STATE OF MAINE PUBLIC UTILITIES COMMISSION Docket No. 2018-00194 February 22, 2019 PUBLIC UTILITIES COMMISSION Investigation into Rates and Revenue Requirements of Central Maine Power Company I. BENCH ANALYSIS INTRODUCTION A. Background On May 29, 2018, the Commission received a Complaint filed by Hebert C. Adams and 12 other individuals (the Adams Complaint) pursuant to 35-A M.R.S. § 1302 against Central Maine Power Company (CMP or the Company). Herbert C. Adams et al., 10-person Complaint Requesting Investigation to Determine if Central Maine Power Company and Its Parent Companies Are Making Excessive Returns on Investments, Docket No. 2018-00123, (May 29, 2018). The Adams Complaint requested that the Commission, on its own initiative, open a rate case to determine “if CMP and its parent companies are making excessive returns on investment, and if so that therefore CMP retail rates should reflect lower rates.” Adams Complaint at 1. Additionally, the Complaint alleged that CMP and its parent companies intended to recover “an inordinate amount of costs associated with the October 2017 windstorm” which should be denied. Adams Complaint at 1. On June 8, 2018, CMP filed its Response and Motion to Dismiss the Adams Complaint. On July 24, 2018, the Commission issued an Order dismissing the Adams Complaint as it applied to CMP’s parent companies and to the recovery of costs Bench Analysis 2 Docket No. 2018-00194 associated with the October 2017 windstorm, but granted the relief sought by the Complainants with respect to initiating a proceeding to examine CMP’s return on its investment. Herbert C. Adams et al., 10-person Complaint Requesting Investigation to Determine if Central Maine Power Company and Its Parent Companies Are Making Excessive Returns on Investments, Docket No. 2018-00123, Order Dismissing Complaint in Part and Granting Complaint in Part at 3 (July 24, 2018) (Adams Order). In so holding, the Commission found that CMP’s return on equity (ROE), when recalculated to reflect a 50% common equity ratio as used in the Company’s last rate case, was 9.87% for 2014; 7.73% for 2015; 11.52% for 2016; and 13.01% for 2017. Adams Order at 4. The Commission found that the returns for both 2016 and 2017 were significantly higher than the ROE utilized to set rates in CMP’s most recently completed rate case, Central Maine Power Co., Request for New Alternative Rate Plan (“ARP 2014”), Docket No. 2013-00168, Order Approving Stipulation (Aug. 25, 2014) (hereinafter “2013-00168 Order Approving Stipulation”).1 Adams Order at 4. The Commission concluded, therefore, that it was appropriate and consistent with the Commission’s duties set forth in 35-A M.R.S. § 301 to examine the justness and reasonableness of CMP’s current rates as part of a full review of CMP’s revenue requirements. Id. The Commission directed CMP to submit a general rate case filing consistent with the provisions of Chapter 120 of the Commission’s Rules no later than October 15, 2018. Id. 1 The ROE utilized in Docket No. 2013-00168 was 9.45%. Bench Analysis B. 3 Docket No. 2018-00194 CMP’s Filing As the Commission directed in its Adams Order, CMP submitted its direct case in this matter on October 15, 2018. CMP’s direct case consisted of the following testimonies: the Policy Testimony of Steve Adams and Eric Stinneford (Policy Dir.); the Revenue Requirements Testimony and Exhibits of Eric Stinneford, Peter Cohen, and Kirk Pelletier (Rev. Req. Dir.); the Operations and Capital Investment Testimony and Exhibits of Christian Bilcheck, Mary Lou Palmieri, William Ransom III, Juan Martinez Martinez, Kevin Elwell, Pamela Kelly, and Michael Zaffina (Ops. and Cap. Dir.); the Tax Testimony and Exhibits of Steve Adams, David Beber, and Mark Kelly (Tax Dir.); the Return on Equity and Capital Structure Testimony and Exhibits of Ann Bulkley (ROE Dir.); and the Sales and Customer Forecast Testimony and Exhibits of Michael Purtell and John Hastings (Sales Dir.). In its direct case, the Company alleges that, rather than over-earning, as the Commission indicated in its Adams Order, using a July 1, 2017 through June 30, 2018, test year, the Company is under-earning. The Company argues that, for a rate effective year beginning July 1, 2019, the Company requires a rate increase of $22.9 million. The Company stated, however, that it was cognizant of the impact of rate increases to customers and, therefore, proposed to fully mitigate its rate increase by accelerating the amortization of the Company’s Unprotected Excess Deferred Income Tax (EDIT) liability created as a result of Tax Cost and Jobs Act (TCJA) of 2017 in the rate effective year. Policy Dir. at 8. For subsequent years, the Company recommended mitigation of the increase by continuing the accelerated amortization of the EDIT Bench Analysis 4 Docket No. 2018-00194 regulatory liability in combination with an acceleration of the Cost of Removal Liability which is currently on the Company’s books at $46.6 million with an amortization recovery schedule of 33.6 years. Policy Dir. at 13–16. The Company provided the following breakdown of its revenue requirement deficiency: Table 12 Central Maine Power – Distribution Rate Case Revenue Requirement Drivers in millions of dollars Description Amount $ (2.4) $ (10.6) 3. 4. 5. 6. 7. 8. 9. Sales Growth Remove One-Time Test Year Adjustments (e.g. TCJA 2018 Deferral, RDM Amortization) Total Revenue Payroll Expense Storm Allowance (lower TY to RY level of $10.0M) Service Company Charges – Remove Cap Vegetation Management – Enhanced Other O&M Total Operating Expenses 10. 11. 12. 13. 14. 15. Depreciation Property and Other Taxes Income Taxes Amortization of Excess Deferred Income Taxes Rate Base Return on Investment (10.0% ROE / 55% Equity) 1. 2. 16. Total Before Mitigation $ (13.0) $ 7.1 (8.5) 5.4 5.0 2.1 $ 11.1 5.9 5.8 0.5 (6.9) 8.0 11.5 $ 17. Impact of Mitigation (EDIT) 18. Total After Mitigation 2 See Policy Test, at 13, Table 2. 22.9 (22.9) $ 0.00 Bench Analysis 5 Docket No. 2018-00194 The revenue requirement deficiency the Company identified is actually significantly worse than portrayed in Table 1. The deficiency assumes that revenues are based on current rate levels, which include a significant number of one-time adjustments, such as recovery of the October 2017 Storm expense, which were approved by the Commission in Central Maine Power Co., Annual Compliance Filing, Docket No. 2018-00069, Order Approving Stipulation (June 29, 2018), and which are scheduled to be removed from rates on July 1, 2019. When these one-time adjustments are removed, the Company’s proposed deficiency increases to $36.6 million. ODR-002-002. C. The Commission’s Investigation in Docket No. 2018-00052 and the Liberty Audit Due to a large number of customer complaints and varied problems customers were experiencing with their bills, as well as ongoing call answer problems, the Commission opened a summary investigation of CMP’s metering, billing and customer communication issues on March 1, 2018, to determine if CMP was meeting its obligation to provide reasonable, adequate and reliable service to its customers. Pub. Utils. Comm’n, Investigation of Central Maine Power Co. Metering, Billing and Customer Communication Issues, Docket No. 2018-00052, Notice of Investigation (Mar. 1, 2018). The Notice of Investigation set forth the following issues for investigation: Metering Issues • Are CMP’s meters accurately reading customer usage? Bench Analysis • 6 Docket No. 2018-00194 Are CMP’s AMI meters accurately communicating with CMP’s new CIS system? Billing Issues • Are CMP’s bills accurately reflecting usage? • Are CMP’s bills utilizing the correct rates? • Are CMP’s bills accurately calculating the total bill? • Is the Company able to identify billing error problems related to its new CIS billing system? • What problems has the Company identified to date? • Is the Company properly addressing such problems? Customer Communication • Is CMP answering and responding to calls from customers within a reasonable timeframe? • How has CMP reacted and responded to customer complaints on high bills? • Is CMP providing reasonable and adequate responses to customers’ calls? On March 22, 2018, given the complexity of the metering and billing issues, the Commission initiated a forensic audit of the Company’s customer billing system pursuant to the provisions of 35-A M.R.S. §113. Docket No. 2018-00052, Order Initiating Audit (Mar. 22, 2018). The Commission contracted with The Liberty Consulting Group (Liberty) to complete the forensic audit. On July 10, 2018, the Commission concluded that, based on Liberty’s expertise and the interrelationship between metering, billing and customer communications issues, the audit it initially Bench Analysis 7 Docket No. 2018-00194 ordered should be expanded to include the customer communication issues identified in the Commission’s March 1, 2018 Notice of Investigation. Docket No. 2018-00052, Order Modifying Scope of Audit at 1 (July 10, 2018). On December 8, 2018, the Liberty Group issued its Final Report on all aspects of its audit (hereinafter the “Liberty Report”). The Liberty Report included the following findings regarding CMP’s customer service issues: • Significant gaps in SmartCare testing and training, and in the transition to it, produced in its initial phase of operation an unnecessarily large number of errors, which required lengthy manual correction before bill issuance. • A shortage of personnel contributed to the inability to eliminate errors before the SmartCare go-live date. Continuing shortages of experienced personnel after go-live unduly delayed fixes to the errors, caused customers significant difficulty in reaching CMP representatives and in obtaining answers to questions and concerns, and created excessive delays in resolving billing problems. Customer performance metrics fell below norms and remained so for some time, including to today. • The extent and degree of performance degradation contributed strongly to a level of customer frustration, doubt, and skepticism that was already high due to uncharacteristically large bills in winter 2017-2018. In an Order dated January 14, 2019, the Commission concluded that the customer communication and service issues that were the subject of the Liberty Report warranted further investigation as part of a formal adjudicatory proceeding and that this further investigation of customer communication and service issues would be done as part of this case. In so holding, the Commission noted that addressing customer service issues in the rate case was likely the most expeditious means of doing so since the Commission could incorporate any determinations regarding the reasonableness or prudence of CMP’s customer service in its order at the end of the rate case. Docket Bench Analysis 8 Docket No. 2018-00194 No. 2018-00052, Order and Notice of Investigation at 9 (Jan. 14, 2019). As part of that same Order, the Commission concluded that further investigation of CMP’s metering and billing issues was warranted and would be done in a separate adjudicatory proceeding. The Commission identified bill errors or “exceptions” as one overlap area between the two investigations. The Commission stated that billing error issues involving customer service communication and delays in billing customers would be addressed in the rate case. Id. at 10. D. Summary of Bench Analysis Findings and Proposal In accordance with the February 14, 2019 Procedural Order (Scheduling), the Staff now submits its Bench Analysis on all revenue requirement issues in this case, including customer service issues.3 The Bench Analysis provides the Commission Staff’s view and proposals on issues at this point in time and based on the information currently available to the Staff. The Staff’s ultimate recommendations to the Commission on revenue requirement issues will be contained in an Examiners’ Report scheduled to be issued on September 12, 2019. As part of this Bench Analysis, the Staff has performed a traditional review of the Company’s proposed revenue requirements, including the separation of transmission and distribution costs, additions to rate base, rate year expense adjustments and projections, and the cost of capital to applied rate base. This traditional review has 3 As provided in the Procedural Order, the Staff may, depending on the information provided by the Company in its March 5, 2018 responses to data requests, update its plant addition analysis on April 2, 2019. Bench Analysis 9 Docket No. 2018-00194 yielded a $20 million reduction to the Company’s proposed revenue requirement deficiency. In addition, as part of its preparation of this Bench Analysis, the Commission Staff has reviewed the Liberty Report and its findings and conclusions as they relate to customer service and communication issues. The Commission Staff has found The Liberty Report to be consistent with the Staff’s own observations and findings regarding customer service performance. The Staff presents a detailed account of its customer service observations and experiences in Section III. As discussed there, it is Staff’s view at this time that starting in 2016 and through the present time, CMP’s performance has significantly and consistently fallen below standards reasonably expected of a utility to provide adequate service, and substantially departs from the regular and accepted practice of both the Company itself as well as other utilities in general. Based on these substantial shortcomings, the Staff is proposing that a management efficiency adjustment be made as part of rate setting process in this proceeding which would reduce the Company’s ROE by 75 to 100 basis points. This management efficiency adjustment would further reduce the Company’s revenue requirement deficiency by $4.8 million to $6.5 million, based on Staff’s proposed rate base. II. REVENUE REQUIREMENT A. Transmission/Distribution Allocations The Company has applied applicable customer, wage, and plant allocators based on year-end 2017 results to allocate costs between transmission and distribution. EXM-003-005. Staff proposes that allocators based on 2018 data better reflect the rate Bench Analysis 10 Docket No. 2018-00194 year cost split between transmission and distribution costs and, once available, should be used to calculate the revenue requirement. In proposing so, the Staff notes that a review of historical allocators from 2013– 2017 shows a downward trend in the distribution portion of the customer and wage allocators. EXM-003-001, Att. 1. For the purposes of this Bench Analysis, Staff has projected customer and wage distribution allocators based on the average annual change between 2013 and 2017 and used these projections as a proxy for 2018 values. The customer allocator has decreased by an average of 3.44% per year, which, when applied to the distribution portion of the 2017 allocator of 53.87%, results in a distribution rate year allocator of 52.01%. Similarly, the wage allocator has decreased an average of 1.48% per year, which, when applied to the 2017 allocator of 75.69%, results in a distribution rate year allocator of 74.57%. There has been greater variability in the plant allocation factor, and, therefore, Staff has used the 2017 value of 42.84% as a proxy for 2018. Each of the three allocation factors will be updated once actual 2018 allocators are known. B. Rate Base 1. Plant Addition Attrition Analysis a. CMP’s Position CMP asserts that a major driver of its requested rate year rate increase is the growth in rate base. CMP calculates rate year plant additions based on a five-year Bench Analysis 11 Docket No. 2018-00194 historical average growth factor. According to CMP, rate relief is required, in part, because of the increase in rate base of approximately $75.1 million for the 12 months ending 6/30/2019 (the interim year) and $78.5 million for the rate year ending on 6/30/2020. Policy Dir. at 13, Table 2. In calculating projected distribution plant additions, CMP uses the “attrition technique” used by the Commission Staff in Docket No. 2013-00168. CMP notes that the attrition technique is a trend analysis that calculates a compound average growth rate (CAGR) based on historical plant balances, adjusted for atypical additions. Rev. Req. Dir. at 26. In performing its analysis, CMP states that it used June 30 Distribution and General and Intangible plant balances for the years 2013–2018. Id. at 27. CMP made various adjustments including removal of $24,698,000 related to the October 2017 Storm, a downward adjustment of the General and Intangible (G&I) balances by amounts related to the Energy Control Center (ECC), and a reduction of the June 30, 2018 G&I balance of $55,497,000 related to CMP’s new Customer Relations Management & Billing (CRM&B) system. Id. CMP combined the adjusted distribution plant and G&I balances to produce total adjusted year-ending June 30 distribution plant balances. CMP used the total adjusted June 30 distribution plant balances were to calculate a 5-year CAGR of 4.54%. Id. This CAGR was applied to the June 30, 2018 adjusted plant balance to project a June 30, 2019 balance of $1,728,457,000 and a June 30, 2020 balance of $1,806,947,000. These projections are the result of expected plant additions of $75,081,000 in the interim year and $78,490,000 in the rate year. Id. at 27– 28. According to CMP, these additions reflect typical distribution-related plant Bench Analysis 12 Docket No. 2018-00194 investments and do not reflect investments in new or atypical projects or programs. Id. at 28. CMP’s attrition calculation is presented in Table 2 below: Table 2 4 Distribution Plant Additions Using Staff's Attrition Technique ($000,000) A Distrib Plant 1. 2. 3. 4. 5. 6. 6/30/2013 6/30/2014 6/30/2015 6/30/2016 6/30/2017 6/30/2018 1,128.9 1,189.3 1,218.3 1,257.3 1,320.6 1,417.0 B Oct 2017 Storm $ (24.7) C D E G&I Plant ECC CRM & B $ 261.7 267.1 293.4 309.6 326.7 417.7 $ (18.8) (18.6) (18.1) (18.1) (16.3) (17.2) 7. 5-yr CAGR $ (55.5) F G&I Transmis $ (47.7) (49.2) (56.0) (69.5) (75.5) (83.9) G Total $1,324.2 1,388.5 1,437.6 1,479.3 1,555.6 1,653.4 H Plant Additions $ 64.3 49.1 41.7 76.3 97.8 4.54% 8. 6/30/2019 9. 6/30/2020 1,795.2 1,873.6 b. $ 75.1 78.5 Staff’s Proposal While the Commission Staff does not take issue with CMP’s use of what has been referred to as the “Staff’s attrition technique” in conducting plant additions analysis in this case, Commission Staff believes that CMP’s analysis requires several adjustments. In calculating its 2018 additions, the Company removed $24.7 million associated with the October 2017 storm restoration and $55.5 million associated with the CRM&B system. Even with these reductions, however, the plant additions for 2018 totaled $97.8 4 Source: Rev. Req. Dir. at 20, Table 20. Bench Analysis 13 Docket No. 2018-00194 million. CMP’s calculated 2018 plant additions of approximately $97.8 million is significantly higher than, and is an outlier when compared to plant additions in previous years. Specifically, the 2018 additions were 52% greater than 2014; almost 100% greater than 2015; 134% greater than 2016; and 28% greater than 2017. In order to provide a more accurate projection of distribution plant additions, Commission Staff believes it is appropriate to exclude 2018 from the trend analysis because it is so much higher than previous years. It appears anomalous, and therefore distorts the CAGR calculation and provides an inaccurate estimate of projected future plant additions. Removal of the June 30, 2018 amounts results in a 4-year CAGR calculation of 4.11%. When this CAGR is applied to the June 30, 2018 adjusted plant balance, it projects a June 30, 2019 balance of approximately $1,721.4 million and a June 30, 2020 balance of approximately $1,792.1 million. These projections are the result of expected plant additions of approximately $68 million in the interim year and approximately $70.7 million in the rate year. Bench Analysis 14 Docket No. 2018-00194 Table 3 1 2 3 4 5 6 Revised Table 20- Distribution Plant Additions Using Staff's Attrition Technique ($000,000) Remove 2017 /2018 A B C D E F G H Oct Distrib. 2017 G&I G&I to Plant Plant Storm Plant ECC CRM&B Transmis. Total Additions 6/30/2013 1128.9 261.7 -18.8 -47.7 1324.1 6/30/2014 1189.3 267.1 -18.6 -49.2 1388.6 64.5 6/30/2015 1218.3 293.4 -18.1 -56 1437.6 49 6/30/2016 1257.3 309.6 -18.1 -69.5 1479.3 41.7 6/30/2017 1320.6 326.7 -16.3 -75.5 1555.5 76.2 6/30/2018 1417 -24.7 417.7 -17.2 -55.5 -83.9 1653.4 97.9 7 4-yr CAGR 8 9 6/30/2019 6/30/2020 4.11% 4.11% 1,721.4 1,792.1 The Commission Staff believes that these results provide a far more accurate projection of future expected distribution plant additions. The Staff also believes that certain additional adjustments, similar to those that CMP made in 2018 for the October Storm and the CRM&B investments, should be made for plant additions in other years based on the extraordinary and non-recurring nature of such investments. Staff has identified certain items that it believes are candidates for exclusion, and has requested that the Company identify the plant addition amounts and year of inclusion in rate base for such items. EXM-009-001. Staff may update its attrition analysis based on the Company’s response to this request, as well as the Company’s supplemental response to ODR-002-012 as part of its April 2, 2019 filing in this case. 68.0 70.7 Bench Analysis 2. 15 Docket No. 2018-00194 Resiliency Adjustment For the reasons discussed in Section IV, the Staff is not including a discrete rate base adjustment for Resiliency Investments projected to occur in 2019 and 2020. Removing such investments reduces 2019 plant additions by $4.0 million for the interim year and $12.0 million for the rate effective year. 3. Capitalized Bonus Compensation a. Company’s Position As discussed in Appendix A to this Bench Analysis, the Company has reduced its variable or bonus compensation expense by 18.6%, an amount which it attributes to shareholder benefits. Rev. Req. Dir. Exh. RRP-3-9 Schedule D. The Company did not, however, make any shareholder benefit adjustment for that portion of variable compensation which is capitalized. Dec. 4, 2018 Tech. Conf. Tr. at 43. b. Staff’s Position Consistent with the Company’s proposal of only including that portion of variable compensation associated with customer benefit, the Staff proposes a reduction in plant additions associated with capitalized shareholder benefit variable compensation. Table 4 below presents the amounts paid out in total variable compensation, broken down by amounts expensed and capitalized for the 2013–2017 time period. Bench Analysis 16 Docket No. 2018-00194 Table 4 5 Year Expensed Capitalized Total Pct Capitalized 2013 1,423 666 2,089 32% 2014 2,568 1,161 3,729 31% 2015 2,752 1,581 4,333 36% 2016 2,314 1,304 3,618 36% 2017 3,183 2,109 5,292 40% Average 35% Dollars in Thousands Based on the amounts projected for variable compensation in the transition year and the rate year, and the average amount of variable compensation capitalized during the 2013–2017 periods, the Staff proposes that $358,000 of rate base plant additions in the transition year and $374,000 in the rate year be removed from rate base as being attributable to the shareholder portion of capitalized bonus compensation. A discussion of the calculation supporting this adjustment is provided in confidential Appendix A to this Bench Analysis. 