ISBN 978-1-869455-02-6 Project no. 16.05/15095 Public version Review of the state of competition in the New Zealand Dairy Industry Final Report The Commission: Sue Begg, Deputy Chair Dr Stephen Gale Anna Rawlings Graham Crombie Date of publication: 2440192 1 March 2016 2 Associated documents Publication date Reference Title 6 November 2015 ISBN 978‐1‐869454‐ 75‐3 Review of the state of competition in the New Zealand dairy industry: Draft report 20 July 2015 ISBN 978‐1‐869454‐ 62‐3 Consultation on substantive issues – review of competition in the dairy industry 12 June 2015 ISBN 978-1-86945457-9 Review of the state of competition in the New Zealand dairy industry: Consultation paper – process and approach Commerce Commission Wellington, New Zealand Confidential material in this report has been removed. Its location in the document is denoted by [ ]. 2440192 3 Contents SUMMARY OF FINDINGS........................................................................................ 6 Purpose of report .............................................................................................. 6 Stage 1: Assessment of the state of competition with and without the DIRA Regulation ............................................................................................ 10 Stage 2: Our assessment of the efficiency costs and benefits of the DIRA Regulation ............................................................................................ 20 Stage 3: Resetting the expiry triggers and transition pathways to deregulation we identified ................................................................... 23 Summary of our key findings ........................................................................... 25 CHAPTER 1: INTRODUCTION ............................................................................. 26 Purpose of this report...................................................................................... 26 Structure of this report .................................................................................... 26 Scope of this report ......................................................................................... 27 CHAPTER 2: BACKGROUND TO DIRA AND THE RAW MILK REGULATIONS........... 29 Purpose of this chapter ................................................................................... 29 History and scope of the DIRA Regulation ...................................................... 29 CHAPTER 3: PROCESS AND FRAMEWORK FOR EVALUATING AND REPORTING ON THE STATE OF COMPETITION ......................................................... 33 Purpose of this chapter ................................................................................... 33 Our interpretation of ‘sufficient’ and ‘insufficient’ ......................................... 33 The conceptual framework for our analysis .................................................... 34 Options to explore if we conclude that the DIRA Regulation enhances efficiency .............................................................................................. 36 Consultation on our evaluation approach ....................................................... 37 Identifying the relevant domestic dairy markets ............................................ 40 How we carried out our analysis of the state of competition......................... 42 We considered expiry triggers and pathways to deregulation ....................... 44 CHAPTER 4: STATE OF COMPETITION UNDER THE DIRA REGULATION ................ 46 Purpose of this chapter ................................................................................... 46 Key findings on the state of competition in farm gate market ....................... 47 Key findings on the state of competition in the factory gate market ............. 47 Farm gate market—assessing the current state of competition .................... 49 Farm gate market—conditions of entry and expansion.................................. 49 Farm gate market—DIRA Regulation has lowered barriers to entry and expansion ............................................................................................. 51 Farm gate market—independent rivalry ......................................................... 59 2440192 4 Farm gate market—Fonterra’s co-operative structure ................................... 71 Factory gate market—current state of competition ....................................... 71 Overview of the factory gate market—raw milk ............................................. 72 Factory gate market—current customers for DIRA milk ................................. 73 Factory gate market—barriers to entry .......................................................... 76 Factory gate market—independent rivalry ..................................................... 77 Factory gate market—constraints on Fonterra’s market power .................... 78 Wholesale and retail supply of fresh processed milk—any new information on state of competition? ...................................................................... 78 CHAPTER 5: STATE OF COMPETITION WITHOUT THE DIRA REGULATION ........... 81 Purpose of this chapter ................................................................................... 81 Key findings on future competition in the farm gate market ......................... 81 Key findings on future competition in the factory gate market...................... 82 Key findings on Fonterra’s incentive and ability to render competing IPs less competitively effective ......................................................................... 82 Fonterra’s buyer side market power in relation to farmers ........................... 83 Fonterra’s impact on IPs without the DIRA Regulation ................................... 86 Fonterra’s market power at the factory gate .................................................. 87 Does Fonterra have market power at the factory gate? ................................. 92 The factory gate milk price may vary across regions ...................................... 94 Fonterra’s incentive and ability to render IPs less competitively effective .. 100 CHAPTER 6: OUR ASSESSMENT OF THE EFFICIENCY COSTS AND BENEFITS OF THE DIRA REGULATION ....................................................................... 116 Purpose of this chapter ................................................................................. 116 Key findings on the sufficiency competition ................................................. 116 The efficiency costs and benefits that accrue from the DIRA Regulation ..... 118 Asymmetric risk to removing DIRA ................................................................ 134 CHAPTER 7: PATHWAYS TO DEREGULATION ................................................... 136 Purpose of this chapter ................................................................................. 136 Our recommendations .................................................................................. 136 Why we identified transition pathways to deregulation .............................. 137 We recommend resetting the market share thresholds of the DIRA ........... 138 A time limit provision is appropriate to accompany the market share thresholds ........................................................................................... 144 We do not recommend additional or alternative expiry triggers ................. 146 Options for transition pathways to deregulation .......................................... 148 ATTACHMENT A: TERMS OF REFERENCE FOR A REPORT ON THE STATE OF COMPETITION IN THE NEW ZEALAND DAIRY INDUSTRY................ 164 2440192 5 Purpose of the terms of reference ................................................................ 164 Context .......................................................................................................... 164 Objectives and scope ..................................................................................... 165 ATTACHMENT B: SUMMARY OF THE KEY PROVISIONS OF THE DIRA REGULATION AND THE RAW MILK REGULATIONS.............................................. 167 ATTACHMENT C: 2012 RAW MILK REGULATIONS AMENDMENTS ...................... 171 Purpose of this attachment ........................................................................... 171 The 2012 DIRA amendments focused on the Raw Milk Regulations ............ 171 What has happened since the restriction and monthly limit amendments . 172 ATTACHMENT D: MARKET STRUCTURE ............................................................. 174 Purpose of this attachment ........................................................................... 174 Milk production and growth .......................................................................... 175 Sources of Fonterra’s milk ............................................................................. 180 Market shares and processing capacities of Fonterra and IPs ...................... 181 IP growth and entry over time ...................................................................... 183 Other co-operatives ....................................................................................... 183 Other IPs ........................................................................................................ 184 Own-source IPs subject to the sunset clause ................................................ 184 ATTACHMENT E: MARKET DEFINITION ............................................................. 187 Purpose of this attachment ........................................................................... 187 Hypothetical monopolist, or SSNIP test ........................................................ 187 Farm gate markets for raw milk .................................................................... 188 Factory gate market for raw milk .................................................................. 189 ATTACHMENT F: EVIDENCE THAT INFORMED OUR CONCLUSIONS ON EXCESS CAPACITY AND RESETTING MARKET SHARE THRESHOLDS ............ 193 Purpose of this attachment ........................................................................... 193 Evidence to support out recommended market share threshold expiry trigger ................................................................................................. 198 Historic observation and judgement on the future to project market shares198 2440192 6 Summary of findings Purpose of report X1 X2 This report details the Commerce Commission’s (the Commission’s) findings on the state of competition in the New Zealand dairy industry. It assesses: X1.1 the competitiveness and contestability of the relevant markets with and without regulation under subparts 5 and 5A1 of the Dairy Industry Restructuring Act 2001 (the DIRA) and the Dairy Industry Restructuring (Raw Milk) Regulations 2012 (Raw Milk Regulations) 2 (together the DIRA Regulation). In particular, it considers Fonterra’s ability and incentive to exercise substantial market power under each scenario; X1.2 whether the relevant markets would be more efficient without all or some of the DIRA Regulations. In particular it considers the efficiency costs and benefits of the DIRA Regulations and the risks associated with removing the DIRA Regulations too soon or too late; and X1.3 the market conditions which would trigger further review of the state of competition in the New Zealand dairy industry in the future and potential options for a pathway to deregulation of the industry. The Minister for Primary Industries (the Minister) requested this report under section 148A of the DIRA, which set 1 June 2015 as a trigger for a review of the state of competition in the dairy industry. The Minister issued us with terms of reference which determined the scope of our report and informed our approach. 3 Key findings X3 We recommend keeping the DIRA Regulation in place at this time. We do not think that competition is sufficient to ensure the efficient and contestable operation of the relevant dairy markets if the DIRA Regulation was removed. X4 In particular, it is not clear that removing subparts 5 and 5A would improve efficiency, and our analysis suggests that efficiency could be reduced by immediately removing the Raw Milk Regulations. 1 Subpart 5 relates to the open entry and exit provisions (including the 20% Rule) and subpart 5A relates to the milk price regime. 2 The Raw Milk Regulations relate to the requirement on Fonterra to supply factory gate raw milk to IPs at the DIRA milk price. 3 Ministry for Primary Industries “Terms of reference for a report on the state of competition in the New Zealand dairy industry” (3 June 2015). Available at: http://www.comcom.govt.nz/regulatedindustries/dairy-industry/report-on-the-state-of-competition-in-the-new-zealand-dairy-industry/. 2440192 7 X5 The efficiency costs and benefits of the DIRA Regulation are finely balanced; however the risks of removing it too soon outweigh the risks of it remaining longer than it should. X6 We consider that competition in the factory gate market4 is very limited. Without the DIRA Regulation Fonterra would be able to increase the price of raw milk it sells to other domestic processors, referred to in this report as independent processors (IPs). This would likely result in higher prices for dairy products in downstream domestic markets. X7 Fonterra also has buyer side market power in the purchase of raw milk at the farm gate. Buyer side market power gives Fonterra the ability to depress the price paid to farmer suppliers for raw milk below competitive levels. However, we do not think that Fonterra would exercise this market power against its farmers despite its ability to do so. X8 Fonterra may have incentives to restrict exit and re-entry of farmer suppliers without the DIRA Regulation. Any such restriction of farmer suppliers may increase the barriers to entry to the farm gate market by new IPs, or expansion by existing IPs seeking to obtain raw milk from those farmer suppliers. X9 We consider that entry and expansion of IPs in the farm gate market may assist with the development of competition in the factory gate market. Given our concern with the lack of competition in the factory gate market, we therefore do not recommend removing the open entry and exit provisions at this time. X10 We consider the pathway to deregulation would be smoothed by facilitating the development of a factory gate market for non-DIRA milk.5 We recommend that the Minister considers the following options for changing the Raw Milk Regulations to facilitate such a development: X10.1 reduce IPs’ entitlements to DIRA milk;6 X10.2 tighten the terms and conditions for the supply of DIRA milk to IPs by Fonterra; or X10.3 reduce Goodman Fielder’s entitlement to DIRA milk. 4 The factory gate market is the market in which Fonterra (and a limited number of other processors) supply raw milk they have collected from farmers to other processors and some food and beverage manufacturers. 5 Non-DIRA milk is milk supplied to IPs outside the provisions of the Raw Milk Regulation. 6 DIRA milk is milk supplied to IPs by Fonterra under the terms of the Raw Milk Regulations. 