cost of credit. § 1041.11 Registered information systems. (a) Definitions. (1) Consumer report has the same meaning as in section 603(d) of the Fair Credit Reporting Act, 15 U.S.C. 1681a(d). (2) Federal consumer financial law has the same meaning as in section 1002(14) of the Dodd-Frank Wall Street Reform and Consumer Protection Act, 12 U.S.C. 5481(14). (b) Eligibility criteria for information systems. An entity is eligible to be a provisionally registered information system pursuant to paragraph (d)(1) of this section or a registered information system pursuant to paragraph (c)(2) or (d)(2) of this section only if the Bureau determines that the following conditions are satisfied: (1) Receiving capability. The entity possesses the technical capability to receive information lenders must furnish pursuant to § 1041.10 immediately upon the furnishing of such information and uses reasonable data standards that facilitate the timely and accurate transmission and processing of information in a manner that does not impose unreasonable costs or burdens on lenders. (2) Reporting capability. The entity possesses the technical capability to generate a consumer report containing, as applicable for each unique consumer, all information described in § 1041.10 substantially simultaneous to receiving the information from a lender. (3) Performance. The entity will perform or performs in a manner that facilitates compliance with and furthers the purposes of this part. 1549 (4) Federal consumer financial law compliance program. The entity has developed, implemented, and maintains a program reasonably designed to ensure compliance with all applicable Federal consumer financial laws, which includes written policies and procedures, comprehensive training, and monitoring to detect and to promptly correct compliance weaknesses. (5) Independent assessment of Federal consumer financial law compliance program. The entity provides to the Bureau in its application for provisional registration or registration a written assessment of the Federal consumer financial law compliance program described in paragraph (b)(4) of this section and such assessment: (i) Sets forth a detailed summary of the Federal consumer financial law compliance program that the entity has implemented and maintains; (ii) Explains how the Federal consumer financial law compliance program is appropriate for the entity’s size and complexity, the nature and scope of its activities, and risks to consumers presented by such activities; (iii) Certifies that, in the opinion of the assessor, the Federal consumer financial law compliance program is operating with sufficient effectiveness to provide reasonable assurance that the entity is fulfilling its obligations under all Federal consumer financial laws; and (iv) Certifies that the assessment has been conducted by a qualified, objective, independent third-party individual or entity that uses procedures and standards generally accepted in the profession, adheres to professional and business ethics, performs all duties objectively, and is free from any conflicts of interest that might compromise the assessor’s independent judgment in performing assessments. 1550 (6) Information security program. The entity has developed, implemented, and maintains a comprehensive information security program that complies with the Standards for Safeguarding Customer Information, 16 CFR part 314. (7) Independent assessment of information security program. (i) The entity provides to the Bureau in its application for provisional registration or registration and on at least a biennial basis thereafter, a written assessment of the information security program described in paragraph (b)(6) of this section and such assessment: (A) Sets forth the administrative, technical, and physical safeguards that the entity has implemented and maintains; (B) Explains how such safeguards are appropriate to the entity’s size and complexity, the nature and scope of its activities, and the sensitivity of the customer information at issue; (C) Explains how the safeguards that have been implemented meet or exceed the protections required by the Standards for Safeguarding Customer Information, 16 CFR part 314; (D) Certifies that, in the opinion of the assessor, the information security program is operating with sufficient effectiveness to provide reasonable assurance that the entity is fulfilling its obligations under the Standards for Safeguarding Customer Information, 16 CFR part 314; and (E) Certifies that the assessment has been conducted by a qualified, objective, independent third-party individual or entity that uses procedures and standards generally accepted in the profession, adheres to professional and business ethics, performs all duties objectively, and is free from any conflicts of interest that might compromise the assessor’s independent judgment in performing assessments. 1551 (ii) Each written assessment obtained and provided to the Bureau on at least a biennial basis pursuant to paragraph (b)(7)(i) of this section must be completed and provided to the Bureau within 60 days after the end of the period to which the assessment applies. (8) Bureau supervisory authority. The entity acknowledges it is, or consents to being, subject to the Bureau’s supervisory authority. (c) Registration of information systems prior to effective date of § 1041.10—(1) Preliminary approval. Prior to the effective date of § 1041.10, the Bureau may preliminarily approve an entity for registration only if the entity submits an application for preliminary approval to the Bureau by the deadline set forth in paragraph (c)(3)(i) of this section containing information sufficient for the Bureau to determine that the entity is reasonably likely to satisfy the conditions set forth in paragraph (b) of this section by the deadline set forth in paragraph (c)(3)(ii) of this section. The assessments described in paragraphs (b)(5) and (b)(7) of this section need not be included with an application for preliminary approval for registration or completed prior to the submission of the application. The Bureau may require additional information and documentation to facilitate this determination. (2) Registration. Prior to the effective date of § 1041.10, the Bureau may approve the application of an entity to be a registered information system only if: (i) The entity received preliminary approval pursuant to paragraph (c)(1) of this section; and (ii) The entity submits an application to the Bureau by the deadline set forth in paragraph (c)(3)(ii) of this section that contains information and documentation sufficient for the Bureau to determine that the entity satisfies the conditions set forth in paragraph (b) of this section. The Bureau may require additional information and documentation to facilitate this determination or 1552 otherwise to assess whether registration of the entity would pose an unreasonable risk to consumers. (3) Deadlines. (i) The deadline to submit an application for preliminary approval for registration pursuant to paragraph (c)(1) of this section is 90 days from the effective date of this section. (ii) The deadline to submit an application to be a registered information system pursuant to paragraph (c)(2) of this section is 120 days from the date preliminary approval for registration is granted. (iii) The Bureau may waive the deadlines set forth in this paragraph. (d) Registration of information systems on or after effective date of § 1041.10—(1) Provisional registration. On or after the effective date of § 1041.10, the Bureau may approve an entity to be a provisionally registered information system only if the entity submits an application to the Bureau that contains information and documentation sufficient for the Bureau to determine that the entity satisfies the conditions set forth in paragraph (b) of this section. The Bureau may require additional information and documentation to facilitate this determination or otherwise to assess whether provisional registration of the entity would pose an unreasonable risk to consumers. (2) Registration. An information system that is provisionally registered pursuant to paragraph (d)(1) of this section shall automatically become a registered information system pursuant to this paragraph (d)(2) upon the expiration of the 240-day period commencing on the date the information system is provisionally registered. For purposes of this paragraph, an information system is provisionally registered on the date that the Bureau publishes notice of the provisional registration on the Bureau’s website. 1553 (e) Applications. Applications for preliminary approval, registration, and provisional registration shall be submitted in the form required by the Bureau and shall include, in addition to the information described in paragraph (c) or (d) of this section, as applicable, the following information: (1) The name under which the applicant conducts business, including any “doing business as” or other trade name; (2)The applicant’s main business address, mailing address if it is different from the main business address, telephone number, electronic mail address, and Internet website; and (3) The name and contact information (including telephone number and electronic mail address) of the person authorized to communicate with the Bureau on the applicant’s behalf concerning the application. (f) Denial of application. The Bureau will deny the application of an entity seeking preliminary approval for registration under paragraph (c)(1) of this section, registration under paragraph (c)(2) of this section, or provisional registration under paragraph (d)(1) of this section, if the Bureau determines, as applicable, that: (1) The entity does not satisfy the conditions set forth in paragraph (b) of this section, or, in the case of an entity seeking preliminary approval for registration, is not reasonably likely to satisfy the conditions as of the deadline set forth in paragraph (c)(3)(ii) of this section; (2) The entity’s application is untimely or materially inaccurate or incomplete; or (3) Preliminary approval, provisional registration, or registration of the entity would pose an unreasonable risk to consumers. (g) Notice of material change. An entity that is a provisionally registered or registered information system must provide to the Bureau in writing a description of any material change to 1554 information contained in its application for registration submitted pursuant to paragraph (c)(2) of this section or provisional registration submitted pursuant to paragraph (d)(1) of this section, or to information previously provided to the Bureau pursuant to this paragraph, within 14 days of such change. (h) Suspension and Revocation. (1) The Bureau will suspend or revoke an entity’s preliminary approval for registration pursuant to paragraph (c)(1) of this section, provisional registration pursuant to paragraph (d)(1) of this section, or registration pursuant to paragraphs (c)(2) or (d)(2) of this section if the Bureau determines: (i) That the entity has not satisfied or no longer satisfies the conditions described in paragraph (b) of this section or has not complied with the requirement described in paragraph (f) of this section; or (ii) That preliminary approval, provisional registration, or registration of the entity poses an unreasonable risk to consumers. (2) The Bureau may require additional information and documentation from an entity if it has reason to believe suspension or revocation under paragraph (h)(1) of this section may be warranted. (3) Except in cases of willfulness or those in which the public interest requires otherwise, prior to suspension or revocation under paragraph (h)(1) of this section, the Bureau will provide written notice of the facts or conduct that may warrant the suspension or revocation and an opportunity for the entity or information system to demonstrate or achieve compliance with this section or otherwise address the Bureau’s concerns. 1555 (4) The Bureau will revoke an entity’s preliminary approval for registration, provisional registration, or registration if the entity submits a written request to the Bureau that its preliminary approval, provisional registration, or registration be revoked. (5) For purposes of §§ 1041.5 and 1041.6, suspension or revocation of an information system’s registration is effective five days after the date that the Bureau publishes notice of the suspension or revocation on the Bureau’s website. For purposes of § 1041.10(b)(1), suspension or revocation of an information system’s provisional registration or registration is effective on the date that the Bureau publishes notice of the suspension or revocation on the Bureau’s website. The Bureau will also publish notice of a suspension or revocation in the Federal Register. (6) In the event that a provisional registration or registration of an information system is suspended, the Bureau will provide instructions concerning the scope and terms of the suspension on its website and in the notice of suspension published in the Federal Register. (i) Administrative appeals—(1) Grounds for administrative appeals. An entity may appeal a determination of the Bureau that: (i) Denies the application of an entity seeking preliminary approval for registration under paragraph (c)(1) of this section, registration under paragraph (c)(2) of this section, or provisional registration under paragraph (d)(1) of this section; or (ii) Suspends or revokes the entity’s preliminary approval for registration pursuant to paragraph (c)(1) of this section, provisional registration pursuant to paragraph (d)(1) of this section, or registration pursuant to paragraphs (c)(2) or (d)(2) of this section. 1556 (2) Time limits for filing administrative appeals. An appeal must be submitted on a date that is within 30 business days of the date of the determination. The Bureau may extend this time for good cause. (3) Form and content of administrative appeals. An appeal shall be made by electronic means as follows: (i) The appeal shall be submitted as set forth on the Bureau’s website. The appeal shall be labeled “Information System Registration Appeal.” (ii) The appeal shall set forth contact information for the appellant including, to the extent available, a mailing address, telephone number, or email address at which the Bureau may contact the appellant regarding the appeal; (iii) The appeal shall specify the date of the letter of determination, and enclose a copy of the determination being appealed; and (iv) The appeal shall include a description of the issues in dispute, specify the legal and factual basis for appealing the determination, and include appropriate supporting information. (4) Appeals process. The filing and pendency of an appeal does not by itself suspend the determination that is the subject of the appeal during the appeals process. Notwithstanding the foregoing, the Bureau may, in its discretion, suspend the determination that is the subject of the appeal during the appeals process. (5) Decisions to grant or deny administrative appeals. The Bureau shall decide whether to affirm the determination (in whole or in part) or to reverse the determination (in whole or in part) and shall notify the appellant of this decision in writing. 1557 § 1041.12 Compliance program and record retention. (a) Compliance program. A lender making a covered loan must develop and follow written policies and procedures that are reasonably designed to ensure compliance with the requirements in this part. These written policies and procedures must be appropriate to the size and complexity of the lender and its affiliates, and the nature and scope of the covered loan lending activities of the lender and its affiliates. (b) Record retention. A lender must retain evidence of compliance with this part for 36 months after the date on which a covered loan ceases to be an outstanding loan. (1) Retention of loan agreement and documentation obtained in connection with originating a covered short-term or covered longer-term balloon-payment loan. To comply with the requirements in this paragraph (b), a lender must retain or be able to reproduce an image of the loan agreement and documentation obtained in connection with a covered short-term or covered longer-term balloon-payment loan, including the following documentation, as applicable: (i) Consumer report from an information system that has been registered for 180 days or more pursuant to § 1041.11(c)(2) or is registered with the Bureau pursuant to § 1041.11(d)(2); (ii) Verification evidence, as described in § 1041.5(c)(2)(ii); (iii) Written statement obtained from the consumer, as described in § 1041.5(c)(2)(i); (2) Electronic records in tabular format regarding origination calculations and determinations for a covered short-term or covered longer-term balloon-payment loan under § 1041.5. To comply with the requirements in this paragraph (b), a lender must retain electronic records in tabular format that include the following information for a covered loan made under § 1041.5: 1558 (i) The projection made by the lender of the amount of a consumer’s net income during the relevant monthly period; (ii) The projections made by the lender of the amounts of a consumer’s major financial obligations during the relevant monthly period; (iii) Calculated residual income or debt-to-income ratio during the relevant monthly period; (iv) Estimated basic living expenses for the consumer during the relevant monthly period; and (v) Other consumer-specific information considered in making the ability-to-repay determination. (3) Electronic records in tabular format regarding type, terms, and performance of covered short-term or covered longer-term balloon-payment loan. To comply with the requirements in this paragraph (b), a lender must retain electronic records in tabular format that include the following information for a covered short-term or covered longer-term balloonpayment loan: (i) As applicable, the information listed in § 1041.10(c)(1)(i) through (viii) and § 1041.10(c)(2); (ii) Whether the lender obtained vehicle security from the consumer; (iii) The loan number in a loan sequence of covered short-term loans, covered longerterm balloon-payment loans, or a combination thereof. (iv) For any full payment on the loan that was not received or transferred by the contractual due date, the number of days such payment was past due, up to a maximum of 180 days; 1559 (v) For a loan with vehicle security: whether repossession of the vehicle was initiated; (vi) Date of last or final payment received; and (vii) The information listed in § 1041.10(c)(3). (4) Retention of records relating to payment practices for covered loans. To comply with the requirements in this paragraph (b), a lender must retain or be able to reproduce an image of the following documentation, as applicable, in connection with a covered loan: (i) Leveraged payment mechanism(s) obtained by the lender from the consumer; (ii) Authorization of additional payment transfer, as described in § 1041.8(c)(3)(iii); and (iii) Underlying one-time electronic transfer authorization or underlying signature check, as described in § 1041.8(d)(2). (5) Electronic records in tabular format regarding payment practices for covered shortterm or covered longer-term balloon-payment loan. To comply with the requirements in this paragraph (b), a lender must retain electronic records in tabular format that include the following information for covered loans: (i) History of payments received and attempted payment transfers, as defined in § 1041.8(a)(1), including: (A) Date of receipt of payment or attempted payment transfer; (B) Amount of payment due; (C) Amount of attempted payment transfer; (D) Amount of payment received or transferred; and (E) Payment channel used for attempted payment transfer. (ii) If an attempt to transfer funds from a consumer’s account is subject to the prohibition in § 1041.8(b)(1), whether the lender or service provider obtained authorization to initiate a 1560 payment transfer from the consumer in accordance with the requirements in § 1041.8(c) or (d). § 1041.13 Prohibition against evasion. A lender must not take any action with the intent of evading the requirements of this part. § 1041.14 Severability. The provisions of this part are separate and severable from one another. If any provision is stayed or determined to be invalid, the remaining provisions shall continue in effect. Appendix A to Part 1041—Model Forms 1561 Loan EDD-5555555 wilowteningoom Notioe of restrictions on future loans It you are unsure whether you will be able to pay $35iji?? by Noyemher 12th, 2016, you should not take out this loan- After you repay this loan, any similar loan you take out within the next 30 days will have to be smaller. This restriction is required by federal law. Huimum amount that you will be able to borrow Loan :1 (ms loan} semen Loan :2 season Loan ?3 $1 Loan :4 um allowed 1562 A-E Model Form for Third 1041 -5 Loan mm EDD-5555555 Himmler-linumm Notice of borrowing limits on this loan and future loans {Slur records show that you haye had two similar loans without taking a 3??oay break- Under federal law, this loan must be smaller than your prior loans- And after you repay this loan, you will not be able to take out another similar loan for at least 3t] clays. 1563 Mandel Form for First Payment Iil'llith-drawal Notice under mm EDD-5555555 wilowierlin-goorn Upcoming Withdrawal Notice from Willow Lending [in November12, 201 5, Willow Lending will attempt to withdraw a payment of irom your account ending in $102!. The payment will be withdrawn by check, using cheek #999- It this payment is not successful, we will add a 5112] returned payment tee to your balance on loan #5432- li'lr'illow Lending at if you have questions or need to stop this withdrawal. The institution where you have your also may be able to assist you. Paymem breakdown Principal now ?ll] Interest 5&0 Total payment amount $313 When you make this payment, your principal balance wil stay the same and you will not be cheer to paying ofl your loan- 1564 A?d Model Form for Unusual Wr?idrawal Notice under Hm sue-5555555 wilowleningoom Alert: Unusual Withdrawal from Willow Lending {in Noyember12,2ll1?, Willow Lending will attempt to withdraw a payment of $Bl] from your ending in 11022. This electronic withdrawal will be made by ACH transfer. This payment is unusual because it is larger than your originally sohedLIed payment. The previous withtiawal was initiated on November a 2015, lot $50. If this payment is no! successful, we will add a $112] returned payment fee to your balance on loan ?5432- (irritant Willow Lending at 1-300-555-5555 if you have questions or need to stop this withdrawal. The insiitution where you have your also may be able to assist you- Payment breakdown Principal Fees $50 hteresi $20 ?e 5&0 Total paymem amoLnt 1565 Model Form for Gatsumer Flights Notice under 1041.9[c] mm BUG-5555555 wilowtenrirrgoom Notice: Willow Lending is no longer permitted to withdraw loan payments from your account {Jur last two attempts to withdraw payment on your loan #5432 from your account ending in {1022 were retumed because your account did not contain enough funds to cover the payment. To protect your account, federal law prohibits us from trying to withdraw payment again without your permission- We may contact you to talk about your payment choices going forward. Prevlous payment attempts Fees charged by Payment due date Date of atten'pt Amount 1Il'tl'illow Lending November I 2016 November T, 2015 $34] 5113 returned payment tee November 2016 November 10, 2016 $51] $111] returned payment tee The mtsumer Financial Protection Bureau created this notice to inform you of your rights under federal law. The is a federal government agency buit to protect consumers. To learn about your rights as a borrower, visit 1566 Model Claus-e for First Payment 'u'?lhtiawal Eleeuenii: Shim Metiee under {applicable it] email Upeeming Wl?'ld'awal Helice from Willew Lending Rudy Upeeming Wl?'ld'awal Helice from Willew Lending On New 12, 2e16, we will attempt te withdraw a payment at $80 tram your account ending in [1022. 1I.I'ie'w the details at 1567 A-T Medal Claus-e fur Unusual Wimdrawal shun hluliee under 1041 stalemate} Sumac! {applicable in email en?d Alert Unusual Withdrawal Willuw Lend'ng 300'! Alert Unusual Withdrawal Willuw Lend'ng On New 12, 2:115, we will attempt ts withdraw a payment at $30 tram your aesuunt ending in [1022. This payment is musual because it is larger than yet: driginally schedtled payment 1I.I'iew the details at willuwlendingsumixdxa?i?ksw. 1568 Model Clause for Consumer Flights Eleotrorio Short Holioe under Eugen! {applicable to email Notice: 1liltillow Lending is no longer permitted to withdraw loan payments from you Bed! Notice: 1ll'll'illow Lending is no longer permitted to withdraw loan payments from you lDur last two attempts to withdraw payment from your account ending in DDEE were returned To protect your account, federal law prohihis us from trying to withdraw payment agair without your permission- 1Ir'iew the details at willowlendingoomoroxa??rsw. 1569 Supplement I to Part 1041—Official Interpretations Section 1041. Section 1041.2—Definitions. 2(a)(3) Closed-end credit. 1. In general. Institutions may rely on 12 CFR 1026.2(a)(10) and its related commentary in determining the meaning of closed-end credit, but without regard to whether the credit is consumer credit, as that term is defined in 12 CFR 1026.2(a)(12), or is extended to a consumer, as that term is defined in 12 CFR 1026.2(a)(11). 2(a)(5) Consummation. 1. New loan. When a contractual obligation on the consumer’s part is created is a matter to be determined under applicable law. A contractual commitment agreement, for example, that under applicable law binds the consumer to the loan terms would be consummation. Consummation, however, does not occur merely because the consumer has made some financial investment in the transaction (for example, by paying a non-refundable fee) unless applicable law holds otherwise. 2. Modification of existing loan that triggers underwriting requirements. A modification of an existing loan that increases the amount of an existing loan triggers underwriting requirements under § 1041.5 in certain circumstances. If the outstanding amount of an existing loan is increased, or if the total amount available under an open-end credit plan is increased, the modification is consummated as of the time that the consumer becomes contractually obligated on such a modification or increase. In those cases, the modification must comply with the requirements of § 1041.5(b). A loan modification does not trigger underwriting requirements under § 1041.5 if the modification reduces the outstanding amount or the total amount available 1570 under an open-end credit plan, or if the modification results only in the consumer receiving additional time in which to repay the loan. For example, providing a cost-free “off-ramp” or repayment plan to a consumer who cannot repay a loan during the allotted term of the loan is a modification of an existing loan—not a new loan—that results only in the consumer receiving additional time in which to repay the loan. Thus, providing a no-cost repayment plan does not constitute a modification that increases the amount of an existing loan. 2(a)(11) Credit. 1. In general. Institutions may rely on 12 CFR 1026.2(a)(14) and its related commentary in determining the meaning of credit. 2(a)(12) Electronic fund transfer. 1. In general. Institutions may rely on 12 CFR 1005.3(b) and its related commentary in determining the meaning of electronic fund transfer. 2(a)(13) Lender. 1. Regularly extends credit. The test for determining whether a person regularly extends credit for personal, family, or household purposes is explained in Regulation Z, 12 CFR 1026.2(a)(17)(v). Any loan to a consumer primarily for personal, family, or household purposes, whether or not the loan is a covered loan under part 1041, counts toward the numeric threshold for determining whether a person regularly extends credit. 2(a)(16) Open-end credit. 1. In general. Institutions may rely on 12 CFR 1026.2(a)(20) and its related commentary in determining the meaning of open-end credit, but without regard to whether the credit permits a finance charge to be imposed from time to time on an outstanding balance as defined in 12 CFR 1026.4. Also, for the purposes of defining open-end credit under part 1041, the term credit, as 1571 defined in § 1041.2(a)(11), is substituted for the term consumer credit, as defined in 12 CFR 1026.2(a)(12); the term lender, as defined in § 1041.2(a)(13), is substituted for the term creditor, as defined in 12 CFR 1026.2(a)(17); and the term consumer, as defined in § 1041.2(a)(4), is substituted for the term consumer, as defined in 12 CFR 1026.2(a)(11). See generally § 1041.2(b). 2(a)(17) Outstanding loan. 1. Payments owed to third parties. A loan is an outstanding loan if it meets all the criteria set forth in § 1041.2(a)(17), regardless of whether the consumer is required to pay the lender, an affiliate of the lender, or a service provider. A lender selling the loan or the loan servicing rights to a third party does not affect whether a loan is an outstanding loan under § 1041.2(a)(17). 2. Stale loans. A loan is generally an outstanding loan if the consumer has a legal obligation to repay the loan, even if the consumer is delinquent or if the consumer is in a repayment plan or workout arrangement. However, a loan that the consumer otherwise has a legal obligation to repay is not an outstanding loan for purposes of part 1041 if the consumer has not made any payment on the loan within the previous 180-day period. A loan ceases to be an outstanding loan as of: the earliest of the date the consumer repays the loan in full, the date the consumer is released from the legal obligation to repay, the date the loan is otherwise legally discharged, or the date that is 180 days following the last payment that the consumer has made on the loan, even if the payment is not a regularly scheduled payment in a scheduled amount. If the consumer does not make any payments on a loan and none of these other events occur, the loan ceases to be outstanding 180 days after consummation. A loan cannot become an outstanding loan due to any events that occur after the consumer repays the loan in full, the consumer is released from the legal obligation to repay, the loan is otherwise legally discharged, 1572 180 days following the last payment that the consumer has made on the loan, or 180 days after consummation of a loan on which the consumer makes no payments. 2(a)(18) Service provider. 1. Credit access businesses and credit services organizations. Persons who provide a material service to lenders in connection with the lenders’ offering or provision of covered loans are service providers, subject to the specific limitations in section 1002(26) of the Dodd-Frank Act. Accordingly, credit access businesses and credit service organizations that provide a material service to lenders during the course of obtaining for consumers, or assisting consumers in obtaining, loans from lenders, are service providers, subject to the specific limitations in section 1002(26) of the Dodd-Frank Act. 2(a)(19) Vehicle security. 