4. Power Tax Regulatory Asset a. Background In 2012, CMP created a regulatory asset (Power Tax Regulatory Asset or PTRA) due to the implementation of a tax accounting software program, Power Tax, which 5 Source: ODR-001-037 Bench Analysis 17 Docket No. 2018-00194 CMP indicated provided a more sophisticated tool for ensuring compliance with the Normalization Rules contained in Section 168 of the Internal Revenue Code Tax Dir. at 12–14. CMP noted that it did not have an automated method of capturing the actual book depreciation associated with the asset basis that was subject to normalization and the manual methods used by CMP over a number of years inadvertently resulted in an assumed shorter life for most assets than was actually being used to depreciate those assets for book purposes. Docket No. 2103-00168, CMP Rev. Req. Reb. at 24. As a result, CMP determined that it had not recorded sufficient deferred income taxes and also that it had flowed through excess benefits to ratepayers. CMP argued that it interpreted its inadvertent flow-through to be in violation of the Normalization Rules adopted by the IRS to ensure that regulatory authorities do not flow through the tax benefits associated with accelerated tax depreciation into utility rates but rather allow the utility to get the use of those benefits as an interest free loan from the Government. The PTRA represented the level of deferred taxes that CMP calculated it had inadvertently flowed through to ratepayers. CMP initially asked for recovery of the PTRA in Docket No. 2013-00168. However, the Commission Staff questioned the requested recovery for the years when CMP was operating under an approved Alternate Rate Plan (ARP). The Commission’s Order Approving the Stipulation in Docket No. 2013-00168 allowed CMP to reflect in rates the impact of the Power Tax adjustment on a going forward basis, limiting the level of regulatory asset. As agreed to in the Stipulation approved by the Commission, the issue of whether CMP was required to record a regulatory asset for the period that it was under an ARP was referred to the IRS in the form of a request for Private Letter Ruling (PLR). In its PLR, the IRS found Bench Analysis 18 Docket No. 2018-00194 that the Commission needed to take some corrective action to address the flow-through of deferred taxes which was detected when CMP adopted its new Power Tax accounting software. CMP has continued to accumulate a balance in its PTRA and has consistently asked for recovery of the deferred amounts. While the Commission has agreed to allow recovery of the PTRA based upon the outcome of the PLR, it has also required that CMP fully support the level of recovery requested. Central Maine Power Company, Request for Approval Annual Compliance Filing, Docket No. 2016-00035, Order Initiating Audit at 4 (Sept. 8, 2016). b. CMP’s Position CMP’s filing includes the recovery of the PTRA with a proposed amortization period of 24.5 years which CMP states represents the average remaining book life of CMP’s distribution plant. CMP estimates the PTRA at $11.8 million at December 2017, adjusted to $13.8 million at June 2019. CMP proposes to include the unamortized PTRA in rate base and the annual amortization of $564,000 in its revenue requirement. Rev. Req. Dir. at 25. c. The Staff’s Proposal The PTRA resulted from CMP’s adoption of the Power Tax software and the discovery that CMP’s previous accounting methods had caused tax benefits to be inadvertently flowed through to customers. Because of the complexity of the subject matter and the confusion surrounding CMP’s changing recovery request, the Commission hired the accounting firm of BerryDunn to review the Company’s PTRA. BerryDunn’s Power Tax Adjustment Report (PT Report) was filed on January 12, 2018 Bench Analysis 19 Docket No. 2018-00194 in Docket No. 2016-00035. Based on its review of the information that it had obtained from CMP, BerryDunn concluded that: Based on the information provided by CMP and the procedures performed by Berry Dunn, we were not able to determine whether the regulatory asset calculated by CMP was reasonable. This is primarily due to the fact that we were not able to validate the acquisition value used in the calculation of the regulatory asset. CMP’s Tax Panel notes that further developments have happened since the audit findings, specifically that CMP re-examined the $790 million acquisition value that BerryDunn had questioned and has developed a reconciliation between that and the FERC and tax values. At the time of the rate filing, CMP had not yet provided this information to BerryDunn. Tax Dir. at 14. Since the Company filed its case, BerryDunn advised the Staff that it will no longer be providing service to the Commission in this issue. Rev. Req. Ex. 1. The Staff is in the process of retaining another accounting firm to assess the additional information that CMP plans to submit regarding validation of the plant acquisition value. This has not yet been accomplished, however. Therefore, Staff cannot confirm the PTRA calculation at this time. In addition to the acquisition value issue, the PT Report also noted that two items CMP had included as part of the Power Tax Adjustment were unrelated to the adoption of the Power Tax software. First, during the 2014 rate case, CMP should have corrected the book depreciation expense used in the calculation of deferred income taxes to reflect the book depreciation rates agreed to and reflected in the final rates. CMP included an adjustment to correct this error as a separate line in the calculation of Bench Analysis 20 Docket No. 2018-00194 the Power Tax adjustment. Second, CMP assumed that it had been collecting 2,028,599 annually in rates for the deferred income taxes but, as a result of the review done in response to the Power Tax Audit CMP noted that it was only collecting $1,521,449 due to the averaging of two years in the 2014 rate case. CMP also included amounts to correct this error as part of the Power Tax Adjustment. PT Report at 8. CMP’s filing included only one total number for the Power Tax Adjustment. Based upon the PT Report, it appears that the Power Tax Adjustment includes at least three components, with only one being related to the implementation of the Power Tax software. In its rebuttal testimony, CMP should provide a breakdown of the Power Tax Adjustment into at least four categories: the adoption of the Power Tax software; the use of the incorrect depreciation rates when calculating deferred taxes in the 2014 rate proceeding; the error resulting from the incorrect assumptions of what had been previously allowed in rates in the 2014 rate proceeding; and any carrying costs associated with each of the components. CMP should also fully explain why it considers all of these items to be part of the Power Tax Adjustment and why the IRS’s Private Letter Ruling of October 15, 2015 is applicable to errors not related to the adoption of the Power Tax software. Staff does not believe that it is appropriate to include any amounts in rates until a conclusion is reached that the Power Tax adjustment proposed is reasonable and related to the Commission’s prior approval of the Power Tax accounting order. Therefore, at this time, the Staff recommends that the PTRA be removed from rate base and that the $564,000 amortization for the PTRA also be excluded from revenue requirements. Bench Analysis C. 21 Docket No. 2018-00194 Expense 1. Payroll a. CMP’s Position As noted in Table 1 above, CMP attributes $7.1 million of its $22.9 million shortfall to an increase in payroll expenses. CMP’s calculation of its rate year distribution O&M payroll projection of $46,975,000 is set forth below: Table 5 6 CMP Rate Year Payroll Projection 1. Total Payroll 2. Percent Allocated to Distribution 3. Distribution Payroll 4. Percent Allocated to O & M 5. Distribution O & M Payroll 6. Overtime 7. Total Payroll 8. Payroll Inflator Test Year to Rate Year $ $ 59,082 79.91% 47,213 $ 73.85% 34,866 $ 5,615 40,481 105.19% 9. Inflated Payroll 10. Plus Impact of Staffing Changes 11. Plus Variable Compensation $ 42,584 1,896 2,495 12. Total Distribution Payroll $ 46,975 CMP notes that its rate year payroll forecast begins with a base staffing level of 808 full time equivalent (FTE) employees as of June 30, 2018. CMP adjusts the test year number based on an assumption that it will fill vacancies by December 2018 and 6 Source: Rev. Req. Dir. at 17, Table 10. Bench Analysis 22 Docket No. 2018-00194 that it will add additional employees during the 6/30/2018 through 6/30/2020 time period so that its total workforce at the end of 6/30/2020 is 883 FTEs. b. Staff’s Position i. Percent Allocated to Distribution Based on the proposed adjustment to the transmission/distribution allocators discussed in Section II.A., the staff proposes to adjust the Percent Allocated to Distribution from 79.91% to 79.54%. This reduces the base payroll distribution amount to $46,994,000. ii. Percent Allocated to O & M Payroll costs are either capitalized to plant or expensed to O&M. In its test year, CMP allocated 73.85% of payroll to O&M. Rev. Req. Dir. RRP 3-9 Schedule A. This percentage was based upon the percentage of expensed historical test year payroll dollars over the total payroll costs for the same period. EXM-003-046, Att. 1. The historic percentage of payroll expensed for the period 2013 through 2017 ranged from a low of 67.76% to a high of 72.66%, all lower than the percentage used to calculate the rate year costs. In addition, except for the historical test year ending June 30, 2018, the percentage of payroll expensed has continually decreased since 2014. EXM-003-046, Att. 2. For costs that fluctuate from year to year, a normalized amount based on an average of actuals should be used to calculate rate year expense rather than using a single test year. The average payroll expensed for the five-year period Bench Analysis 23 Docket No. 2018-00194 2013 to 2017 is 70.33%.7 Staff recommends that this normalized factor be used to calculate the O&M payroll for the rate year. Applying the normalized allocation factor to total distribution payroll reduces the base distribution O&M payroll expense to $33,051,000. iii. Impact of Staffing Changes From year-end 2013 to June 2018, CMP’s employee headcount (HC) has gone from 913 to 808. A breakdown of this overall decrease of 105 employees by CMP Line of Business is presented in Table 6 below: Table 6 8 Line of Business December-13 HC Total Salary June-18 HC Total Salary 1 Asset Management & Planning 2 Customer Service 3 Electric T & D 4 Energy Supply/Services 5 Engineering & Delivery 6 General Services 7 Operations Technologies & Business Trans 8 Process & Technology 9 Projects 10 Smart Grids 11 Corporate Functions 62 239 449 0 28 46 17 0 0 0 72 $ 4,392,577 $12,565,625 $29,096,299 $ 0 $ 2,319,961 $ 2,467,463 $ 1,090,792 $ 0 $ 0 $ 0 $ 5,670,902 24 110 518 7 0 23 0 36 25 36 29 12 Grand Total 913 $ 57,603,622 808 $ 2,125,182 $ 6,232,118 $37,788,804 $ 558,022 $ 0 $ 1,609,777 $ 0 $ 2,863,052 $ 2,387,678 $ 2,875,475 $ 2,641,998 $ 59,082,111 As Table 6 shows, one of the most dramatic changes in staffing levels has occurred in the Customer Service function, which decreased from 239 positions in 2013 7 Staff did not include the historic test year ending June 30, 2018, in this calculation as it includes a period already represented in the historic calendar years. 8 Source: EXM-003-056, Att. 1. Bench Analysis 24 Docket No. 2018-00194 to 110 positions in June 2018, with the vast majority of the decrease (116 positions) occurring in the 2017 to 2018 time period. CMP explained that a significant part of the decrease, was a result of transfers to positions to the Electric T&D category. Table 7 provides a breakdown of the 116 position decrease in the customer service headcount: Table 7 9 TRANSFERS Transfer to Electric T&D Transfer to Smart Grids Transfer to HR Transfer to Projects -84 -2 -1 -1 TURNOVER Voluntary Involuntary Retirements Hires Sum of Actual Changes -11 0 -18 1 -116 In other words, out of the total 116 position decrease, there was a real reduction of 28 positions during the 2017 to 2018 time period. The Company projects it will fill 12 customer service vacancies between June 30, 2018 and December 30, 2018, and it will add another 36 positions between December 30, 2018, and June 30, 2019. In addition, from June 30, 2019 to June 30, 2020 the Company projects that it will decrease the Customer Service staffing level to 145 by eliminating positions that will have been added to catch up for the backlog of work. As discussed in Section III, the Company’s customer service levels have been, and remain, below reasonable levels. Much of the decline in customer service levels 9 Source: ODR-001-017. Bench Analysis 25 Docket No. 2018-00194 can be traced back to insufficient customer service staffing. See Section III.E. The Staff believes that the revenue requirement set in this case should reflect the customer service staffing levels CMP needs in order to provide reasonable service. The amount, however, should not include amounts needed to catch up from backlogs CMP created as a result of imprudent management. In addition, the Company’s assumption that it will completely fill all of its current vacancies and have no vacancies at all in the rate effective period does not seem reasonable. The Staff proposes that the adjustment for the incremental customer service staff to be added in the rate year be based on the number of employees (28) needed to get back to pre-reduction levels as set forth in ODR-001-017. Finally, the Staff does not believe the Company’s assumption that all customer service vacancies and new positions will be compensated at current average compensation levels, which include supervisory salaries, is reasonable. EXM-003-055. Instead, the Staff recommends that the additional compensation for the new employees be based on the actual average salaries of the new employees hired, not that of existing and upper-level employees. ODR-001-039. Based on these adjustments, Staff recommends that the Company’s Proposed Staffing Changes adjustment be reduced to $1,078,000 from the Company’s proposed $1,896,000. iv. Bonus Compensation A detailed description of the Company’s proposal for bonus compensation (which the Company calls “variable” compensation) and the Staff’s analysis of CMP’s proposal Bench Analysis 26 Docket No. 2018-00194 is provided in Appendix A (confidential and public/redacted versions) to this Bench Analysis. v. Summary of Adjustments Applying the proposed adjustments discussed above reduces the Company’s Rate Year Payroll Projection from $46,975,000 to $43,998,000. The Staff’s Revised Rate Year Payroll calculation is presented in Table 8 below: Table 8 (dollars in thousands) Rate Year Payroll Projection 1. Total Payroll 2. Percent Allocated to Distribution 3. Distribution Payroll 4. Percent Allocated to O & M 5. Distribution O & M Payroll $ $ 59,082 79.54% 46,993 $ 70.33% 33,050 $ $ 5,615 38,665 105.19% 9. Inflated Payroll 10. Plum Impact of Staffing Changes 11. Plus Variable Compensation $ 40,673 1,079 2,246 12. Total Distribution Payroll $ 43,998 6. Overtime 7. Total Payroll 8. Payroll Inflator HTY to RY 2. Rate Case Expense a. CMP’s Position Bench Analysis 27 Docket No. 2018-00194 CMP has estimated rate case expenses for the current case of $2.135 million and is proposing to recover these costs over a one-year period. CMP provided the following breakdown of its rate case expenses: Table 9: CMP’s Estimate of Rate Case Expenses10 (Dollars in Thousands) Consultancy Costs ROE & Capital Structure $ 150 Pension and OPEB Studies 200 Policy 200 Taxes 200 Rate Design 250 Shared Services 250 Subtotal Consultancy Costs $1,250 Legal Costs $ Total Rate Case Expense $2,135 885 As the above table shows, CMP includes consultant costs in six areas. When asked to provide copies of all contracts with those consultants, CMP provided copies of only two contracts—one for ROE & Capital Structure and one for Pension and OPEB Studies. EXM-003-087. At the December 4, 2018 technical conference, the Staff asked CMP if there were other contracts that it had not filed. In response, Company witness Peter Cohen stated that “there are no contracts that you didn’t get. The nature of this 10 Source: Rev. Req. Dir. at 24, Table 19. Bench Analysis 28 Docket No. 2018-00194 case resulted in a time period in which we didn’t have all the contracts lined up in order to respond. We’re working on it.” Tr. 85 (Dec. 4, 2018). b. Staff’s Analysis and Proposals Staff proposes several adjustments to CMP’s rate case expense and has additional concerns related to certain of the consultants’ costs that CMP has included in its rate case expenses. First, CMP has not provided any detail to support the $200,000 estimate for consulting expenses related to its policy panel. In EXM-003-086, Staff requested an explanation for the $200,000 included for consulting expense for Policy since the Policy Panel Testimony did not include a consultant. CMP stated in its response that the consulting expense associated with the Policy Panel is attributed to Northbridge Group (“Northbridge”), whose services are strategic and advisory in nature. CMP has not provided a copy of any contract with Northbridge or other details about CMP’s expectations regarding Northbridge’s services. CMP’s response that Northbridge provided services that are “strategic and advisory in nature” is vague and provides little support for this expense, especially considering the fact that the Policy Panel Testimony was presented by two very senior officials, one at Avangrid and one at CMP, with extensive regulatory experience. The Staff recommends that no portion of the $200,000 for this item to be allowed in revenue requirements at this time for this expense because CMP has not supported this cost. The Staff also has concerns related to the Pension and OPEB consultant costs included as rate case expenses. In EXM-003-095, the Staff asked for the names of Bench Analysis 29 Docket No. 2018-00194 consultants CMP used for the Pension and OPEB studies and for CMP to identify the tasks that would otherwise be done in the ordinary course of business for Pension and OPEB studies. In response, CMP noted that it engaged with Aon plc (Aon) and New England Pension Consultants (NEPC) to support this rate case. Aon was engaged to prepare pension, OPEB, and 401(k) projections for the period covering 2018–2023 and NEPC was engaged to provide expected long-term risk and return assumptions over a 5- to 7-year and a 30-year period. EXM-003-085. CMP provided a copy of the contract with Aon but did not provide a contract with NEPC. The Aon contract provided that CMP would pay Aon a fee that is less than half of the amount CMP estimated for the Pension and OPEB Studies portion of its rate case expense. EXM-003-087 Att. 2 (Confidential) at 2. The tasks CMP described in its response to EXM-003-085 and within the contract with Aon appear to be very similar to the tasks that would be necessary to determine the Pension and OPEB costs to be recorded in accordance with standard accounting policies. Therefore, it is possible that CMP is requesting recovery of these costs twice—once as part of its normal outside services expense and once as part of rate case expenses. CMP should provide in its rebuttal testimony a better explanation of how determining the projected healthcare and pension expenses for future periods in this rate proceeding is different than what is done to determine annual healthcare and pension expense. In addition, CMP should also provide a copy of the contract with NEPC and provide support for the estimated rate case expenses not reflected in the Aon contract. As CMP has provided support for approximately 50% of the Pension and Bench Analysis 30 Docket No. 2018-00194 OPEB consultant costs, the Staff recommends that only $100,000 of the estimated $200,000 be included as part of the rate case expenses at this time. CMP also has not provided any detail or contracts related to the services to be provided for Taxes, Rate Design, or Shared Services. 11 While the Staff does not recommend at this point that the revenue requirement be adjusted to remove these costs, the Staff may do so if CMP does not provide further support for the estimated costs. Such support would include a more detailed description of the services provided; copies of the contracts detailing the services to be performed; support for the proposition that the services are being performed specifically for regulatory purposes; and the estimated total cost of the contract. Lastly, CMP has proposed to recover the rate case expenses over a one-year period. The recovery of rate case expenses is governed by Chapter 85 of the Commission’s Rules, which states in part: No public utility shall recover from its ratepayers any regulatory proceeding expense unless such expense has been found by the Commission to have been reasonable. The Commission will set regulatory proceeding expenses on a normalized test year basis. Ch. 85, § 3(A). CMP’s proposal to recover the rate case expenses over a one-year period effectively assumes that these costs will recur every year, given that normalization adjusts amounts to the level that would be incurred during a normal year. 11 Staff assumes that the Shared Services cost is related to the affiliated services review and does not represent rate case services performed by the Avangrid Service Company. Chapter 85 allows for recovery as a regulatory proceeding expense only those services not provided by the utilities’ own employees; Staff considers Service Company costs on an equal footing with services provided by direct employees of CMP. Bench Analysis 31 Docket No. 2018-00194 It has been approximately five years since CMP’s last full rate case and the current filing was at the direction of the Commission. Therefore, the Staff recommends, in keeping with the requirements of Chapter 85, that rate case expenses be normalized over four years, which is more representative of the typical duration between rate cases than the one-year period CMP used. In summary, Staff recommends removing $300,000 from CMP’s rate case expenses for the consultant costs for Policy and for Pension and OPEB Studies. This results in an estimated rate case expense of $1.835 million. When normalized over a four-year period, the annual revenue requirement impact of this total expense is $458,750. This is a reduction of $1,676,250 from what CMP proposed to include in its rate year revenue requirement. 3. Medical Expense The Company estimates that its medical expenses will increase by $867,000 from the test year to the rate year. As part of this overall projection, the Company projected administrative costs growing from $467,000 in 2018 to $600,000 in 2020. Rev. Req. Dir. at 22, Table 16. As a general matter, Staff believes that administrative medical costs should increase at the rate of inflation. The Company explained that the increase to administrative costs in this case was based on certain known and measurable changes. ODR-001-052. As part of its Rebuttal case, the Company should provide support for such changes. Bench Analysis 4. 32 Docket No. 2018-00194 Application of Overall Inflation Rate a. CMP’s Position As a “default” inflator for cost and revenue projections where a specific forecast was not used, CMP calculated the projected increase in the Gross Domestic Product Price Index (GDP-PI) using confidential source data and projections from IHS Markit U.S. Economic Outlook—July 2018, adjusted to remove the effects of healthcare’s impact on GDP. Using these inflation estimates for 2018-2020, the Company then develops a compounded inflator for the rate year of 1.052252. Rev. Req. Dir. at 16, Table 9. b. Staff’s Position Staff has updated the general inflation rate to reflect more recent data and projections and relied on publicly available projections of GDP-PI, specifically The Budget and Economic Outlook: 2019 to 2029 (CBO Outlook) published by the Congressional Budget Office (CBO) in January 2019. The CBO’s GDP-PI projections are 2.2% (estimated) for 2018, 2.1% for 2019, and 2.0% for 2020. CBO Outlook, Table E-1. Using the same methodologies as CMP, Staff derives a compounded inflator for the rate year of 1.042553. Applying this revised inflation rate to CMP’s inflation-based costs reduces the rate year inflation adjustment by approximately $330,000. 