2440192 8 X11 We recognise that a well-functioning factory gate market may not emerge and if that is the case, the pathway to deregulation would need to be re-assessed. X12 We think that keeping the open entry and exit provisions in place during any reforms to the Raw Milk Regulations may help smooth the transition: X12.1 They may help ensure the development of a factory gate market by facilitating further entry to, or expansion in, the farm gate market by IPs that could then enter the factory gate market. X12.2 They may also allow for IPs to increase their own-sourcing of raw milk direct from farmers, which would mainly protect new and less developed IPs. X13 We consider the Minister should explore further the option of removing open entry for new conversions. Open entry for new conversions contributes little to supporting competition in the farm gate market and potentially imposes costs on Fonterra. However we have not found evidence to suggest that these provisions are currently imposing material costs. X14 We do not recommend including any additional or alternative expiry triggers. However we recommend resetting the expiry triggers in DIRA so that the next review of competition is triggered by whichever comes first: X14.1 the time limit provision: 2021/2022 season; or X14.2 the market share threshold: IPs have collected 30% of available raw milk from the farm gate market (considered separately for the North and South islands). Background to DIRA and Raw Milk Regulations X15 7 DIRA authorised the amalgamation of New Zealand's two largest dairy co-operatives at the time7 with the New Zealand Dairy Board to form Fonterra – a single cooperative company. New Zealand Dairy Group and Kiwi Co-operative Dairies. 2440192 9 X16 DIRA also introduced legislation to promote the efficient operation of dairy markets in New Zealand by regulating the activities of Fonterra. The regulation currently includes: X16.1 subpart 5 of the DIRA – which sets out provisions allowing farmer suppliers open entry and exit to supply Fonterra; X16.2 subpart 5A of the DIRA – which sets out Fonterra’s obligations in relation to setting the base milk price (the price it pays farmers for supplying milk). This includes the requirement on Fonterra to maintain a Milk Price Manual which the Commission must review and report on. We refer to this as the ‘milk price regime’; and X16.3 Dairy Industry Restructuring (Raw Milk) Regulations 2012 – which sets out the requirement for Fonterra to supply raw milk to IPs.8 X17 Fonterra is the only firm that is subject to regulation under DIRA. X18 The DIRA Regulation was designed to be temporary. To reflect this, a review of competition in the dairy market is triggered when either a volume limit9 or time limit is reached.10 The time limit was reached on 1 June 2015, triggering the current review of competition. Assessment framework and process X19 We set out and sought feedback on our assessment framework and evaluation approach in June 2015.11 Most submissions were supportive of it. The three key elements of our framework are set out below. The outcome of our analysis of stage 2 determined the need to consider stage 3: X19.1 stage 1: assess the state of competition both with and without the DIRA Regulation. This included considering whether there was any new information in relation to wholesale markets which altered the conclusions we reached in our 2011 preliminary inquiry into domestic milk markets;12 8 These regulations are not contained in the DIRA, but were made under section 115 of the DIRA. 9 This is when a 20% farm gate market share is held by IPs in the North Island or the South Island (or both). 10 Dairy Industry Restructuring Act (2001), s 147. 11 Commerce Commission “Review of the state of competition in the New Zealand dairy industry: Consultation paper – process and approach” 12 June 2015. 12 Ministry for Primary Industries “Milk markets: Consideration of whether to initiate a Commerce Act Part 4 inquiry into milk prices” 2 August 2011 . 2440192 10 X19.2 stage 2: if competition is sufficient to ensure contestable and efficient markets, recommend deregulation. If not, assess the efficiency costs and benefits of the DIRA Regulation to determine if it is appropriate to remove some of the DIRA Regulation; and X19.3 stage 3: if appropriate, reset the expiry triggers and identify transition pathways to deregulation. X20 To review the state of competition without DIRA Regulation we considered whether Fonterra would have market power if DIRA Regulation was removed. This is because if a firm has market power it may use that to lessen the extent of competition in a market. Fonterra has market power if it has the ability to raise prices above competitive levels, or engage in other activity that impacts the competitiveness of the market. Considering Fonterra's market power with and without DIRA Regulation enables us to assess whether the state of competition in the relevant markets is sufficient to ensure the efficient and contestable operation of those markets without DIRA Regulation. X21 In assessing competition we focused on two relevant markets: the farm gate market and the factory gate market (for DIRA and non-DIRA milk). Stage 1: Assessment of the state of competition with and without the DIRA Regulation Farm gate – market definition X22 The farm gate market is the market in which dairy processors compete to secure the supply of raw milk from farmers. Fonterra faces competition in this market from IPs that source their milk directly from farmers (‘own-source’ IPs). X23 Fonterra’s national market share of milk collected at the farm gate has declined from around 96% in the 2001/02 season to around 85% in the 2014/15 season.13 However its market share figures vary regionally. Farm gate – key findings X24 13 Table X1 summarises our key findings on farm gate market outcomes with the DIRA Regulation and the likely market outcomes without the DIRA Regulation. Commission analysis based on information provided by Fonterra and market participants. 2440192 11 Table X1 Farm gate market outcomes with and without the DIRA Regulation Farm gate market outcomes With the DIRA Regulation  DIRA has helped limit Fonterra’s market power and reduce barriers to entry and expansion for IPs.  Competition varies across regions. Without the DIRA Regulation  Fonterra would have the ability but little or no incentive to exercise market power against farmers by offering them lower prices for raw milk.  Fonterra may have the incentive to restrict IPs from accessing farmers but we consider it to be limited in its ability to do so.  Fonterra would be unlikely to make significant changes to how the milk price is set. Farm gate – market outcomes under the DIRA Regulation The DIRA Regulation has helped limit Fonterra’s market power and reduce barriers to entry and expansion for IPs X25 14 Since 2001, when the first DIRA Regulation came into effect there has been significant new entry and expansion. We consider the DIRA Regulation to have contributed to the growth in IPs by helping to reduce the barriers to entry and expansion that they face. This results in more choice for existing dairy farmers and new conversions. In particular: X25.1 the open entry and exit provisions have helped IPs access raw milk from farmer suppliers by reducing the risks to farmers who leave Fonterra to supply IPs. Farmers can freely exit Fonterra to supply an IP, and they can easily leave the IP and return to supply Fonterra if they choose. The open entry and exit provisions are seen by many IPs as the cornerstone of DIRA; X25.2 access to DIRA milk along with the open entry and exit provisions has helped IPs achieve a commercially viable scale during their early years by assuring them access to a prescribed volume of raw milk for processing; and X25.3 the 20% Rule14has been important for small IPs, particularly small cheesemakers, by enabling farmers to supply an IP while remaining a Fonterra supplier. The 20% Rule is part of subpart 5. Fonterra is required to allow its farmer suppliers to supply IPs up to 20% of their milk production. 2440192 12 X26 Fonterra’s price setting disclosure obligations under the DIRA Regulation are widely accepted as increasing transparency of raw milk pricing. We consider this to have contributed to reduced barriers to entry and expansion for IPs who seek to compete with the price Fonterra pays for raw milk at the farm gate. Competition varies between regions X27 The level of competition Fonterra faces for the collection of milk varies across different regions—in some regions there is no, or very limited farm gate competition. Canterbury is the most competitive region, where Fonterra has a market share of [ ]. In comparison, Northland is the least competitive region where Fonterra has a market share of [ ].15 X28 The most significant competition is in regions where there has been increased milk production.16 Increased milk production has contributed to the growth of IPs. In these regions Fonterra faces competition at the farm gate both from existing IPs and from the threat of new IPs entering the market. Farm gate – market outcomes without the DIRA Regulation Fonterra would have the ability but little or no incentive to exercise market power against farmers by paying them lower prices for raw milk X29 Without the DIRA Regulation, Fonterra would have the ability to exercise market power over farmers by lowering the price paid to them for raw milk below the competitive level. This may result in farmer exit or cutting back of supply.17 X30 However, we think that Fonterra would have little or no incentive to exercise its market power in this way despite its ability to do so. This is because: X30.1 Fonterra is a co-operative owned by its farmer suppliers. This ownership structure protects farmer suppliers from Fonterra’s market power in relation to milk purchases, because the suppliers ultimately control the decisions of the company. X30.2 Even without the milk price regime under DIRA, we consider it likely to be in Fonterra’s interests to provide a degree of transparency and independent oversight of base milk price setting. This would provide assurance to farmer shareholders, external shareholders, and contract suppliers that the farm 15 We note that Fonterra does not have any market share in the West Coast, where Westland is the sole IP. 16 These are Canterbury, Southland, Taranaki and Waikato. 17 Cutting back of supply in this way is known as ‘retrenchment’. 2440192 13 gate price of milk is set at an efficient level, which is important for the success of Fonterra’s Trading Amongst Farmers (TAF) scheme.18 X31 In general, we do not think that Fonterra would likely have the ability to offer new or returning farmers lower prices for the milk that they supply to Fonterra than the price that it offers to existing farmer suppliers. Farmers would generally only supply to Fonterra if the prices and terms offered were as good as, or better than, those offered by competing IPs.19 Where there is little or no competition, Fonterra’s national pricing policy would to some extent help protect new or returning farmers from lower prices. Fonterra may have the incentive to restrict IPs from accessing farmers but we consider it to be limited in its ability to do so X32 Our analysis suggests that without the DIRA Regulation, Fonterra may have an incentive to restrict IPs from accessing farmers and therefore obtaining milk at the farm gate (referred to as foreclosure of rival IPs). Fonterra could attempt to do this by either: X32.1 increasing the price paid to farmers for raw milk at the farm gate so that IPs cannot match or better Fonterra’s price; X32.2 locking farmers into long-term contracts; or X32.3 exercising the board’s discretion under the constitution to prevent exit or re-entry. 18 Trading Among Farmers (TAF) was introduced into the DIRA regulations through the Dairy Industry Restructuring Amendment Act 2012 Sections 109A-N. Fonterra introduced TAF primarily to reduce redemption risk created by the open entry and exit provisions. In creating TAF Fonterra effectively changed its capital structure from having redeemable capital (shares that Farmers can sell back to Fonterra) to having permanent capital (shares traded between farmers and outside investors). 19 Without the protections of open entry and exit provisions, farmers who lose contracts with IPs could possibly no longer rely on Fonterra as the “processor of last resort”. 2440192 14 X33 X34 Fonterra’s ability to render rivals less competitive by increasing the farm gate milk price is largely the same with and without regulation. Even with the regulation Fonterra could pay a higher price than that which results under the Milk Price Manual.20 However, Fonterra’s ability, both with and without the regulation, is constrained due to: X33.1 the importance to Fonterra of its TAF scheme;21 X33.2 a need to protect the divergent interests of its supplier shareholders, outside investors and contract suppliers;22 and X33.3 a need to also fund its capital programme and maintain a sustainable business.23 Fonterra’s ability to exercise market power over its rivals by locking its farmer suppliers into long-term contracts is not clear. It will likely be limited in the extent to which it can exercise market power against well-established IPs through long-term contracts or constraints on farmer exit. This is because such IPs already have relationships (including long-term contracts) with their own suppliers. However Fonterra is better placed to adversely impact new and less well-established IPs. There is a risk it may seek to do so in order to maintain and grow its co-operative base and its farmers may be less inclined to view such contracts as being outside their best interests. Fonterra would be unlikely to make significant changes to how the milk price is set X35 We do not think that Fonterra would make significant changes to how the milk price is set if the regulations were removed. This is because it is in Fonterra’s interests to assure its stakeholders that the milk price is set at an efficient level.24 X36 Removal of the milk price regulations may technically give Fonterra more discretion to raise the price of farm gate milk. However, we think that the milk price monitoring regime has improved transparency, and even without regulation, stakeholders would seek continued disclosure of information. This disclosure, and 20 This is because Fonterra is free to deviate from the milk price produced by the calculation under the milk price manual, although under DIRA it is required to publicly disclose its reasons for doing so. 21 In order for the TAF scheme to be successful, investors need to have confidence in the milk price. This means that it is important that they are confident that the milk price is calculated in a transparent way. 22 Fonterra needs to assure outside investors and shareholders who hold ‘dry’ shares that dividends will not be compromised through setting an inefficient milk price. Fonterra is therefore constrained in its ability to increase the milk price above the competitive level at the expense of the dividend paid. 23 This may be difficult to sustain if it Fonterra was also paying a market foreclosing milk price. 