1. An interest in a consumer’s motor vehicle as a condition of credit. Subject to the exclusion described in § 1041.3(d)(1), a lender’s or service provider’s interest in a consumer’s motor vehicle constitutes vehicle security only to the extent that the security interest is obtained in connection with the credit. If a party obtains such a security interest in a consumer’s motor vehicle for a reason that is unrelated to an extension of credit, the security interest does not constitute vehicle security. For example, if a mechanic performs work on a consumer’s motor vehicle and a mechanic’s lien attaches to the consumer’s motor vehicle by operation of law because the consumer did not timely pay the mechanic’s bill, the mechanic does not obtain vehicle security for the purposes of § 1041.2(a)(19). 2(b) Rule of Construction. 1. Incorporation of terms from underlying statutes and regulations. For purposes of this part, where definitions are incorporated from other statutes or regulations, users may as 1573 applicable rely on embedded definitions, appendices, and commentary for those other laws. For example, 12 CFR 1005.2(b) and its related commentary determine the meaning of account under § 1041.2(a)(1). However, where part 1041 defines the same term or a parallel term in a way that creates a substantive distinction, the part 1041 definition shall control. See, for example, the definition of open-end credit in § 1041.2(a)(16), which is generally determined according to 12 CFR 1026.2(a)(20) and its related commentary but without regard to whether the credit is consumer credit, as that term is defined in 12 CFR 1026.2(a)(12), or is extended to a consumer, as that term is defined in 12 CFR 1026.2(a)(11), because part 1041 provides a different and arguably broader definition of consumer in § 1041.2(a)(4). Section 1041.3—Scope of Coverage; exclusions; conditional exemptions 3(b) Covered loans. 1. Credit structure. The term covered loan includes open-end credit and closed-end credit, regardless of the form or structure of the credit. 2. Primary purpose. Under § 1041.3(b), a loan is not a covered loan unless it is extended primarily for personal, family, or household purposes. Institutions may rely on 12 CFR 1026.3(a) and its related commentary in determining the primary purpose of a loan. Paragraph 3(b)(1). 1. Closed-end credit that does not provide for multiple advances to consumers. A loan does not provide for multiple advances to a consumer if the loan provides for full disbursement of the loan proceeds only through disbursement on a single specific date. 2. Loans that provide for multiple advances to consumers. Both open-end credit and closed-end credit may provide for multiple advances to consumers. Open-end credit can have a fixed expiration date, as long as during the plan’s existence the consumer may use credit, repay, 1574 and reuse the credit. Likewise, closed-end credit may consist of a series of advances. For example: i. Under a closed-end commitment, the lender might agree to lend a total of $1,000 in a series of advances as needed by the consumer. When a consumer has borrowed the full $1,000, no more is advanced under that particular agreement, even if there has been repayment of a portion of the debt. 3. Facts and circumstances test for determining whether loan is substantially repayable within 45 days. Substantially repayable means that the substantial majority of the loan or advance is required to be repaid within 45 days of consummation or advance, as the case may be. Application of the standard depends on the specific facts and circumstances of each loan, including the timing and size of the scheduled payments. A loan or advance is not substantially repayable within 45 days of consummation or advance merely because a consumer chooses to repay within 45 days when the loan terms do not require the consumer to do so. 4. Deposit advance products. A loan or advance is substantially repayable within 45 days of consummation or advance if the lender has the right to be repaid through a sweep or withdrawal of any qualifying electronic deposit made into the consumer’s account within 45 days of consummation or advance. A loan or advance described in this paragraph is substantially repayable within 45 days of consummation or advance even if no qualifying electronic deposit is actually made into or withdrawn by the lender from the consumer’s account. 5. Loans with alternative, ambiguous, or unusual payment schedules. If a consumer, under any applicable law, would breach the terms of the agreement between the consumer and the lender or service provider by not substantially repaying the entire amount of the loan or advance within 45 days of consummation or advance, as the case may be, the loan is a covered 1575 short-term loan under § 1041.3(b)(1). For loans or advances that are not required to be repaid within 45 days of consummation or advance, if the consumer, under applicable law, would not breach the terms of the agreement between the consumer and the lender by not substantially repaying the loan or advance in full within 45 days, the loan is a covered longer-term balloonpayment loan under § 1041.3(b)(2) or a covered longer-term loan under § 1041.3(b)(3) if the loan otherwise satisfies the criteria specified in § 1041.3(b)(2) or 1041.3(b)(3), respectively. Paragraph 3(b)(2). 1. Closed-end credit that does not provide for multiple advances to consumers. See comments 3(b)(1)-1 and 3(b)(1)-2. 2. Payments more than twice as large as other payments. For purposes of § 1041.3(b)(2)(i) and (ii), all required payments of principal and any charges (or charges only, depending on the loan features) due under the loan are used to determine whether a particular payment is more than twice as large as another payment, regardless of whether the payments have changed during the loan term due to rate adjustments or other payment changes permitted or required under the loan. 3. Charges excluded. Charges for actual unanticipated late payments, for exceeding a credit limit, or for delinquency, default, or a similar occurrence that may be added to a payment are excluded from the determination of whether the loan is repayable in a single payment or a particular payment is more than twice as large as another payment. Likewise, sums that are accelerated and due upon default are excluded from the determination of whether the loan is repayable in a single payment or a particular payment is more than twice as large as another payment. 1576 4. Multiple-advance structures. Loans that provide for more than one advance are considered to be a covered longer-term balloon-payment loan under § 1041.3(b)(2)(ii) if either: i. The consumer is required to repay substantially the entire amount of an advance more than 45 days after the advance is made or is required to make at least one payment on the advance that is more than twice as large as any other payment; or ii. A loan with multiple advances is structured such that paying the required minimum payment may not fully amortize the outstanding balance by a specified date or time, and the amount of the final payment to repay the outstanding balance at such time could be more than twice the amount of other minimum payments under the plan. For example, the lender extends an open-end credit plan with a $500 credit limit, monthly billing cycles, and a minimum payment due each billing cycle that is equal to 10% of the outstanding principal. Fees or interest on the plan are equal to 10% of the outstanding principal per month, so that if a consumer pays nothing other than the minimum payment amount, the outstanding principal remains the same. All outstanding amounts must be repaid within six months of the advance. The credit plan is a covered loan under § 1041.3(b)(2)(ii) because if the consumer drew the entire amount at one time and then made only minimum payments, the sixth payment would be more than twice the amount of the minimum payment required ($50). Paragraph 3(b)(3). 1. Conditions for coverage of a longer-term loan. A loan that is not a covered short-term loan or a covered longer-term balloon-payment loan is a covered longer-term loan only if it satisfies both the cost of credit requirement of § 1041.3(b)(3)(i) and leveraged payment mechanism requirement of § 1041.3(b)(3)(ii). If the requirements of § 1041.3(b)(3) are met, and the loan is not otherwise excluded or conditionally exempted from coverage by § 1041.3(d), (e), 1577 or (f), the loan is a covered longer-term loan. For example, a 60-day loan that is not a covered longer-term balloon-payment loan is not a covered longer-term loan if the cost of credit as measured pursuant to § 1041.2(a)(6) is less than or equal to a rate of 36 percent per annum even if the lender or service provider obtains a leveraged payment mechanism. 2. No balance during a billing cycle. Under § 1041.2(a)(6)(ii)(B), the cost of credit for open-end credit must be calculated according to the rules for calculating the effective annual percentage rate for a billing cycle as set forth in Regulation Z, 12 CFR 1026.14(c) and (d), which provide that the annual percentage rate cannot be calculated for billing cycles in which there is a finance charge but no other balance. Accordingly, pursuant to § 1041.2(a)(6)(ii)(B), the cost of credit could not be calculated for such billing cycles. Section 1041.3(b)(3)(i)(B)(1) provides that, for such billing cycles, an open-end credit plan is determined to have exceeded the threshold set forth in that paragraph if there is no balance other than a finance charge imposed by the lender. 3. Timing for coverage determination. A loan may become a covered longer-term loan at any such time as both of the requirements of § 1041.3(b)(3)(i) and (ii) are met. For example: i. A lender originates a closed-end loan that is not a longer-term balloon-payment loan to be repaid within six months of consummation with a cost of credit equal to 60 percent. At the time of consummation, the loan is not a covered longer-term loan because it does not have a leveraged payment mechanism. After two weeks, the lender obtains a leveraged payment mechanism. The loan is now a covered longer-term loan because it meets both of the requirements of § 1041.3(b)(3)(i) and (ii). ii. A lender extends an open-end credit plan with monthly billing cycles and a leveraged payment mechanism. At consummation and again at the end of the first billing cycle, the plan is 1578 not a covered longer-term loan because its cost of credit is below 36 percent. In the second billing cycle, the plan’s cost of credit is 45 percent because several fees are triggered in addition to interest on the principal balance. The plan is now a covered longer-term loan because it meets both of the requirements of § 1041.3(b)(3)(i) and (ii). Beginning on the first day of the third billing cycle, and thereafter for the duration of the plan, the lender must therefore comply with the requirements of this part including by, for example, providing a first withdrawal notice before initiating the first payment transfer on or after the first day of the third billing cycle. The requirements to provide certain payment withdrawal notices under § 1041.9 have been structured so that the notices can be provided in the same mailing as the periodic statements that are required by Regulation Z, 12 CFR 1026.7(b). See, e.g., § 1041.9(b)(3)(i)(D). Paragraph 3(b)(3)(ii). 1. Timing. The condition in § 1041.3(b)(3)(ii) is satisfied if a lender or service provider obtains a leveraged payment mechanism before, at the same time as, or after the consumer receives the entire amount of funds that the consumer is entitled to receive under the loan, regardless of the means by which the lender or service provider obtains a leveraged payment mechanism. 2. Leveraged payment mechanism in contract. The condition in § 1041.3(b)(3)(ii) is satisfied if a loan agreement authorizes the lender to elect to obtain a leveraged payment mechanism, regardless of the time at which the lender actually obtains a leveraged payment mechanism. The following are examples of situations in which a lender obtains a leveraged payment mechanism under § 1041.3(b)(3)(ii): 1579 i. Future authorization. A loan agreement provides that the consumer, at some future date, must authorize the lender or service provider to debit the consumer’s account on a recurring basis. ii. Delinquency or default provisions. A loan agreement provides that the consumer must authorize the lender or service provider to debit the consumer’s account on a one-time or a recurring basis if the consumer becomes delinquent or defaults on the loan. Paragraph 3(c). 1. Initiating a transfer of money from a consumer’s account. A lender or service provider obtains the ability to initiate a transfer of money when that person can collect payment, or otherwise withdraw funds, from a consumer’s account, either on a single occasion or on a recurring basis, without the consumer taking further action. Generally, when a lender or service provider has the ability to “pull” funds or initiate a transfer from the consumer’s account, that person has a leveraged payment mechanism. However, a “push” transaction from the consumer to the lender or service provider does not in itself give the lender or service provider a leveraged payment mechanism. 2. Lender-initiated transfers. The following are examples of situations in which a lender or service provider has the ability to initiate a transfer of money from a consumer’s account: i. Check. A lender or service provider obtains a check, draft, or similar paper instrument written by the consumer, other than a single immediate payment transfer at the consumer’s request as described in § 1041.3(c) and comment 3(c)-3. ii. Electronic fund transfer authorization. The consumer authorizes a lender or service provider to initiate an electronic fund transfer from the consumer’s account in advance of the 1580 transfer, other than a single immediate payment transfer at the consumer’s request as described in § 1041.3(c) and comment 3(c)-3. iii. Remotely created checks and remotely created payment orders. A lender or service provider has authorization to create or present a remotely created check (as defined by Regulation CC, 12 CFR 229.2(fff)), remotely created payment order (as defined in 16 CFR 310.2(cc)), or similar instrument drafted on the consumer’s account. iv. Transfer by account-holding institution. A lender or service provider that is an account-holding institution has a right to initiate a transfer of funds between the consumer’s account and an account of the lender or affiliate, including, but not limited to, an account-holding institution’s right of set-off. 3. Single immediate payment transfer at the consumer’s request excluded. A single immediate payment transfer at the consumer’s request, as defined in § 1041.8(a)(2), is excluded from the definition of leveraged payment mechanism. Accordingly, if the loan or other agreement between the consumer and the lender or service provider does not otherwise provide for the lender or service provider to initiate a transfer without further consumer action, the lender or service provider can initiate a single immediate payment transfer at the consumer’s request without causing the loan to become a covered loan under § 1041.3(b)(3). See § 1041.8(a)(2) and related commentary for guidance on what constitutes a single immediate payment transfer at the consumer’s request. 4. Transfers not initiated by the lender. A lender or service provider does not initiate a transfer of money from a consumer’s account if the consumer authorizes a third party, such as a bank’s automatic bill pay service, to initiate a transfer of money from the consumer’s account to a lender or service provider. 1581 3(d) Exclusions. 3(d)(1) Certain purchase money security interest loans. 1. “Sole purpose” test. The requirements of part 1041 do not apply to loans made solely and expressly to finance the consumer’s initial purchase of a good in which the lender takes a security interest as a condition of the credit. For example, the requirements of this part would not apply to a transaction in which a lender makes a loan to a consumer for the express purpose of initially purchasing a motor vehicle, television, household appliance, or furniture in which the lender takes a security interest and the amount financed is approximately equal to, or less than, the cost of acquiring the good, even if the cost of credit exceeds 36 percent per annum and the lender also obtains a leveraged payment mechanism. A loan is made solely and expressly to finance the consumer’s initial purchase of a good even if the amount financed under the loan includes Federal, State, or local taxes or amounts required to be paid under applicable State and Federal licensing and registration requirements. This exclusion does not apply to refinances of credit extended for the purchase of a good. 3(d)(2) Real estate secured credit. 1. Real estate and dwellings. The requirements of part 1041 do not apply to credit secured by any real property, or by any personal property, such as a mobile home, used or expected to be used as a dwelling if the lender records or otherwise perfects the security interest within the term of the loan, even if the cost of credit exceeds 36 percent per annum and the lender or servicer provider also obtains a leveraged payment mechanism. If the lender does not record or perfect the security interest during the term of the loan, however, the credit is not excluded from the requirements of part 1041 under § 1041.3(d)(2). 1582 3(d)(5) Non-recourse pawn loans. 1. Lender possession required and no recourse permitted. A pawn loan must satisfy two conditions to be excluded from the requirements of part 1041 under § 1041.3(d)(5). First, the lender must have sole physical possession and use of the property securing the pawned property at all times during the entire term of the loan. If the consumer retains either possession or use of the property, however limited the consumer’s possession or use of the property might be, the loan is not excluded from the requirements of part 1041 under § 1041.3(d)(5). Second, the lender must have no recourse if the consumer does not elect to redeem the pawned item and repay the loan other than retaining the pawned property to dispose of according to State or local law. If any consumer, or if any co-signor, guarantor, or similar person, is personally liable for the difference between the outstanding balance on the loan and the value of the pawned property, the loan is not excluded from the requirements of part 1041 under § 1041.3(d)(5). 3(d)(6) Overdraft services. 1. Definitions. Institutions may rely on 12 CFR 1005.17(a) and its related commentary in determining whether credit is an overdraft service or an overdraft line of credit that is excluded from the requirements of part 1041 under § 1041.3(d)(6). 3(d)(7) Wage advance programs. 1. Advances of wages under § 1041.3(d)(7) must be offered by an employer, as defined in the Fair Labor Standards Act, 29 U.S.C. 203(d), or by the employer’s business partner to the employer’s employees pursuant to a wage advance program. For example, an advance program might be offered by a company that provides payroll card services or accounting services to the employer, or by the employer with the assistance of such a company. Similarly, an advance program might be offered by a company that provides consumer financial products and services 1583 as part of the employer’s benefits program, such that the company would have information regarding the wages accrued by the employee. 3(d)(7)(i). 1. Under the exclusion in § 1041.3(d)(7)(i), the advance must be made only against accrued wages. To qualify for that exclusion, the amount advanced must not exceed the amount of the employee’s accrued wages. Accrued wages are wages that the employee is entitled to receive under State law in the event of separation from the employer for work performed for the employer, but for which the employee has yet to be paid. 3(d)(7)(ii)(B). 1. Under § 1041.3(d)(7)(ii)(B), the entity advancing the funds is required to warrant that it has no legal or contractual claim or remedy against the consumer based on the consumer’s failure to repay in the event the amount advanced is not repaid in full. This provision does not prevent the entity from obtaining a one-time authorization to seek repayment from the consumer’s transaction account. 3(d)(8) No-cost advances. 1. Under § 1041.3(d)(8)(i), the entity advancing the funds is required to warrant that it has no legal or contractual claim or remedy against the consumer based on the consumer’s failure to repay in the event the amount advanced is not repaid in full. This provision does not prevent the entity from obtaining a one-time authorization to seek repayment from the consumer’s transaction account. 3(e) Alternative loans. 1. General. Section 1041.3(e) conditionally exempts from part 1041 alternative covered loans that satisfy the conditions and requirements set forth in § 1041.3(e). Nothing in 1584 § 1041.3(e) provides lenders with an exemption from the requirements of other applicable laws, including State laws. The conditions for an alternative loan made under § 1041.3(e) largely track the conditions set forth by the National Credit Union Administration at 12 CFR 701.21(c)(7)(iii) for a Payday Alternative Loan made by a Federal credit union. All lenders, including Federal credit unions and persons that are not Federal credit unions, are permitted to make loans under § 1041.3(e), provided that such loans are permissible under other applicable laws, including State laws. 3(e)(1) Loan term conditions. Paragraph 3(e)(1)(iv). 1. Substantially equal payments. Under § 1041.3(e)(1)(iv), payments are substantially equal in amount if the amount of each scheduled payment on the loan is equal to or within a small variation of the others. For example, if a loan is repayable in six biweekly payments and the amount of each scheduled payment is within 1 percent of the amount of the other payments, the loan is repayable in substantially equal payments. In determining whether a loan is repayable in substantially equal payments, a lender may disregard the effects of collecting the payments in whole cents. 2. Substantially equal intervals. The intervals for scheduled payments are substantially equal if the payment schedule requires repayment on the same date each month or in the same number of days of the prior scheduled payment. For example, a loan for which payment is due every 15 days has payments due in substantially equal intervals. A loan for which payment is due on the 15th day of each month also has payments due in substantially equal intervals. In determining whether payments fall due in substantially equal intervals, a lender may disregard that dates of scheduled payments may be slightly changed because the scheduled date is not a 1585 business day, that months have different numbers of days, and the occurrence of leap years. Section 1041.3(e)(1)(iv) does not prevent a lender from accepting prepayment on a loan made under § 1041.3(e). 3. Amortization. Section 1041.3(e)(1)(iv) requires that the scheduled payments fully amortize the loan over the contractual period and prohibits lenders from making loans under § 1041.3(e) with interest-only payments or with a payment schedule that front-loads payments of interest and fees. While under § 1041.3(e)(1)(iv) the payment amount must be substantially equal for each scheduled payment, the amount of the payment that goes to principal and to interest will vary. The amount of payment applied to interest will be greater for earlier payments when there is a larger principal outstanding. Paragraph 3(e)(1)(v). 1. Cost of credit. Under § 1041.3(e)(1)(v), the lender must not impose any charges other than the rate and application fees permissible for Federal credit unions to charge under 12 CFR 701.21(c)(7)(iii). Under 12 CFR 701.21(c)(7)(iii), application fees must reflect the actual costs associated with processing the application and must not exceed $20. 3(e)(2) Borrowing history condition. 1. Relevant records. A lender may make an alternative covered loan under § 1041.3(e) only if the lender determines from its records that the consumer’s borrowing history on alternative covered loans made under § 1041.3(e) meets the criteria set forth in § 1041.3(e)(2). The lender is not required to obtain information about a consumer’s borrowing history from other persons, such as by obtaining a consumer report from an information system that has been registered for 180 days or more pursuant to § 1041.11(c)(2) or is registered with the Bureau pursuant to § 1041.11(d)(2). 1586 2. Determining 180-day period. For purposes of counting the number of loans made under § 1041.3(e)(2), the 180-day period begins on the date that is 180 days prior to the consummation date of the loan to be made under § 1041.3(e) and ends on the consummation date of such loan. 3. Total number of loans made under § 1041.3(e)(2). Section 1041.3(e)(2) excludes loans from the conditional exemption in § 1041.3(e) if the loan would result in the consumer being indebted on more than three outstanding loans made under § 1041.3(e) from the lender in any consecutive 180-day period. See § 1041.2(a)(17) for the definition of outstanding loan. Under § 1041.3(e)(2), the lender is required to determine from its records the consumer’s borrowing history on alternative covered loans made under § 1041.3(e) by the lender. The lender must use this information about borrowing history to determine whether the loan would result in the consumer being indebted on more than three outstanding loans made under § 1041.3(e) from the lender in a consecutive 180-day period, determined in the manner described in comment 3(e)(2)-2. Section 1041.3(e) does not prevent lenders from making a covered loan subject to the requirements of part 1041. 4. Example. For example, assume that a lender seeks to make an alternative loan under § 1041.3(e) to a consumer and the loan does not qualify for the safe harbor under § 1041.3(e)(4). The lender checks its own records and determines that during the 180 days preceding the consummation date of the prospective loan, the consumer was indebted on two outstanding loans made under § 1041.3(e) from the lender. The loan, if made, would be the third loan made under § 1041.3(e) on which the consumer would be indebted during the 180-day period and, therefore, would be exempt from this part under § 1041.3(e). If, however, the lender determined that the consumer was indebted on three outstanding loans under § 1041.3(e) from the lender during the 1587 180 days preceding the consummation date of the prospective loan, the condition in § 1041.3(e)(2) would not be satisfied and the loan would not be an alternative loan subject to the exemption under § 1041.3(e) but would instead be a covered loan subject to the requirements of part 1041. 3(e)(3) Income documentation condition. 1. General. Section 1041.3(e)(3) requires lenders to maintain policies and procedures for documenting proof of recurring income and to comply with those policies and procedures when making alternative loans under § 1041.3(e). Section 1041.3(e)(3) does not require lenders to undertake the same income documentation procedures required by § 1041.5(c)(2). For the purposes of § 1041.3(e)(3), lenders may establish any procedure for documenting recurring income that satisfies the lender’s own underwriting obligations. For example, lenders may choose to use the procedure contained in the National Credit Union Administration’s guidance at 12 CFR 701.21(c)(7)(iii) on Payday Alternative Loan programs recommending that Federal credit unions document consumer income by obtaining two recent paycheck stubs. 3(f) Accommodation Lending. 1. General. Section 1041.3(f) provides a conditional exemption for covered loans if, at the time of origination: (1) the lender and its affiliates collectively have made 2,500 or fewer covered loans in the current calendar year and made 2,500 or fewer covered loans in the preceding calendar year; and (2) during the most recent completed tax year in which the lender was in operation, if applicable, the lender and any affiliates that were in operation and used the same tax year derived no more than 10 percent of their receipts from covered loans, or if the lender was not in operation in a prior tax year, the lender reasonably anticipates that the lender and any of its affiliates that use the same tax year will, during the current tax year, derive no 1588 more than 10 percent of their combined receipts from covered loans. For example, assume a lender begins operation in January 2019, uses the calendar year as its tax year, and has no affiliates. In 2019, the lender could originate up to 2,500 covered loans that are not subject to the requirements of this part if at the time of each origination it reasonably anticipates that no more than 10 percent of its receipts during the current tax year will derive from covered loans. In 2020, the lender could originate up to 2,500 covered loans that are not subject to the requirements of this part if the lender made 2,500 or fewer covered loans in 2019 and the lender derived no more than 10 percent of its receipts in the 2019 tax year from covered loans. Section 1041.3(f) provides that covered longer-term loans for which all transfers meet the conditions in § 1041.8(a)(1)(ii), and receipts from such loans, are not included for the purpose of determining whether the conditions of § 1041.3(f)(1) and (f)(2) have been satisfied. For example, a bank that makes a covered longer-term loan using a loan agreement that includes the conditions in § 1041.8(a)(1)(ii) does not need to include that loan, or the receipts from that loan, in determining whether it is below the 2,500 loan threshold or the 10 percent of receipts threshold in § 1041.3(f)(1) and (f)(2). 2. Reasonable anticipation of receipts for current tax year. A lender and its affiliates can look to receipts to date in forecasting their total receipts for the current tax year, but are expected to make reasonable adjustments to account for an upcoming substantial change in business plans or other relevant and known factors. Section 1041.4— Identification of unfair and abusive practice 1. General. A lender who complies with § 1041.5 in making a covered short-term loan or a covered longer-term balloon-payment loan has not engaged in the unfair and abusive practice under § 1041.4. A lender who complies with § 1041.6 in making a covered short-term loan has 1589 not committed the unfair and abusive practice under § 1041.4 and is not subject to § 1041.5. Section 1041.5—Ability-to-Repay Determination Required 5(a) Definitions. 5(a)(1) Basic living expenses. 1. General. Under § 1041.5(b), a lender must make a reasonable determination that the consumer has the ability to repay a covered short-term loan or covered longer-term balloonpayment loan according to its terms. The consumer’s ability to meet basic living expenses is part of the broader ability-to-repay determination under § 1041.5(b). See comment 5(b)-1 for additional clarification. The lender’s estimate of basic living expenses must be reasonable. The lender may make a reasonable estimate of basic living expenses without making an individualized determination. See comment 5(b)-2.i.C for additional clarification. 