5. Other O&M – Injuries and Damages During the test year, CMP recorded $500,000 in FERC Account 925, Injuries and Damages, related to “Class Action Complaint” to establish a reserve for customer Bench Analysis 33 Docket No. 2018-00194 underground service drop liabilities. EXM-003-021, Att. 1. CMP confirmed that this amount was included as part of its revenue requirements. Tr. 152 (Dec. 3, 2018). At the December technical conference, CMP agreed that this was likely a non-recurring item and should not be included in revenue requirements. Id. at 153. As a non-recurring item, Staff recommends that this item be removed from other O&M costs. CMP allocated these costs between transmission and distribution using the distribution wage allocator of 75.69%, so the test year should be reduced by $378,450 ($500,000 × 75.69%). In addition, to the extent that CMP’s rate model applies an inflation adjuster to a base amount that includes these costs, the related adjustment should also be reduced accordingly. 6. Depreciation Expense and Tax Basis Repair Allowance CMP has included depreciation expense of $46,306,000 and $48,587,000 in the interim year and rate year, respectively. Rev. Req. Dir. at 31. CMP calculated these amounts based upon the plant balances estimated for the rate year using either a total average depreciate rate for plant additions derived from the attrition analysis or a depreciate rate of 2.32% for the Company’s resiliency program. Based on Staff’s proposed changes to the plant balances used to calculate rate base, depreciation expense should be adjusted accordingly. Similarly, CMP calculated the Tax Basis Repair Allowance based upon estimated plant additions of $90.5 million, resulting in an estimated deduction of $30.7 million. Tax Dir. at 9. Similarly, to the extent plant additions are adjusted during the course of this Bench Analysis 34 Docket No. 2018-00194 proceeding, the tax repair allowance reflected in the calculation of the revenue requirements should also be adjusted accordingly. 7. Affiliate Service Charges a. CMP’s Position Another major driver behind the Company’s claimed $22.9 million deficiency is the $5.4 million dollar increase in distribution expense associated with affiliate service costs. In particular, the Company proposes removing the current cap on charges to CMP by its affiliates, primarily Avangrid Services Corporation (ASC). The current annual cap on service company charges to Maine utilities is $32.5 million, with CMP’s share of the cap set at $31.4 million. According to the Company, during the test year, CMP should have been charged $42.8 million but, due to the cap, CMP was charged only $31.4 million, of which $16.2 million was charged to distribution and $15.2 million was charged to transmission. Rev. Req. Dir. at 13. At the full $42.8 million level of costs, $21.5 million would be charged to distribution under CMP’s proposal. Id. CMP has included this full amount in its revenue requirement calculations. In accordance with the Company’s commitments made in the Stipulation in Docket No. 2013-00168, CMP also notes that it has retained Thomas Flaherty from PwC to perform a market study to demonstrate the justness and reasonableness of these charges. Policy Dir. at 11. CMP also requests that the Commission eliminate the annual cap on service company charges. CMP notes that none of the jurisdictions where CMP’s affiliate utility companies operate have placed a cap on affiliate charges and that the practice of Bench Analysis 35 Docket No. 2018-00194 receiving services from a common service company has become commonplace within the utility industry. Id. at 12. For the reasons discussed below, the Staff does not support the inclusion of the additional $5.4 million for affiliate service charges proposed by the Company nor does it support the elimination of the annual cap on affiliate service company charges. b. Staff’s Proposal The issue of a cap on affiliate service charges to CMP and its relationship to market rate cost information has a long history which originates in 2001. Central Maine Power Co. et. al., Request for Approval of Affiliated Interest Transaction for Two Service Agreements with Energy East Management Corporation, Docket No. 2001-00178, Order Approving Stipulation (July 10, 2001) (hereinafter “Docket No. 2001-00178”). In that case, CMP sought approval of service agreements with its affiliate Energy East Management Corporation, a subsidiary of CMP’s parent corporation at the time. Pursuant to those agreements, affiliate charges would be made using fully distributed costs rather than market rates, thus requiring a waiver of the provisions of Chapter 820 of the Commission’s Rules. Under the Stipulation in that case, the Company and the Office of the Public Advocate agreed to the waiver of Chapter 820’s market rate requirement, subject to a cap on the amount that could be charged to CMP, initially at an amount not to exceed $7 million but which could be increased through a notice filing to $10 million. The Stipulation went on to provide that: For ratemaking purposes, each of the applicants will provide appropriate market information (which shall mean market rates for such services or, of the applicants conclude that no market rates are available, the explanation supporting the unavailability of market rates) to demonstrate that the costs Bench Analysis 36 Docket No. 2018-00194 billed under these agreements are just and reasonable. Such market information shall only be required if and to the extent that an applicant is seeking (or another party is requesting) a rate change (whether in a general rate proceeding, pursuant to a bottom-end earnings sharing mechanism, or as a result of a mandated cost) that includes costs billed under the agreements approved herein. In such a proceeding seeking a rate change, any other party is free to contest the reasonableness of the costs incurred under the agreements approved herein and the applicant seeking to include such costs in its rate change shall have the burden of proof as to the reasonableness of such costs. Id. at 4–5. In approving the Stipulation, the Commission concluded that a waiver of Chapter 820’s requirements for market based pricing was appropriate for several reasons. First, Energy East’s fully distributed cost methodology was created under SEC requirements and subject to SEC scrutiny. Second, CMP was at the time under an ARP and, thus, CMP had a direct incentive to minimize costs. However, because rates may be influenced by affiliate transaction costs, CMP was required to provide market prices for services or a specific explanation of why market prices cannot be provided. Finally, the stipulation provided a cap on the amount that a utility could be billed for affiliate charges in any one calendar year without further Commission approval. Id. at 6. Despite there being several rate cases in the intervening years, CMP has yet to file the market rate information called for in Docket No. 2011-00178. In CMP’s last rate case, Docket No. 2013-00168, the Company submitted a benchmark cost comparison which was opposed by the Commission Staff as being inadequate and not in compliance with the requirements of providing a market rate study. The Stipulation that resolved that case provided that the charges to CMP by IUMC, its service company affiliate at the time, were reasonable and should be allowed in rates. However, the Bench Analysis 37 Docket No. 2018-00194 parties also agreed that CMP, the Commission Staff, and other interested parties would engage in a collaborative process “to refine what information CMP must provide in future ratemaking proceedings to demonstrate the justness and reasonableness of any charges.” Docket No. 2013-00168, supra., Stipulation at 960. Despite Staff’s frequent suggestions as to what was needed to conduct the market rate study required by Docket No. 2001-00178, the collaborative process went nowhere. The Company’s filing in this case did not contain a study but instead suggested a collaborative effort to identity the functions to be studied—which was the exact thing the Staff attempted to do with the Company years ago. As the transcripts from the market rate study scoping sessions in this case indicate, progress on developing the parameters of the study has been extremely slow and CMP has yet to provide a proposed list of functions to be studied; four months following its initial filing in this case; seven months following the Commission’s Order initiating this case; and four and a half years following the Stipulation in Docket No. 2013-00168. Even if PwC’s market rate study is completed in sufficient time to be considered in this case, Mr. Flaherty has acknowledged that the study will only answer the question of whether the charges to CMP by ASC are reasonable when compared to market rates. It will not answer the question of whether those services are cost effective when compared to the utility itself performing such services. The lack of such analysis in the anticipated PwC Study or in CMP’s filing in this case makes it difficult if not impossible to judge the reasonableness of CMP’s requested increase in affiliate service charges. The Commission and the parties have been disadvantaged by CMP’s failure to request an increase in the affiliate service charge cap prior to accepting a level of services Bench Analysis 38 Docket No. 2018-00194 which, when billed, will be approximately $10 million more than the existing cap and $31 million more than the cap imposed when CMP’s Chapter 820 waiver request was first granted. In summary, the Staff does not believe that CMP has met its burden of proof to demonstrate the reasonableness of the $5,376,000 of distribution-related charges that exceed the current cap. See Central Maine Power Co., et. al., Petition to Increase the Annual Dollar Limit for the Iberdrola USA Management Corporation Support Services Agreement with Certain Iberdrola Affiliates, Docket No. 2012-00530, Order Approving Stipulation (July 2, 2013). Furthermore, the Staff does not support the elimination of the affiliate service charges cap. The cap provides a useful tool in monitoring the level and reasonableness of the affiliate service costs CMP incurs. D. Cost of Capital 1. CMP Request and Testimony CMP’s cost of equity witness, Ann Bulkley, develops a current cost of equity in the range of 10.0% to 10.5% and recommends a return on equity (ROE) of 10.3%. In its filing, however, the Company requests a 10.0% ROE combined with a common equity ratio of 55%. In developing her recommendation, Ms. Bulkley begins with the 49 domestic U.S. utilities classified by Value Line as Electric Utilities and Natural Gas Distribution Companies and applies screening criteria to establish a risk-comparable proxy group of 21 publicly traded companies. Using market information for this proxy group, she then Bench Analysis 39 Docket No. 2018-00194 estimates CMP’s ROE, using two Constant Growth Discounted Cash Flow (DCF) structures, one which uses current dividends and stock prices and one which uses Value Line projected dividends and stock prices. In addition, she uses two risk premium approaches, the Capital Asset Pricing Model (CAPM) and a Bond Yield Risk Premium model. Ms. Bulkley also recommends a common equity ratio of 55% for CMP, citing the Company’s historical year-end equity ratio over the past five years and provides a comparison to the authorized equity ratios in other jurisdictions for the utilities owned by the companies included in her proxy group. Additionally, she refers to the concerns raised by credit rating agencies as a result of the Tax Cuts and Jobs Act of 2017 as support for a higher equity ratio than CMP may have relied on in prior rate cases. The Company computes a pre-tax weighted average cost of capital (WACC) of 9.62% and an after-tax WACC of 7.48%, using a common equity ratio of 55.00%, longterm debt equal to 42.16% of total capital, short-term debt at 2.82% of total capital and preferred stock at 0.02% of capital, each component at the cost as shown in Table 10 and with the equity returns grossed up to reflect the Company’s current combined federal and state income tax rate. Bench Analysis 40 Docket No. 2018-00194 Table 10 CMP As Filed Capital Structure, Costs and ROE Capitalization Percentage Common Equity Preferred Stock Long Term Debt Short Term Debt 55.00% 0.02% 42.16% 2.82% 100.00% Total 2. Cost 10.00% 6.00% 4.45% 3.50% After -Tax Weighted Pre-Tax Weighted Cost Cost 5.50% 0.00% 1.88% 7.64% 0.00% 1.88% 0.10% 0.10% 7.48% 9.62% Hope-Bluefield Standard Two United States Supreme Court decisions of more than 70 years ago, known as the Bluefield and Hope cases, provide the standards for measuring the reasonableness of a utility’s allowed ROE. Taken together, the Hope-Bluefield decisions establish that: A public utility is entitled to such rates as will permit it to earn a return on the value of the property which it employs for the convenience of the public equal to that generally being made…on investments in other business undertakings which are attended by corresponding risks and uncertainties…The return should be reasonably sufficient to assure confidence in the financial soundness of the utility, and should be adequate, under efficient and economical management, to maintain and support its credit and enable it to raise the money necessary for the proper discharge of its public duties... Bluefield Water Works and Improvement Co. v. Public Service Commission of West Virginia, 262 U.S. 679 (1923). Additionally, the idea of associating the allowed return to a common equity owner with those available from other companies of comparable risk was established in the Hope decision: Bench Analysis 41 Docket No. 2018-00194 [T]he return to the equity owner should be commensurate with the return on investment in other enterprises having corresponding risks. That return, moreover, should be sufficient to assure confidence in the financial integrity of the enterprise, so as to maintain its credit and to attract capital. Federal Power Commission v. Hope Natural Gas Co., 320 U.S. 591 (1944). Thus, determining an appropriate ROE for a regulated utility is one that involves determining a market-based cost of equity. For a company that is not publicly traded such as CMP, the cost of equity is determined to be the return investors expect from alternative investments that present no more and no less risk. In practice, estimating the cost of equity involves developing a comparable group of companies, for which market-based information is available, that are in lines of business that present similar financial risks, and using economic and financial models to set an appropriate ROE. 3. Proxy Group Selection In her testimony, Ms. Bulkley followed a customary approach for selecting a proxy group of publicly traded utilities which are representative of the risks and prospects faced by CMP. Beginning with a group of 49 domestic utilities classified by Value Line as Electric Utilities and Natural Gas Distribution Companies, she then applied screening criteria to include only companies which: a. Consistently pay quarterly cash dividends; b. Have investment grade, long-term issuer ratings from both S&P and Moody’s; c. Have positive long-term earnings growth forecasts from at least two utility industry equity analysts; d. “Owned generation” comprises less than 60.0% of the Company’s MWh sales to ultimate customers; Bench Analysis 42 Docket No. 2018-00194 e. Derive more than 70.0% of their total operating income from regulated operations f. Were not parties to a merger or transformative transaction during the analytical periods relied on. Staff is generally in agreement with the initial group of electric and gas utilities and the screening criteria used by Ms. Bulkley. The selection of proxy members based on publicly-traded companies that pay dividends is essential to a discounted cash flow (DCF) analysis. In addition, companies that are covered by more than one analyst and have an investment grade credit rating provide reasonable assurance that the marketbased analysis that underlies a return on equity determination reflects market information. Eliminating companies that have been a party to a recent merger transaction helps ensure that the ROE range determined based on the proxy group is not unduly influenced by significant events that affect an individual member of the group and is consistent with the criteria previously employed by the Commission in selecting an appropriate proxy group. While generally in agreement with the screening criteria chosen, Staff does not agree with two of the Company’s screens, specifically the “owned generation” screen and the requirement that the utility derive no less than 70% of its total operating income from regulated operations. As explained at the December 3, 2018 technical conference, Ms. Bulkley employed the “owned generation” screen as a way to “[N]arrow the proxy group to be more comparable to the subject company which does not have any generation. And so we’re establishing the generation screening criteria to eliminate companies that have a substantial amount of generation in their portfolios.” Tr. 18 (Dec. 3, 2019). Using FERC Form 1 data from 2015-2017, Ms. Bulkley calculated the megawatt-hours generated by Bench Analysis 43 Docket No. 2018-00194 a regulated company as a percentage of total load served by that same company and eliminated any company for which that percentage was greater than 60%. Without further examination of the specific operations of the utilities in the eliminated group and the underlying data, Staff does not find this screen to be particularly enlightening for purposes of arriving at a risk-comparable proxy group of companies that have lower levels of generation risk. Specifically, the calculation done by Ms. Bulkley relies on summary information from the FERC Form 1 related to total MWh generated by a company and total MWh of load served by that company. There appears to be no additional examination to answer questions such as: the disposition of energy not used to serve load; the commercial terms by which the company acquired the balance of the energy needed to serve load; or the significance of a 60% cut-off as it relates to generation risk. Consequently, Staff does not employ this screening criteria. Ms. Bulkley’s final screening criterion eliminates companies which derive less than 70% of their total operating income from regulated activities. Staff has concerns about using a 70% threshold for regulated operating income. The application of a 70% regulated operating income screen has the effect of including companies that derive a significant portion (up to 30%) of their operating income from non-regulated or competitive business enterprises that do not bear similar risk profiles to CMP. In past rate cases, Staff has expressed a preference for a higher threshold of 90% for this screen, but has used a lower threshold because of concerns about having too small a proxy group. Emera Maine, Request for Approval of Proposed Rate Increase, Docket No. 2017-00198, Bench Analysis at 64 (Dec. 21, 2017). Here, the use of a higher 90% threshold does not result in an unreasonably small proxy group and, thus, Staff has Bench Analysis 44 Docket No. 2018-00194 decided to employ that threshold. Finally, based on recent merger activity, two companies that would otherwise be included in the proxy group should be removed. On February 12, 2019, American Electric Power Corp (AEP) signed a purchase agreement through a subsidiary to acquire Sempra Renewables and its 724 MW of wind generation for approximately $1 billion. On January 17, 2019, Duke Energy (DUK) proposed purchasing South Carolinaowned utility Santee Cooper. With these adjustments, Staff’s final proxy group includes 22 companies, somewhat more than half of which were also in Ms. Bulkley’s proxy group, as shown in Table 11. Bench Analysis 45 Docket No. 2018-00194 Table 11 Staff Final Proxy Group Alliant Energy Corporation Ameren Corporation Atmos Energy Corporation CMS Energy Corporation Consolidated Edison, Inc. DTE Energy Company Edison International El Paso Electric Company Evergy, Inc. Eversource Energy IDACORP, Inc. New Jersey Resources Corporation NiSource, Inc. Northwest Natural Gas Company NorthWestern Corporation ONE Gas, Inc. PNM Resources, Inc. Portland General Electric Company PPL Corporation Public Service Enterprise Group, Inc Spire, Inc. Xcel Energy Inc. 4. Bulkley Final Proxy Group ALLETE, Inc. Alliant Energy Corporation Atmos Energy Corporation Black Hills Corporation CMS Energy Corporation Consolidated Edison, Inc. Edison International Eversource Energy FirstEnergy Corporation Hawaiian Electric Industries, Inc. New Jersey Resources Corporation NorthWestern Corporation ONE Gas, Inc. Otter Tail Corporation Portland General Electric Company PPL Corporation Public Service Enterprise Group Inc. Sempra Energy South Jersey Industries, Inc. Southwest Gas Corporation Spire, Inc. Constant Growth DCF Model for Estimating Cost of Equity Consistent with past Commission practice and orders,12 the Staff first employs a 12 See, e.g. Central Maine Power Company, Proposed Increase In Rates, Docket No. 92-345, Order at 31 (Dec. 14, 1993). Bench Analysis 46 Docket No. 2018-00194 discounted cash flow (DCF) approach to the cost of equity analysis. The DCF model is commonly used for estimating the cost of common equity for public utilities and is based on the financial theory that the value or price of any security is the discounted present value of all future cash flows. As explained in materials published by the Society of Utility and Regulatory Financial Analysts.13 The DCF model is based upon two fundamental principles. First, DCF is based on the postulate that investors value an asset on the basis of the future cash flows (i.e., dividends and ultimate sales in the case of common stocks) they expect to receive from owning the asset. The second DCF principle is that investors value a dollar received in the future less than a dollar received today (i.e., the “time value of money”). Within their context, the current price of a company’s stock is equal to the present value equivalent of the expected dividends and the proceeds from eventually selling the stock. The discount rate that equates the future anticipated dividends and the future anticipated selling price with the current market price is the cost of common equity. In its very simplest form, a DCF estimate of the cost of equity capital uses the formula K=D/P + g where: K= D/P = g= cost of equity capital dividend yield (dividend/stock price) long-term expected growth rate Generally, the market based data (market prices, current dividends and the resulting dividend yield) required to conduct any DCF analysis are readily available. As presented in her testimony, Ms. Bulkley performed a Constant Growth DCF Parcell, David C. The Cost of Capital—A Practitioner’s Guide, Society of Utility and Regulatory Financial Analysts, 2010 Edition. 