24 Fonterra’s milk price setting process and governance of the milk price is currently codified in its constitution. 2440192 15 potential criticisms from stakeholders, would continue to constrain Fonterra’s setting of the milk price. As a result, we do not think there would be a significant increase in the price of farm gate milk or a significant change in the level of transparency of Fonterra’s milk price setting process. Factory gate – market definition X37 The factory gate market is the market in which Fonterra (and a limited number of other processors) supply raw milk they have collected from farmers to other processors and some food and beverage manufacturers. It is very small in comparison to the farm gate market–only [ ] of Fonterra’s farm gate milk was traded at the factory gate in the 2014/15 dairy season. X38 The factory gate market consists of DIRA milk sold under the Raw Milk Regulations and non-DIRA milk sold outside of them. By definition Fonterra is the only processor that supplies milk to the factory gate market for DIRA milk. Fonterra also supplies the vast majority of non-DIRA factory gate milk. X39 There are also two distinct types of customers for factory gate milk: X39.1 large IPs that also source their milk directly from the farm gate (‘ownsource’ IPs) where factory gate supply is an interim measure to help the IP achieve scale quickly; and X39.2 IPs that, for the most part, do not source milk from the farm gate (this includes Goodman Fielder and all major domestic dairy product manufacturers). Factory gate – key findings X40 The factory gate market is a regional market. None of these regional factory gate markets are characterised by effective competition. With the current DIRA Regulation there is limited participation by IPs in the factory gate market for nonDIRA milk. X41 Further to this, we are concerned that without the DIRA Regulation, Fonterra could exercise market power to raise the price for milk it supplies to IPs at the factory gate (factory gate price) above competitive levels. This could adversely affect the price of outputs in downstream domestic markets. X42 Table X2 summarises our key findings of factory gate market outcomes under the DIRA Regulation and the likely market outcomes without the regulations. 2440192 16 Table X2 Factory gate market outcomes with and without the DIRA Regulation Factory gate market outcomes With the DIRA Regulation  Factory gate DIRA milk supports IP entry to the farm gate and downstream markets.    Without the DIRA Regulation  DIRA milk may hinder the development of competition in the factory gate market. Fonterra is likely to raise factory gate prices above competitive levels – this may vary across regions and/or by buyer.  There is a separate, small and highly concentrated factory gate market for non-DIRA milk. Factory gate price increases would have a negative impact on downstream markets.  Fonterra would have the incentive but limited ability to render domestic downstream rivals less competitive.  Fonterra would have limited incentive and limited ability to render IPs who export less competitive. The price for non-DIRA milk generally exceeds the DIRA milk price. Factory gate – market outcomes under the DIRA Regulation Factory gate DIRA milk supports IP entry to the farm gate and downstream markets X43 Fonterra is the only supplier of DIRA milk. The supply of DIRA milk by Fonterra to IPs at a price linked to the farm gate milk price paid by Fonterra to farmers constrains Fonterra’s market power at the factory gate and supports IP entry into the farm gate and downstream markets. This helps to support competition in farm gate and downstream domestic markets: X43.1 Farm gate: The availability of DIRA milk under the Raw Milk Regulations guarantees supply of raw milk to IPs at the factory gate. This milk supplements IPs’ own supply from farmers, helping them to operate more efficiently and better enable them to reach efficient operating capacity. It helps new entrants to establish themselves and, in so doing, build up the reputation needed to attract own supply at the farm gate. X43.2 Downstream domestic markets: many IPs that serve these markets are too small to source milk directly from the farm gate. DIRA milk helps ensure that IPs who manufacture dairy products for the domestic markets are able to secure a supply of raw milk at a competitive price. 2440192 17 X44 A well-developed farm gate market characterised by a number of own-source IPs could support the future development of a factory gate market. This is because once an IP has a good farm gate supply there are unlikely to be significant barriers to selling that milk at the factory gate. Therefore the development of competition in the farm gate market can be viewed as a pre-requisite for a functioning factory gate market. DIRA milk may hinder the development of competition in the factory gate market X45 Access to DIRA milk may be hindering the development of a factory gate market for non-DIRA milk in the following ways: X45.1 If the price does not reflect the opportunity cost of supplying the DIRA milk25 IPs may not be willing to supply the small volumes that some factory gate customers require. X45.2 The benefits IPs receive from being able to access DIRA milk (for example, the DIRA milk price, the tolerance limits and guaranteed supply) may reduce the incentive to seek supply outside the DIRA. There is a separate, small and highly concentrated factory gate market for non-DIRA milk X46 Most factory gate milk is DIRA milk. However, a small amount of non-DIRA milk is traded. We consider this to be a separate factory gate market. The factory gate market for non-DIRA milk is highly concentrated. Fonterra is the dominant supplier with relatively small volumes supplied by other IPs. X47 Many of the larger IPs that could potentially supply milk in the non-DIRA factory gate market are unwilling to do so at the DIRA milk price, but they may be willing to supply at the non-DIRA milk price.26 The price for non-DIRA milk generally exceeds the DIRA milk price X48 The non-DIRA price is very volatile. It has ranged between [ ] higher than the DIRA price. When the price of non-DIRA milk exceeds Fonterra’s opportunity cost, Fonterra may be exercising market power. 25 We think that the DIRA price, for the most part, reflects Fonterra’s average opportunity cost of supplying raw milk at the factory gate. However at times it may be above or below opportunity cost. 26 “Interview with Commerce Commission: Review of the state of competition in the New Zealand dairy industry” [ ] 25 June 2015; [ ], 9 July 2015; [ ], 21 July 2015. [ ] 2440192 18 Factory gate – market outcomes without the DIRA Regulation Fonterra is likely to raise factory gate prices above competitive level – this may vary across regions and/or by buyer X49 Without the DIRA Regulation, the price Fonterra charges at the factory gate could increase. The non-DIRA price is likely indicative of the price it would charge. This price rise could occur for two reasons: X49.1 to cover any opportunity costs not currently covered by the DIRA price; and X49.2 as an exercise in market power by Fonterra. X50 It would be difficult to isolate the effect of each type of price rise. We would not be concerned about a price increase to cover any opportunity costs not currently covered by the DIRA price. We think that the DIRA price, for the most part, reflects Fonterra’s average opportunity cost27 of supplying raw milk at the factory gate although at times it may be above or below opportunity cost. Therefore a price increase without the DIRA Regulation could be a reflection of Fonterra’s market power. X51 A factory gate price increase would result in less milk being bought by IPs at the factory gate. This would lead to a decrease in the output of such IPs, which in turn could adversely impact the price, quantity and, potentially, the quality and variety of dairy products sold in domestic downstream markets. The extent of the adverse impact is unknown and would likely vary by product market. X52 The factory gate price is likely to vary across regions and buyers. IPs that have no viable alternative source of milk at the factory or farm gates would likely face higher prices than those where there are viable alternatives. X53 The price increase faced may also vary by buyer. Buyers who are sensitive to price increases28 are less likely to be able to sustain a price increase. Fonterra therefore may be able to profitably price discriminate across buyers. In this case, buyers with few or no alternatives and little ability to sustain a price increase may face an increase they can pay rather than an increase that forces them to exit the market. 27 The volatility of commodity prices makes it hard to be definitive about opportunity costs. 28 For example buyers who face vigorous downstream markets or are otherwise low margin buyers. 2440192 19 Factory gate price increases would have a negative impact on downstream markets X54 An increase in factory gate prices above competitive levels would adversely affect downstream markets. The negative impact would occur through two main channels: X54.1 price, availability, and quality of products: increased costs for downstream domestic producers could affect the price and quantity of the products they sell. It may also affect the quality and variety of their products. The extent of that adverse effect is unknown; and X54.2 exclusion of producers: producers may be rendered less competitively effective if they cannot afford the price increase that reflects market power. This effect may be limited if Fonterra price discriminates across price sensitive customers. Fonterra would have the incentive, but limited ability, to render domestic downstream rivals less competitive X55 The domestic downstream dairy markets are almost entirely supplied by New Zealand dairy processors. While Fonterra remains the dominant player, smaller IPs have made significant in-roads in certain product categories. The strength of competition Fonterra faces in domestic downstream markets from IPs varies by product market. X56 Our analysis suggests that without the DIRA Regulation, while Fonterra may have an incentive to render downstream competitors less competitive, its ability to do so would be limited. Many, but not all, of the IPs affected would be likely to have alternatives at the factory gate or be able to increase their own supply at the farm gate. X57 There are some IPs, including those that rely on the 20% rule, that would not have an alternative to Fonterra. Fonterra would probably have the ability to exclude these IPs. However, given the small collective market share of such IPs, Fonterra would have little incentive to exclude them. Fonterra would have limited incentive and limited ability to render IPs who export less competitive X58 The principal constraint Fonterra faces in export markets is from global players. The export market is generally understood to be a competitive, international market. Fonterra, while being a large player, does not have significant market power in this market. Even if there is some incentive for Fonterra to render domestic IPs who supply the export market less competitive, it would have limited ability to do so. 2440192 20 Wholesale and retail markets We have not found any new information to alter the conclusion reached in our 2011 preliminary inquiry into domestic milk markets X59 In 2011 the Commission considered whether to initiate an inquiry under Part 4 of the Commerce Act 1986 into milk prices. In relation to the wholesale and retail dairy markets, we concluded at that time, that a full pricing enquiry was not warranted. Our current review has found no new information that would alter the conclusions reached in our 2011 preliminary inquiry into domestic milk markets. Stage 2: Our assessment of the efficiency costs and benefits of the DIRA Regulation Risks outweigh benefits X60 We have found that the potential efficiency benefits of the DIRA Regulation 29 are of a similar scale to the potential efficiency costs.30 X61 However, we think there is asymmetric risk attached to deregulation at this stage. Full deregulation may risk harming the growth of competition which has been achieved to date.31 This could damage long-term efficiency which offers potentially significant future benefits. We think that the impact of this risk outweighs any impact of the risk that DIRA may remain in place too long. X62 We recommend keeping the DIRA Regulation in place at this time. We do not think that competition is sufficient to ensure the efficient and contestable operation of the relevant dairy markets if the DIRA Regulation was removed. X63 In particular, it is not clear that immediately removing subparts 5 and 5A would improve efficiency, and our analysis suggests that efficiency could be reduced by immediately removing the Raw Milk Regulations. X64 Table X3, summarises our assessment of the costs and benefits and Table X4 summarises our assessment of the risks associated with removing the DIRA Regulation too early and with removing them too late. 29 In our analysis we assume that the efficiencies of regulation result from restraining Fonterra from being able to fully exercise market power. 30 It should be noted that this analysis is very sensitive to assumptions. 31 Due to the resulting market power that Fonterra would have. See Chapter 5 for more detail. 2440192 21 Table X3 Efficiency costs and benefits of the DIRA Regulation Cost32 Benefit No evidence to suggest these regulations result in material costs at this time We consider the open entry and exit provisions to be an important safeguard and the milk price regime to be beneficial Regulation Subpart 5: Open entry and exit provisions (including the 20% Rule); and  Subpart 5A: Milk price regime   Raw Milk Regulations 32 May cause some excess investment in capacity by Fonterra but we are not convinced that this is material. Cost to Fonterra of allowing farmers free exit limited to date because milk growth has replaced milk lost to IPs. Open entry for new conversions may impose costs on Fonterra without providing significant competition benefits.  May incentivise inefficient dairy conversion but no evidence that this is significant.  The cost of the 20% rule appears to be negligible.  There are costs associated with the milk price regime but some costs likely to exist without the DIRA Regulation.  Help IPs enter and expand in farm gate market. These own-source IPs are potential entrants to the factory gate market.  Mitigate Fonterra’s ability to exercise market power against IPs.  Help smaller IPs and niche food manufacturers to access raw milk.  The 20% rule helps foster downstream competition.  The Milk Price Manual Improves credibility of farm gate price and TAF.  