2. Expenditures included in basic living expenses. Section 1041.5(a)(1) defines basic living expenses as expenditures, other than payments for major financial obligations, that the consumer makes for goods and services necessary to maintain the consumer’s health, welfare, and ability to produce income, and the health and welfare of the members of the consumer’s household who are financially dependent on the consumer. Examples of basic living expenses include food, utilities not paid as part of rental housing expenses, transportation, out-of-pocket medical expenses, phone and Internet services, and childcare. Basic living expenses do not include expenditures for discretionary personal and household goods or services, such as newspaper subscriptions, or vacation activities. If the consumer is responsible for payment of household goods and services on behalf of the consumer’s dependents, those expenditures are included in basic living expenses. As part of its reasonable ability-to-repay determination, the lender may reasonably consider whether another person (e.g., a spouse or adult family member 1590 living with the consumer) is regularly contributing toward the consumer’s payment of basic living expenses (see comment 5(b)-2.i.C.2). 5(a)(2) Debt-to-income ratio. 1. General. Section 1041.5(a)(2) defines debt-to-income ratio as the ratio, expressed as a percentage, of the sum of the amounts that the lender projects will be payable by the consumer for major financial obligations during the relevant monthly period and the payments under the covered short-term loan or covered longer-term balloon-payment loan during the relevant monthly period, to the monthly net income that the lender projects the consumer will receive during the relevant monthly period, all of which projected amounts are determined in accordance with § 1041.5(c). See § 1041.5(b)(2)(i) and associated commentary for further clarification on the use of debt-to-income methodology to determine ability to repay. For covered longer-term balloon-payment loans, where the relevant monthly period may fall well into the future relative to the consummation of the loan, the lender must calculate the debt-to-income ratio using the projections made under § 1041.5(c) and in so doing must make reasonable assumptions about the consumer’s net income and major financial obligations during the relevant monthly period compared to the period covered by the verification evidence. For example, the lender cannot assume, absent a reasonable basis, that there will be a substantial increase in net income or decrease in major financial obligations between consummation and the relevant monthly period. For further clarification, see comment 5(c)(1)-1 regarding the consistency between the consumer’s written statement and verification evidence and comment 5(c)(2)(ii)(A)-2 regarding what constitutes sufficient history of net income for purposes of verification evidence. 1591 5(a)(3) Major financial obligations. 1. General. Section 1041.5(a)(3) defines major financial obligations as a consumer’s housing expense, required payments due under debt obligations (including, without limitation, outstanding covered loans), child support obligations, and alimony obligations. Housing expense includes the total periodic amount that the consumer pays for housing during the relevant monthly period, such as the amount the consumer pays to a landlord for rent or to a creditor for a mortgage (including principal, interest, and any escrowed amounts if required). Debt obligations for purposes of § 1041.5(a)(3) do not include amounts due or past due for medical bills, utilities, and other items that are generally defined as basic living expenses under § 1041.5(a)(1). The amount of a payment required under a debt obligation includes the amount the consumer must pay when due to avoid delinquency under the debt obligation in the absence of any affirmative act by the consumer to extend, delay, or restructure the repayment schedule. Thus, this would include periodic or lump-sum payments for automobile loans, student loans, and other covered and non-covered loans, and minimum monthly credit card payments due during the relevant monthly period. It also includes any delinquent amounts on such obligations that are due as of the relevant monthly period, except where an obligation on a covered short-term loan or a covered longer-term balloon-payment loan is no longer outstanding or where the obligation is listed as charged off on a national consumer report. For example, if the consumer has a periodic automobile loan payment from a prior period that is past due and the automobile finance company adds the past due payment to the next regularly scheduled periodic payment which falls during the relevant monthly period, then the past due periodic payment is a major financial obligation. 1592 2. Motor vehicle leases. For purposes of this rule, motor vehicle leases shall be treated as a debt obligation. 5(a)(5) Net income. 1. General. Section 1041.5(a)(5) defines a consumer’s net income to mean the total amount that a consumer receives after the payer has deducted amounts for taxes withheld by the consumer, other obligations, and voluntary contributions (but before deductions of any amounts for payments under a prospective covered short-term loan or covered longer-term balloonpayment loan or for any major financial obligation); provided that, a lender may elect to include in the consumer’s net income the amount of any income of another person to which a consumer has a reasonable expectation of access (see comment 5(a)(5)-3). Net income includes income that is regularly received by the consumer as take-home pay, whether the consumer is treated as an employee or independent contractor. Net income also includes income regularly received by the consumer from other sources, such as child support or alimony received by the consumer and any payments received by the consumer from retirement, social security, disability, or other government benefits, or annuity plans. Lenders may include in net income irregular or seasonal income, such as tips, bonuses, and overtime pay. Net income does not include one-time payments anticipated to be received in the future from non-standard sources, such as legal settlements, tax refunds, jury prizes, or remittances, unless there is verification evidence of the amount and expected timing of such income. If the consumer receives a traditional pay check but the verification evidence obtained under § 1041.5(c)(2) shows payment of gross income or otherwise is unclear about whether deductions for the consumer’s taxes, other obligations, or voluntary contributions have been made, or if the consumer is not paid via a traditional pay check, then the lender may draw reasonable conclusions from the information provided and is 1593 not required to inquire further about deductions for the consumer’s taxes, other obligations, or voluntary contributions. 2. Other obligations and voluntary contributions. An example of other obligations is a consumer’s portion of payments for premiums for employer-sponsored health insurance plans. An example of a voluntary contribution is a consumer’s contribution to a defined contribution plan meeting the requirements of Internal Revenue Code section 401(a), 26 U.S.C. 401(a). The lender may inquire about and reasonably consider whether voluntary contributions will be discontinued prior to the relevant monthly period, in which case they would not be deducted from the amount of net income that is projected. 3. Reasonable expectation of access to another person’s income. Under § 1041.5(a)(5), a lender may elect to include in the consumer’s net income the amount of any income of another person to which the consumer has a reasonable expectation of access. The income of any other person is considered net income to which the consumer has a reasonable expectation of access if the consumer has direct access to those funds on a regular basis through a transaction account in which the consumer is an accountholder or cardholder. If the lender elects to include any income of another person to which the consumer has a reasonable expectation of access, then as part of the lender’s obligation to make a reasonable projection of the consumer’s net income during the applicable period, the lender must obtain verification evidence demonstrating that the consumer has a reasonable expectation of access to the portion of the other person’s income that the lender includes within its net income projection. See § 1041.5(c)(2)(ii)(A) and associated commentary. The following examples illustrate when a consumer has reasonable expectation of access to the income of another person for purposes of § 1041.5(a)(5): 1594 i. The consumer’s spouse has a salary or income that is deposited regularly into a joint account the spouse shares with the consumer. The consumer has a reasonable expectation of access to the spouse’s income. ii. The consumer shares a household with a sibling. The sibling’s salary or other income is deposited into an account in which the consumer does not have access. However, the sibling regularly transfers a portion of that income from the sibling’s deposit account into the consumer’s deposit account. The consumer has a reasonable expectation of access to that portion of the sibling’s income. iii. The consumer’s spouse has a salary or other income that is deposited into an account to which the consumer does not have access, and the spouse does not regularly transfer a portion of that income into the consumer’s account. The consumer does not have a reasonable expectation of access to the spouse’s income. iv. The consumer does not have a joint bank account with his spouse, nor does the spouse make regular deposits into the consumer’s individual deposit account. However, the spouse regularly pays for a portion of the consumer’s basic living expenses. The consumer does not have a reasonable expectation of access to the spouse’s income. However, regular contributions toward payment of the consumer’s basic living expenses may be considered by the lender as a consumer-specific factor that is relevant if the lender makes an individualized estimate of basic living expenses (see comment 5(b)-2.i.C.2 for further clarification). 5(a)(6) Payment under the covered short-term loan or covered longer-term balloonpayment loan. Paragraphs 5(a)(6)(i) and (ii). 1595 1. General. Section 1041.5(a)(6)(i) defines payment under a covered short-term loan or covered longer-term balloon-payment loan as the combined dollar amount payable by the consumer at a particular time following consummation in connection with the loan, assuming that the consumer has made preceding required payments and in the absence of any affirmative act by the consumer to extend or restructure the repayment schedule or to suspend, cancel, or delay payment for any product, service, or membership provided in connection with the covered short-term loan or covered longer-term balloon-payment loan. Section 1041.5(a)(6)(ii) clarifies that it includes all principal, interest, charges, and fees. A lender may not exclude a portion of the payment simply because a consumer could avoid or delay paying a portion of the payment, such as by requesting forbearance for that portion or by cancelling a service provided in exchange for that portion. For example: i. Assume that in connection with a covered longer-term balloon-payment loan, a consumer would owe a periodic payment on a particular date of $100 to the lender, which consists of $15 in finance charges, $80 in principal, and a $5 service fee, and the consumer also owes $10 as a credit insurance premium to a separate insurance company. Assume further that under the terms of the loan or other agreements entered into in connection with the loan, the consumer has the right to cancel the credit insurance at any time and avoid paying the $10 credit insurance premium. The payment under the loan is $110. ii. Assume that in connection with a covered short-term loan, a consumer would owe on a particular date $25 in finance charges to the lender. Under the terms of the loan, the consumer has the option of paying $50 in principal on that date, in which case the lender would charge $20 in finance charges instead. The payment under the loan is $25. 1596 iii. Assume that in connection with a covered short-term loan, a consumer would owe on a particular date $25 in finance charges to the lender and $70 in principal. Under the terms of the loan, the consumer has the option of logging into her account on the lender’s website and selecting an option to defer the due date of the $70 payment toward principal. The payment under the covered loan is $95. Paragraph 5(a)(6)(iii). 1. General. Section 1041.5(a)(6)(iii) provides assumptions that a lender must make in calculating the payment under § 1041.5(a)(6) for a covered short-term loan or covered longerterm balloon-payment loan that is a line of credit (regardless of the extent to which available credit will be replenished as the consumer repays earlier advances). For a line of credit, the amount and timing of the consumer’s actual payments after consummation may depend on the consumer’s utilization of the credit or on amounts that the consumer has repaid prior to the payments in question. Section 1041.5(a)(6)(iii) requires the lender to calculate the total loan payment assuming that the consumer will utilize the full amount of credit under the loan as soon as the credit is available and that the consumer will make only minimum required payments for as long as permitted under the loan agreement. Lenders should use the same test with the same assumptions when they make a new ability-to-repay determination under § 1041.5(b)(1)(ii) prior to an advance under the line of credit that is more than 90 days after the date of a prior ability-torepay determination for the line of credit, in order to determine whether the consumer still has the ability to repay the current credit line. 5(a)(8) Residual income. 1. General. Under § 1041.5(a)(8), residual income is defined as the sum of net income that the lender projects the consumer will receive during the relevant monthly period, minus the 1597 sum of amounts that the lender projects will be payable by the consumer for major financial obligations during the relevant monthly period and payments under the covered short-term loan or covered longer-term balloon-payment loan during the relevant monthly period, all of which projected amounts are determined in accordance with § 1041.5(c). See § 1041.5(b)(2)(ii) and associated commentary for further clarification on the use of residual income methodology to determine ability to repay. For covered longer-term balloon-payment loans, where the relevant monthly period may fall well into the future relative to the consummation of the loan, the lender must calculate the residual income using the projections made under § 1041.5(c) and in so doing must make reasonable assumptions about the consumer’s net income and major financial obligations during the relevant monthly period compared to the period covered by the verification evidence. For example, the lender cannot assume, absent a reasonable basis, that there will be a substantial increase in net income or decrease in major financial obligations between consummation and the relevant monthly period. For further clarification, see comment 5(c)(1)-1 regarding the consistency between the consumer’s written statement and verification evidence and comment 5(c)(2)(ii)(A)-2 regarding what constitutes sufficient history of net income for purposes of verification evidence. 5(b) Reasonable determination required. 1. Overview. Section 1041.5(b) prohibits a lender from making a covered short-term loan (other than a covered short-term loan described in § 1041.6) or a covered longer-term balloonpayment loan or increasing the amount of credit available on such loan unless it first makes a reasonable determination that the consumer will have the ability to repay the loan according to its terms. For discussion of loan modifications, see comment 2(a)(5)-2. Section 1041.5(b) provides minimum standards that the lender’s determination must meet to constitute a reasonable 1598 determination. Section 1041.5(b)(2) provides that a lender’s ability-to-repay determination for a covered short-term loan is reasonable only if the lender reasonably concludes that, based on the estimates of the consumer’s basic living expenses for the relevant monthly period and the calculation of the consumer’s residual income or the debt-to-income ratio for the relevant monthly period, as applicable, the consumer can pay for major financial obligations, make any payments under the loan, and meet basic living expenses during the periods specified in § 1041.5(b)(2). For covered short-term loans, the periods are the shorter of the term of the loan or the period ending 45 days after consummation of the loan, and 30 days after having made the highest payment on the loan. For covered longer-term balloon-payment loans, the periods are the relevant monthly period, and 30 days after having made the highest payment on the loan. Thus, the rule requires lenders to make a debt-to-income ratio or residual income calculation for the relevant monthly period—the calendar month in which the highest payments are due on the covered short-term loan or covered longer-term balloon payment loan—and to use the results of the calculation to make reasonable inferences and draw a reasonable conclusion about whether the consumer can make loan payments, pay for major financial obligations, and meet basic living expenses during the periods specified in § 1041.5(b)(2). This analysis is designed to determine whether the consumer has the ability to repay the loan according to its terms. See § 1041.5(b)(2)(i) and 1041.5(b)(2)(ii) and corresponding commentary. 2. Reasonable determination. To comply with the requirements of § 1041.5(b), a lender’s determination that a consumer will have the ability to repay a covered short-term loan or covered longer-term balloon-payment loan must be reasonable in all respects. i. To be reasonable, a lender’s determination of a consumer’s ability to repay a covered short-term loan or covered longer-term balloon-payment loan must: 1599 A. Include the reasonable conclusions required in § 1041.5(b)(2), using either the debt-toincome ratio methodology under § 1041.5(b)(2)(i) or the residual income methodology under § 1041.5(b)(2)(ii) as applied to the relevant monthly period; B. Be based on reasonable projections of a consumer’s net income and major financial obligations during the relevant monthly period in accordance with § 1041.5(c); C. Be based on reasonable estimates of basic living expenses during the relevant monthly period. The following provides additional clarification on what constitutes reasonable estimates of basic living expenses: 1. Section 1041.5(a)(1) and (b) do not specify a particular method that a lender must use to determine a consumer’s basic living expenses. A lender is not required to itemize the basic living expenses of each consumer, but may instead arrive at estimates for the amount needed to cover the costs of food, utilities not paid as part of rental housing expenses, transportation, outof-pocket medical expenses, phone and Internet services, and childcare. A lender may reasonably estimate the dollar amount or percentage of net income the consumer will need to meet these basic living expenses based upon such sources as the lender’s own experience in making covered short-term or longer-term balloon-payment loans to similarly-situated consumers, reasonably reliable information available from government surveys or other publications about the basic living expenses of similarly-situated consumers, or some combination thereof. For example, it would be reasonable for the lender to use data about relevant categories of expenses from the Consumer Expenditure Survey of the Bureau of Labor Statistics or the Internal Revenue Code’s Collection Financial Standards, or a combination of the two data sources, to develop non-individualized estimates of food, utilities not paid as part of rental housing expenses, transportation, out-of-pocket medical expenses, phone and internet 1600 services, and childcare for consumers seeking covered short-term or longer-term balloonpayment loans. In using the data from those sources to estimate the amount spent on a particular category, the lender may make reasonable adjustments to arrive at an estimate of basic living expenses, for instance where a data source’s information on a particular type of basic living expenses overlaps with a type of major financial obligation as defined in § 1041.5(a)(3) or where a data source groups expenses into different categories than comment 5(a)(1)-2. 2. If the lender is conducting an individualized estimate by itemizing the consumer’s costs of food, utilities not paid as part of rental housing expenses, transportation, out-of-pocket medical expenses, phone and Internet services, and childcare, the lender may reasonably consider other factors specific to the consumer that are not required to be projected under § 1041.5(c). Such consumer-specific factors could include whether other persons are regularly contributing toward the consumer’s payment of basic living expenses. The lender may consider such consumer-specific factors only when it is reasonable to do so. It is not reasonable for the lender to consider whether other persons are regularly contributing toward the consumer’s payment of basic living expenses if the lender is separately including in its projection of net income any income of another person to which the consumer has a reasonable expectation of access; and D. Be consistent with a lender’s written policies and procedures required under § 1041.12 and grounded in reasonable inferences and conclusions as to a consumer’s ability to repay a covered short-term loan or covered longer-term balloon-payment loan according to its terms in light of information the lender is required to obtain or consider as part of its determination under § 1041.5(b). ii. A determination of ability to repay is not reasonable if it: 1601 A. Relies on an implicit or explicit assumption that the consumer will obtain additional consumer credit to be able to make payments under the covered short-term loan or covered longer-term balloon-payment loan, to make payments under major financial obligations, or to meet basic living expenses; B. Assumes that a consumer needs implausibly low amounts of funds to meet basic living expenses under the residual income methodology or an implausibly low percentage of net income to meet basic living expenses if a lender uses the debt-to-income methodology. For example, assume a consumer seeks a covered short-term loan. The lender uses a debt-to-income methodology to make an ability-to-repay determination. Based on the lender’s projections of the consumer’s net income and major financial obligations under § 1041.5(c), the lender calculates that the consumer’s debt-to-income ratio would be 90 percent, which means that only 10 percent of the consumer’s net income will be remaining to pay for basic living expenses. It is not reasonable for the lender to conclude under § 1041.5(b)(2) that a consumer with a 90 percent debt-to-income ratio would have the ability to repay the loan. See comment 5(b)(2)(i)-3 for additional examples of ability-to-repay determinations using the debt-to-income methodology; or C. For covered longer-term balloon-payment loans, if the lender relies on an assumption that a consumer will accumulate savings while making one or more payments under a covered longer-term balloon-payment loan and that, because of such assumed savings, the consumer will be able to make a subsequent loan payment under the loan. iii. Evidence that a lender’s determinations of ability to repay are not reasonable may include, without limitation, the factors described under paragraphs (A) through (E) of comment 5(b)-2.iii. These factors may be evaluated across a lender’s entire portfolio of covered shortterm loans or covered longer-term balloon-payment loans or with respect to particular products, 1602 geographic regions, particular periods during which the loans were made, or other relevant categorizations. Other relevant categorizations would include, without limitation, loans made in reliance on consumer statements of income in the absence of verification evidence (see comment 5(c)(2)(ii)(A)-4). The factors described under paragraphs (A) through (E) of comment 5(b)-2.iii may be considered either individually or in combination with one another. These factors also are not absolute in their application; instead, they exist on a continuum and may apply to varying degrees. Each of these factors is viewed in the context of the facts and circumstances relevant to whether the lender’s ability-to-repay determinations are reasonable. Relevant evidence may also include a comparison of the following factors on the part of the lender to that of other lenders making covered short-term loans or covered longer-term balloon-payment loans to similarly situated consumers; however, such evidence about comparative performance is not dispositive as to the evaluation of a lender’s ability-to-repay determinations. A. Default rates. This evidence includes defaults during and at the expiration of covered loan sequences as calculated on a per sequence or per consumer basis; B. Re-borrowing rates. This evidence includes the frequency with which the lender makes consumers multiple covered short-term loans or covered longer-term balloon-payment loans within a loan sequence as defined in § 1041.2(a)(14) (i.e., consecutive or concurrent loans taken out within 30 days of a prior loan being outstanding; C. Patterns of lending across loan sequences. This evidence includes the frequency with which the lender makes multiple sequences of covered short-term loans or covered longer-term balloon-payment loans to consumers. This evidence also includes the frequency with which the lender makes consumers new covered short-term loans or covered longer-term balloon-payment 1603 loans immediately or soon after the expiration of a cooling-off period under § 1041.5(d)(2) or the 30-day period that separates one loan sequence from another (see § 1041.2(a)(14)); D. Evidence of delinquencies and collateral impacts. This evidence includes the proportion of consumers who incur late fees, failed presentments, delinquencies, and repossessions of motor vehicles for loans involving vehicle security; and E. Patterns of non-covered lending. This evidence includes the frequency with which the lender makes non-covered loans shortly before or shortly after consumers repay a covered shortterm loan or covered longer-term balloon-payment loan, and the non-covered loan bridges all or a substantial part of either the period between two loans that otherwise would be part of a loan sequence or of a cooling-off period. An example would be where the lender, its affiliate, or a service provider frequently makes 30-day non-recourse pawn loans to consumers shortly before or soon after repayment of covered short-term loans made by the lender, and where the lender then makes additional covered short-term loans to the same consumers soon after repayment of the pawn loans (see § 1041.5(e) and associated commentary). iv. Examples of evidence of the reasonableness of ability-to-repay determinations. The following examples illustrate how the factors described in comment 5(b)-2.iii may constitute evidence about whether lenders’ determinations of ability to repay are reasonable under § 1041.5(b): A. A significant percentage of consumers who obtain covered short-term loans from a lender under § 1041.5 re-borrow within 30 days of repaying their initial loan, re-borrow within 30 days of repaying their second loan, and re-borrow shortly after the end of the cooling-off period that follows the initial loan sequence of three loans. Based on the combination of these 1604 factors, this evidence suggests that the lender’s ability-to-repay determinations are not reasonable. B. A lender frequently makes at or near the maximum number of loans permitted under § 1041.6 to consumers early within a 12-month period (i.e., the loans do not require ability-torepay determinations) and then makes a large number of additional covered short-term loans to those same consumers under § 1041.5 (i.e., the loans require ability-to-repay determinations) later within the 12-month period. Assume that the loans made under § 1041.5 are part of multiple loan sequences of two or three loans each and the sequences begin soon after the expiration of applicable cooling-off periods or 30-day periods that separate one loan sequence from another. This evidence suggests that the lender’s ability-to-repay determinations for the covered short-term loans made under § 1041.5 are not reasonable. The fact that some of the loans in the observed pattern were made under § 1041.6 and thus are conditionally exempted from the ability-to-repay requirements does not mitigate the potential unreasonableness of the ability-to-repay determinations for the covered short-term loans that were made under § 1041.5. C. A lender frequently makes at or near the maximum number of loans permitted under § 1041.6 to consumers early within a 12-month period (i.e., the loans do not require ability-torepay determinations) and then only occasionally makes additional covered short-term loans to those same consumers under § 1041.5 (i.e., the loans require ability-to-repay determinations) later within the 12-month period. Very few of those additional loans are part of loans sequences longer than one loan. Absent other evidence that the ability-to-repay determination is unreasonable (see comment 5(b)-2.iii.A through E), this evidence suggests that the lender’s ability-to-repay determinations for the loans made under § 1041.5 are reasonable. 1605 D. Within a lender’s portfolio of covered short-term loans, a small percentage of loans result in default, consumers generally have short loan sequences (fewer than three loans), and the consumers who take out multiple loan sequences typically do not begin a new loan sequence until several months after the end of a prior loan sequence. There is no evidence of the lender or an affiliate making non-covered loans to consumers to bridge cooling-off periods or the periods between loan sequences. This evidence suggests that the lender’s ability-to-repay determinations are reasonable. 3. Payments under the covered short-term loan or longer-term balloon-payment loan. Under the ability-to-repay requirements in § 1041.5(b)(2)(i) and (ii), a lender must determine the amount of the payments due in connection with the covered short-term loan or covered longerterm balloon-payment loan during the relevant monthly period. See § 1041.5(a)(6) for the definition of payment under a covered short-term loan or covered longer-term balloon-payment loan, including assumptions that the lender must make in calculating the amount of payments under a loan that is a line of credit. Paragraph 5(b)(2). 