13 Bench Analysis 47 Docket No. 2018-00194 analysis using actual market data as of August 31, 2018 and a Projected Constant Growth DCF analysis using projected market data from Value Line. ROE Test. at 43 and Exhibits AEB-5 and AEB-6. She eliminated any results lower than 7.00%14, noting that such a return would not provide a sufficient return to investors as compared to a utility bond return and citing a recent position established by the Minnesota Department of Commerce. ROE Dir. at 47 and EXM-006-009. She then calculated DCF mean low, mean and mean high results using, respectively, the minimum growth rate (i.e., the lowest of the First Call, Zacks, and Value Line earnings growth rates), the average growth rate from all three sources and the highest of the growth rates for each of the proxy group companies.15 Staff, as a rule, has not eliminated market-derived DCF results based on a selected cut-off level. Rather, Staff has preferred to include all market-indicated results in calculating proxy group averages. Table 12 shows the DCF results as presented in the Company’s testimony. Based on this presentation, an indicated ROE would fall in the range of a low of 8.79% to a high of 10.94%. 14 All ROE estimates cited are prior to any adjustment for flotation costs. The flotation cost adjustment is discussed below. 15 Staff notes that the “high” growth rates for individual companies used by Ms. Bulkley to calculate the mean high results range from 3.38% to 12.20% and average 7.06% for the proxy group. The overall average growth rate used in Ms. Bulkley’s DCF analysis is 5.86%. Bench Analysis 48 Docket No. 2018-00194 Table 12 As Presented in Testimony Constant Growth DCF (includes flotation cost adj.) Mean Low Mean Mean High 30-Day Average 8.79% 9.56% 10.51% 90-Day Average 8.88% 9.65% 10.61% 180-Day Average 8.96% 9.73% 10.68% Constant Growth Average 8.88% 9.65% 10.60% Projected DCF (includes flotation cost adj.) 2021-2023 Projection Mean Low Mean Mean High 9.06% 9.86% 10.94% Within Ms. Bulkley’s workpapers were calculations more in-line with Staff’s preferences, which excludes the flotation cost adjustment discussed below, and which includes results of less than 7.00% in the calculation of averages. Exhibits AEB-5 and AEB-6. Using these alternative calculations, Ms. Bulkley’s DCF analysis indicates an ROE within the range of 8.16% to 10.84% as shown in Table 13. Bench Analysis 49 Docket No. 2018-00194 Table 13 Adjusted to Exclude Flotation Cost Adjustment and to Include Results less than 7.00% Mean Low Mean Mean High 30-Day Average 8.16% 9.18% 10.40% 90-Day Average 8.26% 9.28% 10.50% 180-Day Average 8.33% 9.36% 10.58% Constant Growth Average 8.25% 9.27% 10.49% Mean Low Mean Mean High 8.59% 9.62% 10.84% Projected DCF 2021-2023 Projection 5. Staff DCF Analysis-Constant Growth Model Staff also conducted a Constant Growth DCF analysis using historical market data for the proxy group described above. Staff did not include a DCF analysis based on projected data. As shown in Exhibit ROE-1, the actual DCF calculations are largely self-explanatory. The market data required includes the current dividend, the market price for the company’s common shares and an estimate of future growth. Staff used the current dividend for each utility as of February 13, 2019 as reported on Yahoo! Finance and converted it to a forward dividend based on the growth projection and assuming that future dividend increases will be evenly distributed over calendar quarters. Thus, the forward dividend for each company is equal to the current dividend increased by one-half of the growth rate for that company. Staff calculated the dividend yield component of the model by dividing the resulting forward dividend by the share price for each utility. Bench Analysis 50 Docket No. 2018-00194 In recognition of the day-to-day variability in closing share prices over the past several months, Staff did not include results based on a one-day closing market price for each of the proxy group companies. Instead, Staff employed a 50-day moving average of closing share price and a 200-day moving average of closing share price for each utility as reported by Yahoo! Finance on February 13, 2019. This resulted in a range of current dividend yield calculations for the entire proxy group, from a low of 2.30% to a high of 5.58%, with an average dividend yield for the proxy group of just under 3.30%. For the growth component, Staff used the five-year analysts’ growth projections, also as reported on Yahoo! Finance. The growth rate used by Staff ranged from 2.59% to 9.20% and averaged 5.29%. Staff then added the average growth rate to the low, mean and high dividend yields to achieve an indicated range of estimates of ROE. As shown in Table 14 below, Staff’s constant growth DCF model produces an indicated ROE range of 7.58% to 10.86%. Table 14 Constant Growth DCF Model Indicated ROE Indicated ROE Range Constant Growth DCF Indicated ROE Average Analyst Growth (YahooFinance) 6. Low Growth Dividend Rate Yield 5.29% Capital Asset Pricing Model 7.58% Group Average High Dividend Dividend Yield Yield 8.56% 10.86% Bench Analysis 51 Docket No. 2018-00194 As the Commission has previously recognized, results from an analysis using the Capital Asset Pricing Model (CAPM) provide a useful check on the DCF analysis. Central Maine Power Company, Proposed Increase In Rates, Docket No. 92-345, Order at 31 (Dec. 14, 1993).The CAPM is a risk premium approach to estimating ROE in which the cost of equity is the sum of the interest rate on a risk-free bond and a risk premium. The yield on Treasury securities is typically used for the risk-free rate and market or systematic risk is measured by a firm’s beta. As explained in Ms. Bulkley’s testimony, the CAPM “is a risk premium approach that estimates the cost of equity for a given security as a function of a risk-free return plus a risk premium to compensate investors for the non-diversifiable or ‘systematic’ risk of that security.” Bulkley Pref. Dir. at ROE-49. The general form of the CAPM is: K = Rf + β (Rm – Rf) where: Rf = Rm = Β= Rm – Rf = risk free rate return on market beta market risk premium In her CAPM analysis, Ms. Bulkley relied on three sources for the risk-free rate: the current 30-day average yield on 30-year U.S. Treasury bonds of 3.05%; the average projected 30-year U.S. Treasury bond yield for Q4 2018 through Q4 2019 of 3.50% as reported by Blue Chip Financial Forecasts; and the average projected 30-year U.S. Treasury bond yield for 2020 through 2024 of 4.20% as reported by Blue Chip Financial Forecasts. Id. at ROE-50. For the Beta coefficient, she used the average of the Value Line Betas for her proxy group of .676. She then derives an expected market return Bench Analysis 52 Docket No. 2018-00194 component of 15.25% based on the S&P 500 Index using the constant growth DCF formulation. She derives both the current dividend yield and the long-term growth factors for the S&P 500 as a whole by weighting the individual company dividend yield and long-term growth by the proportion of total market capitalization that each company represents. Ms. Bulkley’s CAPM calculations result in an indicated ROE range of 11.30% to 11.67%. Consistent with the Commission’s preference as indicated in Public Utilities Commission, Investigation of Central Maine Power Company’s Stranded Costs, Transmission and Distribution Utility Revenue Requirements, and Rate Design, Docket No. 97-580, Order (Mar. 19, 1999), Staff used a current Treasury rate rather than a forecast of interest rates as the risk-free component.16 Staff calculated the most recent 30-day average17 of the 30-year Treasury rate and used 3.04% as the risk-free rate. Staff then used a Beta of .633 based on the most recently available Value Line beta, adjusted to reflect the revisions to the proxy group and Ms.Bulkley’s market return component of 15.25%. To develop its CAPM range, Staff looks at the highest and lowest betas from within its proxy group, combined with the risk-free rate and market risk premium. The results of Staff’s CAPM analysis indicate an ROE range of 9.15% to 11.59% with an average of 10.77% as shown in Table 15 below. Public Utilities Commission, Investigation of Central Maine Power Company’s Stranded Costs, Transmission and Distribution Utility Revenue Requirements, and Rate Design, Docket No. 97-580, Order (Mar. 19, 1999). 16 17 The average was calculated based on the most recent 30 business days as of February 14, 2019 as reported in Federal Reserve Economic Data (FRED). Bench Analysis 53 Docket No. 2018-00194 Table 15 CAPM Results 7. Bond Yield Plus Risk Premium The Company also presented a Bond Yield Plus Risk Premium (BYRP) analysis in which Ms. Bulkley compares the difference between the allowed ROE and the Treasury yield from 750 electric utility rate cases from 1992 to August 2018. She develops an equity cost rate by regressing the authorized ROE on the 30-year Treasury yield in effect at the time of the rate case and adding the resulting risk premium to current and projected Treasury rates. Her results range from 9.83% to 10.36%. ROE Dir. at 56-57. Staff has reservations about the methodology used by Ms. Bulkley in which she compares authorized ROEs to market information. The analytical approach to determining an appropriate ROE range is based on obtaining and analyzing the information that results from the financial decisions of investors. Although the Commissions making the ROE decisions presumably had market-based analytical information in front of them, the Commission decisions are likely to reflect other specific information related to the utility and the rate case issues presented. Staff does not Bench Analysis 54 Docket No. 2018-00194 present a BYRP analysis. 8. Additional Considerations Several additional issues are raised in the Company’s testimony and Staff’s analysis that are worthy of Staff comment. a. Market conditions. Market expectations are at the core of utility cost of equity analysis and, because the DCF and CAPM analytical approaches are market-based, Staff relies on these measures to reflect the full range of market expectations. The consideration of whether market conditions remain in an anomalous state following the recent recession continues to be an issue in rate cases. Specifically, the question arises whether the Commission should make explicit recognition of, and adjustment for, an anticipated increase in long-term interest rates in establishing an appropriate ROE. In her testimony, Ms. Bulkley states that “the context for setting the authorized ROE for CMP should not be the low interest rate environment of the last few years.” ROE Dir. at 24. Rather, she urges the Commission to rely on projections in considering the analytical results, including projections of future interest rates by Blue Chip Financial Forecasts. Staff does not agree that specific adjustments to the ROE methodology are necessary. The expectation of higher interest rates in the future is well-recognized in the market-place, understood by investors making financial decisions and reflected in current market data. Adjusting the market-based analytical results by incorporating specific projections of interest rates introduces the not insignificant risk that those Bench Analysis 55 Docket No. 2018-00194 projections will be wrong. As Federal Reserve Research Advisor Michael Bauer stated in his 2017 article assessing current predictive tools “Interest rates are inherently difficult to predict, and the simple random walk benchmark has proven hard to beat.”18 As further support for the inaccuracy of interest rate forecasts, in July 2015 Business Insider provided the information and chart shown in Figure 1 which compares forecaster’s predictions of the 10 Year US Treasury Yield and actual yield between 2010 and 2015. The chart shows that forecasters consistently expected to see interest rate increases in excess of what actually occurred.19 “Bridging the Gap: Forecasting Interest Rates with Macro Trends”, Michael D. Bauer, FRBSF Economic Letter, July 31, 2017. Available at: https://www.frbsf.org/economicresearch/publications/economic-letter/2017/july/bridging-gap-forecasting-interest-rateswith-macroeconomic-trends/ 19 “Interest Rate Forecasters Are Shockingly Wrong Almost All of the Time”, Business Insider, July 8, 2015. Available at: https://www.businessinsider.com/interest-rateforecasts-are-wrong-most-of-the-time-2015-7 18 Bench Analysis 56 Docket No. 2018-00194 Figure 1 This upward bias is confirmed by the Congressional Budget Office (CBO) in its report “CBO’s Economic Forecasting Record: 2017 Update.”20 In the update, the CBO observed that its forecasting record was similar to the Blue Chip consensus, noting that when CBO’s projections were inaccurate by large margins, the other two forecasters’ projections tended to have similar errors because the forecasters faced the same challenges, and that there is an upward bias to their interest rate forecasts. Approximately a decade after the “Great Recession” of 2008-2009, and one of “CBO’s Economic Forecasting Record: 2017 Update”, October 2, 2017, https://www.cbo.gov/publication/53090 20 Bench Analysis 57 Docket No. 2018-00194 the longest post-war growth runs, market conditions are known. Investors have understood the interest rate environment for quite some time now, and the relatively low yields reflect their expectations. A year ago, U.S. tax policy suddenly changed, potentially affecting the creditworthiness of an entire industry, and investors have adjusted their expectations. Staff has seen no testimony in this case to change its recommendation of following Commission precedent to rely primarily on market data, rather than projections. b. ROE Determinations As noted by Regulatory Research Associates (RRA), a unit of S&P Global, authorized ROEs for electric utilities have trended lower since the 1980s and continue to decline consistent with the interest rate environment. For all electric utilities, the average authorized ROEs have declined from 9.85% in 2015, 9.77% in 2016, 9.74% in 2017 and 9.59% in 2018. For delivery-only electric utilities, ROE authorizations ranged from 8.69% to 10%, averaging 9.38% in 2018.21 The RRA analysis is useful in demonstrating the long-term trend of declining authorized returns on equity. It also mirrors the long-term decline of interest rates from 1980 to today. While the RRA data does not indicate the market’s preferences, it does provide useful information about how other regulators view the risk and required return for purposes of meeting the HopeBluefield standard. “Perceived lower risk pull down [sic] electric ROEs in 2018,” Regulatory Research Associates, Focus Notes, February 19, 2019. 21 Bench Analysis 58 Docket No. 2018-00194 Figure 2 c. RDM Effects and Rate Design CMP currently has a revenue decoupling mechanism (RDM) in place. Conceptually, regulatory structures such as this reduce risk for the utility by removing uncertainty around sales lost as a result of weather, energy efficiency, or net energy billing. There is no analytical basis for attributing a specific number to the risk reduction which results from an RDM, and the Commission has declined to attribute a specific number. Nevertheless, as it considers the risks a utility bears, the Commission should consider this risk reduction as a factor when weighing the range of acceptable equity return rates. Likewise, to the extent rate design changes are implemented that move a larger portion of rate recovery into fixed charges, the Commission should consider the effect any rate design changes approved in this case have on the Company’s revenue stability and risk. Bench Analysis 9. 59 Docket No. 2018-00194 Flotation Costs The Commission has, in prior rate cases, permitted an upward adjustment to ROE to reflect the costs associated with the sale of new issues of common stock. Though CMP has not previously requested flotation costs, in this case, they do. Ms. Bulkley asserts that failure to allow recovery of past flotation costs may deny CMP the opportunity to earn its required ROR in the future. Ms. Bulkley calculates a Flotation Cost Adjustment based on the costs of issuing equity that were incurred by AVANGRID in its most recent common equity issuance. To do this, she first creates a ratio of the “Total Flotation Costs” to “Gross Equity Issue Before Costs” (3.07%). Then, using the proxy group, she “grosses up” Expected Dividend Yield for each company then carries that grossed up value through the DCF ROE calculation. Lastly, she compares the grossed-up ROE to the non-grossed up value, and that reveals the flotation adjustment. Her calculations in AEB-4 indicate an adjustment of 10 basis points, though her narrative mentions 11 basis points. ROE Test. at 46. Staff does not dispute this flotation cost adjustment and would add 11 basis points to its ROE recommendation. 10. Recommended ROE In determining its ROE recommendation, Staff depends primarily on the DCF analysis and uses other analyses as a check on the range. The DCF analyses presented by the Staff and the Company produce an indicated ROE range with midpoints that generally falls in the 9.00% to 9.50% range as shown in Table 16. The mean result of the CAPM analysis presented by the Company is 11.47% and the Staff’s Bench Analysis 60 Docket No. 2018-00194 CAPM results range from 9.15% to 11.59% with an average of 10.77%. Table 16 Indicated ROE Range Indicated ROE Range (Excludes flotation cost adjustment) Low High Mid-point Bulkley Constant Growth DCF As Adjusted by Staff 30-Day Average 90-Day Average 180-Day Average Average 8.16% 8.26% 8.33% 8.25% 10.40% 10.50% 10.58% 10.49% 9.28% 9.38% 9.46% 9.37% Staff DCF 7.58% 10.86% 9.22% Based on this analysis, Staff does not support an ROE of 10.00% as requested by the Company. A 10.00% ROE would fall closer to the top of the range indicated by the DCF analyses and substantially above the mid-point. As noted, previously, the Commission uses the results of the CAPM analysis as a useful check on the DCF results. In this case, the Staff does not believe that there is sufficient reason to deviate from the mid-point of the DCF range. Additionally, the more recent vintage of the Staff’s analysis (data from February 13, 2019 versus data from August 31, 2018) suggests that more reliance be placed on the Staff’s results. With the addition of a flotation cost adjustment of 11 basis points as discussed, Staff recommends an “unadjusted” ROE of 9.35%. As discussed later, however, the overall management efficiency of CMP may warrant utilizing an ROE at the low end of the range of reasonable results. 11. Capital Structure Bench Analysis 61 Docket No. 2018-00194 The Company proposes a capital structure that consists of 55.0% common equity, 0.02% preferred stock, 42.16% long-term debt and 2.82% short-term debt. Ms. Bulkley, states that she reviewed CMP’s historical capital structure and the capital structures of utility operating subsidiaries of similar proxy companies in determining this proposed ratio. ROE Dir. at 74. Ms. Bulkley further states that the equity ratios of the utility operating subsidiaries in the proxy group ranged from 48.44% to 63.84%, with an average of 55.28%. Id. at 75. Additionally, Ms. Bulkley examined the historical capital structure of CMP and determined that CMP’s year-end equity ratio has consistently been between 53.60% and 59.57% over the past five years, with an average equity ratio of 56.15%. Id. at 76. Moreover, Ms. Bulkley asserts that the TCJA must be considered when determining the equity ratio, noting that the credit rating agencies have expressed concerns about its effect on utility cash flows. Based on this analysis, Ms. Bulkley contends that the proposed equity ratio of 55% is reasonable. Because the intent of the ROE analysis is to assess what return investors require, Staff does not see the approved equity ratios for operating subsidiaries as the appropriate comparative metric. The equity ratios of the operating subsidiaries are a reflection of regulatory, not investor, preference. Rather, it is appropriate to consider the capital structure of the entities for which market data is used, the publicly-traded parent companies. In response to a data request, Ms. Bulkley provided the equity ratios of the parent companies in her proxy group. EXM-006-011. Using data from SNL, she shows in Attachment 1 the common equity ratios for the publicly traded parent companies in her proxy group, which range from 48% to 64% with an average of 54.9%. Bench Analysis 62 Docket No. 2018-00194 Using SNL data, Staff calculated the common equity ratio of the parent companies in Staff’s proxy group. As shown in Table 17, Staff has calculated the common equity ratios based on the parent Companies in the updated proxy group, as this more closely reflects what investors are considering when making investment decisions. The common equity ratios of the Staff proxy group ranged from a low of 28.93% to a high of 62.16% and averaged 48.10%. Table 17 Proxy Group Company Alliant Energy Corporation Ameren Corporation Atmos Energy Corporation CMS Energy Corporation Consolidated Edison, Inc. DTE Energy Company Edison International El Paso Electric Company IDACORP, Inc. New Jersey Resources Corporation Northwest Natural Holding Company NorthWestern Corporation ONE Gas, Inc. PNM Resources, Inc. Portland General Electric Company PPL Corporation Public Service Enterprise Group Incorporated Spire Inc. Xcel Energy Inc. Average Minimum Maximum Common Equity Ratio 45.80% 47.49% 60.85% 28.93% 49.03% 42.89% 49.05% 48.85% 56.32% 52.11% 48.78% 47.76% 62.16% 41.02% 50.28% 36.14% 51.44% 52.08% 42.99% 48.10% 28.93% 62.16% For many years CMP’s rates have been set on the basis of a capital structure Bench Analysis 63 Docket No. 2018-00194 that includes a 50% equity layer. During that time, CMP has been able to maintain its credit rating and to attract capital and the common equity ratios of Staff’s proxy group do not support moving from that level. Staff recommends maintaining CMP’s common equity ratio at 50%. For purposes of calculating a weighted average cost of capital, Staff has re-allocated the 5% reduction in the common equity ratio to long-term debt as shown in Table 17. 12. Cost of Short-term and Long-Term Debt The Company’s filing includes calculation of the expected cost of long and shortterm debt. For long-term debt, CMP incorporates the effects of retirements and sinking fund requirements and new issuances to arrive at a weighted cost of long-term debt for the rate effective period of 4.45%, a reduction from the test-year weighted cost of longterm debt of 4.87%. Rev. Req. Dir., RRP-4, Schedule E. For short-term debt, the Company’s projection of 3.50% reflects expected increases in short term interest rates and short-term borrowing costs over the next several years. Staff finds the projections of the cost of both debt components to be reasonable. 