Our review helps ensure improvements in the transparency of disclosures over time. Evidence suggests these regulations may have efficiency costs Evidence suggests these regulations may have efficiency benefits  Likely to cause Fonterra to maintain excess capacity to manage milk uncertainty. We estimated the cost of this uncertainty to be around $6 million per year.  Constrains Fonterra from exercising market power by increasing the factory gate price. We estimate the efficiency cost of this price increase to be in the order of $3.5 million to over $13 million.  Could be inefficient if Fonterra  does not recover opportunity costs but no evidence this is material.  May hinder development of a factory gate market for non-DIRA Help IPs enter and expand in farm gate market by supplementing supply and ensuring they can reach minimum viable capacity.  IPs that expand in the farm gate The costs associated with the regulations could potentially be mitigated by amending the regulations to remove inefficiencies. We consider options for this in our transition pathways to regulation. 2440192 22 milk. Table X4 market are better placed to supply a factory gate market for non-DIRA milk should one develop. Risks associated with removing the DIRA Regulation too early and too late Risks if the DIRA Regulation removed too early Risks if the DIRA Regulation removed too late  Could harm efficiency and structural  competition gains achieved through regulation.  Fonterra may increase the raw milk price in the factory gate market above the competitive level. In addition to the direct inefficiency caused by this it could have additional adverse impact, for example through: o o o loss of innovation in domestic dairy markets, including the evolution of new and higher value products; insufficient diversity and competition in the production and marketing of New Zealand dairy products; and reduced level of contestability and competition in the farm gate, factory gate and retail markets. Over-reliance on access regulation when no longer needed can weaken incentives for IPs to enter factory gate market, dampening competition.  Fonterra may have to invest in excess capacity to manage unexpected milk volumes from returning farmers or variances in regulated milk deliveries. This may be at the expense of more longterm profitable investments.  Inefficient entry of IPs. The aggregate impact of this could outweigh any benefits from deregulation. The current form of the DIRA Regulation may increase the risks associated with deregulation X65 Although we recommend keeping the DIRA Regulation in place at this time, we found that the Raw Milk Regulations in their current form could be hampering the development of the factory gate market. This increases the risks of deregulation at this time. Removing access to DIRA milk before an effective factory gate market has developed could result in Fonterra increasing the price of raw milk above competitive levels. IPs who are unable to sustain this price increase and do not have alternative sources of raw milk could be rendered unable to compete in downstream domestic markets. X66 We considered this issue when determining our recommended transition pathways to deregulation. 2440192 23 Stage 3: Resetting the expiry triggers and transition pathways to deregulation we identified Resetting the expiry triggers X67 In order to provide certainty for industry, we consider that it is important that there is sufficient time between one review ending and another beginning. This informed out recommendation for resetting the expiry triggers. X68 We think the existing expiry triggers are fit for purpose. We therefore do not recommend including any additional or alternative expiry triggers. However we recommend resetting the expiry triggers in the DIRA Regulation so that the next review of competition is triggered by whichever of the following comes first: X68.1 the time limit provision: 2021/22 season; or X68.2 the market share threshold: IPs have collected 30% of available raw milk from the farm gate market33 (considered separately for the North and South islands). X69 We considered a number of factors when determining the appropriate level to reset the triggers. This includes projected IP farm gate market shares in the North and South islands. In order to allow sufficient time for competition to develop we think it is appropriate for the time limit provision to be reached first. Therefore we set the market share threshold at a level which, based on current projections, we do not expect the market to reach before the 2021/22 season. X70 The market share threshold will therefore act as a backstop to bring forward the review should there be an unexpected acceleration in competition. This prudent approach balances the risk of regulating too long against the risk of removing the regulations too early. Pathways to deregulation X71 We recommend: X71.1 33 taking a staged approach towards any transition pathways to deregulation; and Our main area of concern is the factory gate market. However we have used the farm gate market share as the threshold as it is simple, transparent and well understood. It is also consistent with the existing expiry triggers. It is a valid proxy for competition in the factory gate market because a well-functioning farm gate is indicative of the potential to supply at the factory gate. 2440192 24 X71.2 that the Minister considers the options we have outlined for amending the Raw Milk Regulations. These amendments are aimed at incentivising entry and expansion in the factory gate market for non-DIRA milk. Staged approach towards any transition pathways to deregulation X72 A staged approach to pathways to deregulation gives market participants the opportunity to remove dependency on regulations gradually. It would also help participants maintain confidence in the regulatory regime. Providing a well signalled and simple path to deregulation for market participants may reduce the costs of the transition to deregulation. X73 This transition should involve removing the regulation that contributes least to efficiency and contestability first. This would depend on the state of competition at the time of any decision to remove parts of the DIRA Regulation. Based on our current assessment we expect this to involve: X73.1 amending the Raw Milk Regulations; and X73.2 possibly removing the open entry requirement for new conversions. Amendments to the Raw Milk Regulations X74 Amending the Raw Milk Regulations may allow the development of a robust factory gate market by reducing the dependency of IPs on DIRA milk. There is a risk that removing these regulations too quickly could hamper competition in domestic dairy markets in the short term. Gradually removing the Raw Milk Regulations may help mitigate this risk and the transitional costs to the industry. X75 The transition path should encourage market participants to depend less on the regulations over time. Our recommendation seeks to facilitate the development of an effective factory gate market for non-DIRA milk. Once this is functioning, the risks associated with full deregulation will be significantly lower. X76 Therefore, we recommend that the Minister explores amendments to the Raw Milk Regulations. We consider amendments to these regulations to be an important step in the transition pathway to deregulation. X77 We have identified potential options for amending the Raw Milk Regulations: 34 X77.1 reduce Goodman Fielder’s entitlement to DIRA milk; X77.2 reduce IPs entitlements to milk supplied to them by Fonterra under the terms of the Raw Milk Regulations (DIRA milk);34 or For example, by extending the sunset clauses introduced in 2012 to other IPs. 2440192 25 X77.3 tighten the terms and conditions for the supply of DIRA milk to IPs by Fonterra. Amendments to the open entry and exit provisions X78 We think that keeping the open entry and exit provisions in place during these transitional reforms to the Raw Milk Regulations would help smooth the transition. They would facilitate further entry to the farm gate market by IPs that could then supply the factory gate market. They could also facilitate increased own-sourcing of raw milk, including by IPs facing a reduction in DIRA milk entitlements. X79 However, we consider the option of removing open entry for new conversions should be explored. These provisions contribute little to supporting competition in the farm gate market, and potentially impose costs on Fonterra, even though we have not found evidence to suggest that the costs involved are currently material. X80 We consider that our recommended amendments would promote the efficient operation of New Zealand dairy markets. Summary of our key findings X81 On balance, competition is not yet sufficient to fully deregulate. Our key concern is that competition is very limited in the factory gate market. X82 In the absence of the DIRA Regulation, Fonterra would have substantial market power in the factory gate market. It could use that market power to increase the price of raw milk it sells to processors producing dairy products for the domestic market. That could result in an increase in the retail price of domestic dairy products. X83 Some competition has developed in the farm gate market. The regulations have reduced barriers to entry and constrained Fonterra’s market power. X84 Estimated efficiency costs and benefits of the regulations are the same order of magnitude. However there are a number of risks to deregulation across the industry; these risks should be mitigated by adopting a staged approach to deregulation. X85 We recommend the Minister first considers options that facilitate the development of the factory gate market. 2440192 26 CHAPTER 1: Introduction Purpose of this report 1.1 This report sets out our conclusions on the state of competition in the New Zealand dairy industry. 1.2 The report was requested by the Minister for Primary Industries (the Minister), in consultation with the Minister of Commerce, to assist the Minister to make the decision required under section 148(3) of the Dairy Industry Restructuring Act 2001 (DIRA). 1.3 Within 90 days of receiving our report, section 148(3) of the DIRA requires the Minister35 to give notice of whether the Minister: ... intends to promote the enactment of legislation that resets either or both of the market share thresholds specified in s 147 or to promote the adoption of measures 36 that provide a transition pathway to deregulation, or to promote both. 1.4 In this report, we refer to the provisions in subparts 5 and 5A of the DIRA and the Raw Milk Regulations jointly as ‘the DIRA Regulation’. Under the Raw Milk Regulations, Fonterra must make available up to 5% of the raw milk it collects from farmers to independent dairy processors (IPs) at either an agreed price, or regulated price. We refer to this as ‘DIRA milk’. Some IPs acquire raw milk that is not subject to the Raw Milk Regulations, either from Fonterra or from another IP. We refer to this as ‘non-DIRA milk’. Structure of this report 1.5 In this chapter, we explain the scope of the report. 1.6 In the following chapters, we: 1.6.1 present background information on the DIRA Regulation (chapter 2); 1.6.2 explain the process we followed and describe the framework we used to evaluate the sufficiency of competition (chapter 3); 1.6.3 report on the current state of competition in the farm gate and factory gate markets (chapter 4); 35 Under s 5 of the DIRA, ’Minister’ means the Minister of the Crown who is responsible for the administration of the DIRA. The DIRA is administered by the Ministry for Primary Industries. 36 Section 147 of the DIRA sets out market share threshold triggers that specify when the industry-specific regulations relating to Fonterra’s conduct will cease to operate in each of the North and South islands. 2440192 27 1.6.4 report on what the state of competition in the farm gate and factory gate markets and the downstream domestic market could be without the DIRA Regulation (chapter 5); 1.6.5 outline why we think there is insufficient competition in the dairy industry to remove the DIRA Regulation, and report on the efficiencies and inefficiencies created by the DIRA Regulation (chapter 6); and 1.6.6 lay out our thoughts on resetting the market share thresholds and potential pathways to deregulation (chapter 7). Scope of this report 1.7 The scope of our report is determined by the terms of reference issued to us by the Minister.37 1.8 The objectives of our report under these terms of reference are: 1.9 1.8.1 to assess the state of competition in the New Zealand dairy industry; and 1.8.2 if the state of competition is insufficient, to ascertain whether the market share thresholds should be reset, the options for a transition pathway to deregulation (if any), and whether particular deregulation options (if any) should be pursued. Our report addresses specific questions in the terms of reference: 1.9.1 What is the current state of competition in the relevant New Zealand dairy markets, in particular the farm gate and factory gate markets? 1.9.2 In relation to the wholesale and retail dairy markets, is there new information that would alter the conclusions reached in the Commission’s preliminary inquiry into domestic milk markets in 2011 under Part 4 of the Commerce Act 1986?38 1.9.3 In the absence of the provisions of subparts 5 and 5A of Part 2 of the DIRA and/or the provisions of the Dairy Industry Restructuring (Raw Milk) Regulations 2012: 37 Ministry for Primary Industries “Terms of reference for a report on the state of competition in the New Zealand dairy industry” 3 June 2015 . 38 Ministry for Primary Industries “Milk markets: Consideration of whether to initiate a Commerce Act Part 4 inquiry into milk prices” 2 August 2011 . 2440192 28 1.9.4 1.9.3.1 Is the current state of competition in the relevant New Zealand dairy markets sufficient to ensure the efficient and contestable operation of these markets? 1.9.3.2 Would Fonterra be likely to have both the ability and the incentive to exercise market power against competitors, suppliers, or purchasers in one or more of the relevant New Zealand dairy markets? If the current state of competition is insufficient in one or more of the relevant New Zealand dairy markets: 1.9.4.1 Should either or both the market share thresholds specified in section 147 of the DIRA be reset (and, if so, to what new level(s))? 1.9.4.2 Are there other expiry triggers that should be provided for, either in addition to or in place of the market share thresholds? 1.9.4.3 What options there are for a transition pathway to deregulation, and if there are any, which of the options should be pursued? 1.10 In undertaking our assessment of whether the state of competition in the relevant New Zealand dairy markets is sufficient or insufficient, we examined the impact of the DIRA Regulation on competition and the efficient operation of those dairy markets. 1.11 We found the state of competition is insufficient. As a result, we considered whether markets would be more efficient with or without the regulation. On balance, we concluded that the DIRA Regulation should remain. We make suggestions for potential changes to the current DIRA Regulation as part of transition pathway options that could be pursued. 