1. General. For a covered short-term loan, § 1041.5(b)(2) requires the lender to reasonably conclude that, based on the estimates of the consumer’s basic living expenses for the relevant monthly period and the lender’s calculation of the consumer’s debt-to-income ratio or residual income for the relevant monthly period, as applicable, the consumer can pay major financial obligations, make any payments on the loan, and meet basic living expenses during the shorter of the term of the loan or the period ending 45 days after consummation of the loan, and for 30 days after having made the highest payment on the loan. See § 1041.5(b)(2)(i)(A) (the debt-to-income methodology) and § 1041.5(b)(2)(ii)(A) (the residual income methodology) and 1606 corresponding commentary. For a covered longer-term balloon-payment loan, § 1041.5(b)(2) requires the lender to reasonably conclude that, based on the estimates of the consumer’s basic living expenses for the relevant monthly period and the lender’s calculation of the consumer’s debt-to-income ratio or residual income, as applicable, the consumer can pay major financial obligations, make any payments on the loan, and meet basic living expenses during the relevant monthly period, and for 30 days after having made the highest payment on the loan. See § 1041.5(b)(2)(i)(B) (the debt-to-income methodology) and § 1041.5(b)(2)(ii)(B) (the residual income methodology) and corresponding commentary. If the loan has two or more payments that are equal to each other in amount and higher than all other payments, the date of the highest payment under the loan is considered the later in time of the two or more highest payments. Under § 1041.5(b)(2), lenders must comply with either § 1041.5(b)(2)(i) or (ii) depending on whether they utilize the residual income or debt-to-income ratio methodology. Paragraph 5(b)(2)(i). 1. Relation of periods under § 1041.5(b)(2)(i) to relevant monthly period. Section 1041.5(a)(2) defines debt-to-income ratio as the ratio, expressed as a percentage, of the sum of net income that the lender projects the consumer will receive during the relevant monthly period, minus the sum of amounts that the lender projects will be payable by the consumer for major financial obligations during the relevant monthly period and the payments under the covered short-term loan or covered longer-term balloon-payment loan during the relevant monthly period, all of which projected amounts are determined in accordance with § 1041.5(c). Comment 5(a)(2)-1 clarifies that the relevant monthly period is the calendar month during which the highest sum of payments on the loan is due. The relevant monthly period is not the same period as the periods set forth in § 1041.5(b)(2)(i), which for covered short-term loans are the 1607 shorter of the loan term or 45 days following consummation, and 30 days following the date of the highest payment under the loan, and for covered longer-term balloon-payment loans are the relevant monthly period, and 30 days following the date of the highest payment under the loan. There may be overlap between the relevant monthly period and the periods set forth in § 1041.5(b)(2)(i), but the degree of overlap will depend on the contractual duration of the loan and the consummation and contractual due dates. For example, assume a consumer takes a covered short-term loan of 30 days in duration that is consummated on June 15 and with a single payment due on July 14. The relevant monthly period is the calendar month in which the sum of the highest payments on the loan is due, which is the calendar month of July. This means that a portion of both the loan term (i.e., June 15 to June 30) and the 30-day period following the date of the highest payment on the loan (i.e., August 1 to August 13) are outside of the relevant monthly period. 2. Use of projections for relevant monthly period to comply with § 1041.5(b)(2)(i). The lender is not required under § 1041.5(b)(2)(i) to estimate the consumer’s basic living expenses, make a projection under § 1041.5(c) of the consumer’s net income and major financial obligations, or calculate the consumer’s debt-to-income ratio for any period other than the relevant monthly period. The lender may use the estimates of the consumer’s basic living expenses for the relevant monthly period, the projections about the consumer’s net income and major financial obligations during the relevant monthly period, and the calculation of the consumer’s debt-to-income ratio as a baseline of information from which to make reasonable inferences and draw a reasonable conclusion about whether the consumer will pay major financial obligations, make the payments on the loan, and meet basic living expenses during the periods specified in § 1041.5(b)(2)(i). To make reasonable inferences and draw a reasonable 1608 conclusion, the lender cannot, for example, assume that the consumer will defer payment of major financial obligations and basic living expenses until after the 30-day period that follows the date of the highest payment on the loan, or assume that obligations and expenses (other than payments on the covered loan itself) during the 30-day period will be less than during the relevant monthly period. Nor can the lender assume the consumer will be able to obtain additional credit during the loan term or during the 30-day period that follows the highest payment on the loan. 3. Examples. The following examples illustrate § 1041.5(b)(2)(i): i. Assume a lender considers making a covered short-term loan to a consumer on March 1. The prospective loan would be repayable in a single payment of $385 on March 17. The lender calculates that, based on its projections of the consumer’s net income and major financial obligations during March (i.e., the relevant monthly period), the consumer will have a debt-toincome ratio of 55 percent. The lender complies with the requirement in § 1041.5(b)(2) if, using that debt-to-income ratio, the lender reasonably concludes that the consumer can pay for major financial obligations, make the loan payment, and meet basic living expenses during the loan term and to pay for major financial obligations and meet basic living expenses for 30 days following the contractual due date (i.e., from March 18 to April 16). The lender would not make a reasonable conclusion if the lender were to assume, for example, that the consumer would defer payment of major financial obligations until after April 16 or that the consumer would obtain an additional extension of credit on April 1. ii. Assume a lender considers making a covered longer-term balloon-payment loan to a consumer on March 1. The prospective loan would be repayable in six biweekly payments. The first five of which would be for $100, and the last of which would be for $275, due on May 20. 1609 The highest sum of these payments that would be due within a monthly period would be $375, during the month of May. The lender further calculates that, based on its projections of net income and major financial obligations during the relevant monthly period, the consumer will have a debt-to-income ratio of 50 percent. The lender complies with the requirement in § 1041.5(b)(2)(i) if, applying that debt-to-income ratio, the lender reasonably concludes that the consumer can pay for major financial obligations, make the payments under the loan, and meet basic living expenses during the month in which the highest sum of payments on the loan are due (i.e., during the month of May) and for 30 days following the highest payment on the loan (i.e., from May 21 to June 19). The lender would not make a reasonable conclusion if the lender were to assume, for example, that the consumer would defer payment of major financial obligations until after June 19 or that the consumer would obtain an additional extension of credit on June 1. Paragraph 5(b)(2)(ii). 1. Relation of periods under § 1041.5(b)(2)(ii) to relevant monthly period. Section 1041.5(a)(8) defines residual income as the sum of net income that the lender projects the consumer will receive during the relevant monthly period, minus the sum of the amounts that the lender projects will be payable by the consumer for major financial obligations during the relevant monthly period and payments under the covered short-term loan or covered longer-term balloon-payment loan during the relevant monthly period, all of which projected amounts are determined in accordance with paragraph (c). The relevant monthly period is the calendar month in which the highest sum of payments on the loan is due. The relevant monthly period is not the same period as the periods set forth in § 1041.5(b)(2)(ii), although there may be some overlap. See comment 5(b)(2)(i)-1 for further clarification and an analogous example. 2. Use of projections for relevant monthly period to comply with § 1041.5(b)(2)(ii). The 1610 lender is not required under § 1041.5(b)(2)(ii) to estimate the consumer’s basic living expenses, make a projection under § 1041.5(c) of the consumer’s net income and major financial obligations, or calculate the consumer’s residual income for any period other than the relevant monthly period. The lender may use the estimates of the consumer’s basic living expenses for the relevant monthly period, projections about the consumer’s net income and major financial obligations during the relevant monthly period and the calculation of the consumer’s residual income as a baseline of information on which to make reasonable inferences and draw a reasonable conclusion about whether the consumer will pay major financial obligations, make the payments on the loan, and meet basic living expenses during the periods specified in § 1041.5(b)(2)(ii). See comment 5(b)(2)(i)-2 for further clarification. 3. Examples. The following examples illustrate § 1041.5(b)(2)(ii): i. Assume a lender considers making a covered short-term loan to a consumer on March 1. The prospective loan would be repayable in a single payment of $385 on March 17. The lender calculates that, based on its projections of the consumer’s net income and major financial obligations during March (i.e., the relevant monthly period), the consumer will have $1,000 in residual income for the month. The lender complies with the requirement in § 1041.5(b)(2)(ii) if, based on the calculation of residual income, it reasonably concludes that the consumer will be able to pay major financial obligations, make the loan payment, and meet basic living expenses during the loan term and for 30 days following the contractual due date (i.e., from March 18 to April 16). The lender would not make a reasonable conclusion if the lender were to assume, for example, that the consumer would defer payment of major financial obligations until after April 16, that the consumer would obtain an additional extension of credit on April 1, or that the consumer’s net income will increase in April relative to the relevant monthly period (i.e., 1611 March). ii. Assume a lender considers making a covered longer-term balloon-payment loan to a consumer on March 1. The prospective loan would be repayable in six biweekly payments. The first five payments would be for $100, and the last payment would be for $275, on May 20. The highest sum of these payments that would be due within a monthly period would be $375, during the month of May. The lender further calculates that, based on its projections of net income and major financial obligations during the relevant monthly period (i.e., May), and accounting for the $375 amount, which is the highest sum of loan payments due within a monthly period, the consumer will have $1,200 in residual income. The lender complies with the requirement in § 1041.5(b)(2)(ii) if, based on the calculation of residual income, it reasonably concludes that the consumer will be able to pay major financial obligations, make the loan payments, and meet basic living expenses during the relevant monthly period (i.e., May) and to pay for basic living expenses and major financial obligations for 30 days following the highest payment on the loan (i.e., from May 21 to June 19). The lender would not make a reasonable conclusion if the lender were to assume, for example, that the consumer would be able to defer payment of major financial obligations until after June 19 or that the consumer would obtain an additional extension of credit on June 1, or that the consumer’s net income will increase in June relative to the relevant monthly period (i.e., May). 5(c) Projecting consumer net income and payments for major financial obligations. Paragraph 5(c)(1). 1. General. Section 1041.5(c)(1) requires lenders to consider major financial obligations that are listed in a consumer’s written statement described in § 1041.5(c)(2)(i)(B) even if the obligations do not appear in the national credit report or other verification documentation that 1612 lenders are required to compile under § 1041.5(c)(2)(ii)(B). To be reasonable, § 1041.5(c)(1) provides that a projection of the amount of net income or payments for major financial obligations may be based on a consumer’s written statement of amounts under § 1041.5(c)(2)(i) only as specifically permitted by § 1041.5(c)(2)(ii) or (iii) or to the extent the stated amounts are consistent with the verification evidence that is obtained in accordance with § 1041.5(c)(2)(ii). Section 1041.5(c)(1) further provides that, in determining whether the stated amounts are consistent with the verification evidence, the lender may reasonably consider other reliable evidence the lender obtains from or about the consumer, including any explanations the lender obtains from the consumer. For example: i. Assume that a consumer states that her net income is $900 every two weeks, pursuant to § 1041.5(c)(2)(i)(A). The consumer pay stub the lender obtains as reasonably available verification evidence pursuant to § 1041.5(c)(2)(ii)(A) shows that the consumer received $900 during the preceding pay period. The lender complies with § 1041.5(c)(1) if it makes the determination required under § 1041.5(b) based on a projection of $1,800 in net income for the relevant monthly period because the reasonably available verification evidence supports a projection of $900 in net income every two weeks. ii. Assume that a consumer states that net income is $1,000 every two weeks, pursuant to § 1041.5(c)(2)(i)(A). The lender obtains a copy of the consumer’s recent deposit account transaction records as verification evidence pursuant to § 1041.5(c)(2)(ii)(A). The account transaction records show biweekly take-home pay of $800 during the preceding two-week period. The lender does not comply with § 1041.5(c)(1) if it makes the determination required under § 1041.5(b) based on a net income projection of a $2,000 for the relevant monthly period because this projection is not consistent with the reasonably available verification evidence 1613 (which, rather, is consistent with a total of $1,600 net income for the relevant monthly period). The lender may request additional deposit account transaction records for prior recent pay cycles and may reasonably project $2,000 in net income for the relevant monthly period if such additional evidence is consistent with the consumer’s statement. iii. Assume that a consumer states that net income is $1,000 every two weeks, pursuant to § 1041.5(c)(2)(i)(A). The lender obtains a copy of the consumer’s recent deposit account transaction records as verification evidence pursuant to § 1041.5(c)(2)(ii)(A). The account transaction records show biweekly take-home pay of $800 during the preceding two-week period. Assume also, however, that the consumer states that the consumer supplements his regular payroll income with cash income from a second job, for which verification evidence is not reasonably available because the consumer is paid in cash and does not deposit the cash into the consumer’s bank account, and that the consumer earns between $100 and $300 every two weeks from this job. In this instance, the lender complies with § 1041.5(c)(1) if it makes the determination required under § 1041.5(b) based on a net income projection of $2,000 for the relevant monthly period. The lender’s projection includes both the payroll income from the first job for which verification evidence is reasonably available and the cash income from the second job for which verification evidence is not reasonably available (see comment 5(c)(2)(ii)(A)-3). In such circumstances, the lender may reasonably consider the additional income reflected in the consumer’s written statement pursuant to § 1041.5(c)(2)(ii)(A)(1). iv. Assume that a consumer states that her net income is $1,000 every two weeks, pursuant to § 1041.5(c)(2)(i)(A). The lender obtains electronic records of the consumer’s deposit account transactions as verification evidence pursuant to § 1041.5(c)(2)(ii)(A) showing a biweekly direct deposit $800 during the preceding two-week period and a biweekly direct 1614 deposit of $1,000 during the prior two-week period. The consumer explains that the most recent income was lower than her usual income of $1,000 because she missed two days of work due to illness. The lender complies with § 1041.5(c)(1) if it makes the determination required under § 1041.5(b) based on a projection of $2,000 for the relevant monthly period because it reasonably considers the consumer’s explanation in determining whether the stated amount is consistent with the verification evidence. v. Assume that a consumer states that her net income is $2,000 every two weeks, pursuant to § 1041.5(c)(2)(i)(A). The lender obtains electronic records of the consumer’s deposit account transactions as verification evidence pursuant to § 1041.5(c)(2)(ii)(A) showing no income transactions in the preceding month but showing consistent biweekly direct deposits of $2,000 from ABC Manufacturing prior to that month. The consumer explains that she was temporarily laid off for one month while ABC Manufacturing retooled the plant where she works but that she recently resumed work there. The lender complies with § 1041.5(c)(1) if it makes the determination required under § 1041.5(b) based on a projection of $4,000 for the relevant monthly period because it reasonably considers the consumer’s explanation in determining whether the stated amount is consistent with the verification evidence. vi. Assume that a consumer states that she owes a child support payment of $200 each month, pursuant to § 1041.5(c)(2)(i)(B). The national consumer report that the lender obtains as verification evidence pursuant to § 1041.5(c)(2)(ii)(C) does not include any child support payment. The lender must consider the child support obligation listed in the written statement. The lender complies with § 1041.5(c)(1) if it reasonably relies on the amount in the consumer’s written statement pursuant to § 1041.5(c)(2)(ii)(C) to make the determination required under § 1041.5(b) based on a projection of a $200 child support payment each month. 1615 vii. Assume that a consumer does not list a student loan in her written statement pursuant to § 1041.5(c)(2)(i)(B), but the national consumer report that the lender obtains as verification evidence pursuant to § 1041.5(c)(2)(ii)(B) lists such a loan with a payment due during the relevant monthly period. The lender does not comply with § 1041.5(c)(1) if it makes the determination required under § 1041.5(b) without including the student loan payment based on the consumer’s failure to list the loan in the written statement or on the consumer’s explanation that the loan has recently been paid off. The lender may obtain and reasonably consider other reliable evidence, such as records from the consumer or an updated national consumer report, and may exclude the student loan payment if such additional evidence is consistent with the consumer’s statement or explanation. viii. Assume that a consumer states that he owes a child support payment of $200 each month, pursuant to § 1041.5(c)(2)(i)(B). The national consumer report that the lender obtains as verification evidence pursuant to § 1041.5(c)(2)(ii)(C) includes the child support payment. The consumer states, further, that his child support payment is deducted out of his paycheck prior to his receipt of take-home pay. The lender obtains a recent pay stub of the consumer as verification evidence which shows a $200 deduction but does not identify the payee or include any other information regarding the nature of the deduction. The lender complies with § 1041.5(c)(1) if it makes the determination required under § 1041.5(b) based on a projection of major financial obligations that does not include the $200 child support payment each month, because it relies on the consumer’s statement that the child support payment is deducted from his paycheck prior to receipt of take-home pay and nothing in the verification evidence is inconsistent with the statement. 1616 2. Consumer-specific factors regarding payment of major financial obligations. Under § 1041.5(c)(1), in projecting major financial obligations the lender may consider consumerspecific factors, such as whether other persons are regularly contributing toward the consumer’s payment of major financial obligations. The lender may consider such consumer-specific factors only when it is reasonable to do so. It is not reasonable for the lender to consider whether other persons are regularly contributing toward the consumer’s payment of major financial obligations if the lender is separately including in its projection of net income any income of another person to which the consumer has a reasonable expectation of access (see comment 5(a)(5)-3). 5(c)(2) Evidence of net income and payments for major financial obligations. Paragraph 5(c)(2)(i). 1. Statements from the consumer. Section 1041.5(c)(2)(i) requires a lender to obtain a consumer’s written statement of the amounts of the consumer’s net income and payments for the consumer’s major financial obligations currently and for the relevant monthly period. Section 1041.5(c)(2)(i) also provides that the written statement from the consumer may include a statement from the consumer about the amount of any income of another person to which the consumer has a reasonable expectation of access. A consumer’s written statement includes a statement the consumer writes on a paper application or enters into an electronic record, or an oral consumer statement that the lender records and retains or memorializes in writing or electronically and retains. Paragraph 5(c)(2)(ii). 1. Verification requirement. Section 1041.5(c)(2)(ii) establishes requirements for a lender to obtain verification evidence for the amounts of a consumer’s net income and required payments for major financial obligations other than rental housing expense. 1617 Paragraph 5(c)(2)(ii)(A). 1. Income. Section 1041.5(c)(2)(ii)(A) requires a lender to obtain a reliable record (or records) of an income payment (or payments) directly to the consumer covering sufficient history to support the lender’s projection under § 1041.5(c)(1) if a reliable record (or records) of income payment (or payments) is reasonably available. Section 1041.5(c)(2)(ii)(A) also provides that if the lender elects to include as the consumer’s net income for the relevant monthly period the income of another person to which the consumer has a reasonable expectation of access, the lender must obtain verification evidence of that income in the form of a reliable record (or records) demonstrating that the consumer has regular access to that income. Such verification evidence could consist of bank account statements indicating that the consumer has access to a joint bank account in which the other person’s income is deposited, or that the other person regularly deposits income into the consumer’s bank account (see comment 5(a)(5)-3 for further clarification). For purposes of verifying net income, a reliable transaction record includes a facially genuine original, photocopy, or image of a document produced by or on behalf of the payer of income, or an electronic or paper compilation of data included in such a document, stating the amount and date of the income paid to the consumer. A reliable transaction record also includes a facially genuine original, photocopy, or image of an electronic or paper record of depository account transactions, prepaid account transactions (including transactions on a general purpose reloadable prepaid card account, a payroll card account, or a government benefits card account) or money services business check-cashing transactions showing the amount and date of a consumer’s receipt of income. 2. Sufficient history. Under § 1041.5(c)(2)(ii)(A), the lender must obtain a reliable record or records of the consumer’s net income covering sufficient history to support the lender’s 1618 projection under § 1041.5(c). For a covered short-term loan, sufficient history typically would consist of one biweekly pay cycle or one monthly pay cycle, depending on how frequently the consumer is paid. However, if there is inconsistency between the consumer’s written statement regarding net income and the verification evidence which must be reconciled by the lender (see comment 5(c)(1)-1), then depending on the circumstances more than one pay cycle may be needed to constitute sufficient history. For a covered longer-term balloon-payment loan, sufficient history would generally consist of two biweekly pay cycles or two monthly pay cycles, depending on how frequently the consumer is paid. However, depending on the length of the loan, and the need to resolve inconsistency between the consumer’s written statement regarding net income and the verification evidence, more than two pay cycles may be needed to constitute sufficient history. 3. Reasonably available. The lender’s obligation to obtain a reliable record (or records) of income payment (or payments) covering sufficient history to support the lender’s projection under § 1041.5(c)(1) applies if and to the extent a reliable record (or records) is reasonably available. A reliable record of the consumer’s net income is reasonably available if, for example, the consumer’s source of income is from her employment and she possesses or can access a copy of the consumer’s recent pay stub. The consumer’s recent transaction account deposit history is a reliable record (or records) that is reasonably available if the consumer has such an account. With regard to such bank account deposit history, the lender could obtain it directly from the consumer or, at its discretion, with the consumer’s permission via an account aggregator service that obtains and categorizes consumer deposit account and other account transaction data. In situations in which income is neither documented through pay stubs nor transaction account records, the reasonably available standard requires the lender to act in good faith and exercise 1619 due diligence as appropriate for the circumstances to determine whether another reliable record (or records) is reasonably available. 4. Reasonable reliance on consumer’s statement if reliable record not reasonably available. Under § 1041.5(c)(2)(ii)(A), if a lender determines that a reliable record (or records) of some or all of the consumer’s net income is not reasonably available, the lender may reasonably rely on the consumer’s written statement described in § 1041.5(c)(2)(i)(A) for that portion of the consumer’s net income. Section 1041.5(c)(2)(ii)(A) does not permit a lender to rely on a consumer’s written statement that the consumer has a reasonable expectation of access to the income of another person (see comment 5(c)(2)(ii)(A)-1). A lender reasonably relies on the consumer’s written statement if such action is consistent with a lender’s written policies and procedures required under § 1041.12 and there is no indication that the consumer’s stated amount of net income on a particular loan is implausibly high or that the lender is engaged in a pattern of systematically overestimating consumers’ income. Evidence of the lender’s systematic overestimation of consumers’ income could include evidence that the subset of the lender’s portfolio consisting of the loans where the lender relies on the consumers’ statements to project income in the absence of verification evidence perform worse, on a non-trivial level, than other covered loans made by the lender with respect to the factors noted in comment 5(b)-2.iii indicating poor loan performance (e.g., high rates of default, frequent re-borrowings). If the lender periodically reviews the performance of covered short-term loans or covered longer-term balloon-payment loans where the lender has relied on consumers’ written statements of income and uses the results of those reviews to make necessary adjustments to its policies and procedures and future lending decisions, such actions indicate that the lender is reasonably relying on consumers’ statements. Such necessary adjustments could include, for example, the 1620 lender changing its underwriting criteria for covered short-term loans to provide that the lender may not rely on the consumer’s statement of net income in absence of reasonably available verification evidence unless the consumer’s debt-to-income ratio is lower, on a non-trivial level, than that of similarly situated applicants who provide verification evidence of net income. A lender is not required to consider income that cannot be verified other than through the consumer’s written statement. For an illustration of a lender’s reliance on a consumer’s written statement as to a portion of her income for which verification evidence is not reasonably available, see comment 5(c)(1)-1.iii. Paragraph 5(c)(2)(ii)(B). 1. Payments under debt obligations. To verify a consumer’s required payments under debt obligations, § 1041.5(c)(2)(ii)(B) requires a lender to obtain a national consumer report, the records of the lender and its affiliates, and a consumer report obtained from an information system that has been registered for 180 days or more pursuant to § 1041.11(c)(2) or is registered pursuant to § 1041.11(d)(2), if available. A lender satisfies its obligation under § 1041.5(d)(1) to obtain a consumer report from an information system that has been registered for 180 days or more pursuant to § 1041.11(c)(2) or is registered pursuant to § 1041.11(d)(2), if available, when it complies with the requirement in § 1041.5(c)(2)(ii)(B) to obtain this same consumer report. See comment 5(a)(3)-1 regarding the definition of required payments. 2. Deduction of debt obligations prior to consumer’s receipt of take-home pay. If verification evidence shows that a debt obligation is deducted prior to the consumer’s receipt of take-home pay, the lender does not include the debt obligation in the projection of major financial obligations under § 1041.5(c). 