13. Staff Recommended Weighted Average Cost of Capital Combining the Staff recommended ROE of 9.35% and a capital structure that includes a 50% equity layer produces a pre-tax WACC of 8.70% and after-tax WACC of 6.87%as shown in Table 18. Bench Analysis 64 Docket No. 2018-00194 Table 18 Staff Recommended Capital Structure, Costs and ROE Capitalization Percentage Common Equity Preferred Stock Long Term Debt Short Term Debt Total III. 50.00% 0.02% 47.16% 2.82% 100.00% Cost 9.35% 6.00% 4.45% 3.50% After -Tax Weighted Cost Pre-Tax Weighted Cost 4.68% 0.00% 2.10% 0.10% 6.50% 0.00% 2.10% 0.10% 6.87% 8.70% CUSTOMER SERVICE ISSUES A. Liberty Audit Findings The Liberty Report contained the following findings regarding CMP’s SmartCare implementation and go-live: • SmartCare has functioned largely as planned, but defects in its operation have substantially contributed to very large numbers of exceptions to expected billing processes. Management has been taking corrective actions to address the billing issues, but resulting exceptions and delays have been substantial, and still affect some 10,000 customers each month. SmartCare’s generation of an excessive number of incorrect bills after go-live has served as the main contributor to delays, as management worked to correct errors found in billings to be issued. • The resources and time needed for system corrections to address process exceptions have delayed bills. They have also required time-consuming manual bill corrections, and generated bill cancellations, rebills and deposit and refund issues. Without knowledge of the status of fixes and with the need for manual corrections increasing, CMP’s customer representatives have frequently been unable to give customers meaningful answers about cause, corrections and time solutions, further increasing concern and frustration. Bench Analysis • 65 Docket No. 2018-00194 Performance weaknesses in managing CIS development and go-live processes contributed substantially to the high level of billing and customerservice challenges experienced. The specific gaps in effective management Liberty found included: inadequate testing, insufficient project staffing, lack of a sufficiently strong focus on quality, insufficiencies in project reporting for identifying and planning to address problems, and inadequate project readiness. With regards to customer service and complaint handling following SmartCare implementation, Liberty found: • The transition to SmartCare introduced problems that drove customer satisfaction below targets and customer complaints well above normal levels. Management experienced undue difficulty in eliminating underlying causes of billing complaints and in timely resolving individual customer issues. Continued high rates of complaint generation and extensive delay in resolving then have continued. Delays in resolving billing issues added to already elevated levels of customer concern, creating a cycle of escalating customer impatience and skepticism. • Insufficient Billing Group staffing resulted from an inability to identify correctly the time needed to stabilize SmartCare. Insufficient numbers of staff slowed identification and correction of errors at their system sources and as part of manual efforts required to correct individual bills held for issuance due to known problems or identified by customers after receiving them. • In the months following SmartCare go-live, many bills were also delayed (for December and January particularly) due to CMP’s pre-issuance erroridentification processes. Management has addressed sources of these errors and reduced them, but more than 100,000 accounts have experienced billing error and CMP is still reporting customer-affecting errors. Bill presentation errors, avoidable had management tested presentation well before go-live, further contributed to customer confusion, promoting doubt about billing accuracy. • CMP experienced a 22 percent increase in times to handle calls following SmartCare go-live and high rates of calls abandoned by customers before response by the Company. CMP has also failed to meet consistently its target level of service (answering 80 percent of customer calls within 45 Bench Analysis 66 Docket No. 2018-00194 seconds, a target consistent with industry experience) as it continued in the second and third quarters of 2018 to struggle to address billing issues. A lack of sufficient staffing has materially contributed to long answer and call handling times. A lack of sufficient experience and supervision have impaired the ability to resolve specific customer inquiries and complaints and to address systemic issues underlying them. • Numbers of representatives hired have not kept pace with management’s analyses of needs. The extraordinarily high ratio of representatives to supervisors impaired efforts to apply experience in addressing the root causes of customer inquiries and complaints, to direct inexperienced representatives, and to provide coaching and training needed to give representatives the skills and confidence to function independently in helping customers with questions and issues. Liberty Report – Executive Summary at 8 – 10. B. Chronology of CMP’s Customer Service Issues The Commission Staff’s concern about CMP’s customer service predates both the issuance of the Liberty Report and the implementation of SmartCare and, in fact, go back as far as 2016. In a letter dated April 14, 2016 to Leona Michelson at CMP from Susan Cottle, the Deputy Director of the Consumer Assistance and Safety Division (CASD), Ms. Cottle expressed concerns regarding CMP Call Center Representatives’ (CSRs) lack of understanding regarding the Commission’s Arrearage Management Program (AMP) and CMP’s obligation under Commission rules to administer the AMP.22 On April 19, 2014, An Act to Assist Electric Utility Ratepayers, was enacted into law (Act) (P.L 2013, ch. 556 ). The Act directs each electric utility to implement an Arrearage Management Program (AMP) to assist eligible low-income residential customers who are in arrears on their electricity bills. The Act defines an AMP as a plan “under which a transmission and distribution [electric] utility works with an eligible lowincome residential customer to establish an affordable payment plan and provide credit to that customer toward the customer’s accumulated arrears as long as that customer remains in compliance with the terms of the program.” Chapter 317 of the 22 Bench Analysis 67 Docket No. 2018-00194 The letter asked CMP about its training programs for CSRs and requested copies of training materials. In its response to Ms. Cottle’s letter, CMP stated that it was committed to a successful AMP and that it would continue to reinforce AMP requirements with its agents.23 Commission Staff’s concerns about CMP CSRs continued during the summer of 2016. In a letter dated August 4, 2016 from Susan Cottle to Ann Brooks with CMP, Ms. Cottle expressed concern regarding the significant increase in violations committed by CMP’s CSRs in their interactions with customers.24 The violations primarily related to the level of understanding (or lack thereof) by CMP’s CSRs regarding the Company’s obligations to provide certain information to customers pursuant to the Commission’s Consumer Protection rules. MPUC Rules , ch. 815 § 13(D) The letter noted that there had been a three-fold increase in the number of violations cited by the CASD against CMP between February 1 and August 1 of 2016 in comparison to the same period the previous year, with the majority of these violations relating to CMP staff failing to provide proper information to customers and in some situations, providing incorrect information to customers. For example, in six situations, customers who had been disconnected for non-payment were given the wrong amounts they needed to pay to have their service reconnected. The CASD again requested information regarding CMP’s training procedures for its CSRs and emphasized the need for CMP staff to be familiar with the Commission’s rules establishes the process by which each electric transmission and distribution utility will implement and AMP. 23 A copy of Ms. Cottle’s letter is attached as Customer Service Exhibit 1. 24 A copy of Ms. Cottle’s letter is attached as Customer Service Exhibit 2. Bench Analysis 68 Docket No. 2018-00194 requirements of the Commission’s Consumer Protection Rules and customer rights and responsibilities under those rules. Problems with the quality of information being provided by CMP’s CSRs continued through 2016 and into the spring of 2017. In a letter dated May 26, 2017 from the Commission’s General Counsel to CMP’s Vice President of Customer Service, the General Counsel stated that the Commission had serious concerns with the performance of CMP’s call center over the past year.25 The letter stated that on numerous occasions in the past year, the CASD had brought concerns to CMP’s attention regarding the information that CMP’s call center staff was providing, or failing to provide, to applicants and customers in violation of Commission Rules. The letter further stated that since the start of the summer collection period on April 18, 2017, the CASD had investigated numerous instances where CMP staff had either provided inaccurate information to customers or had failed to follow the disconnection requirements of Chapter 815. These instances included: 25 • Disconnection calls being made to hundreds of customers before the disconnection date stated on the respective disconnection notices. • Erroneous (too high) amounts being quoted to customers to avoid or remedy disconnection. • A CMP representative refusing to accept payment from a customer under the threat of disconnection when the amount the customer committed to pay was consistent with the amount necessary to prevent disconnection and was well within the effective period of the disconnection notice. A copy of this letter is attached as Customer Service Exhibit 3. Bench Analysis 69 Docket No. 2018-00194 • A disconnected customer who was eligible for a payment arrangement being told she would have to pay her entire account balance before being reconnected. • Continuing failure by CMP staff to inform potentially eligible customers of the existence of the AMP program. • Customers under the threat of disconnection contacting CMP and being told that they would not be disconnected based on information provided during the call and then subsequently being disconnected. • CMP requiring a deposit from a customer in violation of Chapter 815. • Repeated failure of CMP staff to advise customers with a dispute of their right to contact the CASD. The letter concluded that due to the pervasiveness of the problems, as well as the length of time that the problems had continued, it appeared to Commission Staff that the problems were systemic in nature. The letter also stated that in previous communications regarding these violations, CMP stated that it had taken steps to address the problem. Given the ongoing and repeated nature of the violations, it seemed, however, that steps taken by CMP were insufficient. During the summer of 2017, the CASD began receiving calls from customers who reported that they could not reach CMP to discuss their credit and collection issues. Many of these customers had received disconnection notices and were instructed in the notice to contact CMP immediately to resolve the billing problem. This was especially concerning because customers were required to contact CMP prior to the disconnection date stated in the notice, which in some cases was only three days in the future, in order to resolve the problem and avoid disconnection. Relevant to this issue, Section 13(B) of Chapter 815 of the Commission’s Rules requires utilities to have Bench Analysis 70 Docket No. 2018-00194 an adequate number of properly trained employees available during business hours to respond to questions from applicants and customers, resolve customer disputes, and respond to requests for new service. This section further requires utilities to provide customers the opportunity to speak with a live person without spending an unreasonable amount of time on hold. In response to these complaints, the CASD sent a violation letter to CMP on August 1, 2017.26 The letter stated that the CASD had received a call from a customer under the threat of disconnection who reported that she called CMP several times over the past three days but each time she received a recorded message indicating all the phone lines were busy. Further, the customer said she received a message that told her to call again and the call was terminated. The customer was not provided an opportunity to hold for a representative or the opportunity to leave a message to have her call returned. The letter further stated that the CASD had been receiving this same type of complaint from numerous customers over past few months and put the Company on notice that the CASD considered this a violation of Chapter 815, section 13(B) of the Commission’s rules. In its response to the violation letter, the Company explained that when its telephone carrier is unable to deliver a call to the call center due to high call volumes, the customer will receive a “courtesy message,” in lieu of a busy signal. The letter further stated that from March 5, 2017 through July 29, 2017, 14% of the calls made to CMP’s credit and collections call center received the courtesy message. CMP also 26 A copy of Ms. Cottle’s letter is attached as Customer Service Exhibit 4. Bench Analysis 71 Docket No. 2018-00194 attributed part of this problem to the need to prepare and train call center staff for the implementation of its new SmartCare System, which began on July 6, 2017.27 On August 11, 2017, CMP’s vice president of Customer Service met with Commission Staff to discuss the Company’s ongoing problems associated with answering customer calls and providing customers with accurate information. During that meeting, Commission Staff emphasized the importance of properly handling customer calls and the fact that this problem had continued for over a year. It appeared to Commission Staff, at that point in time, that CMP both understood the seriousness of the problems and that the continuation of the problems was unacceptable. Despite this meeting and the previous three violation letters sent to the Company in relation to the company’s poor call center performance, throughout the remainder of 2017 the CASD continued to receive calls from customers reporting that they either could not reach CMP to discuss their problems or did not receive a helpful response from CMP Staff when they did reach the Company. On November 14, 2017, the CASD sent an email to the Company expressing concern about the large number of customers who continued to report that they could not reach the Company to discuss their credit and collections concerns and requested call statistics from the Company for October 1, 2017 through November 14, 2017.28 Staff notes that the “courtesy message” is due to insufficient call capacity at CMP’s call centers, which is a separate and distinct issue from a lack of sufficient staff to answer calls. Further, the courtesy message problem began in earnest in March of 2017, several months prior to the SmartCare training session that began on July 6, 2017. 27 28 A copy of that email is attached as Customer Service Exhibit 5. Bench Analysis 72 Docket No. 2018-00194 The CASD sent another letter to the Company dated January 10, 2018 stating that a review of the call data provided in its response to the November 14th letter showed that 63,970 out of 273,685 calls, or one out of every four calls, received the “courtesy message” instead of being able to speak with a live person.29 The letter acknowledged that while the October Storm certainly exacerbated this problem; however, the call statistics for the month before the Storm indicated that thousands of customers could not reach CMP to discuss their credit and collection problem and instead received the “courtesy message”. Thus, the problem could not be blamed on the Storm and was instead a result of CMP’s management practices. The January 10th letter also stated that in addition to the problems relating to the customer call center, that callers to CMP’s “contractor line” were reporting the same type of problems.30 Contractors reported that they were being placed on hold for extended periods of time – often hours – while waiting to speak to a live person at CMP. This was causing significant problems for contractors and customers who were trying to have their electricity connected before the onset of winter. Further, customers and contractors also began reporting that the Company was often not showing up for field appointments or was cancelling long scheduled filed appointments for no apparent reason. 29 30 A copy of Ms. Cottle’s letter is attached as Customer Service Exhibit 6. The Company maintains a phone line, 866-225-4200, for contractors to call to arrange for a field visit. The purpose of the line is to avoid the need for contractors to call the general information line so they may reach a live person to address their needs in a reasonable amount of time. Bench Analysis 73 Docket No. 2018-00194 The January 10, 2018 letter pointed out that these call center problems had begun in early 2016 and that the Commission Staff has had numerous discussions with CMP personnel regarding its call answering problems and the need for CMP improve its call answer service. The letter asked CMP to provide a date by when it would begin meeting its obligation to answer and respond to customer credit and collection calls and what actions the Company would take to ensure that it answered calls to its contractor line within a reasonable amount of time. The letter also required the Company to report to the CASD on a monthly basis key call answer statistics to both its credit and collections line as well as its contractor line beginning with January of 2018. In a response dated January 25, 2018, CMP stated that it believed it was meeting its obligation under Chapter 815 to have an adequate number of trained staff to answer its credit and collections line and that no violations were occurring. The Company went on to state that 2017 was an extraordinary year and the October Storm created challenges for it and its call center. The Company also stated that the implementation of its new customer care system posed challenges for its call center agents and this, combined with the after effects of the October Storm, resulted in higher call volumes that exacerbated these challenges. In response to the CASD question regarding when CMP would begin to meet its obligation to answer customer credit and collection calls, the Company stated that call volumes have peaks and valleys and the Company did not believe it practical or prudent to staff to meet peak call volumes. Further, the Company stated that it expects the number of customers receiving the “courtesy message” to decline, and again noted that it was not prudent or practical to manage a call center to be staffed to “take every call at every moment.” Regarding the Bench Analysis 74 Docket No. 2018-00194 problems noted with the contractor’s line, the Company stated it was planning on increasing the number of staff taking dedicated calls to the contractor’s line between 7:00 am and 7:30 am and it planned on training more agents for the new service/contractor queue beginning in February of 2018. In addition to the call answer problems, the CASD also began receiving a large number of complaints from customers in December of 2017 and January of 2018 relating to billing concerns. These concerns included customers receiving higher than usual bills, customers not receiving proper credits on their bills, customers on payment arrangements having the arrangements changed without notice, and customers not receiving bills at all. Two of the more prevalent billing complaints received by the CASD were high usage complaints and customers not receiving bills (referred to by CMP as “delayed bills”). Regarding the high usage complaints, due to the excessively large number of high use complaints being filed with the CASD, the CASD requested in early February of 2018 that CMP establish a specialized group of individuals who were familiar with CMP’s billing and metering processes to investigate and resolve customer complaints of high usage. Under this process, when the CASD receives a customer complaint of high usage, the CASD forwards the complaint to the specialized team, which in turn reviews the customer’s account to ensure that bills are accurate and attempts to ascertain the cause of the high bill. If CMP’s team is able to determine the cause of the increased usage or bill amount and resolves the matter to the customer’s satisfaction, the CASD will consider the matter closed. If CMP is not able to ascertain the cause of the increased usage or otherwise resolve the matter to the customer’s satisfaction, the Bench Analysis 75 Docket No. 2018-00194 matter is referred back to the CASD. The customer is then notified that their complaint will be incorporated into and addressed through the Commission’s high bill investigation. The intent of this process is to ensure that customer complaints that involve customer specific issues or situations are resolved when possible, while incorporating complaints where a customer’s problem of high usage cannot be resolved into the Commission’s billing investigation for further consideration. When the specialized team was first established, the CASD and CMP agreed on a two-business day time-period for CMP’s specialized team to contact each customer. Due to the large number of customers being referred to the specialized team, the contact period was extended to two weeks at CMP’s request. Even with this extension, CMP was unable to consistently meet the contact goal. In September of 2018, the Director of CASD sent a letter to CMP expressing concern that at times it was taking CMP’s team over three months to contact customers. The letter also stated that at that time, only 45% of the total number of customers referred to CMP’s team had been contacted by CMP. The letter further stated that this delay had resulted in a large backlog of outstanding customer complaints, as well as a significant number of calls to the CASD from referred customers concerned about the time it was taking to hear from CMP. In a subsequent letter sent February 6, 2019 to CMP, the CASD again notified CMP that the time it was taking the Company to address customer complaints or contact customers who have complained, continued to be insufficient and stated that the CASD viewed this failure to contact customers in a timely manner as a violation of the Commission’s rules. Bench Analysis 76 Docket No. 2018-00194 Regarding the delayed bill issues, customers were contacting the CASD stating that they had not received a bill from CMP for a number of months. Customers were concerned that they would end up receiving a large make-up bill at some point in the future that they would be unable to pay. According to CMP, it identified bills with a potential problem and “held the bill” until such time that the billing problem could be identified and rectified. C. Call Center Performance Issues Liberty found that a significant degradation of CMP’s Call Center performance occurred with the deployment of SmartCare in October of 2017. Further, it also found that CMP had been having problems answering customer calls before the implementation of SmartCare. Liberty found that when training for SmartCare began in July 2017, many customers dialing CMP received a courtesy message (were blocked) because the number of customers attempting to reach CMP exceeded the capacity of trunk lines connecting the contact center to the telephone network.31 As seen in Figure 3 below, the worst blockage occurred from July 2017 to February 2018, during the pre- and post SmartCare go-live periods. During this period, approximately 25% of callers trying to reach CMP received the “courtesy message.” EXM 01-019, Attachment 1. 