2440192 29 CHAPTER 2: Background to DIRA and the Raw Milk Regulations Purpose of this chapter 2.1 This chapter sets out the history and scope of subparts 5 and 5A of Part 2 of the DIRA and Raw Milk Regulations (together the DIRA Regulation) and how they have been amended. History and scope of the DIRA Regulation DIRA and merger that formed Fonterra 2.2 The DIRA authorised the merger of the two largest dairy co-operatives at the time (New Zealand Dairy Group and Kiwi Co-operative Dairies) with the New Zealand Dairy Board to form Fonterra—a single co-operative company. 2.3 The authorisation recognised that Fonterra would have a dominant market position in a number of domestic New Zealand dairy markets as a result of this merger. 39 The DIRA therefore also contains measures designed to reduce the risks associated with that dominant position. Accordingly, one of the key purposes of the DIRA is “to promote the efficient operation of dairy markets in New Zealand by regulating the activities of Fonterra40 to ensure New Zealand markets for dairy goods and services are contestable”.41 DIRA rules to ensure Fonterra’s contestability 2.4 Subpart 5 of Part 2 of the DIRA imposed certain obligations on Fonterra from its inception. The purpose of the subpart is “to promote the efficient operation of dairy markets in New Zealand” (section 70). The subpart promotes a number of principles under section 71, including:42 2.4.1 IPs must be able to obtain raw milk and other dairy goods and services necessary for them to compete in domestic dairy markets; 2.4.2 Fonterra must accept applications by new entrants and shareholding farmers to supply it with milk, as shareholding farmers; and 39 “Facilitation of the Proposed Dairy Industry Merger: Paper One” 9 April 2001, Hon Jim Sutton, Minister of Agriculture, paragraph 19. 40 References to ‘new co-op’ in the legislation have been replaced with ‘Fonterra’ in this report. 41 Dairy Industry Restructuring Act 2001, s 4(f). 42 Dairy Industry Restructuring Act 2001, ss 71(a) to (c). 2440192 30 2.4.3 2.5 Fonterra must not discriminate between new entrants and shareholding farmers whose circumstances are the same. Subpart 5 of the DIRA also sets out various rules that govern Fonterra’s behaviour in order to ensure contestability, including: 2.5.1 Fonterra’s obligation to accept applications to supply Fonterra with milk,43 and the right of shareholding farmers to cease or reduce44 the supply of milk (the open entry and exit regime); 2.5.2 a rule allowing Fonterra shareholding farmers to supply up to 20% of their weekly production to IPs45 (the ‘20% rule’); 2.5.3 Fonterra's obligation not to discriminate between new entrants and shareholding farmers whose circumstances are the same, or between a shareholding farmer who exercises an entitlement under subpart 5 and a shareholding farmer who does not;46 2.5.4 a rule ensuring that, at any time, at least a third of the milk solids produced within a 160km radius47 of any point in New Zealand is supplied either under contracts with IPs, or under short term contracts48 with Fonterra (the ‘33% rule’); and 2.5.5 a rule providing that a shareholding farmer who withdraws from Fonterra may require Fonterra to sell the vat situated on their farm to either the shareholding farmer or an IP.49 Raw Milk Regulations 2.6 The DIRA also provides for the making of regulations relating to milk.50 The Raw Milk Regulations51include requirements that require Fonterra to supply up to 795 million litres of milk to IPs at a regulated price. An IP is limited to 50 million litres 43 Dairy Industry Restructuring Act 2001, ss 73-76. 44 Dairy Industry Restructuring Act 2001, ss 97. 45 Dairy Industry Restructuring Act 2001, s 108. 46 Dairy Industry Restructuring Act 2001, ss 106. 47 Dairy Industry Restructuring Act 2001, s 107. 48 Contracts that either expire, or may be terminated by the supplier at the end at the end of the current season. 49 Dairy Industry Restructuring Act 2001, s 109. 50 Dairy Industry Restructuring Act 2001, s 115. 51 Subject to maximum monthly limits. Dairy Industry Restructuring (Raw Milk) Regulations (2012), ss 6-8 2440192 31 per season at the regulated price, except for Goodman Fielder, whose limit is 250 million litres per season. Subpart 5A of Part 2 of the DIRA 2.7 The provisions under subpart 5 of the DIRA and the Raw Milk Regulations are augmented by the milk price monitoring regime under subpart 5A of Part 2 of the DIRA that was introduced on 27 July 2012. 2.8 Subpart 5A is intended to bolster the provisions in subpart 5 that are directed at ensuring contestability and efficiency in New Zealand dairy markets. It does so by, among other things, providing incentives for Fonterra to act efficiently whilst providing contestability in the farm gate market through the setting of an efficient farm gate milk price.52 2.9 The Commission’s reviews assess whether the Milk Price Manual and the farm gate milk price that Fonterra sets using the manual provide the incentive for Fonterra to operate efficiently while not precluding efficient IPs from competing. 2.10 The milk price monitoring regime promotes greater transparency of Fonterra’s farm milk price setting processes, and greater confidence in the consistency of Fonterra’s base milk price with contestable market outcomes. Regular review of the DIRA Regulation 2.11 Between 2008 and 2013, the Ministry of Agriculture & Forestry (MAF, now part of the Ministry for Primary Industries), in consultation with the dairy industry, reviewed the DIRA. This review resulted in legislative amendments designed to refine the provisions in place at the time, including: 2.11.1 the requirement for a report on the state of competition53 when IPs reach a market share threshold of 20% or more of milk solids collected in a season in either the North or South island. Unless legislation is passed to keep aspects of the DIRA Regulation, it will expire in the relevant island after the threshold is triggered; 2.11.2 the requirement that the report to the Minister on the state of competition in the dairy industry could lead to resetting the current thresholds and/or adopting measures to promote a transitional pathway to deregulation; 52 Dairy Industry Restructuring Act 2001, s 150A. 53 Dairy Industry Restructuring Act 2001, s 148. 2440192 32 2.12 54 2.11.3 changes to the DIRA to enable Fonterra to change its capital structure54 and introduce the Trading Amongst Farmers scheme (TAF); 2.11.4 amendments to the Raw Milk Regulations introducing restrictions relating to the access to DIRA milk and revised monthly limits; and 2.11.5 the introduction of subpart 5A which enshrined Fonterra’s existing milk price governance and transparency processes in the DIRA and the introduction of a new farm gate milk price monitoring regime undertaken by the Commerce Commission. Further details of the DIRA Regulation and amendments to the Raw Milk Regulations are contained in Attachments A and C. Dairy Industry Restructuring Act 2001, ss 109A-109N. 2440192 33 CHAPTER 3: Process and framework for evaluating and reporting on the state of competition Purpose of this chapter 3.1 This chapter explains the process we followed and the framework we used to consider and report on the state of competition in the New Zealand dairy industry, as requested by the Minister of Primary Industries (the Minister) under section 148A of the DIRA. 3.2 In this chapter we outline: 3.2.1 our interpretation of ‘sufficient’ and ‘insufficient’ in the terms of reference and section 148 of the DIRA; 3.2.2 the conceptual framework we used to answer the questions in the terms of reference; 3.2.3 options to explore if we conclude that the DIRA Regulation enhances efficiency; 3.2.4 the consultations on our proposed evaluation approach; 3.2.5 the market definitions for the relevant domestic dairy markets; 3.2.6 how we analysed the state of competition; and 3.2.7 our approach when considering pathways to deregulation. Our interpretation of ‘sufficient’ and ‘insufficient’ 3.3 The terms of reference required us to determine whether or not the state of competition in the relevant New Zealand dairy markets is sufficient to ensure efficient and contestable markets in the absence of the DIRA Regulation.55 3.4 If the state of competition is insufficient, the terms of reference further require us to consider market share thresholds for deregulation, other potential triggers for deregulation, options for a transition pathway to deregulation, and what options for deregulation should be pursued. 3.5 We, therefore, had to decide what the terms ‘sufficient’ and ‘insufficient’ meant in order to conduct our review. 55 Under s 148(2)(d)(ii) of the DIRA, we are required to determine whether or not the state of competition in the relevant New Zealand dairy markets is ‘insufficient’. 2440192 34 3.6 We consider that the efficiency purpose is the overriding goal of the DIRA measures that are designed to reduce the risks associated with Fonterra’s dominant position. Our approach was therefore driven by the efficiency purpose of the DIRA, and in particular the obligations it imposes upon Fonterra to promote the efficient operation of New Zealand dairy markets.56 3.7 Contestability and workable competition are sufficient but not necessary conditions for markets to be more efficient without the DIRA Regulation. 3.8 Accordingly, while our review recognises the link between contestability and the efficient operation of New Zealand dairy markets, our interpretation is that the state of competition will only be sufficient if the relevant New Zealand dairy markets would be more efficient without the DIRA Regulation. 3.9 In determining whether the state of competition is sufficient or insufficient we therefore also carried out a balancing exercise to assess whether the relevant New Zealand dairy markets would be more efficient with or without some or all of the DIRA Regulation. The conceptual framework for our analysis 3.10 56 This section describes the conceptual framework we used to carry out our review, see figure 3.1. Dairy Industry Restructuring Act 2001, s 4(f), s 70, s 71 and s 150A (1). 2440192 35 Figure 3.1 Conceptual framework Conceptual framework Is competition sufficient to ensure contestable and efficient markets? No Will removing all DIRA regulation improve market efficiency? No Will removing some DIRA regulation improve market efficiency? Yes Yes Yes Recommend deregulation Recommend deregulation Recommend removing relevant regulations No Should the market share thresholds should be reset? Are there alternative expiry triggers? Consider options for pathways to deregulation Which options for pathways to deregulation should be pursued? Is competition sufficient to ensure efficient markets? 3.11 We first assessed the relevant markets by analysing their competitiveness and contestability with the DIRA Regulation. 3.12 We then compared the level of competitiveness and contestability that exists with the DIRA Regulation with what the level would most likely be without that regulation. To do this, we examined Fonterra’s ability and incentive to exercise significant market power under both scenarios. 3.13 If we had concluded that the degree of competition or contestability would be sufficient to ensure efficient markets without the DIRA Regulation—due to Fonterra not having the ability and incentive to exercise significant market power in the relevant markets—we would have recommended removing the DIRA Regulation because the markets would be more efficient without the regulation. Will removing some or all of the DIRA Regulation improve market efficiency? 3.14 We concluded that the level of competition or contestability would not ensure the efficient operation of the markets in the absence of DIRA Regulation, because we 2440192 36 found that Fonterra would have some ability and incentive to exercise significant market power in the relevant markets in the absence of the regulation. 3.15 We therefore went on to assess whether the relevant markets would be more efficient without all or some of the DIRA Regulation. We considered the extent to which the different regulations contributed to efficiency and assessed whether the efficiency gains from the regulation were outweighed by any efficiency losses. This is an important balancing exercise in situations where the level of competition or contestability without the regulation is close to ensuring the efficient operation of the markets in the absence of regulation. 3.16 If we had found that the relevant markets would be more efficient without all or some of the DIRA Regulation, then those regulations would no longer promote the purpose of the DIRA. We would then have recommended removing the relevant regulations. In those circumstances, we would have identified the regulation that is not promoting efficiency and explained why the market outcomes would be better without that regulation. Options to explore if we conclude that the DIRA Regulation enhances efficiency 3.17 We concluded that the potential efficiency costs and benefits were of a similar order of magnitude. However, after taking account of the risk of adverse consequences from removing the DIRA Regulation too early, we concluded, on balance, that the markets would be more efficient with the DIRA Regulation, and so we examined: 3.17.1 whether or not the market share thresholds should be reset; 3.17.2 if there are alternative expiry triggers that may be appropriate (other than the market share thresholds); 3.17.3 options for pathways to deregulation; and 3.17.4 which options for pathways to deregulation should be pursued, including recommendations on which options could be further considered by the Ministry of Primary Industries as part of its review processes on the DIRA Regulation. 2440192 37 Consultation on our evaluation approach 3.18 On 12 June 2015, we published a paper for public consultation setting out our proposed approach to the review.57 The main purpose of this consultation was to obtain feedback from dairy industry stakeholders on our proposed evaluation approach. 3.19 We received 11 submissions and two cross-submissions on our proposed approach to evaluating the issues. Submissions came from Fonterra, IPs, and Federated Farmers. 3.20 We considered all information and submissions we received in finalising our evaluation approach. 3.21 Most submitters were supportive of our proposed evaluation approach. The key arguments of submitters relating to our evaluation approach are set out below. Submissions on our approach to assessing whether the state of competition is ‘sufficient’ 3.22 Fonterra, Westland, Goodman Fielder, Open Country, and Federated Farmers all supported our efficiency approach to assessing the ‘sufficiency’ of the state of competition.58 Open Country’s submission also noted that in practice the workable competition and efficiency tests generally involve assessing the same substance.59 3.23 Miraka disagreed with our approach to assessing the ‘sufficiency’ of the state of competition and our interpretation of the DIRA purpose. Miraka asserted that the primary purpose of the DIRA is to ensure contestable markets and not efficiency.60 3.24 We maintained our focus on efficiency when assessing the ‘sufficiency’ of the state of competition. As discussed in paragraph 3.6, we believe efficiency is the primary 57 Commerce Commission “Review of the state of competition in the New Zealand dairy industry: Consultation paper—process and approach” 12 June 2015. 