1621 3. Inconsistent information. If the consumer reports and lender and affiliate records do not include a debt obligation listed in the consumer’s written statement described in § 1041.5(c)(2)(ii)(B), the lender must consider the debt obligation listed in the consumer’s written statement to make a reasonable projection of the amount of payments for debt obligations. The lender may reasonably rely on the written statement in determining the amount of the required payment for the debt obligation. If the reports and records include a debt obligation that is not listed in the consumer’s written statement, the lender must consider the debt obligation listed in the report or record unless it obtains additional verification evidence confirming that the obligation has been paid off or otherwise released. A lender is not responsible for information about a major financial obligation that is not owed to the lender, its affiliates, or its service providers if such obligation is not listed in a consumer’s written statement, a national consumer report, or a consumer report from an information system that has been registered for 180 days or more pursuant to § 1041.11(c)(2) or is registered pursuant to § 1041.11(d)(2). Paragraph 5(c)(2)(ii)(C). 1. Payments under child support or alimony obligations. Section 1041.5(c)(2)(ii)(B) requires a lender to obtain a national consumer report to verify a consumer’s required payments under child support obligations or alimony obligations under § 1041.5(c)(2)(ii)(C). A lender may use the same national consumer report to satisfy the verification requirements under both § 1041.5(c)(2)(ii)(B) and (C). See comment 5(c)(2)(ii)(B) for clarification on the interplay between this obligation and § 1041.5(d)(1). If the report does not include a child support or alimony obligation listed in the consumer’s written statement described in § 1041.5(c)(2)(i)(B), the lender must consider the obligation listed in the consumer’s written statement to make a 1622 reasonable projection of the amount of payments for the child support or alimony obligation. The lender may reasonably rely on the written statement in determining the amount of the required payment for the obligation. 2. Deduction of child support or alimony obligations prior to consumer’s receipt of takehome pay. If verification evidence shows that a child support or alimony obligation is deducted prior to the consumer’s receipt of take-home pay, the lender does not include the child support or alimony obligation in the projection of major financial obligations under § 1041.5(c). For an illustration, see comment 5(c)(1)-1.viii. Paragraph 5(c)(2)(ii)(D). 1. Exception to obligation to obtain consumer report. Section 1041.5(c)(2)(ii)(D) provides that notwithstanding § 1041.5(c)(2)(ii)(B) and (C), a lender is not required to obtain a consumer report to verify debt obligations and child support obligations if during the preceding 90 days: the lender or its affiliate has obtained a national consumer report for the consumer, retained the report under § 1041.12(b)(1)(ii) and checked it again in connection with the new loan; and the consumer did not complete a loan sequence of three loans under § 1041.5 and trigger the 30-day cooling-off period under § 1041.5(d)(2) since the previous report was obtained. To illustrate how the two conditions relate to each other, assume a consumer obtains a sequence of three covered short-term loans under § 1041.5, with each loan being 15 days in duration, the first loan consummating on June 1, and the final loan no longer being outstanding as of July 15. The lender obtained a consumer report on May 30 as part of its ability-to-repay determination for the first loan in the sequence. Under § 1041.5(c)(2)(ii)(D), the lender is not required to obtain a consumer report for the second and third loan in the sequence. Because the consumer took a three-loan sequence, the consumer is subject to a 30-day cooling-off period 1623 which expires on August 15 pursuant to § 1041.5(d)(2). If the consumer returns to the lender for another covered short-term loan under § 1041.5 on August 15, the lender must obtain a consumer report under § 1041.5(c)(2)(ii)(B) and (C) to verify debt obligations and child support even though fewer than 90 days has elapsed since the lender previously obtained a consumer report for the consumer because the consumer completed a three-loan sequence and triggered the 30day cooling-off period since the previous report was obtained. 2. Conflicts between consumer’s written statement and national consumer report. A lender is not required to obtain a new national consumer report if the conditions under § 1041.5(c)(2)(ii)(D) are met; however, there may be circumstances in which a lender would voluntarily obtain a new national consumer report to resolve potential conflicts between a consumer’s written statement and a national consumer report obtained in the previous 90 days. See comments 5(c)(1)-1.vii and 5(c)(2)(ii)(B)-3. 1624 Paragraph 5(c)(2)(iii). 1. Rental housing expense. Section 1041.5(c)(2)(iii) provides that for the consumer’s housing expense other than a payment for a debt obligation that appears on a national consumer report obtained pursuant to § 1041.5(c)(2)(ii)(B) (i.e., with respect to lease or other rental housing payments), the lender may reasonably rely on the consumer’s statement described in § 1041.5(c)(2)(i)(B). A lender reasonably relies on the consumer’s written statement if such actions are consistent with a lender’s written policies and procedures required under § 1041.12, and there is no evidence that the stated amount for rental housing expense on a particular loan is implausibly low or that there is a pattern of the lender underestimating consumers’ rental housing expense. 2. Mortgage obligations. For a housing expense under a debt obligation (i.e., a mortgage), a lender generally must verify the obligation by obtaining a national consumer report that includes the housing expense under a debt obligation pursuant to § 1041.5(c)(2)(ii)(B). Under § 1041.5(c)(2)(ii)(D), however, a lender is not required to obtain a national consumer report if, during the preceding 90 days, the lender or its affiliate has obtained a national consumer report for the consumer and the consumer has not taken out a loan sequence of three loans under § 1041.5 and triggered the 30-day cooling-off period under § 1041.5(d)(2) since the previous report was obtained (see comment 5(c)(2)(ii)(D)-1). 5(d) Additional limitations on lending—covered short-term loans and covered longerterm balloon-payment loans. Paragraph 5(d). 1. General. Section 1041.5(d) specifies certain circumstances in which making a new covered short-term loan or a covered longer-term balloon-payment loan under § 1041.5 during or 1625 after a sequence of covered short-term loans, covered longer-term balloon-payment loans, or a combination of covered short-term loans and covered longer-term balloon-payment loans is prohibited during a mandatory cooling-off period. The prohibitions apply to making a covered short-term loan or covered longer-term balloon-payment loan under § 1041.5. 2. Application to rollovers. The prohibitions in § 1041.5(d) apply to new covered shortterm loans or covered longer-term balloon-payment loans under § 1041.5, as well as to loans that are a rollover of a prior loan (or what is termed a “renewal” in some States). Rollovers are defined as a matter of State law but typically involve deferral of repayment of the principal amount of a short-term loan for a period of time in exchange for a fee. In the event that a lender is permitted under State law to roll over a loan, the rollover would be treated as applicable as a new covered short-term loan or covered longer-term balloon-payment loan that, depending on when it occurs in the sequence, would be subject to the prohibitions in § 1041.5(d). For example, assume that a lender is permitted under applicable State law to roll over a covered short-term loan and the lender makes a covered short-term loan with $500 in principal and a 14day contractual duration. Assume that the consumer returns to the lender on day 14 (the repayment date of the first loan), the lender reasonably determines that the consumer has the ability to repay a new loan, and the consumer is offered the opportunity to roll over the first loan for an additional 14 days for a $75 fee. The rollover would be the second loan in a loan sequence, as defined under § 1041.2(a)(14), because fewer than 30 days would have elapsed between consummation of the new covered short-term loan (the rollover) and the consumer having had a covered short-term loan made under § 1041.5 outstanding. Assume that the consumer returns on day 28 (the repayment date of the first rollover, i.e., the second loan in the sequence) and the lender again reasonably determines that the consumer has the ability to repay a 1626 new loan and offers to roll over the loan again for an additional 14 days for a $75 fee. The second rollover would be the third loan in a loan sequence. If the consumer were to return on day 42 (the repayment date of the second rollover, which is the third loan in the sequence) and attempt to roll over the loan again, that rollover would be considered the fourth loan in the loan sequence. Therefore, that rollover would be prohibited and the consumer could not obtain another covered short-term loan or covered longer-term balloon-payment loan until the expiration of the 30-day cooling-off period, which begins after the consumer repays the second rollover (i.e., the third loan in the sequence). 5(d)(1) Borrowing history review. 1. Relationship to § 1041.5(c)(2)(ii)(B) and (C). A lender satisfies its obligation under § 1041.5(d)(1) to obtain a consumer report from an information system that has been registered for 180 days or more pursuant to § 1041.11(c)(2) or is registered pursuant to § 1041.11(d)(2), if available, when it complies with the requirement in § 1041.5(c)(2)(ii)(B) and (C) to obtain this same consumer report. 2. Availability of information systems that have been registered for 180 days or more pursuant to § 1041.11(c)(2) or are registered pursuant to § 1041.11(d)(2). If no information systems that have been registered for 180 days or more pursuant to § 1041.11(c)(2) or are registered pursuant to § 1041.11(d)(2) are available at the time that the lender is required to obtain the information about the consumer’s borrowing history, the lender is nonetheless required to obtain information about the consumer’s borrowing history from the records of the lender and its affiliates and to obtain the consumer’s statement about the amount and timing of payments of major financial obligations as required under § 1041.5(c)(2)(i)(B) (which would include information on current debt obligations including any outstanding covered loans). A 1627 lender may be unable to obtain a consumer report from an information system that has been registered for 180 days or more pursuant to § 1041.11(c)(2) or that is registered pursuant to § 1041.11(d)(2) if, for example, all registered information systems are temporarily unavailable. 5(d)(2) Prohibition on loan sequences of more than three covered short-term loans or covered longer-term balloon-payment loans made under § 1041.5. 1. Prohibition. Section 1041.5(d)(2) prohibits a lender from making a fourth covered short-term loan or covered longer-term balloon-payment loan under § 1041.5 in a loan sequence of covered short-term loans, covered longer-term balloon-payment loans, or a combination of covered short-term loans and covered longer-term balloon-payment loans made under § 1041.5. See § 1041.2(a)(14) for the definition of a loan sequence. 2. Examples. The following examples illustrate application of the prohibition under § 1041.5(d)(2): i. Assume that a lender makes a covered short-term loan to a consumer under the requirements of § 1041.5 on February 1 with a contractual due date of February 15, the consumer repays the loan on February 15, and the consumer returns to the lender on March 1 for another loan. Assume that the second loan is a covered short-term loan with a contractual due date of March 15. The second loan would be part of the same loan sequence as the first loan because 30 or fewer days have elapsed since repayment of the first loan. Assume that the lender makes the second loan, the consumer repays the loan on March 15, and the consumer returns to the lender on April 1 for another loan. Assume that the third loan is a covered short-term loan with a contractual due date of April 15. The third loan would be part of the same loan sequence as the first and second loans because 30 or fewer days have elapsed since repayment of the second loan. Assume that the lender makes the third loan and the consumer repays the loan on April 15. 1628 Assume that all loans are reported to a registered information system. The consumer would not be eligible for another covered short-term loan or covered longer-term balloon-payment loan under § 1041.5(d) from any lender until a 30-day cooling-off period following April 15 has elapsed, that is, starting on May 16. The consumer also would not be eligible for another covered short-term loan under § 1041.6 during the same 30-day cooling-off period. See § 1041.6(c)(2) and accompanying commentary. ii. Assume that a lender makes a covered short-term loan to a consumer under the requirements of § 1041.5 on February 1 with a contractual due date of February 15, the consumer repays the loan on February 15, and the consumer returns to the lender on March 1 for another loan. Assume that the second loan is a covered longer-term balloon-payment loan that has biweekly installment payments followed by a final balloon payment on the contractual due date of May 1. The second loan would be part of the same loan sequence as the first loan because 30 or fewer days have elapsed since repayment of the first loan. Assume that the lender makes the second loan, the consumer repays the loan in full as of May 1, and the consumer returns to the lender on May 15 for another loan. Assume that the third loan is a covered short-term loan with a contractual due date of May 30. The third loan would be part of the same loan sequence as the first and second loans because 30 or fewer days have elapsed since repayment of the second loan. Assume that the lender makes the third loan and the consumer repays the loan on May 30. Assume that all loans are reported to a registered information system. The consumer would not be eligible to receive another covered short-term loan or covered longer-term balloon-payment loan under § 1041.5(d) from any lender until a 30-day cooling-off period following May 30 has elapsed, that is until after June 29. The consumer also would not be eligible for another covered 1629 short-term loan under § 1041.6 during the same 30-day cooling-off period. See § 1041.6(c)(2) and accompanying commentary. 5(e) Prohibition against evasion. 1. General. Section 1041.5(e) provides that a lender must not take any action with the intent of evading the requirements of § 1041.5. In determining whether a lender has taken action with the intent of evading the requirements of § 1041.5, the form, characterization, label, structure, or written documentation of the lender’s action shall not be dispositive. Rather, the actual substance of the lender’s action as well as other relevant facts and circumstances will determine whether the lender’s action was taken with the intent of evading the requirements of § 1041.5. If the lender’s action is taken solely for legitimate business purposes, it is not taken with the intent of evading the requirements of § 1041.5. By contrast, if a consideration of all relevant facts and circumstances reveals a purpose that is not a legitimate business purpose, the lender’s action may have been taken with the intent of evading the requirements of § 1041.5. A lender action that is taken with the intent of evading the requirements of part 1041 may be knowing or reckless. Fraud, deceit, or other unlawful or illegitimate activity may be one fact or circumstance that is relevant to the determination of whether a lender’s action was taken with the intent of evading the requirements of § 1041.5, but fraud, deceit, or other unlawful or illegitimate activity is not a prerequisite to such a finding. 2. Illustrative example—lender action that may have been taken with the intent of evading the requirements of the rule. The following example illustrates a lender action that, depending on the relevant facts and circumstances, may have been taken with the intent of evading the requirements of § 1041.5 and thus may have violated § 1041.5(e): 1630 i. A storefront payday lender makes covered short-term loans to consumers with a contractual duration of 14 days and a lump-sum repayment structure. The lender’s policies and procedures provide for a standard loan contract including a “recurring late fee” as a lender remedy that is automatically triggered in the event of the consumer’s delinquency (i.e., if the consumer does not pay the entire lump-sum amount on the contractual due date, with no grace period), and in the loan contract the consumer grants the lender authorization to initiate a recurring ACH in the event such remedy is triggered. Assume that the recurring late fee is to be paid biweekly while the loan remains outstanding and is substantially equal to or greater than the fee that the lender charges on transactions that are considered rollovers under applicable State law. The practice of imposing a recurring late fee by contract differs from the lender’s prior practice of contacting the consumer on or about the contractual due date requesting that the consumer visit the store to discuss payment options including rollovers. Assume that as a matter of practice, if a consumer does not repay the first loan in a sequence when it is due, the lender charges recurring late fees for 60 days unless the consumer repays the outstanding balance. Such a period is roughly equivalent to two 14-day loan cycles or two rollovers following the initial loan in the sequence, plus a 30-day cooling-off period. See § 1041.5(d)(2) and related commentary. Depending on the relevant facts and circumstances, this action may have been taken with the intent of evading the requirements of § 1041.5. By charging the recurring late fee for 60 days after the initial loan was due, the lender avoided its obligation under § 1041.5(b) to make an ability-to-repay determination for the second and third loans in the sequence and to comply with the mandatory cooling-off period in § 1041.5(d)(2) after the third loan was no longer outstanding. 1631 Section 1041.6 Conditional Exemption for Certain Covered Short-term Loans 6(a) Conditional exemption for certain covered short-term loans. 1. General. Under § 1041.6(a), a lender that complies with § 1041.6(b) through (e) can make a covered short-term loan without complying with the otherwise applicable requirements under § 1041.5. A lender who complies with § 1041.6 in making a covered short-term loan has not committed the unfair and abusive practice under § 1041.4 and is not subject to § 1041.5. However, nothing in § 1041.6 provides lenders with an exemption to the requirements of other applicable laws, including subpart C of part 1041 and State laws. 2. Obtaining consumer borrowing history information. Under § 1041.6(a), the lender must determine prior to making a covered short-term loan under § 1041.6 that requirements under § 1041.6(b) and (c) are satisfied. In particular, § 1041.6(a) requires the lender to obtain information about the consumer’s borrowing history from the records of the lender and the records of the lender’s affiliates. (This information about borrowing history with the lender and its affiliates is also important to help a lender avoid violations of § 1041.6(d)). Furthermore, § 1041.6(a) requires the lender to obtain a consumer report from an information system that has been registered for 180 days or more pursuant to § 1041.11(c)(2) or is registered pursuant to § 1041.11(d)(2). If no information systems have been registered for 180 days or more pursuant to § 1041.11(c)(2) or are registered pursuant to § 1041.11(d)(2) and available as of the time the lender is required to obtain the report, the lender cannot comply with the requirements in § 1041.6(b) and (c). A lender may be unable to obtain such a consumer report if, for example: (i) no information systems have been registered for 180 days or more pursuant to § 1041.11(c)(2) or are registered pursuant to § 1041.11(d)(2); or (ii) if information systems have been registered for 180 days or more pursuant to § 1041.11(c)(2) or are registered pursuant to § 1041.11(d)(2) but 1632 all such registered information systems are temporarily unavailable. Under these circumstances, a lender cannot make a covered short-term loan under § 1041.6. 3. Consumer reports. A lender is not responsible for inaccurate or incomplete information contained in a consumer report from an information system that has been registered for 180 days or more pursuant to § 1041.11(c)(2) or is registered pursuant to § 1041.11(d)(2). 6(b) Loan term requirements. Paragraph 6(b)(1). 1. Loan sequence. Section 1041.2(a)(14) defines a loan sequence. For further clarification and examples regarding the definition of loan sequence, see § 1041.2(a)(14). 2. Principal amount limitations—general. For a covered short-term loan made under § 1041.6, different principal amount limitations apply under § 1041.6(b)(1) depending on whether the loan is the first, second, or third loan in a loan sequence. The principal amount limitations apply regardless of whether any or all of the loans are made by the same lender, an affiliate, or unaffiliated lenders. Under § 1041.6(b)(1)(i), for the first loan in a loan sequence, the principal amount must be no greater than $500. Under § 1041.6(b)(1)(ii), for the second loan in a loan sequence, the principal amount must be no greater than two-thirds of the principal amount of the first loan in the loan sequence. Under § 1041.6(b)(1)(iii), for the third loan in a loan sequence, the principal amount must be no greater than one-third of the principal amount of the first loan in the loan sequence. 3. Application to rollovers. The principal amount limitations under § 1041.6 apply to rollovers of the first or second loan in a loan sequence as well as new loans that are counted as part of the same loan sequence. Rollovers are defined as a matter of State law but typically involve deferral of repayment of the principal amount of a short-term loan for a period of time in 1633 exchange for a fee. In the event the lender is permitted under State law to make rollovers, the lender may, in a manner otherwise consistent with applicable State law and § 1041.6, roll over a covered short-term loan made under § 1041.6, but the rollover would be treated as the next loan in the loan sequence, as applicable, and would therefore be subject to the principal amount limitations set forth in § 1041.6(b)(1) as well as other limitations in § 1041.6. For example, assume that a lender is permitted under applicable State law to make a rollover. If the consumer obtains a first loan in a loan sequence under § 1041.6 with a principal amount of $300, under § 1041.6(b)(1)(ii), the lender may allow the consumer to roll over that loan so long as the consumer repays at least $100, so that the principal of the loan that is rolled over would be no greater than $200. Similarly, under § 1041.6(b)(1)(iii), the lender may allow the consumer to roll over the second loan in the loan sequence as permitted by State law, so long as the consumer repays at least an additional $100, so that the principal of the loan that is rolled over would be no greater than $100. 4. Example. Assume that a consumer who otherwise satisfies the requirements of § 1041.6 seeks a covered short-term loan and that the lender chooses to make the loan without meeting all the specified underwriting criteria required in § 1041.5. Under § 1041.6(b)(1)(i), the principal amount of the loan must not exceed $500. Assume that the consumer obtains a covered short-term loan under § 1041.6 with a principal amount of $450, the loan is contractually due in 14 days, and the consumer repays the loan on the contractual due date. Assume that the consumer returns to the lender 10 days after the repayment of the first loan to take out a second covered short-term loan under § 1041.6. Under § 1041.6(b)(1)(ii), the principal amount of the second loan may not exceed $300. Assume, further, that the consumer is then made a covered short-term loan under § 1041.6 with a principal amount of $300, the loan is contractually due in 1634 14 days, and the consumer repays the loan on the contractual due date. If the consumer returns to the lender 25 days after the repayment of the second loan to take out a third covered shortterm loan under § 1041.6, under § 1041.6(b)(1)(iii), the principal amount of the third loan may not exceed $150. These same limitations would apply if the consumer went to a different, unaffiliated lender for the second or third loan. If, however, the consumer does not return to the lender seeking a new loan under § 1041.6 until 32 days after the date on which the second loan in the loan sequence was repaid, the subsequent loan would not be part of the prior loan sequence and instead would be the first loan in a new loan sequence. Therefore, if otherwise permissible under § 1041.6, that loan would be subject to the $500 principal amount limitation under § 1041.6(b)(1)(i). Paragraph 6(b)(2). 1. Equal payments and amortization for loans with multiple payments. Section 1041.6(b)(2) provides that for a loan with multiple payments, the loan must amortize completely during the term of the loan and the payment schedule must allocate a consumer’s payments to the outstanding principal and interest and fees as they accrue only by applying a fixed periodic rate of interest to the outstanding balance of the unpaid loan principal during every repayment period for the term of the loan. For example, if the loan has a contractual duration of 30 days with two scheduled biweekly payments, under § 1041.6(b)(2) the lender cannot require the consumer to pay interest only for the first scheduled biweekly payment and the full principal balance at the second scheduled biweekly payment. Rather, the two scheduled payments must be equal in amount and amortize over the course of the loan term in the manner required under § 1041.6(b)(2). 1635 Paragraph 6(b)(3). 1. Inapplicability of conditional exemption to a loan with vehicle security. Section 1041.6(b)(3) prohibits a lender from making a covered-short-term loan under § 1041.6 with vehicle security. If the lender or its service provider take vehicle security in connection with a covered short-term loan, the loan must be originated in compliance with all of the requirements under § 1041.5, including the ability-to-repay determination. Paragraph 6(b)(4). 1. Inapplicability of conditional exemption to an open-end loan. Section 1041.6(b)(4) prohibits a lender from making a covered short-term loan under § 1041.6 structured as an openend loan under § 1041.2(a)(16). If a covered short-term loan is structured as an open-end loan, the loan must be originated in compliance with all of the requirements under § 1041.5. 6(c) Borrowing history requirements. Paragraph 6(c)(1). 1. Preceding loans. Section 1041.6(c)(1) provides that prior to making a covered shortterm loan under § 1041.6, the lender must determine that more than 30 days has elapsed since the consumer had an outstanding loan that was either a covered short-term loan (as defined in § 1041.2(a)(10)) made under § 1041.5 or a covered longer-term balloon-payment loan (as defined in § 1041.2(a)(7)) made under § 1041.5. This requirement applies regardless of whether this prior loan was made by the same lender, an affiliate, or an unaffiliated lender. For example, assume that a lender makes a covered short-term loan to a consumer under § 1041.5, that the loan has a contractual duration of 14 days, and that the consumer repays the loan on the contractual due date. If the consumer returns for a second loan 20 days after repaying the loan, the lender cannot make a covered short-term loan under § 1041.6. 1636 Paragraph 6(c)(2). 1. Loan sequence limitation. Section 1041.6(c)(2) provides that a lender cannot make a covered short-term loan under § 1041.6 if the loan would result in the consumer having a loan sequence of more than three covered short-term loans under § 1041.6 made by any lender. This requirement applies regardless of whether any or all of the loans in the loan sequence are made by the same lender, an affiliate, or unaffiliated lenders. See comments 6(b)(1)-1 and -2 for further clarification on the definition of loan sequence, as well as § 1041.2(a)(14) and accompanying commentary. For example, assume that a consumer obtains a covered short-term loan under the requirements of § 1041.6 on February 1 that has a contractual due date of February 15, that the consumer repays the loan on February 15, and that the consumer returns to the lender on March 1 for another loan under § 1041.6. The second loan under § 1041.6 would be part of the same loan sequence because 30 or fewer days have elapsed since repayment of the first loan. Assume that the lender makes the second loan with a contractual due date of March 15, that the consumer repays the loan on March 15, and that the consumer returns to the lender on April 1 for another loan under § 1041.6. The third loan under § 1041.6 would be part of the same loan sequence as the first and second loans because fewer than 30 days have elapsed since repayment of the second loan. Assume that the lender makes the third loan, which has a contractual due date of April 15 and that the consumer repays the loan on April 15. The consumer would not be permitted to receive another covered short-term loan under § 1041.6 until the 30-day period following April 15 has elapsed, that is until after May 15, assuming the other requirements under § 1041.6 are satisfied. The consumer would also be prohibited from obtaining other forms of credit from the same lender or its affiliate for 30 days under § 1041.6(d); see comment 6(d)-1. Loans that are rollovers count toward the sequence limitation 1637 under § 1041.6(c)(2). For further clarification on how the requirements of § 1041.6 apply to rollovers, see comment 6(b)(1)-3. Paragraph 6(c)(3). 1. Consecutive 12-month period. Section 1041.6(c)(3) requires that a covered short-term loan made under § 1041.6 not result in the consumer having more than six covered short-term loans outstanding during a consecutive 12-month period or having covered short-term loans outstanding for an aggregate period of more than 90 days during a consecutive 12-month period. The consecutive 12-month period begins on the date that is 12 months prior to the proposed contractual due date of the new covered short-term loan to be made under § 1041.6 and ends on the proposed contractual due date. The lender must review the consumer’s borrowing history on covered short-term loans for the 12 months preceding the consummation date of the new covered short-term loan less the period of proposed contractual indebtedness on that loan. For example, for a new covered short-term loan to be made under § 1041.6 with a proposed contractual term of 14 days, the lender must review the consumer’s borrowing history during the 351 days preceding the consummation date of the new loan. The lender also must consider the making of the new loan and the days of proposed contractual indebtedness on that loan to determine whether the requirement under § 1041.6(c)(3) regarding the total number of covered short-term loans and total time of indebtedness on covered short-term loans during a consecutive 12-month period is satisfied. Paragraph 6(c)(3)(i). 