31 CMP began tracking blocked calls in July 2017 and since then, nearly 350,000 customer calls have been blocked, with a peak of 112,710 in November 2017, following the October 2017 storm event. The Liberty Report also stated that Call Center best practice eliminates instances of blocked calls through the use of overflow services or additional capacity. Liberty Report at. 89. Bench Analysis 77 Docket No. 2018-00194 Figure 332 Call Blocking Liberty also found that the numbers of callers abandoning their calls before reaching a live person also increased significantly after go-live, peaking in late March and early April 2018. Figure 433 Calls Abandoned The Liberty Report stated that typical industry abandonment rates range between 5 and 10 percent; consequently, the abandonment rates experienced by the Company far exceeded industry standards. Liberty Report at 89. Further, when “Blocked Calls,” which are situations where the customers cannot get through to the Company, are 32 Data Source: Liberty Report at 89. 33 Data Source: Liberty Report at 90. Bench Analysis 78 Docket No. 2018-00194 considered along with Calls Abandoned, situations where customers that do get through to the Company are put on hold for long periods of time and abandon the call, it becomes clear that a large number of CMP customers were simply unable to reach the Company from July of 2017 through January of 2018. The Liberty Report also found that CMP’s Call Center was not consistently meeting its call answer goal of 80% of calls answered within 45 seconds prior to the implementation of SmartCare in October of 2017 and experienced a significant drop in performance after the implementation of SmartCare.34 This trend is observed in Figure 5 below. Figure 5 35 % of Calls Answered in 45 seconds Staff notes that while Liberty reviewed CMP’s call center performance using the Company’s goal of answering 80% of calls within 45 seconds, Staff does not view this level of performance as representing “reasonable and adequate service.” CMP’s call answer metric contained in its previous ARP was “80% of calls answered in 30 Staff notes that CMP’s benchmark for reasonable performance for it “% of calls answered” service quality metric under its previous ARP was “80% of calls answered in 30 seconds.” 35 Data Source: Liberty Report at 90. 34 Bench Analysis 79 Docket No. 2018-00194 seconds.” Staff continue to view this level of performance as “reasonable and adequate” with regards to CMP’s call answer performance. Figure 6 below depicts the Company’s call answer rate, as well as its call abandonment rate, for its “business line,” for the past four years. The table shows that CMP has failed to meet the “80% of calls answered in 30 seconds” benchmark for the past four years. The Figure also shows that CMP’s call abandonment rate over this same period has exceeded the 5% to 10% industry standard noted by Liberty since 2016. Figure 6 36 Call Answer Statistics 2015 - 2018 100% 80% 80% 70% 60% 67% 57% 47% 40% 20% 0% 11% 7% 2015 2016 % ans. ≤30 secs 13% 2017 20% 2018 abandon. rate These call answer statistics indicate then that CMP’s problems at its call center have been serious, long lasting, and substantially predate the Company’s implementation of SmartCare. 36 Data Source: EXM 001-019, Attachment; Call Answer Statistics provided to CASD. Bench Analysis 80 Docket No. 2018-00194 D. Delayed Bills/New Customer Accounts As discussed above, the Company experienced significant problems after the implementation of its SmartCare System with certain customers going extended periods of time and not receiving a bill for service. The Company referred to this problem as “delayed billing.” The Liberty Report explained that “delayed billing” is the situation where bills are held-up (not issued) by billing exceptions or defects. Liberty reviewed billing data for six months, beginning in November of 2017, to determine the number of bills held up by billing exceptions or defects. Liberty’s findings can be seen in Figure 7 below. Figure 7 37 Delayed Bills Due to Outsorts and Out-of-Balance As review of Figure 8 below shows that the delayed bill problem has worsened since last April and has continued into 2019. It greatly concerns Commission Staff that the problem of delayed bills has continued for over a year since the implementation of SmareCare without significant improvement. One year should be more than adequate 37 Data source: Liberty Report at 84. Bench Analysis 81 Docket No. 2018-00194 time to either make the necessary modifications to SmartCare to reduce the number of exceptions or to hire adequate staff to work the exceptions. It does not appear that the Company has taken either of these actions. Figure 8 38 In addition to the “delayed billing problem,” the Company also has experienced delays in establishing new accounts for some customers since November 2017. At a technical conference held on December 3, 2018, Company witnesses testified that approximately 3,400 new customer accounts had not been established and consequently were not being billed for service.39 [t]here was a new system put in place within our SAP system to establish new accounts. The group that was charged with doing that experienced (sic) a significant number of retirements. It’s taken time to backfill those positions and 38 39 Data Source: weekly “Delayed Bill” report filed with the CASD. If the account was not created for more than 30 days after the meter was installed, the Company considered this as a “delayed bill” situation. 12/3 Trans. at 119. Bench Analysis 82 Docket No. 2018-00194 get the new employees trained to do the work, and that has resulted in a delay in the backlog of this work that is longer than we would like. Tr. 110 (Dec. 3, 2018) The Company witness also testified that another contributing factor to the problem was an increase in the number of new accounts that this group has been asked to process in the last two years. Id. at 111. Commission Staff asked the witness how that statement meshed with EXM 002-027 that shows that new connections statistics since 2016 look relatively stable. In the Company’s response to Staff’s question, the Company stated that it is on pace in 2016, 2017 and 2018 to have an average of 5,284 new service orders per year. ODR 001-020. This response does not demonstrate an increase in the number of new accounts processed in the last two years. Further, even if the number of new accounts has been increasing significantly over the past few years, this is a good “problem” to have and an effectively managed utility would hire the appropriate number of staff to address the increased workload. A “high level” estimate provided by the Company of the total amount of unbilled revenue associated with the 3,400 accounts is approximately $729,675. ODR 001-019. Bench Analysis 83 Docket No. 2018-00194 Figure 9 40 E. Staffing Issues Related to Customer Service Performance In its report, Liberty noted that CMP had reorganized customer service functions in December 2017, after the appointment of Avangrid’s Vice President of Customer Service, moving CMP from a decentralized, company-functional organization to the structure in place today. The reorganization included moving responsibility for customer service functions from the CMP Vice President of Customer Service to Avangrid. At the same time, responsibilities for Meter Operations and Meter Services moved from the CMP Customer Service organization to Avangrid’s Operations Technologies and Regional Operations. Liberty Report at 74. Liberty further reported that Avangrid’s Director of Customer Care oversees contact centers, key accounts, vendor management, and customer service technologies and that CMP’s Contact Center Manager reports to the Director of Customer Care. 40 Data source: weekly “Delayed Bill” report filed with the CASD. Bench Analysis 84 Docket No. 2018-00194 Further, Avangrid’s Director of Customer System Operations and Support oversees billing, payment and collections. CMP’s Billing Supervisor reports to Avangrid’s Billing Manager, who oversees billing operations within all Avangrid operating companies. Id. at 75. Liberty reported that Avangrid combined CMP’s Portland and Augusta contact centers in 2015 and offered a voluntary separation package to affected employees. Later, in September of 2017, Avangrid again offered four voluntary separation offers for certain non-union, Customer Service organization employees. This offer applied to those working on Avangrid’s SmartCare, Click (field service order), and New York Meter Services projects. Depending upon the plan selected, employees agreeing to the offer could work through either November 30, 2017 or April 30, 2018. These voluntary separation offers came as part of a large-scale Avangrid reorganization. Id. at 75 and 76. The Company’s Customer Service employees became aware of the separation offers the weekend before SmartCare’s go-live date in late October of 2017. Thirteen employees accepted the offer, departing the Company in April 2018. These departures came at a time when call volumes, customer complaints, and billing exceptions were at very high levels. Further, six of the departed employees had been members of the Company’s Customer Service and Call Center management team. Id. at 76. Liberty found that staff departures in April of 2018 as a result of the voluntary separation package detrimentally impacted CMP’s Customer Service employees responsible for billing functions at the Company. At the end of April 2018, with a backlog of billing exceptions growing, two billing analysts accepted the early separation Bench Analysis 85 Docket No. 2018-00194 option and left the company. Their departures left the billing group below normal staffing levels at a time of significantly increased workload. Id. at 77. Liberty further found that CMP’s management underestimated the level of billing work following SmartCare deployment and failed to staff the billing group adequately to meet the increasing volume of billing exceptions and manual work in the months following SmartCare golive. As discussed above, the Commission Staff has been concerned about CMP’s call answer performance since 2016 and believes that the reduction in Customer Service positions was a contributing factor in the degradation of service. Liberty’s findings support this conclusion. The Liberty Report found that: Management’s failure to maintain a hiring pace sufficient to meet growing Call Center needs contributed materially to CMP’s inability to properly handle customer calls. With a 2016 analysis in hand showing a need and recommendation for adding 45 customer service representatives, management sought approval for only 14. As a result, call answering performance was below standard for much of 2016 and 2017. Id.at 98. The Liberty Report also stated that staffing could be a factor in CMP’s degrading call center performance. The Report noted that over the past three years, CMP has not always staffed the Call Center to levels recommended by Call Center management. In February of 2016, a staffing analysis completed by Call Center management called for adding 35 permanent representatives to maintain adequate service levels for calls handled by employees in Maine. The analysis used projected call volumes and the expectation of answering 80 percent of calls within 45 seconds. Table 19 below shows Bench Analysis 86 Docket No. 2018-00194 the number of staff additions recommended by the staffing analysis versus the actual number of staff hired. Table 19 41 Period Recommended Hired 1Q 2016 35 14 2Q 2017 24 29 1Q 2018 29 20 3Q 2018 27 25 Based on this, the Liberty Report concluded that: Management’s failure to maintain a hiring pace sufficient to meet growing Call Center needs contributed materially to CMP’s inability to properly handle customer calls. With a 2016 analysis in hand showing a need and recommendation for adding 45 customer service representatives, management sought approval for only 14. As a result, call answering performance was below standard for much of 2016 and 2017. Liberty Report at 98. The Report further concluded that: [a]n insufficient number of supervisors has also impaired CMP’s ability to respond effectively to customer inquiries and complaints. The ratio of supervisors to representatives doubled (to 25:1)42 during the last several months of testing, training, and preparation for SmartCare go-live. Management did not add supervisors to cover increased numbers of representatives hired and it lost three experienced ones needed to provide experience, training, and direction at critical times leading to and following the switchover to SmartCare. Id. at 98. 41 Data Source: Liberty Report at 91. Staff notes that the ratio of supervisors to representatives should be 1:25, not 25:1 as stated in the Liberty Report. 42 Bench Analysis F. 87 Docket No. 2018-00194 Accounts Receivable As with the Company’s call answer performance, Commission Staff has also been concerned about CMP’s credit and collection performance since 2016. Although the Company’s actual bad debt (write-off amounts) has been reasonable over the past few years and the bad debt figure contained in the Company’s test year in this case also seems reasonable, Commission Staff is concerned that the Company has not been as diligent with its credit and collection activities as it could have been in the past few years and that this lack of diligence may be contributing to a mounting level of accounts receivable for the Company. A review of Figure 10 below shows a significant decrease in the number of Winter Requests to Disconnect43 (WRTD) and Variances44 submitted by the Company Between November 15 and April 15, electric and gas utilities are prohibited from disconnecting customers without first receiving permission from the CAD. During this time period, utilities must make significant attempts to personally contact customers who are behind on their bills to negotiate a payment arrangement prior to seeking permission to disconnect. In situations where the utility cannot make contact or is not able to negotiate a reasonable payment arrangement with a customer after making contact, the utility may submit a request to disconnect the customer’s service to the CAD. In these situations, the CAD also attempts contact with the customer for the purpose of establishing a reasonable payment arrangement. Whether or not the CAD is able to contact the customer, it will ensure that the customer is on a reasonable payment arrangement. 43 44 A utility may request that the CAD grant a waiver from any provision of Chapter 815 (consumer protection rules) in any case involving an individual applicant or customer whose conduct and known financial condition pose a clear danger of substantial losses to the utility. A request for waiver under this subsection must be made to the CAD. The request may be written or oral, but an oral request must be followed promptly by a written confirmation. The written request or confirmation shall include a detailed statement of the facts alleged by the utility in support of the request. The utility shall immediately notify, in writing, the individual applicant or customer whose service would be affected by the proposed waiver, describing the nature and effect of the requested waiver and the facts alleged in support of the request. Bench Analysis 88 Docket No. 2018-00194 since 2015. WRTD’s and Variances are an important part of a robust credit and collections process because these actions ensure repeated and consistent contact on a year-round basis with customers having payment problems. This repeated contact helps customers stay in a routine of making monthly payments and thus keeps them from falling behind on their bills. Further, these actions incorporate the involvement of the CASD in the collections process, ensuring that customers having payment problems are receiving advice and input regarding all the available payment related resources, such as the Low-Income Assistance Program, the Arrearage Management Program, the Oxygen Pump and Ventilator Programs, assistance agencies, as well as the services of the CASD itself. Figure 10 45 WRTD/VARIANCES 350 299 300 250 200 188 196 163 150 84 100 30 50 14 4 0 2017 2018 0 2014 2015 2016 wrtd VARIANCE Figure 11 below shows an increase in aging accounts receivable from $18.3 million (excluding current amount due) in September of 2017 to $40 million in September of 2018. Further, Figure 11 also shows an increase from $6 million in 45 Data Source: Commission Case Management System. Bench Analysis 89 Docket No. 2018-00194 accounts receivable that is older than 90 days in September of 2017 to $22.9 million in September of 2018. This statistic is important because this represents the oldest amounts owed to Company and thus is the least likely amount to be ultimately collected. Figure 11 46 September Accounts Receivable 50.00 40.00 $ millions 40.00 30.00 22.90 20.40 18.30 20.00 10.00 8.20 6.00 0.00 2016 2017 90 days 2018 Total In addition to the delayed bill problem and the diminished attention to credit and collections, another potential cause of the significant increase in accounts receivable that the Company experienced between 2017 and 2018 is a lack of focus on credit and collections activities by the Company, possibly due to problems it experienced after the implementation of SmartCare in November of 2017. During the December 3 Technical Conference, a Company witness stated that the cause of the significant increase in accounts receivable from 2017 to 2018 was the Company “turning off” its dunning (collections) process when the Commission adopted its Interim Payment Policy on April 46 Data Source: EXM 03-043. Bench Analysis 90 Docket No. 2018-00194 11, 2018.47 The Company witness explained that the Company could not immediately implement the Commission's order on the set-aside process because there was a lot of coding that had to go into place to do that that then had to be tested. The Company witness further stated that the coding and testing took a number of months and the Company did not want to take the risk of turning on the dunning process until the coding had been fully implemented and tested. Tr. 166 – 167 (Dec. 3, 2018) Commission Staff does not agree that the Company’s dunning process needed to be “turned off” due to the Commission’s Interim Payment Policy (policy). A key aspect of the policy is that a customer must “dispute” the bill increase with CMP in order for the customer to qualify for disconnection protection under the policy. The Commission intentionally designed the policy in this fashion so that CMP could issue disconnection notices using its existing policies and procedures and manually halt a pending disconnection when and if a customer disputed a bill increase under the policy. Because the vast majority of customers who receive a disconnection notice either pay the outstanding amount or contact the Company to establish a payment arrangement, this should not have been an overly burdensome process. As discussed above with the call center performance, Commission Staff is concerned that lack of adequate staffing, combined with the loss of staff expertise and 47 In an Order dated April 11, 2018, the Commission ordered CMP to identify a class of residential customers whose bills issued after November 1, 2017 reflect an increase in delivery charges of 25% or more over prior usage (as defined in this Order). For such customers, the Company may not threaten disconnection if the customer pays the undisputed amount of the charges for the months in question. The Commission further Ordered CMP to identify all customers that meet the Customer Eligibility Criteria specified in the Order and provide a list of these customers and account numbers to the CASD, with updates for new customers on a weekly basis. Bench Analysis 91 Docket No. 2018-00194 managerial inefficiency has led to an excessively large accounts receivable amount for the Company. Further, a large accounts receivable amount ultimately will impact a Company’s bad debt. The bad debt amount included in the test year in this case seems reasonable, so this is not an immediate problem in this case.48 However, because CMP’s practice is to write off accounts receivable as bad after the debt goes 120 days uncollected past the date the final bill is issued, it may take quite a bit of time before this account receivable problem impacts bad debt. Nonetheless, this is a problem that the Commission may want to consider directing the Company to take appropriate remedies. Further, staff is concerned about the lost revenue associated with accounts not being established for new customers in a timely manner. Regarding this issue, staff recommends that the Company identify the specific amount of uncollected revenue and describe how it plans on accounting for this revenue. G. Customer Service Conclusions and Remedy Under the provisions of 35-A M.R.S. § 301, a public utility is required to provide safe, reasonable and adequate service to its customers. The Commission has held that reasonable and adequate service encompasses more than the delivery of electric energy and incorporates a requirement that the utility, as a monopoly service provider, adequately and reasonably communicate with its customers. This duty includes the duty to take calls from its customers in a reasonably prompt fashion. Emera Maine 48 Staff notes that this may be a more immediate issue for standard offer service uncollectable adders. The extent to which standard offer bad debt is due to the Company’s lack of attention to credit and collections should be a factor in considering any increase to an uncollectable adder. Bench Analysis 92 Docket No. 2018-00194 Request for Approval of Proposed Rate Increase, Docket No. 2015-00360, Order – Part II at 49. (hereinafter Emera Maine). Because the statutory standard of reasonable and adequate service cannot be defined with precision, the Commission has utilized the following criteria in determining whether a utility’s service practices were unreasonable or inadequate: 1. Whether the company’s practice substantially departs from the regular and accepted practice of the company in question as well as that of other utilities in general; 2. Whether benefits to the company of the practice are outweighed by the adverse impact of the practice on its ratepayers; and 3. Whether the company’s practice results in inadequacy of service when considering such factors as the number of customers affected, the duration of the impact, the reason for the company’s action and the departure from historic trends. Emera Maine, supra. at 44, citing Hogan v Hampden Telephone Company, F.L. 2438, 36 PUR 4th 480, 485 (May 16, 1988). Evidence warranting a finding adverse to the utility on any one or more of these criteria is sufficient to support a finding that the practice is unreasonable. Id. In addition, the Commission has noted that a comparison of utility performance to the call answering and call abandonment standards discussed above provides evidence as to whether the utility has met its duty to take calls from its customers in a reasonably prompt fashion. Emera Maine, supra. at 49. Based on the evidence available to date, as fully described in sections III (A) through E, CMP’s customer service from 2016 through the present time has substantially deviated from both the regular and accepted practices of the Company as well as the practices of other capably managed utilities. In addition, the evidence Bench Analysis 93 Docket No. 2018-00194 indicates that the Company’s customer service practices over this time period have resulted in CMP’s customers receiving inadequate service when considering the number of customers affected, the duration of the impact, the reasons for the Company’s action, and the departure from historic trends. Thus, the Staff believes that current evidence warrants a finding that CMP’s customer service has failed to meet both criteria (1) and criteria (2) of the Hogan standard and, therefore, the Company has failed to meet its statutory obligation to provide reasonable and adequate service to its customers as required by 35-A M.R.S. § 301. In coming to this conclusion, the Staff points to the Company’s inability or unwillingness to address these customer service issues despite numerous requests from the Commission as well as notification by the Commission that in some cases the Company’s service was in apparent violation of the Commission’s Rules. In determining just and reasonable rates in this proceeding, the Commission shall, to a level within the Commission’s discretion, consider whether the utility is operating as efficiently as possible and is utilizing sound management practices. 