58 Fonterra "Review of the state of competition in the New Zealand dairy industry: Consultation paper— process and approach" 10 July 2015, paragraph 22; Westland "Review of the state of competition in the New Zealand dairy industry: Consultation paper—process and approach" 10 July 2015, paragraph 11; Goodman Fielder "Review of the state of competition in the New Zealand dairy industry: Consultation paper—process and approach" 10 July 2015, paragraph 1.4; Castalia Strategic Advisors for Open Country Dairy "Review of the State of Competition in the New Zealand Dairy Industry: Comments on the Commission's Approach and Potential Outcomes of Deregulation" 10 July 2015, p.1; Federated Farmers “Submission on: Review of the state of competition in the New Zealand dairy industry Consultation paper—process and approach” p.11, paragraphs 10.2–10.3. 59 Castalia Strategic Advisors for Open Country Dairy "Review of the State of Competition in the New Zealand Dairy Industry: Comments on the Commission's Approach and Potential Outcomes of Deregulation" 10 July 2015, p.1. 60 Miraka "Review of the state of competition in the New Zealand dairy industry: Consultation paper— process and approach" 10 July 2015, paragraphs 2.4.1–2.4.2. 2440192 38 purpose of section 4(f) of the DIRA, and that contestability is the means to achieving efficient markets in the absence of workable competition. Submissions on the scope of the review 3.25 Miraka, Tatua, Synlait, and Open Country were all concerned that the scope of our review was too confined.61 3.26 Miraka said our review should consider changes to the current regulatory framework to ensure that contestable markets are achieved.62 3.27 Tatua said our proposed approach was biased towards deregulation and that we should admit the opportunity to improve the competition and efficiency of the market through changes to the DIRA Regulation, apart from as part of pathways to deregulation or resetting thresholds.63 3.28 Synlait said the review should also consider if different regulation would better promote efficient dairy markets.64 3.29 Open Country asked us to consider options to improve the DIRA that would enhance competition and contestability and therefore the DIRA’s purpose of creating a transition pathway to deregulation.65 3.30 We largely maintained our proposed approach to the scope of the review as we believed our approach was consistent with the efficiency purpose of the DIRA, the proper interpretation of section 148 of the DIRA, and the terms of reference for our report. 61 Miraka "Review of the state of competition in the New Zealand dairy industry: Consultation paper— process and approach" 10 July 2015, paragraph 4.2.2; Tatua "Review of the state of competition in the New Zealand dairy industry: Consultation paper—process and approach" 10 July 2015, paragraph 2.1; Synlait "Review of the state of competition in the New Zealand dairy industry: Consultation paper— process and approach" 13 July 2015, p.1; Castalia Strategic Advisors for Open Country Dairy "Review of the State of Competition in the New Zealand Dairy Industry: Comments on the Commission's Approach and Potential Outcomes of Deregulation" 10 July 2015, p.5. 62 Miraka “Review of the state of competition in the New Zealand dairy industry: Consultation paper— process and approach" 10 July 2015, paragraph 2.4.2. 63 Tatua "Review of the state of competition in the New Zealand dairy industry: Consultation paper—process and approach" 10 July 2015, paragraph 4.3. 64 Synlait "Review of the state of competition in the New Zealand dairy industry: Consultation paper— process and approach" 13 July 2015, p.1. 65 Open Country Dairy "Review of the state of competition in the New Zealand dairy industry: Consultation paper—process and approach" 10 July 2015, p.4. 2440192 39 3.31 We did not attempt to identify whether or not a specific different set of regulations would better promote efficient New Zealand dairy markets, as this is outside of the scope of the terms of reference. 3.32 We have, however, suggested possible changes to the DIRA that would enhance efficiency. 3.33 We also noted aspects of the DIRA Regulation that could be materially improved by the Ministry for Primary Industries as part of its policy processes. Submissions on the focus on Fonterra 3.34 Tatua asserted that our competition assessment should consider all facets of both domestic and global dairy market competition, rather than just focus on how Fonterra might or might not react in a deregulated environment.66 3.35 In contrast to Tatua, Open Country suggested we should focus particularly on how Fonterra might react to different deregulation scenarios.67 3.36 We decided the main focus of our review should remain as proposed, namely: whether Fonterra would be likely to have both the ability and the incentive to exercise market power—or to exercise enhanced market power—against competitors, suppliers, or purchasers in the relevant dairy markets in the absence of some, or all, of the DIRA Regulation. Submissions on testing the premise ‘what if there was no DIRA Regulation’ 3.37 Fonterra, Tatua, and Open Country all supported our intention to test ‘what if there was no DIRA Regulation’.68 3.38 Fonterra noted the risk that the analysis could become overly complex given the number of permutations of ‘what if there was no DIRA Regulation’, potentially obscuring where the efficiency of regulation could be improved.69 66 Tatua "Review of the state of competition in the New Zealand dairy industry: Consultation paper—process and approach" 10 July 2015, paragraph 5.8. 67 Castalia Strategic Advisors for Open Country Dairy "Review of the State of Competition in the New Zealand Dairy Industry: Comments on the Commission's Approach and Potential Outcomes of Deregulation" 10 July 2015, pp.1–2. 68 Fonterra "Review of the state of competition in the New Zealand dairy industry: Consultation paper— process and approach" 10 July 2015, paragraph 23; Castalia Strategic Advisors for Open Country Dairy "Review of the State of Competition in the New Zealand Dairy Industry: Comments on the Commission's Approach and Potential Outcomes of Deregulation" 10 July 2015, pp.1–2.; Tatua "Review of the state of competition in the New Zealand dairy industry: Consultation paper—process and approach" 10 July 2015, paragraph 5.1. 2440192 40 3.39 Tatua and Miraka both sought greater clarity on how we would examine the different counterfactuals given there is no alternate history without the DIRA Regulation.70 3.40 Open Country asked us to consider how removing the DIRA’s components in combination might exacerbate any inefficiencies of deregulation due to interactive, cumulative effects.71 3.41 In assessing the likely state of competition without the DIRA Regulation, we focused on those regulations that are most effective in constraining Fonterra’s market power. Our evaluation of the likely outcomes if the DIRA Regulation was removed was based on our judgement of Fonterra’s ability and incentive to exercise market power in the absence of the different DIRA Regulations. Submissions on assessment of regional markets 3.42 Fonterra and Talleys both suggested we should assess different regions individually.72 3.43 Miraka suggested a regional assessment would only be relevant where a comprehensive analysis had found competition to be generally sufficient. Miraka also suggested it might still be appropriate for the DIRA Regulation to be retained in specific regions even though competition was sufficient in most regions.73 3.44 We examined differences in regional competition. However, we did not precisely define the geographic dimensions of the markets, as this level of detail would not affect our conclusions. Identifying the relevant domestic dairy markets 3.45 We identified the following New Zealand dairy markets: 69 Fonterra "Submission on review of the state of competition in the New Zealand dairy industry – process and approach" 10 July 2015, para 25. 70 Tatua "Review of the state of competition in the New Zealand dairy industry: Consultation paper—process and approach" July 2015, para 3.4; Miraka "Consultation Paper – process and approach: Review of the state of competition in the New Zealand dairy industry" 10 July 2015, para 4.2. 71 Castalia Strategic Advisors for Open Country Dairy "Review of the State of Competition in the New Zealand Dairy Industry: Comments on the Commission's Approach and Potential Outcomes of Deregulation" July 2015, p.5. 72 Fonterra "Submission on review of the state of competition in the New Zealand dairy industry – process and approach" 10 July 2015, para 18.1; Talleys "RE: Commerce Commission Review of the State of Competition in the New Zealand Dairy Industry – July 2015" 22 June 2015, para 7–8. 73 Miraka "Consultation Paper – process and approach: Review of the state of competition in the New Zealand dairy industry (12 June 2015)" 10 July 2015, para 4.1.4. 2440192 41 3.45.1 farm gate markets in which dairy farmers supply raw milk to dairy processors and dairy processors compete to secure farmer supply; 3.45.2 factory gate markets in which dairy processors supply DIRA milk they have collected from farmers to other processors and some food and beverage manufacturers; and 3.45.3 factory gate markets in which dairy processors supply non-DIRA milk they have collected from farmers to other processors and some food and beverage manufacturers. 3.46 A market encompasses actual and potential transactions between sellers and buyers. Market definition seeks to capture the factors that directly shape and constrain rivalry between buyers and between sellers. 3.47 A market is defined in the Commerce Act 1986 as “a market in New Zealand for goods and services as well as other goods and services that, as a matter of fact and commercial common sense, are substitutable for them”.74 In general, the more closely substitutable two products are, the closer the competition and the greater the competitive constraint between the products. 3.48 We define the markets in a way that best isolates the key competition issues relevant to our enquiry to help identify and assess the competitive constraints Fonterra faces and would likely face without the DIRA Regulation. 3.49 In many cases this may not require us to precisely define the boundaries of a market. Accordingly, while we considered differences in regional competition, we did not precisely define the geographic dimensions of the markets as this would not affect the conclusions of our analysis. 3.50 Because the review of competition in relevant New Zealand dairy markets examines Fonterra’s market power, we focused our analysis on Fonterra’s market power as a buyer of raw milk from farmers in the farm gate markets, and as a seller of raw milk to other dairy processors in the factory gate markets. 3.51 In order to identify the key competitive constraints faced by Fonterra, we identified the areas of overlap between Fonterra and IPs, and then considered what, if any, products and geographic regions constituted close substitutes from a supplier’s75 75 Farmers supplying raw milk to Fonterra and IPs. 2440192 42 point of view (farm gate markets) and a customer’s point of view76 (factory gate markets). 3.52 The DIRA Regulation and competition in the farm gate and factory gate markets also affects downstream domestic dairy markets as well as the strength of exportorientated firms. We did not explicitly define these downstream domestic dairy markets as a detailed analysis of the state of competition in these markets is outside the scope of our review. However, our analysis did consider how the DIRA Regulation might impact competition at those levels of the value chain. 3.53 Attachment E, Market definition, sets out our reasons for identifying the above markets as the relevant markets. How we carried out our analysis of the state of competition 3.54 This section describes what we did to answer the questions in the terms of reference after identifying the relevant markets. We: 3.54.1 assessed the current state of competition with the DIRA Regulation; 3.54.2 compared competition with and without the DIRA Regulation; 3.54.3 assessed the impact of the DIRA Regulation on efficiency; and 3.54.4 considered pathways to deregulation. We assessed the current state of competition with the DIRA Regulation 3.55 To review the state of competition without the DIRA Regulation we considered whether Fonterra would have market power if the DIRA Regulation was removed. This is because if a firm has market power it may use that to lessen the extent of competition in a market. Fonterra has market power if it has the ability to raise prices above competitive levels, or engage in other activity that impacts the competitiveness of the market. Considering Fonterra's market power with and without DIRA Regulation enables us to assess whether the state of competition in the relevant markets is sufficient to ensure the efficient and contestable operation of those markets without DIRA Regulation. 3.56 To assess the current state of competition in the farm gate and factory gate markets, we looked at four market indicators of Fonterra’s buying power in the farm gate markets and Fonterra’s selling power in the factory gate markets. The market indicators are: 76 The buyers of milk at the factory gate. 2440192 43 3.56.1 the conditions faced by dairy processors entering or expanding in the relevant markets; 3.56.2 how the DIRA Regulation has lowered barriers to entry and expansion for IPs; 3.56.3 the degree of rivalry between Fonterra and the IPs, as evidenced by changes in dairy processors’ market shares over time, new entry and expansion by IPs, farmer switching (farm gate markets), and dairy processor switching (factory gate markets), Fonterra’s response to competition from IPs (farm gate markets); and 3.56.4 the impact of Fonterra’s co-operative ownership on its incentives to exercise market power. We compared the state of competition with and without the DIRA Regulation 3.57 3.58 The central aspect of our review was to compare the state of competition where the DIRA Regulation continues with the likely state of competition that would exist in the absence of the DIRA Regulation, including: 3.57.1 subpart 5 of Part 2 of the DIRA (which sets out, amongst other things, open entry and exit requirements for Fonterra farmers);77 3.57.2 subpart 5A of Part 2 of the DIRA (which sets out provisions relating to Fonterra’s calculation of the base milk price); and 3.57.3 the Raw Milk Regulations (which provide for the regulated supply of raw milk by Fonterra to IPs).78. To assess what competition in New Zealand’s dairy industry would be like without the DIRA Regulation, we analysed whether without regulation Fonterra would have the ability and the incentive to exercise buyer power or seller power in the following ways: 3.58.1 exercise buyer power against farmer suppliers at the farm gate in the purchase of milk by decreasing prices below competitive levels; 77 In examining the impact of the subpart 5 of Part 2 of the DIRA, we also examined distinct provisions within this subpart separately (eg, the 20% rule). 