1. Total number of covered short-term loans. Section 1041.6(c)(3)(i) provides that a lender cannot make a covered-short term loan under § 1041.6 if the loan would result in the consumer having more than six covered short-term loans outstanding in any consecutive 121638 month period. The requirement counts covered short-term loans made under either § 1041.5 or § 1041.6 toward the limit. This requirement applies regardless of whether any or all of the loans subject to the limitations are made by the same lender, an affiliate, or an unaffiliated lender. Under § 1041.6(c)(3)(i), the lender must use the consumer’s borrowing history to determine whether the loan would result in the consumer having more than six covered short-term loans outstanding during a consecutive 12-month period. A lender may make a loan that would comply with the requirement under § 1041.6(c)(3)(i) even if the six-loan limit would prohibit the consumer from taking out one or two subsequent loans in the sequence. 2. Example. Assume that a lender seeks to make a covered short-term loan to a consumer under § 1041.6 with a contractual duration of 14 days. Assume, further, that the lender determines that during the past 30 days the consumer has not had an outstanding covered shortterm loan and that during the 351 days preceding the consummation date of the new loan the consumer had outstanding a total of five covered short-term loans. The new loan would be the sixth covered short-term loan that was outstanding during a consecutive 12-month period. Therefore, the loan would comply with the requirement regarding the aggregate number of covered short-term loans under § 1041.6. Because the consumer has not had an outstanding covered short-term loan in the preceding 30 days, this loan would be the first loan in a new loan sequence. Assume that a week after repaying this first loan the consumer seeks another covered short-term loan under § 1041.6, also with a contractual duration of 14 days. Under § 1041.6(c)(3)(i), this second loan in the loan sequence cannot be made if it would result in the consumer taking out more than six covered short-term loans in the 351 days preceding the proposed consummation date of this loan. 1639 Paragraph 6(c)(3)(ii). 1. Aggregate period of indebtedness. Section 1041.6(c)(3)(ii) provides that a lender cannot make a covered short-term loan under § 1041.6 if the loan would result in the consumer having covered short-term loans outstanding for an aggregate period of more than 90 days in any consecutive 12-month period. In addition to the proposed contractual duration of the new loan, the aggregate period in which all covered short-term loans made to the consumer during the consecutive 12-month period under either § 1041.5 or § 1041.6 were outstanding is counted toward the limit. This requirement applies regardless of whether any or all of the covered shortterm loans are made by the same lender, an affiliate, or an unaffiliated lender. Under § 1041.6(c)(3)(ii), the lender must use the information it has obtained about the consumer’s borrowing history to determine whether the loan would result in the consumer having covered short-term loans outstanding for an aggregate period of more than 90 days during a consecutive 12-month period. A lender may make a loan that would comply with the requirement under § 1041.6(c)(3)(ii) even if the 90-day limit would prohibit the consumer from taking out one or two subsequent loans in the sequence. 2. Example. Assume that Lender A seeks to make a covered short-term loan under § 1041.6 with a contractual duration of 14 days. Assume, further, that Lender A determines that during the past 30 days the consumer did not have an outstanding covered short-term loan and that during the 351 days preceding the consummation date of the new loan the consumer had outstanding three covered short-term loans made by Lender A and a fourth covered short-term loan made by Lender B. Assume that each of the three loans made by Lender A had a contractual duration of 14 days and that the loan made by Lender B had a contractual duration of 30 days, for an aggregate total of 72 days of contractual indebtedness. Assume, further, that the 1640 consumer repaid each loan on its contractual due date. The new loan, if made, would result in the consumer having covered short-term loans outstanding for an aggregate period of 86 days during the consecutive 12-month period. Therefore, the loan would comply with the requirement regarding aggregate time of indebtedness. Because the consumer has not had an outstanding covered short-term loan in the preceding 30 days, this loan would be the first loan in a new loan sequence. Assume that a week after repaying this first loan the consumer seeks another covered short-term loan under § 1041.6, also with a contractual duration of 14 days. Under § 1041.6(c)(3)(ii), this second loan in the loan sequence cannot be made if it would result in the consumer being in debt on covered short-term loans for more than 90 days in the 351 days preceding the proposed consummation date of this loan. 6(d) Restrictions on making certain covered loans and non-covered loans following a covered short-term loan made under the conditional exemption. 1. General. If a lender makes a covered short-term loan under § 1041.6 to a consumer, § 1041.6(d) prohibits the lender or its affiliate from making a covered short-term loan under § 1041.5, a covered longer-term balloon payment loan under § 1041.5, a covered longer-term loan, or a non-covered loan to the consumer while the covered short-term loan made under § 1041.6 is outstanding and for 30 days thereafter. During this period, a lender or its affiliate could make a subsequent covered short-term loan in accordance with the requirements in § 1041.6. 2. Example. Assume that a lender makes both covered short-term loans under § 1041.6 and non-covered installment loans. Assume, further, that the lender makes on April 1 a covered short-term loan under § 1041.6 to a consumer who has not obtained a covered short-term loan under § 1041.6 in the previous 30 days. Assume that the consumer repays this loan on April 15 1641 and that the consumer returns to the lender on April 30 to seek a non-covered installment loan. Because 30 days have not elapsed since the consumer repaid the loan made under § 1041.6, neither the lender nor its affiliate can make a non-covered installment loan to the consumer on April 30. May 16 is the earliest the lender or its affiliate could make a non-covered installment loan to the consumer. The prohibition in § 1041.6(d) applies to covered short-term loans and covered longer-term balloon payment loans made under § 1041.5 but not to covered short-term loans made under § 1041.6. Section 1041.6(d) would, therefore, not prohibit the consumer from obtaining an additional covered short-term loan under § 1041.6 from the same lender or its affiliate on April 30, provided that such loan complies with the principal amount reduction and other requirements of § 1041.6. The prohibition in § 1041.6(d) on making subsequent noncovered loans applies only to a lender and its affiliates. Section 1041.6(d) would, therefore, not prohibit the consumer from obtaining on April 30 a non-covered installment loan from a lender not affiliated with the lender that made the covered short-term loan on April 1. 6(e) Disclosures. 1. General. Section 1041.6(e) sets forth two main disclosure requirements related to a loan made under the requirements in § 1041.6. The first, set forth in § 1041.6(e)(2)(i), is a notice of the restriction on the principal amount on the loan and restrictions on the number of future loans and the principal amounts of such loans, which is required to be provided to a consumer when the consumer seeks the first loan in a sequence of covered short-term loans made under § 1041.6. The second, set forth in § 1041.6(e)(2)(ii), is a notice of the restriction on the principal amount on the loan and the prohibition on another similar loan for at least 30 days after the loan is repaid, which is required to be provided to a consumer when the consumer seeks the third loan in a sequence of covered short-term loans made under § 1041.6. 1642 6(e)(1) General form of disclosures. 6(e)(1)(i) Clear and conspicuous. 1. Clear and conspicuous standard. Disclosures are clear and conspicuous for purposes of § 1041.6(e) if they are readily understandable by the consumer and their location and type size are readily noticeable to the consumer. 6(e)(1)(ii) In writing or electronic delivery. 1. General. Section 1041.6(e)(1)(ii) requires that disclosures required by § 1041.6 be provided to the consumer in writing or through electronic delivery. 2. E-Sign Act requirements. The notices required by § 1041.6(e)(2)(i) and (e)(2)(ii) may be provided to the consumer in electronic form without regard to the Electronic Signatures in Global and National Commerce Act (E-Sign Act) (15 U.S.C. 7001 et seq.). 6(e)(1)(iii) Retainable. 1. General. Electronic disclosures are retainable for purposes of § 1041.6(e) if they are in a format that is capable of being printed, saved, or emailed by the consumer. 6(e)(1)(iv) Segregation requirements for notices. 1. Segregated additional content. Although segregated additional content that is not required by this section may not appear above, below, or around the required content, this additional content may be delivered through a separate form, such as a separate piece of paper or web page. 6(e)(1)(vi) Model Forms. 1. Safe harbor provided by use of model forms. Although the use of the model forms and clauses is not required, lenders using them will be deemed to be in compliance with the 1643 disclosure requirement with respect to such model forms consistent with section 1032(d) of the Dodd-Frank Wall Street Reform and Consumer Protection Act (12 U.S.C. 5481, et seq.) 6(e)(2) Notice requirements. 6(e)(2)(i) First loan notice. 1. As applicable standard. Due to the requirements in § 1041.6(c)(3), a consumer may not be eligible to complete a three-loan sequence of covered short-term loans under § 1041.6 because additional loans within 30 days of the expected pay-off date for the first loan would violate one or more provisions of § 1041.6(c)(3). Such a consumer may be permitted to obtain only one or two loans in a sequence of covered short-term loans under § 1041.6, as applicable. Under these circumstances, § 1041.6(e)(2)(i) would require the lender to modify the notice in § 1041.6(e)(2)(i) to reflect these limitations on subsequent loans. For example, if a consumer can receive only a sequence of two covered short-term loans under § 1041.6 because of the requirements in § 1041.6(c)(3), the lender would have to modify the notice to list the maximum principal amount on loans 1 and 2 and to indicate that loan 3 would not be permitted. 6(e)(3) Timing. 1. General. Section 1041.6(e)(3) requires a lender to provide the notices required in § 1041.6(e)(2)(i) and (ii) to the consumer before the applicable covered short-term loan under § 1041.6 is consummated. For example, a lender can provide the notice after a consumer has completed a loan application but before the consumer has signed the loan agreement. A lender would not have to provide the notices to a consumer who inquires about a covered short-term loan under § 1041.6 but does not fill out an application to obtain this type of loan. 2. Electronic notices. If a lender delivers a notice required by this section electronically in accordance with § 1041.6(e)(1)(ii), § 1041.6(e)(3) requires a lender to provide the electronic 1644 notice to the consumer before a covered short-term loan under § 1041.6 is consummated. Specifically, § 1041.6(e)(3) requires a lender to present the retainable notice to the consumer before the consumer is contractually obligated on the loan. To comply with § 1041.6(e)(3), a lender could, for example, display a screen on a web browser with the notices required in § 1041.6(e)(2)(i) and (ii), provided the screen can be emailed, printed, or saved, before the covered short-term loan under § 1041.6 has been consummated. Section 1041.7— Identification of unfair and abusive practice 1. General. A lender who complies with § 1041.8 with regard to a covered loan has not committed the unfair and abusive practice under § 1041.7. Section 1041.8—Prohibited Payment Transfer Attempts 8(a) Definitions. 8(a)(1) Payment transfer. 1. Lender-initiated. A lender-initiated debit or withdrawal includes a debit or withdrawal initiated by the lender’s agent, such as a payment processor. 2. Any amount due. The following are examples of funds transfers that are for the purpose of collecting any amount due in connection with a covered loan: i. A transfer for the amount of a scheduled payment due under a loan agreement for a covered loan. ii. A transfer for an amount smaller than the amount of a scheduled payment due under a loan agreement for a covered loan. iii. A transfer for the amount of the entire unpaid loan balance collected pursuant to an acceleration clause in a loan agreement for a covered loan. iv. A transfer for the amount of a late fee or other penalty assessed pursuant to a loan 1645 agreement for a covered loan. 3. Amount purportedly due. A transfer for an amount that the consumer disputes or does not legally owe is a payment transfer if it otherwise meets the definition set forth in § 1041.8(a)(1). 4. Transfers of funds not initiated by the lender. A lender does not initiate a payment transfer when: i. A consumer, on her own initiative or in response to a request or demand from the lender, makes a payment to the lender in cash withdrawn by the consumer from the consumer’s account. ii. A consumer makes a payment via an online or mobile bill payment service offered by the consumer’s account-holding institution. iii. The lender seeks repayment of a covered loan pursuant to a valid court order authorizing the lender to garnish a consumer’s account. Paragraph 8(a)(1)(i)(A). 1. Electronic fund transfer. Any electronic fund transfer meeting the general definition in § 1041.8(a)(1) is a payment transfer, including but not limited to an electronic fund transfer initiated by a debit card or a prepaid card. Paragraph 8(a)(1)(i)(B). 1. Signature check. A transfer of funds by signature check meeting the general definition in § 1041.8(a)(1) is a payment transfer regardless of whether the transaction is processed through the check network or through another network, such as the ACH network. The following example illustrates this concept: A lender processes a consumer’s signature check through the check system to collect a scheduled payment due under a loan agreement for a covered loan. 1646 The check is returned for nonsufficient funds. The lender then converts and processes the check through the ACH system, resulting in a successful payment. Both transfers are payment transfers, because both were initiated by the lender for purposes of collecting an amount due in connection with a covered loan. Paragraph 8(a)(1)(i)(E). 1. Transfer by account-holding institution. Under § 1041.8(a)(1)(i)(E), when the lender is the account holder, a transfer of funds by the account-holding institution from a consumer’s account held at the same institution is a payment transfer if it meets the general definition in § 1041.8(a)(1)(i), unless the transfer of funds meets the conditions in § 1041.8(a)(1)(ii) and is therefore excluded from the definition. See § 1041.8(a)(1)(ii) and related commentary. 2. Examples. Payment transfers initiated by an account-holding institution from a consumer’s account include, but are not limited to, the following: i. Initiating an internal transfer from a consumer’s account to collect a scheduled payment on a covered loan. ii. Sweeping the consumer’s account in response to a delinquency on a covered loan. iii. Exercising a right of offset to collect against an outstanding balance on a covered loan. Paragraph 8(a)(1)(ii) Conditional exclusion for certain transfers by account-holding institutions. 1. General. The exclusion in § 1041.8(a)(1)(ii) applies only to a lender that is also the consumer’s account-holding institution. The exclusion applies only if the conditions in both § 1041.8(a)(1)(ii)(A) and (B) are met with respect to a particular transfer of funds. A lender whose transfer meets the exclusion has not committed the unfair and abusive practice under § 1647 1041.7 and is not subject to §§ 1041.8 or 1041.9 in connection with that transaction, but is subject to subpart C for any transfers that do not meet the exclusion in § 1041.8(a)(1)(ii) and are therefore payment transfers under § 1041.8(a)(1)(i). Paragraph 8(a)(1)(ii)(A). 1. Terms of loan agreement or account agreement. The condition in § 1041.8(a)(1)(ii)(A) is met only if the terms of the loan agreement or account agreement setting forth the restrictions on charging fees are in effect at the time the covered loan is made and remain in effect for the duration of the loan. 2. Fees prohibited. Examples of the types of fees restricted under § 1041.8(a)(1)(ii)(A) include, but are not limited to, nonsufficient fund fees, overdraft fees, and returned-item fees. A lender seeking to initiate transfers of funds pursuant to the exclusion in § 1041.8(a)(1)(ii) may still charge the consumer a late fee for failure to make a timely payment, as permitted under the terms of the loan agreement and other applicable law, notwithstanding that the lender has initiated a transfer of funds meeting the description in § 1041.8(a)(1)(ii)(A) in an attempt to collect the payment. Paragraph 8(a)(1)(ii)(B). 1. General. Under § 1041.8(a)(1)(ii)(B), to be eligible for the exclusion in § 1041.8(a)(1)(ii), a lender may not close the consumer’s account in response to a negative balance that results from a lender-initiated transfer of funds in connection with the covered loan. A lender is not restricted from closing the consumer’s account in response to another event, even if the event occurs after a lender-initiated transfer of funds has brought the account to a negative balance. For example, a lender may close the account at the consumer’s request, for purposes of 1648 complying with other regulatory requirements, or to protect the account from suspected fraudulent use or unauthorized access, and still meet the condition in § 1041.8(a)(1)(ii)(B). 2. Terms of loan agreement or account agreement. The condition in § 1041.8(a)(1)(ii)(B) is met only if the terms of the loan agreement or account agreement providing that the lender will not close the account in the specified circumstances are in effect at the time the covered loan is made and remain in effect for the duration of the loan. 8(a)(2) Single immediate payment transfer at the consumer’s request. Paragraph 8(a)(2)(i). 1. Time of initiation. A one-time electronic fund transfer is initiated at the time that the transfer is sent out of the lender’s control. Thus, the electronic fund transfer is initiated at the time that the lender or its agent sends the transfer to be processed by a third party, such as the lender’s bank. The following example illustrates this concept: A lender obtains a consumer’s authorization for a one-time electronic fund transfer at 2 p.m. and sends the payment entry to its agent, a payment processor, at 5 p.m. on the same day. The agent then sends the payment entry to the lender’s bank for further processing the next business day at 8 a.m. The timing condition in § 1041.8(a)(2)(ii) is satisfied, because the lender’s agent sent the transfer out of its control within one business day after the lender obtained the consumer’s authorization. Paragraph 8(a)(2)(ii). 1. Time of processing. A signature check is processed at the time that the check is sent out of the lender’s control. Thus, the check is processed at the time that the lender or its agent sends the check to be processed by a third party, such as the lender’s bank. For an example illustrating this concept within the context of initiating a one-time electronic fund transfer, see comment 8(a)(2)(i)-1. 1649 2. Check provided by mail. For purposes of § 1041.8(a)(2)(ii), if the consumer provides the check by mail, the check is deemed to be provided on the date that the lender receives it. 8(b) Prohibition on initiating payment transfers from a consumer’s account after two consecutive failed payment transfers. 1. General. When the prohibition in § 1041.8(b) applies, a lender is generally restricted from initiating any further payment transfers from the consumer’s account in connection with any covered loan that the consumer has with the lender at the time the prohibition is triggered, unless the requirements and conditions in either § 1041.8(c) or (d) are satisfied for each such covered loan for which the lender seeks to initiate further payment transfers. The prohibition applies, for example, to payment transfers that might otherwise be initiated to collect payments that later fall due under a loan agreement for a covered loan and to transfers to collect late fees or returned item fees as permitted under the terms of such a loan agreement. In addition, the prohibition applies regardless of whether the lender holds an otherwise valid authorization or instrument from the consumer, including but not limited to an authorization to collect payments by preauthorized electronic fund transfers or a post-dated check. See § 1041.8(c) and (d) and accompanying commentary for guidance on the requirements and conditions that a lender must satisfy to initiate a payment transfer from a consumer’s account after the prohibition applies. 2. Account. The prohibition in § 1041.8(b) applies only to the account from which the lender attempted to initiate the two consecutive failed payment transfers. 3. More than one covered loan. The prohibition in § 1041.8(b) is triggered after the lender has attempted to initiate two consecutive failed payment transfers in connection with any covered loan or covered loans that the consumer has with the lender. Thus, when a consumer has more than one covered loan with the lender, the two consecutive failed payment transfers 1650 need not be initiated in connection with the same loan in order for the prohibition to be triggered, but rather can be initiated in connection with two different loans. For example, the prohibition is triggered if the lender initiates the first failed payment transfer to collect payment on one covered loan and the second consecutive failed payment transfer to collect payment on a different covered loan, assuming that the conditions for a first failed payment transfer, in § 1041.8(b)(2)(i), and second consecutive failed transfer, in § 1041.8(b)(2)(ii), are met. 4. Application to bona fide subsequent loan. If a lender triggers the prohibition in § 1041.8(b), the lender is not prohibited under § 1041.8(b) from initiating a payment transfer in connection with a bona fide subsequent covered loan that was originated after the prohibition was triggered, provided that the lender has not attempted to initiate two consecutive failed payment transfers from the consumer’s account in connection with the bona fide subsequent covered loan. For purposes of § 1041.8(b) only, a bona fide subsequent covered loan does not include a covered loan that refinances or rolls over any covered loan that the consumer has with the lender at the time the prohibition is triggered. 8(b)(1) General. 1. Failed payment transfer. A payment transfer results in a return indicating that the consumer’s account lacks sufficient funds when it is returned unpaid, or is declined, due to nonsufficient funds in the consumer’s account. 2. Date received. The prohibition in § 1041.8(b) applies as of the date on which the lender or its agent, such as a payment processor, receives the return of the second consecutive failed transfer or, if the lender is the consumer’s account-holding institution, the date on which the second consecutive failed payment transfer is initiated. 3. Return for other reason. A transfer that results in a return for a reason other than a 1651 lack of sufficient funds, such as a return made due to an incorrectly entered account number, is not a failed transfer for purposes of § 1041.8(b). 4. Failed payment transfer initiated by a lender that is the consumer’s account-holding institution. When a lender that is the consumer’s account-holding institution initiates a payment transfer for an amount that the account lacks sufficient funds to cover, the payment transfer is a failed payment transfer for purposes of the prohibition in § 1041.8(b), regardless of whether the result is classified or coded in the lender’s internal procedures, processes, or systems as a return for nonsufficient funds or, if applicable, regardless of whether the full amount of the payment transfer is paid out of overdraft. Such a lender does not initiate a failed payment transfer for purposes of the prohibition if the lender merely defers or foregoes debiting or withdrawing payment from an account based on the lender’s observation that the account lacks sufficient funds. 8(b)(2) Consecutive failed payment transfers. 8(b)(2)(i) First failed payment transfer. 1. Examples. The following examples illustrate concepts of first failed payment transfers under § 1041.8(b)(2)(i). All of the examples assume that the consumer has only one covered loan with the lender: i. A lender, having made no other attempts, initiates an electronic fund transfer to collect the first scheduled payment due under a loan agreement for a covered loan, which results in a return for nonsufficient funds. The failed transfer is the first failed payment transfer. The lender, having made no attempts in the interim, re-presents the electronic fund transfer and the re-presentment results in the collection of the full payment. Because the subsequent attempt did not result in a return for nonsufficient funds, the number of consecutive failed payment transfers 1652 resets to zero. The following month, the lender initiates an electronic fund transfer to collect the second scheduled payment due under the covered loan agreement, which results in a return for nonsufficient funds. That failed transfer is a first failed payment transfer. ii. A storefront lender, having made no prior attempts, processes a consumer’s signature check through the check system to collect the first scheduled payment due under a loan agreement for a covered loan. The check is returned for nonsufficient funds. This constitutes the first failed payment transfer. The lender does not thereafter convert and process the check through the ACH system, or initiate any other type of payment transfer, but instead contacts the consumer. At the lender’s request, the consumer comes into the store and makes the full payment in cash withdrawn from the consumer’s account. The number of consecutive failed payment transfers remains at one, because the consumer’s cash payment was not a payment transfer as defined in § 1041.8(a)(2). 8(b)(2)(ii) Second consecutive failed payment transfer. 1. General. Under § 1041.8(b)(2)(ii), a failed payment transfer is the second consecutive failed transfer if the previous payment transfer was a first failed payment transfer. The following examples illustrate this concept: i. Assume that a consumer has only one covered loan with a lender. The lender, having initiated no other payment transfer in connection with the covered loan, initiates an electronic fund transfer to collect the first scheduled payment due under the loan agreement. The transfer is returned for nonsufficient funds. The returned transfer is the first failed payment transfer. The lender next initiates an electronic fund transfer for the following scheduled payment due under the loan agreement for the covered loan, which is also returned for nonsufficient funds. The second returned transfer is the second consecutive failed payment transfer. 1653 ii. Assume that a consumer has two covered loans, Loan A and Loan B, with a lender. Further assume that the lender has initiated no failed payment transfers in connection with either covered loan. On the first of the month, the lender initiates an electronic fund transfer to collect a regularly scheduled payment on Loan A, resulting in a return for nonsufficient funds. The returned transfer is the first failed payment transfer. Two weeks later, the lender, having initiated no further payment transfers in connection with either covered loan, initiates an electronic fund transfer to collect a regularly scheduled payment on Loan B, also resulting in a return for nonsufficient funds. The second returned transfer is the second consecutive failed payment transfer, and the lender is thus prohibited under § 1041.8(b) from initiating further payment transfers in connection with either covered loan. 2. Previous payment transfer. Section 1041.8(b)(2)(ii) provides that a previous payment transfer includes a payment transfer initiated at the same time or on the same day as the first failed payment transfer. The following example illustrates how this concept applies in determining whether the prohibition in § 1041.8(b) is triggered: Assume that a consumer has only one covered loan with a lender. The lender has made no other payment transfers in connection with the covered loan. On Monday at 9 a.m., the lender initiates two electronic fund transfers to collect the first scheduled payment under the loan agreement, each for half of the total amount due. Both transfers are returned for nonsufficient funds. Because each transfer is one of two failed transfers initiated at the same time, the lender has initiated a second consecutive failed payment transfer under § 1041.8(b)(2)(ii), and the prohibition in § 1041.8(b) is therefore triggered. 3. Application to exception in § 1041.8(d). When, after a second consecutive failed payment transfer, a lender initiates a single immediate payment transfer at the consumer’s 1654 request pursuant to the exception in § 1041.8(d), the failed transfer count remains at two, regardless of whether the transfer succeeds or fails. Further, the exception is limited to a single payment transfer. Accordingly, if a payment transfer initiated pursuant to the exception fails, the lender is not permitted to re-initiate the transfer, such as by re-presenting it through the ACH system, unless the lender obtains a new authorization under § 1041.8(c) or (d). 8(b)(2)(iii) Different payment channel. 1. General. Section 8(b)(2)(iii) provides that if a failed payment transfer meets the descriptions set forth in § 1041.8(b)(2)(ii), it is the second consecutive failed transfer regardless of whether the first failed transfer was made through a different payment channel. The following example illustrates this concept: A lender initiates an electronic funds transfer through the ACH system for the purpose of collecting the first payment due under a loan agreement for a covered loan. The transfer results in a return for nonsufficient funds. This constitutes the first failed payment transfer. The lender next processes a remotely created check through the check system for the purpose of collecting the same first payment due. The remotely created check is returned for nonsufficient funds. The second failed attempt is the second consecutive failed attempt because it meets the description set forth in § 1041.8(b)(2)(ii). 