35-A M.R.S. § 301 (4) (B). The Commission has, on a number of occasions in the past, applied a cost of equity adjustment to reflect inefficient management, especially where the management’s inefficiency is difficult to quantify, and even in those instances when the inefficiency has resulted in less expenses in the short term. Emera Maine, supra at 32. In Emera Maine, the Commission found that Emera Maine’s failure to provide adequate call center service to its customers, along with the utility’s failure to adequately inspect its transmission and distribution lines and certain accounting Bench Analysis 94 Docket No. 2018-00194 shortcomings, warranted a cost of equity adjustment of 50 basis points from what the Commission identified as the midpoint of the ROE range. In this case, based on the persistent and substantial nature of CMP’s failure to provide adequate service to its customers, the Staff believes that a 75 to 100 basis point downward adjustment from the mid-point of the ROE range is appropriate. As discussed in Section III(D), the Staff has identified 9.35% as the ROE mid-point and 7.69% as the low end of reasonable ROE results. Staff notes then that its proposed adjustment would result in an ROE of 8.35% to 8.60%, which is still within the reasonable range of ROE results identified by Staff and thus, is within the Commission’s discretion pursuant to the provisions of 35-A M.R.S. § 301. In making this proposal, the Staff notes the following language from the recent Emera Maine decision: It is clearly within the Commission’s discretion to set the ROE within a range of reasonableness indicated by the analyses in the record. In this case setting the ROE at the lower end of that range allows the company to attract capital but at the same time holds shareholders accountable for management’s lack of efficiency. In a competitive market, shareholders make decisions about which companies to invest in based on how well they are managed and the return they generate for investors through the efficiency and effectiveness of their business operations. If management is to receive a signal for greater or lesser efficiency in managing the business practices and operations of the company they control, it is through the return on equity allowed in a rate case such as this. Id. at 84. Based on Staff’s proposed rate base, a 75 basis point to a 100 basis point adjustment to the ROE reduces the revenue requirement in this case by $4.849 million to $6.466 million. Staff recommends that this cost of equity adjustment remain in place until such time as the Company demonstrates that its customer service in the areas of call center operations and billing functions has returned to “reasonable and adequate” Bench Analysis 95 Docket No. 2018-00194 levels. As an aid in determining “reasonable and adequate levels” going forward, Staff recommends that service quality metrics with appropriate service quality benchmarks be established in these key customer service areas that the Company must meet for a certain period of time before it can seek relief from the cost of equity adjustment with the Commission. To measure the adequacy of service being provided by the Company in its call center operations, Staff recommends that the Commission establish a service quality metric that measures the “percentage of calls answered by a live customer service representative,” and that the benchmark for this metric be “80% of calls answered in 30 seconds.” This is the same metric and benchmark for call answer performance that was contained in the Company’s last alternative rate plan. Central Maine Power Company, Chapter 120 Information Post ARP 2000 Transmission and Distribution Utility Revenue Requirements and Rate Design, and Request for Alternative Rate Plan, Docket No. 2007-00215, Order Approving Stipulation (July 1, 2008). Staff further recommends that this metric and benchmark apply to both CMP’s “business line,” as well as its “contractor line” and that performance be measured separately for each line. Staff also recommends that a service quality metric be established that measures the percentage of calls abandoned by callers to both CMP’s business line as well as its contractor line. The benchmark for this metric is “no more than 7% of calls can be abandoned.” The 7% reflects the mid-point between the “5% to 10% of calls abandoned” industry standard stated by Liberty in its Report. Liberty Report at 89. To measure the adequacy of service being provided by and in the area of billing, Staff recommends that the Commission establish a service quality metric that measures Bench Analysis 96 Docket No. 2018-00194 the “percentage of customer bills issued accurately” each year and that the benchmark for this metric be “no more than 0.4% of bills issued in a calendar year can be erroneous.” This is the same metric that was contained in Bangor Hydro Electric Company’s last Alternative Rate Plan. Bangor Hydro Electric Company Request for Approval of Alternative Rate Plan, Docket 2001-410, Order Approving Stipulation (June 11, 2002). A detailed description of the calculation for each of these metrics is contained in Attachment B. As noted above, the Company’s response to customer requests for field services has been inadequate. The Company has been scheduling field appointments several months out from the date of the customer or contractor request, and has often canceled these appointments at the last minute. Because the Company does not currently track the amount of time it takes for it to respond to the different types of requests for field services, it is difficult to establish a meaningful service quality metric. Thus, Staff recommends that the Company make a proposal in its Rebuttal Case for tracking the various types of customer requests for field services and ultimately establish a service quality metric to measure the Company’s performance responding to these requests. IV. OPERATIONS AND CAPITAL INVESTMENT A. Vegetation Management 1. Background In Docket No. 2013-00168, the Commission included in rates amounts to continue CMP’s 5-year cycle trim program as well as amounts for an enhanced Bench Analysis 97 Docket No. 2018-00194 vegetation management program. The Incidental, Hot Spot and Enhanced Line Clearance programs comprise CMP’s enhanced programs. Table 20 below provides the amount provided in rates in Docket No. 2013-00168, broken down by program. Table 20 49 Dollars in Thousands A In Current Rates (a) 1 2 3 4 5 2. Hot Spot Incidental Enhanced Line Clearance Circuit Maintenance Subtotal 1,029 895 1,800 16,649 20,373 CMP’s Position In its current filing, CMP is asking to include $21.8 million for its total vegetation management program. This is based on a test year actual spend amount of $16.9 million plus a projected $5.0 million to fund expanded programs. Rev. Req. Dir. Exh. RRP 3-12. As proposed, CMP’s vegetation management program would continue the 5-year cycle trim and enhanced programs established in 2014. Additional elements of the proposed plan include implementing a ground-to-sky clearance program and more aggressively targeting hazard trees. In its testimony, CMP claims: A CMP distribution workload study conducted in 2010 indicated that CMP had over 69,000 hazard trees on its system at that time. With the implementation of the enhanced tree program in 2014, CMP has been able to selectively remove some hazard trees off of the right-of-way, but 49 Source: Rev. Req. Dir. Exh. RRP 3-12. Bench Analysis 98 Docket No. 2018-00194 the available funding for this program was limited. The Company is seeking an increase in funding to more aggressively target the removal of off right-of-way hazard trees. Ops. and Cap. Dir. at 8. CMP’s request is only slightly more than the $20.3 million currently in rates. The Company is currently evaluating bids for vegetation management services for the period 2019–2023 solicited in a competitive RFP process. As such, CMP has not yet selected a contractor, nor has it finalized the details of its enhanced program. Ops. and Cap. Dir. at 8–9. 3. Staff’s Proposal CMP argues that it needs additional funding to more aggressively target hazard trees than it has in previous years. CMP cites budgetary constraints for not harvesting more hazard trees. However, a review of past spending levels indicates that CMP was not fully utilizing its budget to maximize removals. As illustrated in Table 21 below, CMP failed to exhaust the amount provided in rates both for the Hot Spot and Enhanced Tree Trimming programs, which both include hazard tree removal in each of the past five years. CMP attributes some of the spending shortfall on the number of major storms it has had to respond to over the past two years. EXM-003-077. Bench Analysis 99 Docket No. 2018-00194 Table 21 50 Hot Spot Pruning Incidental Year Actual Spend 2014 2015 2016 2017 2018 YTD/ NOV 2014-2018 YTD Average 2013-168 Allowed in Rates Variance Actual Spend 2013-168 Allowed in Rates Enhanced Tree Trimming Delta Actual Spend 2013-168 Allowed in Rates $965,324 $892,960 $887,252 $658,190 $160,469 $895,000 $895,000 $895,000 $895,000 $895,000 $70,324 ($2,040) ($7,748) ($236,810) ($734,531) $998,187 $1,027,823 $1,002,548 $835,806 $246,269 $1,029,000 $1,029,000 $1,029,000 $1,029,000 $1,029,000 ($30,813) ($1,177) ($26,452) ($193,194) ($782,731) $1,790,608 $1,798,620 $1,777,562 $1,755,167 $1,108,559 $1,800,000 $1,800,000 $1,800,000 $1,800,000 $1,800,000 ($9,392) ($1,380) ($22,438) ($44,833) ($691,441) $712,839 $895,000 ($182,161) $822,127 $1,029,000 ($206,873) $1,646,103 $1,800,000 ($153,897) In response to ODR-001-050, CMP included $2,728,104 for costs associated with Trouble/Storm in its 2018 YTD total vegetation management costs. Staff does not believe that these costs should be reflected in the vegetation spend since it is Staff’s understanding that they would already be included as part of storm cost expense. In its Rebuttal, CMP should provide 2018 year-end totals for vegetation management expense; confirm that vegetation clean-up costs incurred during a storm are reflected in storm expense; explain significant variances from the amounts allowed in rates; and explain any proposed reconciliations for any underspending. The ECI Report used to support the need for the enhanced vegetation management programs initiated in 2014, provided “a comprehensive study that documented the quantity and characteristics of the existing tree and brush workload on the overhead portion of the IBERDROLA USA distribution system.” EXM-002-010, Att. 1. It included methodologies, projections, analysis and recommendations designed to assist Iberdrola USA in optimizing the management of vegetation on its distribution system and appears to have provided a useful benchmark for CMP from which to base 50 Delta Source: EXM-002-002 and ODR-001-050. Bench Analysis 100 Docket No. 2018-00194 an enhanced vegetation program. Given the extent of the targeted work that CMP has been performing over the past several years, as well as CMP’s proposed expansion of its vegetation management program, Staff recommends that CMP revisit this type of study prior to engaging in additional vegetation management programs to both quantify the progress that has been made on CMP’s system and determine the most costeffective ways to enhance system reliability and storm resiliency going forward. Given that the anticipated cost of the program is in line with what is currently in rates, Staff accepts CMP’s proposed adjustment at this time with the understanding that the matter will be revisited once CMP has finalized the price and scope of services.51 B. Storm Costs and Storm Cost Recovery Mechanism 1. Background As part of its Order in Docket No. 2013-00168, the Commission established a multi-tier mechanism for funding storm restoration. Under that multi-tier approach, Tier 1 storms are defined as storms where the incremental restoration costs are less than $3.5 million per event; Tier 2 storms are storms having incremental restoration costs between $3.5 million and $15 million per event; and Tier 3 storms are those where incremental restoration costs are greater than $15 million. Of the $10 million annually allowed in rates for storm recovery, $4 million is designated for Tier 1 storm costs, while the remaining $6 million is placed into a reserve storm account to cover Tier 2 storm 51 The current vegetation management contract ended December 31, 2018. At the time of the technical conference, CMP was currently in negotiation with vendors for the upcoming cycle. Confidential Tr. at 17–19 (Dec. 4, 2018). Bench Analysis 101 Docket No. 2018-00194 costs. CMP annually reconciles its actual incremental, prudently incurred Tier 2 storm costs against the reserve balance. If the reserve balance at the end of the calendar year exceeds $10 million (positive or negative), CMP and its customers share on a 50/50 basis any such overage, with CMP’s share of any negative balance capped at $3 million per year. For Tier 3 storms, the first $15 million of incremental storm costs are subject to Tier 2 treatment and charged against the reserve account. CMP’s exposure for sharing under the Tier 2 storm provisions for any single Tier 3 storm event is capped at $2 million. Tier 3 storm amounts above $15 million are deferred for future recovery. Distribution rates are adjusted on July 1 of the year following the Tier 3 storm for the recovery of deferred amounts over $15 million. Central Maine Power Co., Request for New Alternative Rate Plan (“ARP 2014”), Docket No. 2013-00168 Order Approving Stipulation at 5 (Aug. 25, 2014). 2. CMP’s Position CMP’s rate proposal seeks to maintain the current storm restoration funding levels set in Docket No. 2013-00168. While CMP is not seeking additional funding for its storm recovery costs, CMP is proposing to lower the ceiling for Tier 1 storms from $3.5 million per storm to $1.5 million per storm. Under this suggested change, Tier 2 storms would be defined as any storm with incremental recovery costs between $1.5 million and $15 million. These costs would continue to be charged against the reserve account. CMP claims it needs to lower the threshold of the Tier 1 storms to more equitably balance the costs and collections for storm events based upon recent historical experience. Ops. and Cap. Dir. at 15. CMP claims that “[d]uring the period Bench Analysis 102 Docket No. 2018-00194 from July 2014 to June 2018 CMP experienced $24.0 million of Tier 1 storm events but only collected $16 million from customers.” Id. 3. Staff’s Proposal In Central Maine Power Company, Annual Price Changes Pursuant to Alternative Rate Plan (ARP 2008), Docket No. 2011-00077, Order at 17 (July 27, 2012), the Commission noted that in future storm costs proposals, CMP and the Staff should consider provisions that remove the incentive for the utility to act in ways that increase the number of interruptions, and should also consider alternatives to the binary nature of the then-current storm cost recovery provision, which either fully included or fully excluded all costs related to a storm event. When CMP’s current storm mechanism was established in 2014, the Commission stated: We find that the Stipulation's Storm Cost Recovery Mechanism addresses these concerns and through its sharing mechanism provides CMP with new incentives to control storm costs that may operate more effectively than the prior storm recovery mechanism. We also find that the increased amount included in rates for storm costs in the Stipulation, along with the Stipulation's Storm Reserve Account Mechanism, should reduce the rate volatility which has resulted from extraordinary storms in the past. Central Maine Power Co., Request for New Alternative Rate Plan (“ARP 2014”), Docket No. 2013-00168, Order Approving Stipulation at 13 (Aug. 25, 2014). The Staff believes that the current storm recovery mechanism has generally achieved these goals, but acknowledges that in the past two years CMP has experienced a significant number of Tier 1 storms. In response, the Staff proposes to maintain the storm mechanism at the current category levels but would include an additional $2.4 million in revenue requirements in this case for Tier 1 storms. This approach allows the Company a Bench Analysis 103 Docket No. 2018-00194 reasonable opportunity to recover its prudently incurred storm cost while preserving the incentives to control storm restoration costs and provide for rate stability, consistent with the Commission’s 2014 Order. At the December 7, 2018 Technical Conference, CMP noted that there have been three Tier 2 storms since October 2018. Mr. Adams stated they were reviewing whether CMP would file a proposal to reset the current reserve account balance to zero. Tr. at 39 (Dec. 7, 2018). The Staff will address the resetting of the reserve account balance if and when CMP files a proposal. C. Inspections 1. Distribution Line Inspection (DLI) Distribution line inspections are an integral part of CMP’s reliability program. CMP’s Distribution Line Inspection (DLI) program began in 2010 and the Company is proposing to maintain its current 5-year inspection cycle in this case. Table 22 below illustrates that CMP has successfully completed its planned inspection cycles over the past four years. Bench Analysis 104 Docket No. 2018-00194 Table 2252 Summary of Pole Inspections, 2014–2017 Year Total Planned Pole Inspections Total Poles Inspected % Inspected 2014 134,665 135,456 100.59% 2015 135,567 136,312 100.55% 2016 134,665 138,883 103.13% 2017 132,629 133,001 100.28% While CMP has achieved the established targets for inspecting its pole plant, the Company has not met its goals in correcting issues found in the remedy periods it has set. Table 23 below shows the number of defects identified by year through the inspection process and repair completion dates. The Company utilizes three priority levels for remedying defects identified; L1, L2, and L3. Prior to 2016, the L1 defects were targeted for repair within the year they were identified. In 2016, CMP reclassified L1 defects to better align the designation to issues posing an immediate safety concern and set the target repair date to seven days. L2 defects were reclassified in a manner consistent with the previous definition of L1, those defects which need to be resolved within a year. Finally, L3 defects need to be resolved during the next inspection cycle. Tr. at 28 (Dec. 7, 2018). As the table below demonstrates, CMP has performed well in 52 Source: Ops. and Cap. Dir. at 12. Bench Analysis 105 Docket No. 2018-00194 addressing the L1 defects post-2016. However, the L2 defects are not being resolved within the scheduled period. Table 23 53 Summary of DLI Defects 2014–2018 YTD 2014 2014 2014 2014 2014 2014 2014 LI-1 FOUND LI-1 COMPLETE LI-2 FOUND LI-2 COMPLETE LI-3 FOUND LI-3 COMPLETE TOTAL 502 1281 6782 Completed in 2014 120 215 312 Completed in 2015 232 552 519 Completed in 2016 108 275 2680 Completed in 2017 36 104 1182 Completed in 2018 6 81 381 Total Completed 502 1227 5074 2015 2015 2015 2015 2015 2015 2015 LI-1 FOUND LI-1 COMPLETE LI-2 FOUND LI-2 COMPLETE LI-3 FOUND LI-3 COMPLETE TOTAL 454 1444 7607 Completed in 2015 94 72 170 Completed in 2016 193 469 996 Completed in 2017 107 421 1728 Completed in 2018 17 116 1291 Total Completed 411 1078 4185 2016 2016 2016 2016 2016 2016 2016 LI-1 FOUND LI-1 COMPLETE LI-2 FOUND LI-2 COMPLETE LI-3 FOUND LI-3 COMPLETE TOTAL 36 1136 4674 Completed in 2016 29 207 437 Completed in 2017 5 412 869 Completed in 2018 2 298 559 Total Completed 36 917 1865 2017 2017 2017 2017 2017 2017 2017 LI-1 FOUND LI-1 COMPLETE LI-2 FOUND LI-2 COMPLETE LI-3 FOUND LI-3 COMPLETE TOTAL 4 809 4874 Completed in 2017 2 241 647 Completed in 2018 2 215 409 Total Completed 4 456 1056 2018 2018 2018 2018 2018 2018 2018 LI-1 FOUND LI-1 COMPLETE LI-2 FOUND LI-2 COMPLETE LI-3 FOUND LI-3 COMPLETE TOTAL 2 710 4996 Completed in 2018 0 125 308 TOTAL TOTAL FOUND COMPLETED 8565 6803 TOTAL TOTAL FOUND COMPLETED 9505 5674 TOTAL TOTAL FOUND COMPLETED 5846 2818 TOTAL TOTAL FOUND COMPLETED 5687 1516 TOTAL TOTAL FOUND COMPLETED 5708 433 At the December 7, 2018 technical conference, CMP suggested that many of the unresolved issues were on telephone utility owned poles. Id. at 30. It appears, 53 Source: EXM-002-017, Att. 1. Bench Analysis 106 Docket No. 2018-00194 however, that only 45 of the 219 unresolved 2016 L2 defects and 70 of the 353 L2 defects identified in 2017 are on telephone set poles. ODR-002-005, Att. 1. Because addressing defects has been identified as one part of improving system reliability and resiliency, it is imperative that CMP develop a plan for addressing these uncured defects. As part of its Rebuttal case, CMP should include its plan on how it intends to address this issue. 2. Infrared Inspections a. CMP’s Position CMP proposes to enhance its infrared inspection (IR) program to include an additional 14,000 miles of single phase primary voltage distribution lines. The current IR program inspects all 3-phase and 2-phase distribution lines. Ops. and Cap. Dir. at 12. The estimated cost to inspect all single phase primary voltage lines is $538,000 per year. Id. CMP is requesting to include half of that amount in the rate year revenue requirement as it does not expect to begin the enhanced program until 2020. Id. CMP states that its current IR program annually detects between 300 and 400 deficiencies. Based on this information, CMP projects that the enhanced program will identify an additional 1,500 deficiencies per year.54 Id. at 13. 