78 In examining the impact on the likely future state of competition of the Raw Milk Regulations that require Fonterra to supply raw milk to IPs, we took into account reg 6(3), which provides that from 1 June 2016 IPs that have collected (in three consecutive seasons) 30 million litres or more of raw milk from their own farmer suppliers will no longer be eligible for regulated milk. 2440192 44 3.58.2 exercise seller power against IPs at the factory gate in the sale of milk by increasing prices Fonterra charges IPs for milk; and 3.58.3 prevent IPs from effectively competing (generally referred to as foreclosure) by: 3.58.3.1 restricting IPs from accessing milk from farmers at the farm gate by either increasing the farm gate price or locking farmers into longer term contracts; and 3.58.3.2 raising prices or restricting access to factory gate milk for IPs that sell to customers in domestic downstream dairy markets (such as milk and cream). We assessed the impact of the DIRA Regulation on efficiency 3.59 We also considered the extent to which the DIRA Regulation may have a positive or negative effect on efficiency in the domestic dairy markets. Our assessment included consideration of: 3.59.1 how the different regulations have constrained Fonterra's ability to exercise market power and aided the development of competition with Fonterra; 3.59.2 the potential inefficiencies created by the different parts of the DIRA Regulation, including: 3.59.2.1 whether or not the DIRA Regulation results in Fonterra maintaining excess capacity; 3.59.2.2 whether inefficiency is created in the factory gate market as a result of the DIRA Regulation; 3.59.2.3 whether the DIRA Regulation has incentivised inefficient dairy conversions; and 3.59.2.4 whether there are inefficiencies of the base milk price monitoring regime. We considered expiry triggers and pathways to deregulation 3.60 We considered whether the market share thresholds should be reset, including the potential forms the thresholds may take, and whether the thresholds would trigger the timely assessment of these markets. 3.61 We also considered whether expiry triggers are best used to trigger a competition review or automatic deregulation, and whether the market share expiry thresholds should be replaced or augmented by additional expiry triggers. 2440192 45 3.62 We considered and identified options for transition pathways to deregulation in relation to each core element of the DIRA Regulation. This included consideration of whether a staged approach is appropriate. In considering options, we took account of the risks of deregulating too early and the uncertainties involved. 3.63 We also identified and recommended consideration of the most beneficial routes for a transition pathway to deregulation. 2440192 46 CHAPTER 4: State of competition under the DIRA Regulation Purpose of this chapter 4.1 This chapter provides our assessment of the current state of competition in the farm gate and factory gate markets with the DIRA Regulation in place as required in the terms of reference for our report. 4.2 Our assessment focused on how the DIRA Regulation has affected Fonterra’s ability to exercise market power and the ability of IPs to compete with Fonterra with the regulation in place. 4.3 Sections in this chapter cover: 4.3.1 key findings on the state of competition in farm gate market; 4.3.2 key findings on the state of competition in the factory gate market; 4.3.3 our assessment of the current state of competition in the farm gate market; 4.3.4 conditions of entry and expansion into the farm gate market; 4.3.5 our finding that DIRA Regulation has lowered barriers to entry and expansion in the farm gate market; 4.3.6 independent rivalry in the farm gate market; 4.3.7 Fonterra’s co-operative structure in relation to the farm gate market; 4.3.8 our assessment of the current state of competition in the factory gate market 4.3.9 an overview of the factory gate market for raw milk; 4.3.10 current customers for DIRA milk in the factory gate market; 4.3.11 barriers to entry into the factory gate market; 4.3.12 independent rivalry in the factory gate market; 4.3.13 constraints on Fonterra’s market power in the factory gate market; and 4.3.14 if there is any new information on state of competition in the market for wholesale and retail supply of fresh processed milk. 2440192 47 Key findings on the state of competition in farm gate market 4.4 The farm gate market is the market in which dairy farmers supply raw milk to dairy processors and dairy processors compete to secure farmer supply. 4.5 Overall, we conclude that with the current DIRA Regulation, Fonterra does not have the ability to exercise market power either by lowering the farm gate prices paid by Fonterra to farmers below competitive levels, or engaging in conduct to prevent or hinder rival IPs from accessing raw milk at the farm gate. We consider this is a result of a combination of the DIRA Regulation (which helps ensure contestability), Fonterra’s co-operative nature, and constraints from competitors. 4.6 All farm gate markets for the supply of raw milk are highly concentrated and Fonterra remains the monopsony purchaser of raw milk in a many regional markets. 4.7 There are significant barriers to entry in these markets, in particular, processing plants are large and capital intensive and large-scale entry requires secure farmer suppliers. At the same time farmers also require a guaranteed offtake of their milk before investing in new dairy conversions. 4.8 The DIRA Regulation and Fonterra’s co-operative nature have succeeded in lowering entry barriers for IPs substantially. 4.9 There are signs that the markets are contestable—new IPs have been entering the markets, and existing processors have expanded their operations in several regions. This has led to increased competition for raw milk in those regions. 4.10 The level of competition Fonterra faces varies by region. The most significant competition is in Canterbury, Southland, Taranaki, and Waikato regions, where increased milk production has contributed to growth of IPs and their ability to compete. 4.11 In other regions, including Northland, Wairarapa, and Hawke’s Bay regions, Fonterra faces little or no competition from IPs in the acquisition of milk from farmers. 4.12 Given Fonterra’s continued high market share in many regions, we consider that effective competition is still in the process of emerging in those regional markets. We do not consider that the constraints from Fonterra’s competitors are currently sufficient to ensure effective competition in those regional markets. Key findings on the state of competition in the factory gate market 4.13 The factory gate market is the market in which dairy processors supply raw milk they have collected from farmers to other processors and some food and beverage manufacturers, including the DIRA milk by Fonterra to IPs. As discussed below we consider that there are separate markets for DIRA milk and non-DIRA milk. 2440192 48 4.14 Fonterra’s market power in the factory gate market for DIRA milk is constrained by Fonterra’s requirement to supply DIRA milk to IPs, including Goodman Fielder, under the Raw Milk Regulations.79 Fonterra is the only firm with this obligation. The price of DIRA milk is regulated to reflect the price Fonterra pays farmers for milk plus its average transport costs. 4.15 DIRA milk has successfully supported IP entry by guaranteeing them a supply of raw milk. DIRA milk can be an IP’s only source of raw milk or it may supplement an IP’s own supply from farmers. DIRA milk has also helped ensure that IPs that manufacture dairy products for the domestic markets are able to secure raw milk at a price that approximates that which would be found in a workably competitive market. DIRA milk has also eased the entry of IPs focused on export markets. 4.16 Some IPs acquire raw milk that is not subject to the Raw Milk Regulations. This milk, which might come from either Fonterra or from another IP at the factory gate, is often acquired to supplement DIRA milk purchases. We refer to this as ‘non-DIRA milk’ and we consider this to be supplied in a separate factory gate market from DIRA milk. In the 2014/15 season, non-DIRA milk accounted for [ ] of Fonterra’s total factory gate milk sales.80 The non-DIRA milk market is limited and highly concentrated—only Fonterra, [ ] have supplied milk outside the DIRA Regulation in any significant and/or regular quantities.81 4.17 While IPs large enough to source their own raw milk supply at the farm gate are unlikely to face significant barriers to supplying raw milk at the factory gate, they have told us they are generally not interested in supplying at the factory gate at the DIRA milk price. 4.18 Where Fonterra (or occasionally IPs) supplies other processors that do not qualify for DIRA milk with non-DIRA milk (for all or part of their factory gate requirements), they do so at a price that is substantially higher than the price of DIRA milk.82 The prices charged by Fonterra for non-DIRA milk [ ]. 79 While Fonterra currently supplies Goodman Fielder with milk under contracts it is treated as DIRA milk for purposes of this report. 80 Fonterra “Information request by the Commerce Commission: Tranche 1 response to info request” 11 September 2015, para 12. 81 [ ] 82 As noted above, only [ 2440192 ] supplies non-DIRA milk on a regular basis. 49 4.19 Other IPs that are potential entrants to the factory gate market have indicated they would consider supplying that market at the right price. However, even if such IPs were to enter the factory gate market Fonterra would remain the only seller of raw milk to IPs in a number of regions. 4.20 Given the high concentration levels in the factory gate markets and the limited participation by IPs, we do not think the regional factory gate markets are currently characterised by effective competition. Farm gate market—assessing the current state of competition 4.21 The farm gate markets are the markets in which dairy farmers supply raw milk to dairy processors and dairy processors compete to secure farmer supply. 4.22 Competitive pressure can be applied to Fonterra from existing competitors and potential competition—existing IPs competing with Fonterra for raw milk supply from farmers and the threat of entry by IPs that may do so. 4.23 In order to assess the current state of competition in the farm gate market we looked at: 4.23.1 how entry and expansion conditions influence the likelihood of entry by new competitors or expansion by existing competitors; 4.23.2 the effect of the DIRA Regulation on entry and expansion by IPs; 4.23.3 the degree of independent rivalry between Fonterra and IPs as evidenced by changes in market shares over time, new entry and expansion, and farmer switching; and 4.23.4 the effect of Fonterra’s co-operative structure on its incentives to exercise market power in the acquisition of raw milk from farmers. Farm gate market—conditions of entry and expansion 4.24 One of the methods we used to assess the current state of competition in the farm gate markets was to look at the conditions faced by IPs entering or expanding in these markets. 4.25 An effectively competitive market will generally have no significant barriers to entry or expansion.83 Entry or expansion will generally occur when the expected profits 83 While the proposition that a firm’s market power depends substantially on the level of barriers to entry and expansion in the relevant market is well established in New Zealand competition law (see Southern Cross Medical Care Society v Commerce Commission (2001) 10 TCLR 25), New Zealand’s courts have subsequently highlighted that the question of whether conditions in a market qualify as a barrier to entry, 2440192 50 from entry or expansion are positive. Certain market conditions, commonly referred to as barriers to entry and expansion, can impact the likelihood of positive profits. 4.26 The likelihood of entering or expanding in dairy milk processing in New Zealand can be affected by the following factors, amongst others84: 4.26.1 sunk costs that are not recoverable on exit; 4.26.2 economies of scale and capacity utilisation—to operate efficiently and have low enough costs to compete effectively a new entrant needs: 4.26.3 4.27 4.26.2.1 access to a secure source of raw milk of adequate size; and/or 4.26.2.2 a sufficient share of the downstream market(s) (or a reasonable expectation of reaching such share); the risk that Fonterra engages in strategic behaviour to discourage prospective entrants or expansion. In general, new IP entry into and expansion in the farm gate markets is eased by the DIRA Regulation. For example, entry, and to a lesser degree expansion, is less likely in the absence of the open entry and exit provisions in subpart 5 of Part 2 of the DIRA. IP entry may also be less likely in the absence of the milk price regime in subpart 5A of Part 2 of the DIRA and the Raw Milk Regulations. Sunk costs 4.28 Processing plants are typically large and capital intensive.85 Potential entrants or IPs considering expanding will therefore take account of the risk of such costs not being recouped on exit. However, these assets are not entirely sunk in that there may be other processors that are willing to purchase such plants on a firm’s exit, as however defined, is less important than whether those conditions have the potential to prevent, impede, or slow entry and expansion. 84 There are other factors, for example, access to product distribution channels, and regulatory barriers such as environmental consents and food safety regulations. These are not considered in the report as they are not affected by the DIRA Regulation under consideration. 85 In 2006 the cost of entry to be an acquirer in the raw milk market was approximately $12 million for a processing plant of 100,000 litres per day. Fonterra Co-operative Group Limited and Kapiti Fine Foods Limited and United Milk Limited (Commerce Commission Decision 574, 23 February 2006, paragraph 204). Yashili’s plant commissioned this year with a capacity to process [ ] of milk a day cost around $212 million. Stuff – NZ Farmer (2015) “Opening of Yashilis 212 million pant delayed” (Viewed September 2015); [ ] 2440192 51 demonstrated by Fonterra’s purchase of the New Zealand Dairies Limited (NZDL) plant in 2012 when it went into liquidation.86 Where processing plants are based in areas where milk supply is growing, sunk costs are likely to be relatively low. It is unlikely that the size of sunk costs has been significantly affected by the DIRA Regulation. Economies of scale and capacity utilisation 4.29 IPs are unlikely to be able to operate viably without achieving scale and efficient levels of capacity utilisation. 