8(c) Exception for additional payment transfers authorized by the consumer. 1. General. Section 1041.8(c) sets forth one of two exceptions to the prohibition in § 1041.8(b). Under the exception in § 1041.8(c), a lender is permitted to initiate additional payment transfers from a consumer’s account after the lender’s second consecutive transfer has failed if the additional transfers are authorized by the consumer in accordance with certain requirements and conditions as specified in the rule. In addition to the exception under § 1041.8(c), a lender is permitted to execute a single immediate payment transfers at the 1655 consumer’s request under § 1041.8(d), if certain requirements and conditions are satisfied. 8(c)(1) General. 1. Consumer’s underlying payment authorization or instrument still required. The consumer’s authorization required by § 1041.8(c) is in addition to, and not in lieu of, any separate payment authorization or instrument required to be obtained from the consumer under applicable laws. 8(c)(2) General authorization requirements and conditions. 8(c)(2)(i) Required transfer terms. 1. General. Section 1041.8(c)(2)(i) sets forth the general requirement that, for purposes of the exception in § 1041.8(c), the specific date, amount, and payment channel of each additional payment transfer must be authorized by the consumer, subject to a limited exception in § 1041.8(c)(2)(iii) for payment transfers solely to collect a late fee or returned item fee. Accordingly, for the exception to apply to an additional payment transfer, the transfer’s specific date, amount, and payment channel must be included in the signed authorization obtained from the consumer under § 1041.8(c)(3)(iii). For guidance on the requirements and conditions that apply when obtaining the consumer’s signed authorization, see § 1041.8(c)(3)(iii) and accompanying commentary. 2. Specific date. The requirement that the specific date of each additional payment transfer be authorized by the consumer is satisfied if the consumer authorizes the month, day, and year of each transfer. 3. Amount larger than specific amount. The exception in § 1041.8(c)(2) does not apply if the lender initiates a payment transfer for an amount larger than the specific amount authorized by the consumer. Accordingly, such a transfer would violate the prohibition on additional 1656 payment transfers under § 1041.8(b). 4. Smaller amount. A payment transfer initiated pursuant to § 1041.8(c) is initiated for the specific amount authorized by the consumer if its amount is equal to or smaller than the authorized amount. 8(c)(2)(iii) Special authorization requirements and conditions for payment transfers to collect a late fee or returned item fee. Paragraph 8(c)(2)(iii). 1. General. If a lender obtains the consumer’s authorization to initiate a payment transfer solely to collect a late fee or returned item fee in accordance with the requirements and conditions under § 1041.8(c)(2)(iii), the general requirement in § 1041.8(c)(2) that the consumer authorize the specific date and amount of each additional payment transfer need not be satisfied. 2. Highest amount. The requirement that the consumer’s signed authorization include a statement that specifies the highest amount that may be charged for a late fee or returned item fee is satisfied, for example, if the statement specifies the maximum amount permitted under the loan agreement for a covered loan. 3. Varying fee amounts. If a fee amount may vary due to the remaining loan balance or other factors, the rule requires the lender to assume the factors that result in the highest amount possible in calculating the specified amount. 8(c)(3) Requirements and conditions for obtaining the consumer’s authorization. 8(c)(3)(ii) Provision of payment transfer terms to the consumer. 1. General. A lender is permitted under § 1041.8(c)(3)(ii) to request a consumer’s authorization on or after the day that the lender provides the consumer rights notice required by § 1041.9(c). For the exception in § 1041.8(c)(2) to apply, however, the consumer’s signed 1657 authorization must be obtained no earlier than the date on which the consumer is considered to have received the consumer rights notice, as specified in § 1041.8(c)(3)(iii). 2. Different options. Nothing in § 1041.8(c)(3)(ii) prohibits a lender from providing different options for the consumer to consider with respect to the date, amount, or payment channel of each additional payment transfer for which the lender is requesting authorization. In addition, if a consumer declines a request, nothing in § 1041.8(c)(3)(ii) prohibits a lender from making a follow-up request by providing a different set of terms for the consumer to consider. For example, if the consumer declines an initial request to authorize two recurring payment transfers for a particular amount, the lender may make a follow-up request for the consumer to authorize three recurring payment transfers for a smaller amount. Paragraph 8(c)(3)(ii)(A). 1. Request by email. Under § 1041.8(c)(3)(ii)(A), a lender is permitted to provide the required terms and statements to the consumer in writing or in a retainable form by email if the consumer has consented to receive electronic disclosures in that manner under § 1041.9(a)(4) or agrees to receive the terms and statements by email in the course of a communication initiated by the consumer in response to the consumer rights notice required by § 1041.9(c). The following example illustrates a situation in which the consumer agrees to receive the required terms and statements by email after affirmatively responding to the notice: i. After a lender provides the consumer rights notice in § 1041.9(c) by mail to a consumer who has not consented to receive electronic disclosures under § 1041.9(a)(4), the consumer calls the lender to discuss her options for repaying the loan, including the option of authorizing additional payment transfers pursuant to § 1041.8(c). In the course of the call, the consumer asks the lender to provide the request for the consumer’s authorization via email. Because the 1658 consumer has agreed to receive the request via email in the course of a communication initiated by the consumer in response to the consumer rights notice, the lender is permitted under § 1041.8(c)(3)(ii)(A) to provide the request to the consumer by that method. 2. E-Sign Act does not apply to provision of terms and statements. The required terms and statements may be provided to the consumer electronically in accordance with the requirements for requesting the consumer’s authorization in § 1041.8(c)(3) without regard to the E-Sign Act. However, under § 1041.8(c)(3)(iii)(A), an authorization obtained electronically is valid only if it is signed or otherwise agreed to by the consumer in accordance with the signature requirements in the E-Sign Act. See § 1041.8(c)(3)(iii)(A) and comment 8(c)(3)(iii)(A)-1. 3. Same communication. Nothing in § 1041.8(c)(3)(ii) prohibits a lender from requesting the consumer’s authorization for additional payment transfers and providing the consumer rights notice in the same communication, such as a single written mailing or a single email to the consumer. Nonetheless, the consumer rights notice may be provided to the consumer only in accordance with the requirements and conditions in § 1041.9, including but not limited to the segregation requirements that apply to the notice. Thus, for example, if a lender mails the request for authorization and the notice to the consumer in the same envelope, the lender must provide the notice on a separate piece of paper, as required under § 1041.9. Similarly, a lender could provide the notice to a consumer in the body of an e-mail and attach a document containing the request for authorization. In such cases, it would be permissible for the lender to add language after the text of the notice explaining that the other document is a request for a new authorization. 1659 Paragraph 8(c)(3)(ii)(B). 1. Request by oral telephone communication. Nothing in § 1041.8(c)(3)(ii) prohibits a lender from contacting the consumer by telephone to discuss repayment options, including the option of authorizing additional payment transfers. However, under § 1041.8(c)(3)(ii)(B), a lender is permitted to provide the required terms and statements to the consumer by oral telephone communication for purposes of requesting authorization only if the consumer affirmatively contacts the lender in that manner in response to the consumer rights notice required by § 1041.9(c) and agrees to receive the terms and statements by that method of delivery in the course of, and as part of, the same communication. 8(c)(3)(iii) Signed authorization required. 8(c)(3)(iii)(A) General. 1. E-Sign Act signature requirements. For authorizations obtained electronically, the requirement that the authorization be signed or otherwise agreed to by the consumer is satisfied if the E-Sign Act requirements for electronic records and signatures are met. Thus, for example, the requirement is satisfied by an email from the consumer or by a code entered by the consumer into the consumer’s telephone keypad, assuming that in each case the signature requirements in the E-Sign Act are complied with. 2. Consumer’s affirmative response to the notice. A consumer affirmatively responds to the consumer rights notice that was provided by mail when, for example, the consumer calls the lender on the telephone to discuss repayment options after receiving the notice. 8(c)(3)(iii)(C) Memorialization required. 1. Timing. The memorialization is deemed to be provided to the consumer on the date it is mailed or transmitted. 1660 2. Form of memorialization. The requirement that the memorialization be provided in a retainable form is not satisfied by a copy of a recorded telephone call, notwithstanding that the authorization was obtained in that manner. 3. Electronic delivery. A lender is permitted under § 1041.8(c)(3)(iii)(C) to provide the memorialization to the consumer by email in accordance with the requirements and conditions for requesting authorization in § 1041.8(c)(3)(ii)(A), regardless of whether the lender requested the consumer’s authorization in that manner. For example, if the lender requested the consumer’s authorization by telephone but also has obtained the consumer’s consent to receive electronic disclosures by email under § 1041.9(a)(4), the lender may provide the memorialization to the consumer by email, as specified in § 1041.8(c)(3)(ii)(A). 8(d) Exception for initiating a single immediate payment transfer at the consumer’s request. 1. General. For guidance on the requirements and conditions that must be satisfied for a payment transfer to meet the definition of a single immediate payment transfer at the consumer’s request, see § 1041.8(a)(2) and accompanying commentary. 2. Application of prohibition. A lender is permitted under the exception in § 1041.8(d) to initiate a single payment transfer requested by the consumer only once and thus is prohibited under § 1041.8(b) from re-initiating the payment transfer if it fails, unless the lender subsequently obtains the consumer’s authorization to re-initiate the payment transfer under § 1041.8(c) or (d). However, a lender is permitted to initiate any number of payment transfers from a consumer’s account pursuant to the exception in § 1041.8(d), provided that the requirements and conditions are satisfied for each such transfer. See comment 8(b)(2)(ii)-3 for further guidance on how the prohibition in § 1041.8(b) applies to the exception in § 1041.8(d). 1661 3. Timing. A consumer affirmatively contacts the lender when, for example, the consumer calls the lender after noticing on her bank statement that the lender’s last two payment withdrawal attempts have been returned for nonsufficient funds. 8(e) Prohibition against evasion. 1. General. Section 1041.8(e) provides that a lender must not take any action with the intent of evading the requirements of § 1041.8. In determining whether a lender has taken action with the intent of evading the requirements of § 1041.8, the form, characterization, label, structure, or written documentation of the lender’s action shall not be dispositive. Rather, the actual substance of the lender’s action as well as other relevant facts and circumstances will determine whether the lender’s action was taken with the intent of evading the requirements of § 1041.8. If the lender’s action is taken solely for legitimate business purposes, it is not taken with the intent of evading the requirements of § 1041.8. By contrast, if a consideration of all relevant facts and circumstances reveals a purpose that is not a legitimate business purpose, the lender’s action may have been taken with the intent of evading the requirements of § 1041.8. A lender action that is taken with the intent of evading the requirements of part 1041 may be knowing or reckless. Fraud, deceit, or other unlawful or illegitimate activity may be one fact or circumstance that is relevant to the determination of whether a lender’s action was taken with the intent of evading the requirements of § 1041.8, but fraud, deceit, or other unlawful or illegitimate activity is not a prerequisite to such a finding. 2. Illustrative example. A lender collects payment on its covered loans primarily through recurring electronic fund transfers authorized by consumers at consummation. As a matter of lender policy and practice, after a first attempt to initiate an ACH payment transfer from a consumer’s account for the full payment amount is returned for nonsufficient funds, the lender 1662 initiates a second payment transfer from the account on the following day for $1.00. If the second payment transfer succeeds, the lender immediately splits the amount of the full payment into two separate payment transfers and initiates both payment transfers from the account at the same time, resulting in two returns for nonsufficient funds in the vast majority of cases. The lender developed the policy and began the practice shortly prior to the effective date of the rule that is codified in 12 CFR part 1041. The lender’s prior policy and practice when re-presenting the first failed payment transfer was to re-present for the payment’s full amount. Depending on the relevant facts and circumstances, the lender’s actions may have been taken with the intent of evading the requirements of § 1041.8. Specifically, by initiating a second payment transfer for $1.00 from the consumer’s account the day after a first transfer for the full payment amount fails and, if that payment transfer succeeds, initiating two simultaneous payment transfers from the account for the split amount of the full payment, resulting in two returns for nonsufficient funds in the vast majority of cases, the lender avoided the prohibition in § 1041.8(b) on initiating payment transfers from a consumer’s account after two consecutive payment transfers have failed. Section 1041.9—Disclosure of Payment Transfer Attempts 1. General. Section 1041.9 sets forth two main disclosure requirements related to collecting payments from a consumer’s account in connection with a covered loan. The first, set forth in § 1041.9(b), is a payment notice required to be provided to a consumer in advance of a initiating the first payment withdrawal or an unusual withdrawal from the consumer’s account, subject to certain exceptions. The second, set forth in § 1041.9(c), is a consumer rights notice required to be provided to a consumer after a lender receives notice of a second consecutive failed payment transfer from the consumer’s account, as described in § 1041.8(b). In addition, § 1663 1041.9 requires lenders to provide an electronic short notice in two situations when they are providing the disclosures required by this section through certain forms of electronic delivery. The first, set forth in § 1041.9(b)(4), is an electronic short notice that must be provided along with the payment notice. This provision allows an exception for when the method of electronic delivery is email; for that method, the lender may use the electronic short notice under § 1041.9(b)(4)(ii) or may provide the full notice within the body of the email. The second, set forth in § 1041.9(c)(4), is an electronic short notice that must be provided along with the consumer rights notice. As with the payment notices, this consumer rights notice provision also allows an exception for when the method of electronic delivery is email; for that method, the lender may use the electronic short notice under § 1041.9(c)(4)(ii) or may provide the full notice within the body of the email. 9(a) General form of disclosures. 9(a)(1) Clear and conspicuous. 1. Clear and conspicuous standard. Disclosures are clear and conspicuous for purposes of § 1041.9 if they are readily understandable and their location and type size are readily noticeable to consumers. 9(a)(2) In writing or electronic delivery. 1. Electronic delivery. Section 1041.9(a)(2) allows the disclosures required by § 1041.9 to be provided through electronic delivery as long as the requirements of § 1041.9(a)(4) are satisfied, without regard to the Electronic Signatures in Global and National Commerce Act (ESign Act) (15 U.S.C. 7001 et seq.). 9(a)(3) Retainable. 1. General. Electronic disclosures, to the extent permitted by § 1041.9(a)(4), are 1664 retainable for purposes of § 1041.9 if they are in a format that is capable of being printed, saved, or emailed by the consumer. The general requirement to provide disclosures in a retainable form does not apply when the electronic short notices are provided in via mobile application or text message. For example, the requirement does not apply to an electronic short notice that is provided to the consumer’s mobile telephone as a text message. In contrast, if the access is provided to the consumer via email, the notice must be in a retainable form, regardless of whether the consumer uses a mobile telephone to access the notice. 9(a)(4) Electronic delivery. 1. General. Section 1041.9(a)(4) permits disclosures required by § 1041.9 to be provided through electronic delivery if the consumer consent requirements under § 1041.9(a)(4) are satisfied. 9(a)(4)(i) Consumer consent. 9(a)(4)(i)(A) General. 1. General. Section 1041.9(a)(4)(i) permits disclosures required by § 1041.9 to be provided through electronic delivery if the lender obtains the consumer’s affirmative consent to receive the disclosures through a particular electronic delivery method. This affirmative consent requires lenders to provide consumers with an option to select a particular electronic delivery method. The consent must clearly show the method of electronic delivery that will be used, such as email, text message, or mobile application. Consent provided by checking a box during the origination process may qualify as being in writing. Consent can be obtained for multiple methods of electronic delivery, but the consumer must have affirmatively selected and provided consent for each method. 1665 9(a)(4)(i)(B) Email option required. 1. General. Section § 1041.9(a)(4)(i)(B) provides that when obtaining consumer consent to electronic delivery under § 1041.9(a)(4), a lender must provide the consumer with an option to receive the disclosures through email. The lender may choose to offer email as the only method of electronic delivery under § 1041.9(a)(4). 9(a)(4)(ii) Subsequent loss of consent. 1. General. The prohibition on electronic delivery of disclosures in § 1041.9(a)(4)(ii) applies to the particular electronic method for which consent is lost. When a lender loses a consumer’s consent to receive disclosures via text message, for example, but has not lost the consumer’s consent to receive disclosures via email, the lender may continue to provide disclosures via email, assuming that all of the requirements in § 1041.9(a)(4) are satisfied. 2. Loss of consent applies to all notices. The loss of consent applies to all notices required by § 1041.9. For example, if a consumer revokes consent in response to the electronic short notice text message delivered along with the payment notice under § 1041.9(b)(4)(ii), that revocation also applies to text delivery of the electronic short notice that would be delivered with the consumer rights notice under § 1041.9(c)(4)(ii). Paragraph 9(a)(4)(ii)(A). 1. Revocation. For purposes of § 1041.9(a)(4)(ii)(A), a consumer may revoke consent for any reason and by any reasonable means of communication. Reasonable means of communication may include calling the lender and revoking consent orally, mailing a revocation to an address provided by the lender on its consumer correspondence, sending an email response or clicking on a revocation link provided in an email from the lender, and responding by text message to a text message sent by the lender. 1666 Paragraph 9(a)(4)(ii)(B). 1. Notice. A lender receives notification for purposes of § 1041.9(a)(4)(ii)(B) when the lender receives any information indicating that the consumer did not receive or is unable to receive disclosures in a particular electronic manner. Examples of notice include but are not limited to the following: i. An email returned with a notification that the consumer’s account is no longer active or does not exist. ii. A text message returned with a notification that the consumer’s mobile telephone number is no longer in service. iii. A statement from the consumer that the consumer is unable to access or review disclosures through a particular electronic delivery method. 9(a)(5) Segregation requirements for notices. 1. Segregated additional content. Although segregated additional content that is not required by § 1041.9 may not appear above, below, or around the required content, additional content may be delivered through a separate form, such as a separate piece of paper or web page. 9(a)(7) Model Forms. 1. Safe harbor provided by use of model forms. Although the use of the model forms and clauses is not required, lenders using them will be deemed to be in compliance with the disclosure requirement with respect to such model forms. 9(b) Payment notice. 9(b)(1)(i) First payment withdrawal. 1. First payment withdrawal. Depending on when the payment authorization granted by the consumer is obtained on a covered loan and whether the exception for a single immediate 1667 payment transfer made at the consumer’s request applies, the first payment withdrawal may or may not be the first payment made on a covered loan. When a lender obtains payment authorization during the origination process, the lender may provide the first payment withdrawal notice at that time. A lender that obtains payment authorization after a payment has been made by the consumer in cash, or after initiating a single immediate payment transfer at the consumer’s request, would deliver the notice later in the loan term. If a consumer provides one payment authorization that the lender uses to initiate a first payment withdrawal after a notice as required by § 1041.9(b)(1)(i), but the consumer later changes the authorization or provides an additional authorization, the lender’s exercise of that new authorization would not be the first payment withdrawal; however, it may be an unusual withdrawal under 9(b)(1)(ii). 2. First payment withdrawal is determined when the loan is in covered status. As discussed in comments 3(b)(3)-1 and -3, there may be situations where a longer-term loan is not covered at the time of origination but becomes covered at a later date. The lender’s first attempt to execute a payment transfer after a loan becomes a covered loan under part 1041 is the first payment withdrawal. For example, consider a loan that is not considered covered at the time of origination. If the lender initiates a payment withdrawal during the first and second billing cycles and the loan becomes covered at the end of the second cycle, any lender initiated payment during the third billing cycle is considered a first payment withdrawal under this section. 3. Intervening payments. Unscheduled intervening payments do not change the determination of first payment withdrawal for purposes of the notice requirement. For example, a lender originates a loan on April 1, with a payment scheduled to be withdrawn on May 1. At origination, the lender provides the consumer with a first payment withdrawal notice for May 1. On April 28, the consumer makes the payment due on May 1 in cash. The lender does not 1668 initiate a withdrawal on May 1. The lender initiates a withdrawal for the next scheduled payment June 1. The lender satisfied its notice obligation with the notice provided at origination, so it is not required to send a first payment notice in connection with the June 1 payment although it may have to send an unusual payment notice if the transfer meets one of the conditions in § 1041.9(b)(3)(ii)(C). 9(b)(1)(iii) Exceptions. 9(b)(1)(iii)(A) General. 1. Exception for initial payment transfer applies even if the transfer is unusual. The exception in § 1041.9(b)(1)(iii)(A) applies even if the situation would otherwise trigger the additional disclosure requirements for unusual attempts under § 1041.9(b)(3). For example, if the payment channel of the initial payment transfer after obtaining the consumer’s consent is different than the payment channel used before the prohibition under § 1041.8 was triggered, the exception in § 1041.9(b)(1)(iii)(A) applies. 2. Multiple transfers in advance. If a consumer has affirmatively consented to multiple transfers in advance, the exception in § 1041.9(b)(1)(iii)(A) applies only to the first initial payment transfer of that series. 9(b)(2) First payment withdrawal notice 9(b)(2)(i) Timing. 1. When the lender obtains payment authorization. For all methods of delivery, the earliest point that the lender may provide the first payment withdrawal notice is when the lender obtains the payment authorization. For example, the notice can be provided simultaneously when the lender provides a consumer with a copy of a completed payment authorization, or after providing the authorization copy. The provision allows the lender to provide consumers with the 1669 notice at a convenient time because the lender and consumer are already communicating about the loan, but also allows flexibility for lenders that prefer to provide the notice closer to the payment transfer date. For example, the lender could obtain consumer consent to electronic delivery and deliver the notice through email 4 days before initiating the transfer, or the lender could hand deliver it to the consumer at the end of the loan origination process. 9(b)(2)(i)(A) Mail. 1. General. The six business-day period begins when the lender places the notice in the mail, not when the consumer receives the notice. For example, if a lender places the notice in the mail on Monday, June 1, the lender may initiate the transfer of funds on Tuesday, June 9, if it is the 6th business day following mailing of the notice. 9(b)(2)(i)(B) Electronic delivery. Paragraph 9(b)(2)(i)(B)(1). 1. General. The three-business-day period begins when the lender sends the notice, not when the consumer receives or is deemed to have received the notice. For example, if a lender sends the notice by email on Monday, June 1, the lender may initiate the transfer of funds on Thursday, June 4, the third business day following transmitting the notice. Paragraph 9(b)(2)(i)(B)(2). 1. General. In some circumstances, a lender may lose a consumer’s consent to receive disclosures through a particular electronic delivery method after the lender has provided the notice. In such circumstances, the lender may initiate the transfer for the payment currently due as scheduled. If the lender is scheduled to make a future unusual withdrawal attempt following the one that was disclosed in the previously provided first withdrawal notice, the lender must 1670 provide notice for that unusual withdrawal through alternate means, in accordance with the applicable timing requirements in § 1041.9(b)(3)(i). 2. Alternate Means. The alternate means may include a different electronic delivery method that the consumer has consented to, in person, or by mail, in accordance with the applicable timing requirements in § 1041.9(b)(3)(i). 9(b)(2)(ii) Content requirements. 9(b)(2)(ii)(B) Transfer terms. Paragraph 9(b)(2)(ii)(B)(1) Date. 1. Date. The initiation date is the date that the payment transfer is sent outside of the lender’s control. Accordingly, the initiation date of the transfer is the date that the lender or its agent sends the payment to be processed by a third party. For example, if a lender sends its ACH payments to a payment processor working on the lender’s behalf on Monday, June 1, but the processor does not submit them to its bank and the ACH network until Tuesday, June 2, the date of the payment transfer is Tuesday the 2nd. Paragraph 9(b)(2)(ii)(B)(2) Amount. 1. Amount. The amount of the transfer is the total amount of money that will be transferred from the consumer’s account, regardless of whether the total corresponds to the amount of a regularly scheduled payment. For example, if a single transfer will be initiated for the purpose of collecting a regularly scheduled payment of $50.00 and a late fee of $30.00, the amount that must be disclosed under § 1041.9(b)(2)(ii)(B)(2) is $80.00. Paragraph 9(b)(2)(ii)(B)(5) Payment channel. 1. General. Payment channel refers to the specific payment method, including the network that the transfer will travel through and the form of the transfer. For example, a lender 1671 that uses the consumer’s paper check information to initiate a payment transfer through the ACH network would use the ACH payment channel under § 1041.9(b)(2)(ii)(B)(5). A lender that uses consumer account and routing information to initiate a remotely created check over the check network would use the remotely created check payment channel. A lender that uses a post-dated signature check to initiate a transfer over the check network would use the signature check payment channel. A lender that initiates a payment from a consumer’s prepaid card would specify whether that payment is processed as an ACH transfer, a PIN debit card network payment, or a signature debit card network payment. 2. Illustrative examples. In describing the payment channel in the disclosure, the most common payment channel descriptions include, but are not limited to, ACH transfers, checks, remotely created checks, remotely created payment orders, internal transfers, PIN debit card payments, and signature debit card network payments. 9(b)(2)(ii)(C) Payment breakdown. 9(b)(2)(ii)(C)(2) Principal. 1. General. The amount of the payment that is applied to principal must always be included in the payment breakdown table, even if the amount applied is $0. 9(b)(2)(ii)(C)(4) Fees. 1. General. This field must only be provided if some of the payment amount will be applied to fees. In situations where more than one fee applies, fees may be disclosed separately or aggregated. A lender may use its own term to describe the fee, such as “late payment fee.” 9(b)(2)(ii)(C)(5) Other charges. 