54 The criteria listing to prioritize IR defects identified - IR Priority 1 is temperature rise >=126°F immediate action (less than 1 week), IR Priority 2 is temperature rise 72-125°F repair ASAP (expectation within 6 months), IR Priority 3 is temperature rise 36-71°F repair soon (expectation within 1 year), and IR Priority 4 is temperature rise 18-35°F repair at lesser of next scheduled maintenance or 3 years. ODR-002-007. Bench Analysis 107 b. Docket No. 2018-00194 Staff’s Position As discussed above, CMP has not been completing its repairs of known defects under the existing DLI program. In addition, CMP has not increased its budget to account for the repair of the additional defects it states will be identified by the new IR inspection program. Therefore, it is not clear how much of a reliability improvement will result from the expanded inspection program. Furthermore, CMP currently relies on IR inspections on poor performing single-phase lines when other inspection activities fail to find the problem. At the technical conference, CMP acknowledged that other utilities focus their IR inspections on the three-phase portion of their distribution systems. Tr. at 37–38 (Dec. 7, 2018). In this case, CMP appears to currently be following standard industry practice. Wholesale IR inspections on single phase primary lines would be beyond what is common utility practice and does not appear to be necessary or cost-effective since the Company can inspect problematic circuits on a case by case basis. Accordingly, as noted in Section II.B., Staff has removed $ 269,000 from the Company’s proposed revenue requirement. D. Resiliency Investment 1. Company’s Position As part of its Operations and Capital Investment Testimony, the Company submitted what it referred to as its “Capital Resiliency Plan”. Ops. and Cap. Dir. Exh. CAP-3. In it, the Company describes the impact of Maine weather on the ability to Bench Analysis 108 Docket No. 2018-00194 provide reliable service and the likely impact that climate change will have on service interruptions. The Company states that it endeavors to decrease the number of outages experienced by customers and the duration of those outages by developing a portfolio of programs with distinct projects that will harden the system and increase resiliency. The Company states the following criteria were utilized to select the programs and projects that would provide the most benefit for customers: • Focus on worst performing circuits as identified by reliability metrics including major storms (initially, the period 2015–2017); • Harden the distribution infrastructure and address vegetation management to prevent outages from occurring; • Improve system restoration response options through electrical topology changes should an outage occur to decrease the duration and area affected by the outage; • Deploy automated/sectionalizing devices to minimize the customers affected by the outage; • Coordinate with existing reliability, DLI and modernization initiatives to increase the efficiency and effectiveness of the Plan. Id. at 7. CMP groups its Resiliency Plan into four main areas: 1. Topology – Circuit segmentation to shorten distribution feeder lines and reduce customer exposure, Circuit ties to allow distribution feeder lines to be fed from another feeder or substation to reduce the time of the outage, and New substations located closer to the end user thus reducing exposure to the customer. 2. Hardening – Includes pole strengthening as discussed in Section E, Tree wire as discussed in Section D, and Selective undergrounding which will be developed in the future. 3. Automation – Remote terminal devices (RTUs) to be installed on Reclosers and SCADA Switches with Fault detectors. 4. Vegetation Management – See Section A for details. Bench Analysis 109 Docket No. 2018-00194 Id. at 8–10. CMP provided the following cost estimates for the investments associated with its Resiliency Plan over the 2019–2020 timeframe. Table 24 55 Resiliency Planned Capital Forecast by Project for 2019–2020 Description 2019 Forecast 2020 Forecast Total 1 Circuit Segmentation $1,000 $5,000 $6,000 2 Feeder Ties 2,000 5,000 7,000 3 Circuit Upgrades 2,000 2,000 4,000 4 Incremental Automation 3,000 4,000 7,000 5 Total $8,000 $16,000 $24,000 2. Staff’s Proposal As part of discovery, the Company was asked to provide CMP’s actual Resiliency Plan which formed the basis of Exhibit CAP-3. The Company responded that CAP-3 provided an overview of the Resiliency Plan. The Company stated that the Resiliency Projects were currently undergoing an internal review prior to finalizing them and sequencing them into a ten-year plan. The Company stated that it expected to have more detailed project and circuit information by the first quarter of 2019. EXM-002-050. The Company was also asked to provide all cost benefit analyses performed to develop the Resiliency Plan. The Company responded that it had not yet performed any formal financial cost-benefit analysis to develop CMP’s Resiliency Plan. The Company noted that while there was a preliminary projected impact to reliability as 55 Source: Id. at 11, Figure 5. Bench Analysis 110 Docket No. 2018-00194 measured by circuit SAIFI, specific improvements will be determined after final actions of the different programs are defined. In addition, the Company noted that as the Company vets each planning level proposal into detailed design, a formalized costbenefit analysis will be performed. EXM-002-051. As the Commission most recently held in Emera Maine, Request for Approval of Proposed Rate Increase, Docket No. 2017-00198, Order at 13 (June 28, 2018), in order to be included as part of adjusted test year rate base, new investments must be both “known” and “measurable.” Id. To be known, any change in the test year must be reasonably certain as to whether and when it will occur. To be “measurable,” the amount of the change must be reasonably certain. Id. Given the lack of any specificity of the design, timing, and amounts of the Company’s Resiliency Plan investments, the Staff proposes that they not be included in the calculation of revenue requirements at this time. In addition, it is unclear to the Staff why most of the projects classified as “Resiliency Plan Investments” would not already be picked up as part of the Company’s attrition adjustment. Automation, circuit tying, and circuit segmentation are already CMP programs. Further, the Company’s attrition proposal would add $75.1 million in distribution plant in the interim year and $78.5 million in distribution plant in the rate year. Even under Staff’s more conservative attrition calculation, $68 million would be included for plant additions in the interim year and $70.7 million in the rate year. As such, Staff has removed the $4 million and $12 million of discrete incremental plan additions proposed by the Company for the interim year and the rate year. Bench Analysis 111 Docket No. 2018-00194 The Company did provide some detail on two aspects of its hardening plan: the use of covered or tree wire (as opposed to bare wire), and the use of taller and stronger poles. With regard to the Company’s new proposed pole standards, the Company states that its new standard discontinues the use of 30 foot poles, class 4 poles, and class 5 poles. The Company notes that class 3, 2, or 1 wood poles will now be the standard for use and that smaller class poles will no longer be available at any of the Avangrid companies. The Company provided a table with its testimony that sets out the increase in resistance strength that is achieved by going to higher class and taller poles. The Company estimates an annual incremental cost of $370,000 associated with its new pole standards. Based on the information CMP provided, the Staff believes that the Company’s proposal to adopt a standard calling for taller and stronger poles is justified. The Staff believes, however, that the incremental costs associated with this design change can be easily accommodated as part of the attrition adjustment. With respect to tree wire, the Company proposes to use tree wire in all areas where tree encroachment is possible. The Company explained that this means it will only use bare wire in areas like parking lots and wide open fields. Tr. 67–78 (Dec. 7, 2018). CMP states that during the 2015–2017 period, CMP installed tree wire in 35% of its installations. CMP notes that 92% of its roadside distribution system is subject to tree encroachment and, therefore, presumably will now be the subject of CMP’s tree wire conductoring program. Ops. and Cap. Dir. Exh. CAP-4. For the reasons set forth below, and based on the information available to date, the Staff does not support the wholesale use of tree wire that CMP proposes. Bench Analysis 112 Docket No. 2018-00194 First, the Staff would note that in information collected by CMP’s affiliate indicates that 67% of tree related outages occurred within 10 feet of the right of way. EXM-007-013 Attachment 1. This suggests that the use of tree wire far outside the right of way is of limited value. Second, the incremental cost of the tree wire program is substantial. CMP estimates that the costs of the tree wire program will increase conductor costs by approximately 50%. Based on the information provided in response to EXM-002-075, Staff has calculated the average cost difference between bare wire and tree wire to be $6,530 per circuit mile. CMP states that it has 21,734 circuit miles on its system. Based on this cost difference, and the number of miles now served with bare wire, Staff estimates that the incremental cost of converting to tree wire is approximately $129 million. In recommending that the Company’s blanket tree wire program not be adopted at this time, it is important to note that the Staff does not oppose the use of tree wire. However, given the potential cost of the blanket program, Staff believes that CMP should adopt a more targeted approach as recommended in the Cost-Benefit Analysis of the deployment of Utility Infrastructure Upgrades and Storm Hardening Programs Report prepared by Quanta Technology and provided to the Commission during the course of the recently completed investigation of the October 2017 Storm. Pub. Utils. Comm’n, Investigation into the Response by Public Utilities to the October 2017 Storm, Docket No. 2017-00324, ODR-009-007, Att. A. As the Company has recognized, a great deal of money could potentially be spent on storm hardening or for resiliency; ratepayers’ ability to pay for such investments is not unlimited, however. Docket No. 2017-00324, Tr. 22-23 (Setp. 12, Bench Analysis 113 Docket No. 2018-00194 2018). Therefore, as the state’s utilities and the Commission go forward to address this important topic in light of climate change and its impact on weather-related outages, it is important that investments be made in a cost-effective way that maximizes ratepayer benefits. As part of the targeted hardening approach recommended in their report, Quanta Technology recommended that utilities, as part of storm restoration activities, gather forensic evidence of the specific causes of outages and whether and how they could be prevented. The Staff believes that this would be a worthwhile effort and asks CMP to address this issue in its Rebuttal case. V. TAX ISSUES A. Background On December 22, 2017, the President signed into law the Tax Cut and Jobs Act of 2017 (Tax Act). The Tax Act reduced the top corporate income tax rate from 35% to 21% and required utilities to revalue their Accumulated Deferred Income Taxes (ADIT) based on the reduced corporate tax rate. CMP re-measured its ADIT as of December 31, 2017, using the newly enacted 21% federal income tax rate and determined that, at that time, it had Excess ADITs of approximately $276 million.56 CMP agrees that the Excess ADITs should be preserved and returned back to customers. Tax Dir. at 3. Staff notes that in the footnotes to CMP’s Consolidated Financial Statements for the Years Ended December 31, 2017 and 2016, CMP estimated that the Excess ADIT liability reduction due to the reduction in the corporate federal income tax to be $489 million. This is significantly different than the $276 million included in CMP’s testimony. CMP should provide an explanation for and reconciliation of the difference between the $489 million and $276 million in its rebuttal testimony. 56 Bench Analysis 114 Docket No. 2018-00194 CMP’s overall estimate included distribution and transmission Excess ADITs and consisted of both “Protected” and “Unprotected” balances. The treatment of each of these balances are discussed separately below. B. Protected Excess ADIT Protected excess deferred taxes are subject to the Internal Revenue Code’s normalization rules. These rules require the use of the protected balance (primarily the book-tax timing difference related to using accelerated methods and lives for determining tax depreciation versus the straight-line method for determining book depreciation) to reduce customer rates no faster than will occur under the Average Rate Assumption Method (ARAM). CMP determined that the total transmission and distribution Protected Excess ADIT at December 31, 2017 is $259 million, which equals $360.6 million on a grossed-up basis and comprises $123 million for CMP Distribution and $237.6 million for CMP Transmission. CMP determined that under ARAM the distribution portion of the allowed amortization of the excess ADIT was approximately $3 million annually. CMP also noted that, because ARAM amortization could begin January 2018, it will have accumulated $4,529,000 of amortization not yet returned to ratepayers for the period January 2018 through June 2019. CMP is proposing to return this balance, plus the annual amortization of ongoing Protected Excess ADIT of $3,168,000, during the rate year for a reduction of $7,697,000 in the rate year. Tax. Dir. at 5. Staff accepts CMP’s proposal for the amortization of the Protected Excess ADIT which reduces the rate year revenue requirements; however, we note, as reflected in Bench Analysis 115 Docket No. 2018-00194 Table 1 of CMP’s tax testimony, that the use of the $4,529,000 in this case will result in an increase in rates when rates are reset in the future and this reduction to expense is removed. Tax. Dir. At 6. C. Unprotected Excess ADIT While protected Excess ADITs generally result in liabilities to ratepayers, unprotected Excess ADITs result in both refundable and recoverable amounts. CMP has calculated the balance of its Refundable and Recoverable Unprotected Excess ADITs to be $51.1 million and $59.5 million, respectively. CMP proposes a ten-year amortization period for both the Refundable and Recoverable balances. CMP’s proposal results in an increase in its revenue requirements of $841,000 per year. Tax Dir. at 7, Table 2. Staff agrees with CMP’s proposed ten-year amortization period for both the Refundable and Recoverable Unprotected Excess ADITs. Staff discusses the possible use of the Refundable Excess ADITs to mitigate the rate impact of CMP’s proposed revenue requirement deficiency in Section VII, infra. VI. REVENUE ISSUES B. Sales Forecast CMP used an econometric modeling methodology to develop its sales forecast. To prepare the forecast, CMP used economic variables from IHS Market, and weather data from the NOAA. CMP further provided adjustments for projected energy efficiency savings from EMT and developed its own “behind the meter” forecast for the fiscal effects of net energy billing. All of this resulted in “essentially a flat forecast that CMP Bench Analysis 116 Docket No. 2018-00194 believes is reasonable and consistent with overall sales trends in recent years.” Sales Dir. at 2. In general, forecasting does not provide a precise point of what will happen in the future, but instead provides a likely range of possibilities if the current assumptions hold true and the current trends continue. For this reason, it is important that any forecast have an accurate baseline as a starting point. Staff generally agrees with CMP’s conclusions and the shape of its sales forecast, but is concerned about the starting point of the analysis. During the test year (12 months ending in June 2018) and since the end of the test year, CMP has identified a number of significant billing issues. The full extent of these apparent errors has not been quantified. This lack of clarity could have a significant impact on the accuracy of the sales forecast. The effect of these errors should be quantified prior to using this forecast in any type of revenue decoupling mechanism (RDM). B. Revenue Decoupling Mechanism 1. Background RDMs can benefit a utility and its customers by reducing risks associated with sales. Some of these risks are due to factors beyond the utility’s control such as weather, population, and the economy. RDMs also help to mitigate conflicts between a utility’s incentives and public policy objectives, such as with respect to energy efficiency and net energy billing. CMP’s existing RDM was adopted pursuant to Commission approval of a stipulation in CMP’s last base rate case. Docket No. 2013-00168 Bench Analysis 117 Docket No. 2018-00194 Stipulation. Under CMP’s RDM, revenue targets are established for two broad customer classes, Residential and Commercial/Industrial, against which actual revenues are reconciled. Initial revenue targets were established based on the rate year revenue levels established in Docket No. 2013-00168. The revenue targets have been adjusted each year by 75% of the average annual customer growth (positive or negative) in the applicable rate classes. The RDM adjustments occur annually at the same time as other CMP distribution rate changes, such as changes pursuant to the storm cost provisions established by the Docket No. 2013-000168 Stipulation and other one-time adjustments. To date, there have been three adjustments under CMP’s RDM. The adjustments are summarized in Table 25 below: Table 25 57 CMP RDM Adjustments Effective Date Amount (millions) July 1, 2016 ($10.8) July 1, 2017 ($1.0) July 1, 2018 $0.9 In preparation for its most recent annual distribution rate adjustment case (Docket No. 2018-00069), CMP retained David Segal from the Northbridge Group to 57 See Docket No. 2018-00069, March 30, 2018 Test. of David W. Segal at 14 and June 20, 2018 Stipulation Att. 1 at 1. Bench Analysis 118 Docket No. 2018-00194 conduct a complete review of the RDM. CMP did so because of concerns that RDM targets and, thus, surcharges/credits were being calculated incorrectly. Mr. Segal identified three examples of how, in his view, the RDM had been incorrectly implemented. Docket No. 2018-00069 Segal Dir. at 1–3. Docket No. 2018-00069 was ultimately resolved by a stipulation. The Stipulation established an RDM surcharge to ratepayer bills of $862,000 for the July 2018 rate year. Docket No. 2018-00069 Stipulation (June 18, 2018). Regarding whether an RDM for CMP should remain in place, it is relevant that, as detailed in Section III, infra, the Commission is investigating the extent to which CMP may have over-billed, under-billed, or not billed at all large numbers of customers during this period of time. This investigation follows several incidents that suggest potential billing errors by CMP in recent months. This is relevant to any future RDM in key respects, including whether accurate revenue targets can be established and what effect, if any, an RDM might have on CMP’s incentives to ensure that it is billing customers on an accurate and complete basis.58 2. Discussion In this proceeding, CMP is proposing to extend its existing RDM without modification. As Staff understands CMP’s proposal, the initial revenue targets would be established in a manner similar to the prior RDM—that is, based on the rate year revenue levels set in this proceeding. In support of its rate year sales projections, CMP This issue may also be relevant to the integrity of CMP’s rate year sales projections. Staff’s recommendation with respect to any adjustment to rate year sales to correct for billing errors will be provided in the Examiners’ Report. 58 Bench Analysis 119 Docket No. 2018-00194 has provided the testimony of Mr. Purtell and Mr. Hastings. As Mr. Purtell and Mr. Hastings note, CMP used econometric models to forecast sales. Sales Dir. at 16. CMP’s rate year sales estimates are predicted to grow or decline relative to historic sales. In this case, the accuracy of CMP’s historic sales data and, thus, its sales forecasts, may be in question, given the billing issues noted above. In establishing an RDM, it is important that revenue targets are accurate and not skewed by the types of billing issues that are the subject of the ongoing investigations. Due to the outstanding billing errors, it may not be possible to set accurate targets at this time. If it is not possible to do so, Staff would not support continuation of an RDM. In addition, the Staff is concerned that RDM could mask and mitigate management errors related to billing. For example, as discussed in Section III, infra. CMP has failed to bill large numbers of customers over several months. With an RDM, the Company may not have an incentive to catch and remediate such problems, given that the revenue shortfalls from not billing customers would be automatically recovered from other customers through the RDM absent any steps to preclude such recovery. Finally, if an extended or successor RDM is to be implemented for CMP, Staff recommends that the structure of the RDM be significantly simplified. As noted above, the structure of the existing RDM seems overly complicated and, perhaps because of this, has resulted in errors. For example, the existing RDM requires multiple detailed steps to include and exclude “one-time adjustments”; these steps seem overly complex and unnecessary. The potential for RDM-related errors going forward and the complexities related to one-time adjustments may be mitigated if targets are set and Bench Analysis 120 Docket No. 2018-00194 reconciled against using kWhs and kWs rather than revenue. For example, an RDM adjustment could simply reflect the difference between actual and target kWh given the actual distribution rates in effect during the applicable period. Staff seeks CMP’s response to such an approach in its rebuttal filing. VII. MITIGATION As discussed in Section I, infra., CMP has proposed accelerating the amortization of regulatory liabilities to offset the impact of its proposed revenue requirement deficiency. The Staff has proposed several adjustments and modifications to CMP’s revenue requirement calculations, which may make mitigation unnecessary. As such, the Staff is not specifically proposing mitigation as part of this Bench Analysis. Ultimately, the determination of whether mitigation through the acceleration of the amortization of regulatory liabilities is a wise course will depend on the revenue requirement that ultimately comes out of this process, the potential impact of any revenue requirement deficiency on current rates, and the impact that acceleration of amortizations will have on future rates. If mitigation is appropriate, Staff believes that such mitigation can best be accomplished through the acceleration of the Cost of Removal Liability given the size of that liability and its current amortization period. VIII. CONCLUSION Incorporating the Staff positions discussed above as well as the recommended management efficiency adjustment, Staff calculates a total revenue requirement between $262.645 million and $264.262 million depending on the amount of the Bench Analysis 121 Docket No. 2018-00194 management efficiency adjustment.59 This represents a revenue decrease between $1.979 million (-0.74%) and $3.596 million (-1.35%) from projected rate year revenue levels. As noted earlier, these rate year revenue levels include one-time adjustments which are slated to be removed on July 1, 2019. Dated: February 22, 2019 Respectfully submitted, ________________________ Charles Cohen Hearing Examiner Katie Gray Hearing Examiner Brian George Hearing Examiner On Behalf of Advisory Staff: Faith Huntington Derek Davidson Christine Cook Staff notes that the revenue requirement as set forth in the Bench Analysis was calculated using the models provided by CMP, but revised to reflect the recommended decisions in this Bench Analysis. To the extent the Company believes that there are calculational errors, CMP should identify those errors in its rebuttal case and provide corrections. Staff notes that schedules J and K of Exhibit RRP-2 contain hard coded values which have not been updated. In addition, as a result of the Company providing workpapers with external links broken, Staff believes there is a possibility that not all references to the distribution allocators have been updated as necessary. Staff provides its workpapers as Appendix C. 59 Bench Analysis 122 Docket No. 2018-00194 Lucretia Smith Michael Simmons Sally Merritt Matthew Rolnick Margrethe Heimgartner