1. General. This field must only be provided if some of the payment amount will be applied to other charges. In situations when more than one other charge applies, other charges 1672 may be disclosed separately or aggregated. A lender may use its own term to describe the charge, such as “insurance charge.” 9(b)(3) Unusual withdrawal notice. 9(b)(3)(i) Timing. 1. General. See comments on 9(b)(2) regarding the first payment withdrawal notice. 9(b)(3)(ii) Content requirements. 1. General. If the payment transfer is unusual according to the circumstances described in § 1041.9(b)(3)(ii)(C), the payment notice must contain both the basic payment information required by § 1041.9(b)(2)(ii)(B) through (D) and the description of unusual withdrawal required by § 1041.9(b)(3)(ii)(C). 9(b)(3)(ii)(C) Description of unusual withdrawal. 1. General. An unusual withdrawal notice is required under § 1041.9(b)(3) if one or more conditions are present. The description of an unusual withdrawal informs the consumer of the condition that makes the pending payment transfer unusual. 2. Illustrative example. The lender provides a first payment withdrawal notice at origination. The first payment withdrawal initiated by the lender occurs on March 1, for $75, as a paper check. The second payment is scheduled for April 1, for $75, as an ACH transfer. Before the second payment, the lender provides an unusual withdrawal notice. The notice contains the basic payment information along with an explanation that the withdrawal is unusual because the payment channel has changed from paper check to ACH. Because the amount did not vary, the payment is taking place on the regularly scheduled date, and this is not a re-initiated payment, the only applicable content under § 1041.9(b)(3)(ii)(C) is the different payment channel information. 1673 3. Varying amount. The information about varying amount for closed-end loans in § 1041.9(b)(3)(ii)(C)(1)(i) applies in two circumstances. First, the requirement applies when a transfer is for the purpose of collecting a payment that is not specified by amount on the payment schedule, including, for example, a one-time electronic payment transfer to collect a late fee. Second, the requirement applies when the transfer is for the purpose of collecting a regularly scheduled payment for an amount different from the regularly scheduled payment amount according to the payment schedule. Given existing requirements for open-end credit plans, circumstances that trigger an unusual withdrawal for these plans are more limited according to § 1041.9(b)(3)(ii)(C)(1)(ii). Because the outstanding balance on open-end credit plans may change over time, the minimum payment due on the scheduled payment date may also fluctuate. However, the minimum payment amount due for these open-end credit plans would be disclosed to the consumer according to the periodic statement requirement in Regulation Z. The payment transfer amount would not be considered unusual with regards to an open end credit plan unless the amount deviates from the minimum payment due as disclosed in the periodic statement. The requirement for a first payment withdrawal notice under § 1041.9(b)(2) and the other circumstances that could trigger an unusual withdrawal notice under § 1041.9(b)(3)(ii)(C)(2) through (4), continue to apply. 4. Date other than due date of regularly scheduled payment. The changed date information in § 1041.9(b)(3)(ii)(C)(2) applies in two circumstances. First, the requirement applies when a transfer is for the purpose of collecting a payment that is not specified by date on the payment schedule, including, for example, a one-time electronic payment transfer to collect a late fee. Second, the requirement applies when the transfer is for the purpose of collecting a regularly scheduled payment on a date that differs from the regularly scheduled payment date 1674 according to the payment schedule. 9(b)(4) Electronic delivery. 1. General. If the lender is using a method of electronic delivery other than email, such as text or mobile application, the lender must provide the notice with the electronic short notice as provided in § 1041.9(b)(4)(ii). If the lender is using email as the method of electronic delivery, § 1041.9(b)(4)(iii) allows the lender to determine whether to use the electronic short notice approach or to include the full text of the notice in the body of the email. 9(b)(4)(ii) Electronic short notice. 9(b)(4)(ii)(A) Content. 1. Identifying statement. If the lender is using email as the method of electronic delivery, the identifying statement required in § 1041.9(b)(2)(ii)(A) and 1041.9(b)(3)(ii)(A) must be provided in both the email subject line and the body of the email. 9(c) Consumer Rights Notice. 9(c)(2) Timing. 1. General. Any information provided to the lender or its agent that the payment transfer has failed would trigger the timing requirement provided in § 1041.9(c)(2). For example, if the lender’s agent, a payment processor, learns on Monday, June 1 that an ACH payment transfer initiated by the processor on the lender’s behalf has been returned for non-sufficient funds, the lender would be required to send the consumer rights notice by Thursday, June 4. 9(c)(3) Content. 1. Identifying statement. If the lender is using email as the method of electronic delivery, the identifying statement required in § 1041.9(c)(3)(i) must be provided in both the email subject line and the body of the email. 1675 2. Fees. If the lender is also the consumer’s account-holding institution, this includes all fees charged in relation to the transfer, including any returned payment fees charged to outstanding loan balance and any fees, such as overdraft or insufficient fund fees, charged to the consumer’s account. 9(c)(4) Electronic delivery. 1. General. See comments 9(b)(4)-1 and 9(b)(4)(ii)(A)-1. Section 1041.10 Furnishing Information to Registered Information Systems 10(a) Loans subject to furnishing requirement. 1. Application to rollovers. The furnishing requirements in § 1041.10(a) apply to each covered short-term loan or covered longer-term balloon-payment loan a lender makes, as well as to loans that are a rollover of a prior covered short-term loan or covered longer-term balloonpayment loan (or what is termed a “renewal” in some States). Rollovers are defined as a matter of State law but typically involve deferral of repayment of the principal amount of a short-term loan for a period of time in exchange for a fee. In the event that a lender is permitted under State law to roll over a covered short-term loan or covered longer-term balloon-payment loan and does so in accordance with the requirements of § 1041.5 or § 1041.6, the rollover would be treated, as applicable, as a new covered short-term loan or as a new covered longer-term balloon-payment loan for purposes of § 1041.10. For example, assume that a lender is permitted under applicable State law to roll over a covered short-term loan; the lender makes a covered short-term loan with a 14-day contractual duration; and on day 14 the lender reasonably determines that the consumer has the ability to repay a new loan under § 1041.5 and offers the consumer the opportunity to roll over the first loan for an additional 14 days. If the consumer accepts the rollover, the lender 1676 would report the original loan as no longer outstanding and would report the rollover as a new covered short-term loan. 2. Furnishing through third parties. Section 1041.10(a) requires that, for each covered short-term loan and covered longer-term balloon loan a lender makes, the lender must furnish the information concerning the loan described in § 1041.10(c) to each information system described in § 1041.10(b). A lender may furnish information to a registered information system directly, or may furnish through a third party acting on its behalf, including a registered information system. 10(b) Information systems to which information must be furnished. 1. Provisional registration and registration of information system while loan is outstanding. Pursuant to § 1041.10(b)(1), a lender is only required to furnish information about a covered loan to an information system that, at the time the loan is consummated, has been registered pursuant to § 1041.11(c)(2) for 180 days or more or has been provisionally registered pursuant to § 1041.11(d)(1) for 180 days or more or subsequently has become registered pursuant to § 1041.11(d)(2). For example, if an information system is provisionally registered on March 1, 2020, the obligation to furnish information to that system begins on August 28, 2020, 180 days from the date of provisional registration. A lender is not required to furnish information about a loan consummated on August 27, 2020 to an information system that became provisionally registered on March 1, 2020. 2. Preliminary approval. Section 1041.10(b) requires that lenders furnish information to information systems that are provisionally registered pursuant to § 1041.11(d)(1) and information systems that are registered pursuant to § 1041.11(c)(2) or § 1041.11(d)(2). Lenders are not required to furnish information to entities that have received preliminary approval for registration pursuant to § 1041.11(c)(1) but are not registered pursuant to § 1041.11(c)(2). 1677 10(c) Information to be furnished. 1. Deadline for furnishing under § 1041.10(c)(1) and (3). Section 1041.10(c)(1) requires that a lender furnish specified information no later than the date on which the loan is consummated or as close in time as feasible to the date the loan is consummated. Section 1041.10(c)(3) requires that a lender furnish specified information no later than the date the loan ceases to be an outstanding loan or as close in time as feasible to the date the loan ceases to be an outstanding loan. Under each of § 1041.10(c)(1) and (3), if it is feasible to report on the specified date (such as the consummation date), the specified date is the date by which the information must be furnished. 10(c)(1) Information to be furnished at loan consummation. 1. Type of loan. Section 1041.10(c)(1)(iii) requires that a lender furnish information that identifies a covered loan as either a covered short-term loan or a covered longer-term balloonpayment loan. For example, a lender must identify a covered short-term loan as a covered shortterm loan. 2. Whether a loan is made under § 1041.5 or § 1041.6. Section 1041.10(c)(1)(iv) requires that a lender furnish information that identifies a covered loan as made under § 1041.5 or made under § 1041.6. For example, a lender must identify a loan made under § 1041.5 as a loan made under § 1041.5. 10(c)(2) Information to be furnished while loan is an outstanding loan. 1. Examples. Section 1041.10(c)(2) requires that, during the period that the loan is an outstanding loan, a lender must furnish any update to information previously furnished pursuant to § 1014.10 within a reasonable period of the event that causes the information previously furnished to be out of date. Information previously furnished can become out of date due to 1678 changes in the loan terms or due to actions by the consumer. For example, if a lender extends the term of a closed-end loan, § 1041.10(c)(2) would require the lender to furnish an update to the date that each payment on the loan is due, previously furnished pursuant to § 1041.10(c)(1)(vii)(B), and to the amount due on each payment date, previously furnished pursuant to § 1041.10(c)(1)(vii)(C), to reflect the updated payment dates and amounts. If the amount or minimum amount due on future payment dates changes because the consumer fails to pay the amount due on a scheduled payment date, § 1041.10(c)(2) would require the lender to furnish an update to the amount or minimum amount due on each payment date, previously furnished pursuant to § 1041.10(c)(1)(vii)(C) or (c)(1)(viii)(D), as applicable, to reflect the updated amount or minimum amount due on each payment date. However, if a consumer makes payment on a closed-end loan as agreed and the loan is not modified to change the dates or amounts of future payments on the loan, § 1041.10(c)(2) would not require the lender to furnish an update to information concerning the date that each payment on the loan is due, previously furnished pursuant to § 1041.10(c)(1)(vii)(B), or the amount due on each payment date, previously furnished pursuant to § 1041.10(c)(1)(vii)(C). Section 1041.10(c)(2) does not require a lender to furnish an update to reflect that a payment was made. 2. Changes to information previously furnished pursuant to § 1041.10(c)(2). Section 1041.10(c)(2) requires that, during the period that the loan is an outstanding loan, a lender must furnish any update to information previously furnished pursuant to § 1014.10 within a reasonable period of the event that causes the information previously furnished to be out of date. This requirement extends to information previously furnished pursuant to § 1014.10(c)(2). For example, if a lender furnishes an update to the amount or minimum amount due on each payment date, previously furnished pursuant to § 1041.10(c)(1)(vii)(C) or (c)(1)(viii)(D), as 1679 applicable, and the amount or minimum amount due on each payment date changes again after the update, § 1041.10(c)(2) requires that the lender must furnish an update to the information previously furnished pursuant to § 1041.10(c)(2). Section 1041.11 Registered Information Systems 11(b) Eligibility criteria for registered information systems. 11(b)(2) Reporting capability. 1. Timing. To be eligible for provisional registration or registration, an entity must possess the technical capability to generate a consumer report containing, as applicable for each unique consumer, all information described in § 1041.10 substantially simultaneous to receiving the information from a lender. Technological limitations may cause some slight delay in the appearance on a consumer report of the information furnished pursuant to § 1041.10, but any delay must reasonable. 11(b)(3) Performance. 1. Relationship with other law. To be eligible for provisional registration or registration, an entity must perform in a manner that facilitates compliance with and furthers the purposes of part 1041. However, this requirement does not supersede consumer protection obligations imposed upon a provisionally registered or registered information system by other Federal law or regulation. For example, the Fair Credit Reporting Act requires that, “[w]henever a consumer reporting agency prepares a consumer report it shall follow reasonable procedures to assure maximum possible accuracy of the information concerning the individual about whom the report relates.” 15 U.S.C. 1681e(b). If including information furnished pursuant to § 1041.10 in a consumer report would cause a provisionally registered or registered information system to 1680 violate this requirement, § 1041.11(b)(3) would not require that the information be included in a consumer report. 2. Evidence of ability to perform in a manner that facilitates compliance with and furthers the purposes of this part. Section 1041.11(c)(1) requires that an entity seeking preliminary approval to be a registered information system must submit an application to the Bureau containing information sufficient for the Bureau to determine that the entity is reasonably likely to satisfy the conditions set forth in § 1041.11(b). Section 1041.11(c)(2) and (d)(1) requires that an entity seeking to be a registered information system or a provisionally registered information system must submit an application that contains information and documentation sufficient for the Bureau to determine that the entity satisfies the conditions set forth in § 1041.11(b). In evaluating whether an applicant is reasonably likely to satisfy or satisfies the requirement set forth in § 1041.11(b)(3), the Bureau will consider the extent to which an applicant has experience functioning as a consumer reporting agency. 11(b)(4) Federal consumer financial law compliance program. 1. Policies and procedures. To be eligible for provisional registration or registration, an entity must have policies and procedures that are documented in sufficient detail to implement effectively and maintain its Federal consumer financial law compliance program. The policies and procedures must address compliance with applicable Federal consumer financial laws in a manner reasonably designed to prevent violations and to detect and prevent associated risks of harm to consumers. The entity must also maintain and modify, as needed, the policies and procedures so that all relevant personnel can reference them in their day-to-day activities. 2. Training. To be eligible for provisional registration or registration, an entity must provide specific, comprehensive training to all relevant personnel that reinforces and helps 1681 implement written policies and procedures. Requirements for compliance with Federal consumer financial laws must be incorporated into training for all relevant officers and employees. Compliance training must be current, complete, directed to appropriate individuals based on their roles, effective, and commensurate with the size of the entity and nature and risks to consumers presented by its activity. Compliance training also must be consistent with written policies and procedures and designed to enforce those policies and procedures. 3. Monitoring. To be eligible for provisional registration or registration, an entity must implement an organized and risk-focused monitoring program to promptly identify and correct procedural or training weaknesses so as to provide for a high level of compliance with Federal consumer financial laws. Monitoring must be scheduled and completed so that timely corrective actions are taken where appropriate. 11(b)(5) Independent assessment of Federal consumer financial law compliance program. 1. Assessor qualifications. An objective and independent third-party individual or entity is qualified to perform the assessment required by § 1041.11(b)(5) if the individual or entity has substantial experience in performing assessments of a similar size, scope, or subject matter; has substantial expertise in both the applicable Federal consumer financial laws and in the entity’s or information system’s business; and has the appropriate professional qualifications necessary to perform the required assessment adequately. 2. Written assessment. A written assessment described in § 1041.11(b)(5) need not conform to any particular format or style as long as it succinctly and accurately conveys the required information. 1682 11(b)(7) Independent assessment of information security program. 1. Periodic assessments. Section 1041.11(b)(7) requires that, to maintain its registration, an information system must obtain and provide to the Bureau, on at least a biennial basis, a written assessment of the information security program described in § 1041.11(b)(6). The period covered by each assessment obtained and provided to the Bureau to satisfy this requirement must commence on the day after the last day of the period covered by the previous assessment obtained and provided to the Bureau. 2. Assessor qualifications. Professionals qualified to conduct assessments required under § 1041.11(b)(7) include: a person qualified as a Certified Information System Security Professional (CISSP) or as a Certified Information Systems Auditor (CISA); a person holding Global Information Assurance Certification (GIAC) from the SysAdmin, Audit, Network, Security (SANS) Institute; and an individual or entity with a similar qualification or certification. 3. Written assessment. A written assessment described in § 1041.11(b)(7) need not conform to any particular format or style as long as it succinctly and accurately conveys the required information. 11(c) Registration of information systems prior to effective date of § 1041.10. 11(c)(1) Preliminary approval. 1. In general. An entity seeking to become preliminarily approved for registration pursuant to § 1041.11(c)(1) must submit an application to the Bureau containing information sufficient for the Bureau to determine that the entity is reasonably likely to satisfy the conditions set forth in § 1041.11(b) as of the deadline set forth in § 1041.11(c)(3)(ii). The application must describe the steps the entity plans to take to satisfy the conditions set forth in § 1041.11(b) by the 1683 deadline and the entity’s anticipated timeline for such steps. The entity’s plan must be reasonable and achievable. 11(c)(2) Registration. 1. In general. An entity seeking to become a registered information system pursuant to § 1041.11(c)(2) must submit an application to the Bureau by the deadline set forth in § 1041.11(c)(3)(ii) containing information and documentation adequate for the Bureau to determine that the conditions described in § 1041.11(b) are satisfied. The application must succinctly and accurately convey the required information, and must include the written assessments described in § 1041.11(b)(5) and (7). 11(d) Registration of information systems on or after effective date of § 1041.10. 11(d)(1) Provisional registration. 1. In general. An entity seeking to become a provisionally registered information system pursuant to § 1041.11(d)(1) must submit an application to the Bureau containing information and documentation adequate for the Bureau to determine that the conditions described in § 1041.11(b) are satisfied. The application must succinctly and accurately convey the required information, and must include the written assessments described in § 1041.11(b)(5) and (7). Section 1041.12—Compliance Program and Record Retention 12(a) Compliance program. 1. General. Section 1041.12(a) requires a lender making a covered loan to develop and follow written policies and procedures that are reasonably designed to ensure compliance with the applicable requirements in part 1041. These written policies and procedures must provide guidance to a lender’s employees on how to comply with the requirements in part 1041. In particular, under § 1041.12(a), a lender must develop and follow detailed written policies and 1684 procedures reasonably designed to achieve compliance, as applicable, with the ability-to-repay requirements in § 1041.5, alternative requirements in § 1041.6, payments requirements in §§ 1041.8 and 1041.9, and requirements on furnishing loan information to registered and provisionally registered information systems in § 1041.10. The provisions and commentary in each section listed above provide guidance on what specific directions and other information a lender must include in its written policies and procedures. 2. Examples. The written policies and procedures a lender must develop and follow under § 1041.12(a) depend on the types of loans that the lender makes. A lender that makes a covered loan under § 1041.5 must develop and follow written policies and procedures to ensure compliance with the ability-to-repay requirements, including on projecting a consumer’s net income and payments on major financial obligations, and estimating a consumer’s basic living expenses. Among other written policies and procedures, a lender that makes a covered loan under § 1041.5 or § 1041.6 must develop and follow written policies and procedures to furnish loan information to registered and provisionally registered information systems in accordance with § 1041.10. A lender that makes a covered loan subject to the requirements in § 1041.6 or § 1041.9 must develop and follow written policies and procedures to provide the required disclosures to consumers. 12(b) Record retention. 1. General. Section 1041.12(b) requires a lender to retain various categories of documentation and information in connection with the underwriting and performance of covered short-term loans and covered longer-term balloon payment loans, as well as payment practices in connection with covered loans generally. The items listed are non-exhaustive as to the records 1685 that may need to be retained as evidence of compliance with part 1041 concerning loan origination and underwriting, terms and performance, and payment practices. 12(b)(1) Retention of loan agreement and documentation obtained in connection with originating a covered short-term or covered longer-term balloon-payment loan. 1. Methods of retaining loan agreement and documentation obtained for a covered short-term or covered longer-term balloon-payment loan. Section 1041.12(b)(1) requires a lender either to retain the loan agreement and documentation obtained in connection with a covered short-term or covered longer-term balloon-payment loan in original form or to be able to reproduce an image of the loan agreement and documentation accurately. For example, if the lender uses a consumer’s pay stub to verify the consumer’s net income, § 1041.12(b)(1) requires the lender to either retain a paper copy of the pay stub itself or be able to reproduce an image of the pay stub, and not merely the net income information that was contained in the pay stub. For documentation that the lender receives electronically, such as a consumer report from a registered information system, the lender may retain either the electronic version or a printout of the report. 12(b)(2) Electronic records in tabular format regarding origination calculations and determinations for a covered short-term or longer-term balloon-payment loan under § 1041.5. 1. Electronic records in tabular format. Section 1041.12(b)(2) requires a lender to retain records regarding origination calculations and determinations for a covered loan in electronic, tabular format. Tabular format means a format in which the individual data elements comprising the record can be transmitted, analyzed, and processed by a computer program, such as a widely used spreadsheet or database program. Data formats for image reproductions, such as PDF, and document formats used by word processing programs are not tabular formats. A lender does not 1686 have to retain the records required in § 1041.12(b)(2) in a single, combined spreadsheet or database with the records required in § 1041.12(b)(3) and (b)(5). Section 1041.12(b)(2), however, requires a lender to be able to associate the records for a particular covered short-term or covered longer-term balloon payment loan in § 1041.12(b)(2) with unique loan and consumer identifiers in § 1041.12(b)(3). 12(b)(3) Electronic records in tabular format regarding type, terms, and performance of covered short-term or covered longer-term balloon-payment loans. 1. Electronic records in tabular format. Section 1041.12(b)(3) requires a lender to retain records regarding loan type, terms, and performance of covered short-term or covered longerterm balloon-payment loans for a covered loan in electronic, tabular format. See comment 12(b)(2)-1 for a description of how to retain electronic records in tabular format. A lender does not have to retain the records required in § 1041.12(b)(3) in a single, combined spreadsheet or database with the records required in § 1041.12(b)(2). Section 1041.12(b)(3), however, requires a lender to be able to associate the records for a particular covered short-term or covered longerterm balloon payment loan in § 1041.12(b)(2) and (b)(5) with unique loan and consumer identifiers in § 1041.12(b)(3). Paragraph 12(b)(3)(iv). 1. Maximum number of days, up to 180 days, any full payment was past due. Section 1041.12(b)(3)(iv) requires a lender that makes a covered loan to retain information regarding the number of days any full payment is past due beyond the payment schedule established in the loan agreement, up to 180 days. For this purpose, a full payment is defined as principal, interest, and any charges. If a consumer makes a partial payment on the contractual due date and the remainder of the payment 10 days later, the lender must record the full payment 1687 as being 10 days past due. If a consumer fails to make a full payment on a covered loan more than 180 days after the contractual due date, the lender must only record the full payment as being 180 days past due. 12(b)(4) Retention of records relating to payment practices for covered loans. 1. Methods of retaining documentation. Section 1041.12(b)(4) requires a lender either to retain certain payment-related information in connection with covered loans in original form or to be able to reproduce an image of such documents accurately. For example, § 1041.12(b)(4) requires the lender to either retain a paper copy of the leveraged payment mechanism obtained in connection with a covered longer-term loan or to be able to reproduce an image of the mechanism. For documentation that the lender receives electronically, the lender may retain either the electronic version or a printout. 12(b)(5) Electronic records in tabular format regarding payment practices for covered loans. 1. Electronic records in tabular format. Section 1041.12(b)(5) requires a lender to retain records regarding payment practices in electronic, tabular format. See comment 12(b)(2)-1 for a description of how to retain electronic records in tabular format. A lender does not have to retain the records required in § 1041.12(b)(5) in a single, combined spreadsheet or database with the records required in § 1041.12(b)(2) and (b)(3). Section 1041.12(b)(5), however, requires a lender to be able to associate the records for a particular covered short-term or covered longerterm balloon payment loan in § 1041.12(b)(5) with unique loan and consumer identifiers in § 1041.12(b)(3). 1688 Section 1041.13—Prohibition against Evasion 1. Lender action taken with the intent of evading the requirements of the rule. Section 1041.13 provides that a lender must not take any action with the intent of evading the requirements of part 1041. In determining whether a lender has taken action with the intent of evading the requirements of part 1041, the form, characterization, label, structure, or written documentation of the lender’s action shall not be dispositive. Rather, the actual substance of the lender’s action as well as other relevant facts and circumstances will determine whether the lender’s action was taken with the intent of evading the requirements of part 1041. If the lender’s action is taken solely for legitimate business purposes, it is not taken with the intent of evading the requirements of part 1041. By contrast, if a consideration of all relevant facts and circumstances reveals the presence of a purpose that is not a legitimate business purpose, the lender’s action may have been taken with the intent of evading the requirements of part 1041. A lender action that is taken with the intent of evading the requirements of part 1041 may be knowing or reckless. Fraud, deceit, or other unlawful or illegitimate activity may be one fact or circumstance that is relevant to the determination of whether a lender’s action was taken with the intent of evading the requirements of part 1041, but fraud, deceit, or other unlawful or illegitimate activity is not a prerequisite to such a finding. 1689 !THIS SIGNATURE PAGE PERTAINS TO THE RULE TITLED "PAYDAY, VEHICLE TITLE, AND CERTAIN HIGH-COST INSTALLMENT LOANS"] Dated: October !/_, 2017. Richard Cordray, Director, Bureau of Consumer Financial Protection.