4.30 Although we have limited evidence on the minimum viable scale required by entrants, we consider that there are likely to be substantial economies of scale in dairy processing given the nature of the costs involved. Efficient levels of capacity utilisation is also important for viability. Feedback from [ ] stated that a plant would need to be operating at two-thirds capacity at least by the end of the second year of operation to make a case for investment.87 4.31 Large-scale entry and efficient levels of capacity utilisation require secure access to adequate supplies of raw milk. A critical condition for entry is therefore sufficient secure access to raw milk. 4.32 A secure source of milk supply may also be required to obtain funding for investment in new production facilities.88 4.33 Barriers to expansion are generally lower than barriers to entry as farmers are likely to be willing to commit to supply raw milk to an IP that has been established for a number of years and has built up a reputation. 4.34 We also note that some of the more recent IP ‘new entrants’ are part-owned by large international food/dairy companies and this may provide adequate access to finance and farmers with some comfort about the sustainability of these processors. Farm gate market—DIRA Regulation has lowered barriers to entry and expansion 4.35 The DIRA Regulation has lowered the barriers to IPs entering into and expanding in the farm gate market and assisted IPs to access farmer suppliers and meet the milk 86 Commerce Commission “Fonterra Limited and New Zealand Dairies Limited (in receivership) [2012] NZCC 21” 6 September 2012. 87 [ 88 [ 2440192 ] ] 52 requirements for a minimum viable operation on entry and/or expansion. The DIRA Regulation remains important for IP entry, but is less important to enabling expansion by well-established IPs. 4.36 We discuss the extent to which each of the following key regulations have assisted entry and/or expansion to the farm gate market: 4.36.1 access to DIRA milk; 4.36.2 open entry and exit provisions; 4.36.3 right to supply IPs—the 20% rule 4.36.4 non-discrimination rule 4.36.5 regulation of supply contracts for raw milk—the 33% rule; 4.36.6 sale of milk vats on supplier exit; and 4.36.7 milk price regime. DIRA milk89 4.37 As NERA notes, the majority of entrants have used a combination of DIRA milk and direct supply from farmers.90 This may suggest that DIRA milk is useful in facilitating minimum viable capacity utilisation during the one to three years following entry. 4.38 Once an IP’s reputation has been built, it is in a better position to attract farmers and increase production to full capacity and DIRA milk is no longer necessary. Although some IPs told us DIRA milk was not a necessary requirement to their entry considerations,91 others indicated that it was important for prospective new entrants.92 On balance, we consider that DIRA milk has facilitated entry and expansion. 89 Dairy Industry Restructuring (Raw Milk) Regulations 2012, s 4. 90 NERA Economic Consulting for Fonterra "Assessment of Competition in Raw Milk Markets and Costs and Benefits of the DIRA provisions" 17 August 2015, p.38. 91 [ ]Farmers Weekly News (2015) "New dairy plant goes organic” (Viewed February 2015). We note that according to media reports, the Chinese investors in He Run have withdrawn and that the proposed dairy factory will no longer be built. 92 [ ] 2440192 53 Open entry and exit93 4.39 The open entry and exit provisions oblige Fonterra to accept applications to supply it with milk, and provide a right for shareholding farmers to cease or reduce the supply of milk to Fonterra. 4.40 Therefore, the effect of the open entry and exit provisions of the DIRA has been to ensure that Fonterra cannot lock-in its supplier shareholders through the use of long-term contracts or exclusivity requirements. 4.41 Winning farmer suppliers is important for an IP’s success in entering and expanding in the farm gate market.94 It may be difficult for a new entrant to sign up enough farmers to meet their minimum viable capacity utilisation requirements or to achieve ‘critical mass’ in the absence of the open exit and entry provisions. This is because farmers will not typically commit to IPs that do not have a track record as the farmers need a guaranteed offtake of their supply. However, the processors need guaranteed farmers before they invest in expensive plant. NERA refers to this as the ‘catch-22’ situation.95 4.42 The open entry and exit provisions reduce the risk to farmers of switching to an IP that does not have a reputation by assuring that they can return to Fonterra if the arrangement with the IP does not work out. Open entry also facilitates new dairy conversions (and an increase in the milk supply) as farmers have a guarantee that Fonterra will take their milk (subject to limited exclusions relating to minimal supply and transport costs)96. 4.43 Almost all IPs emphasised the importance of these provisions in relation to their ability and decisions to enter. A number of the IPs noted that the open entry and exit provisions give farmers the confidence to leave Fonterra because they know they can always re-enter at a later stage.97 Fonterra also stated that exiting 93 Dairy Industry Restructuring Act 2001, s 73 and s 97. 94 Although non-DIRA milk supplied at the factory gate could theoretically support entry or expansion, this milk is supplied at higher prices than DIRA milk. DIRA milk may also be more costly than sourcing milk directly from farmers in some cases, because IPs can attract suppliers that are located close to their plants with transport costs that are lower than Fonterra’s average transport costs used to calculate the DIRA milk price. As such, factory gate milk might not always be a good alternative to sourcing directly from farmers in the longer term for the larger entrants that can achieve economies of collection. [ ] 95 NERA Economic Consulting for Fonterra "Assessment of Competition in Raw Milk Markets and Costs and Benefits of the DIRA provisions" 17 August 2015, p.9. 96 Dairy Industry Restructuring Act 2001, s 94 and s 95. 97 For example: Open Country Dairy "Cross-submission on consultation on substantive issues—review of the state of competition in the New Zealand dairy industry" 31 August 2015, paragraph 4.4. 2440192 54 suppliers often cite their right to return to Fonterra as a factor in their decision to exit.98 4.44 Most of the IPs consider the open entry and exit provisions to be the core of the DIRA regime. 4.45 We consider that these provisions have been important in promoting entry and expansion by IPs. However, they are less important for well-established IPs. Right to supply IPs—the 20% rule99 4.46 The 20% rule prevents Fonterra from requiring its supplier shareholders to supply it with all the milk they produce, by allowing a Fonterra supplier shareholder to sell up to 20% of their raw milk production to IPs. 4.47 Most of the large IPs told us they do not use the 20% rule as it is impractical and it requires duplicate testing on the farm and duplicate vats.100 However, some large IPs do purchase a small amount of milk under this provision.101 4.48 In contrast to large IPs, this provision appears to be valuable for a significant number of the small IPs that supply dairy products to the domestic market. These include, in particular, smaller cheesemakers.102 4.49 These small IPs do not typically have access to DIRA milk because they do not meet Fonterra’s minimum volume delivery requirement. In addition, purchasing at the factory gate could lead to lower quality product being produced because the milk may not be as fresh or have the particular characteristics the IP wants. 98 Fonterra "Review of the state of competition in the New Zealand dairy industry: Consultation paper— process and approach" 17 August 2015, paragraph 29.6. 99 Dairy Industry Restructuring Act 2001, s 108. 100 For example: Westland "Review of the state of competition in the New Zealand dairy industry: Consultation paper—process and approach" 10 July 2015; [ ] 101 [ ] 102 For example: Grinning Gecko Cheese "Consultation on substantive issues—review of the state of competition in the New Zealand dairy industry" 14 August 2015, paragraph 3; [ ] Mercer Cheese "Information request by the Commerce Commission: Review of the state of competition in the New Zealand dairy industry" 13 August 2015; Over the Moon Dairy "Consultation on substantive issues—review of the state of competition in the New Zealand dairy industry" 13 August 2015, pp.1–2. 2440192 55 4.50 In the absence of the 20% rule, some existing cheesemakers may be excluded from the market or prevented from operating as they would not have access to milk at the factory gate and are often not large enough to take all of even a small farm’s milk. 4.51 We consider that without access to milk under the 20% rule, these small IPs would face significantly higher barriers to entry into and expansion in the farm gate markets.103 Non-discrimination rule104 4.52 The non-discrimination rule ensures that Fonterra cannot discriminate between a new entrant105 and other shareholding farmers. It does so by stipulating that the terms of supply that apply to a new entrant must be the same as those that apply to a shareholding farmer in the same circumstances, and may only differ to reflect different circumstances. 4.53 For example, Fonterra, when purchasing NZDL when it went into receivership, upon the NZDL farmers’ applications, offered to buy milk from the farmers on less favourable terms than other shareholding farmers. The court found that the treatment of the NZDL farmer suppliers as compared to shareholding farmers was a breach of section 106 of DIRA, as the different terms were not objectively justified by different circumstances.106 In the absence of the DIRA Regulation, there is an enhanced risk of such asymmetric treatment, which could in turn dissuade farmers from leaving Fonterra in the first place. 4.54 It also prevents Fonterra from discriminating against a shareholding farmer who exercises an entitlement under subpart 5. 4.55 The rule means that Fonterra cannot discourage farmer switching by discriminating against suppliers who previously exited and then re-entered Fonterra, or who reduced their supply to Fonterra under the 20% rule. 4.56 The rule is therefore important in bolstering the open entry and exit rule and the 20% rule. 103 Clause 3.20 of Fonterra’s constitution provides for suppliers to supply up to 20% of their output to IPs. However, it is unclear whether this clause would be retained or how it would be applied in the absence of the 20% rule, as Fonterra’s Board has a wide discretion to set the conditions under which split supply will be accepted (which is currently subject to the DIRA), and this clause was specifically included in Fonterra’s constitution under the authorisation of the amalgamation which created Fonterra. 104 Dairy Industry Restructuring Act 2001, s 106. 105 S 5 of the Dairy Industry Restructuring Act 2001 defines a “new entrant” as “a dairy farmer who is not a shareholding farmer who applies to become a shareholding farmer under s 73”. 106 MacIntyre and Williamson Partnership v Fonterra Co-Operative Group Limited [2015] NZHC 3012. 2440192 56 Regulation of supply contracts for raw milk—the 33% rule107 4.57 Under the DIRA Regulation, Fonterra is free, subject to the 33% rule, to arrange milk supply from non-shareholder suppliers outside the open entry and exit regime under normal commercial contracts. This means Fonterra could lock-in farmers who are willing to commit to supply for long terms and who are not supplier shareholders. 4.58 This rule ensures that, at any time, at least a third of the milk solids produced within a 160km radius of any point in New Zealand is supplied either under contracts with IPs, or under short term contracts with Fonterra.108 It therefore constrains Fonterra’s ability to tie up all regional supply outside the open entry and exit regime for periods longer than one season and ensures that IPs are able to compete with Fonterra for the supply of a significant volume of raw milk on at least an annual basis. 4.59 An IP noted that this rule is difficult to measure and impractical given that it still allows Fonterra to impose long-term notice provisions on two-thirds of raw milk supply within a 160km radius of its plant.109 4.60 The value of this rule appears unclear given the relatively low proportion of Fonterra supply from non-shareholder suppliers (currently around 8%) and the open entry and exit provisions that ensure IPs are able to compete for supply from Fonterra’s supplier shareholders. Sale of milk vats on supplier exit110 4.61 Since the refrigerated milk storage vats on Fonterra’s shareholder suppliers’ farms are usually owned by Fonterra, the need to transfer ownership of the vats is a potential barrier to farmers switching from Fonterra to an IP. 4.62 The ‘sale of milk vats on supplier exit’ rule therefore provides that a shareholding farmer who withdraws from Fonterra may require Fonterra to sell the vat situated on their farm to either the shareholding farmer or an IP. It also provides a process for determining the price of the vat in the event that the parties cannot agree. 107 Dairy Industry Restructuring Act 2001, s 107. 108 Under s 107, 33% or more of the milk solids produced in a region must be supplied to an IP or supplied under contract to Fonterra that expires or can be terminated by the farmer supplier at the end of the current season without penalty to the farmer supplier and on expiry or termination all the farmer supplier’s obligations to supply milk to Fonterra are extinguished so that they are free to supply to an IP. 109 [ 110 Dairy Industry Restructuring Act 2001, s 109. 2440192 ] 57 4.63 Most IPs do not consider this rule to be significant in relation to their ability to obtain raw milk supply as they consider that Fonterra would either sell the vat or the exiting farmers would purchase a new vat.111 Only one large IP considered this to be a valuable provision as it was easier to purchase an existing vat than to install a new vat.112 4.64 We consider that the ‘sale of milk vats’ rule is less important than the other rules discussed above, as it generally does not appear to be a significant consideration in relation to switching. Milk price regime113 4.65 The purpose of the milk price regime as set out in subpart 5A is to promote the setting of the base (farm gate milk) price that provides an incentive to Fonterra to operate efficiently while providing contestability in the market for the purchase of milk from farmers.114 4.66 The milk price regime rules codify the setting of the base milk price through the Milk Price Manual and Fonterra’s governance arrangements for determining a base milk price, which are also currently included in its constitution. 4.67 The rules provide for: 4.67.1 Fonterra’s setting of the base milk price; 4.67.2 an independent milk price panel; 4.67.3 Fonterra’s disclosure obligations in relation to its base milk price setting; and 111 [ ] 112 [ 113 Dairy Industry Restructuring Act 2001, subpart 5A. The level of the farm gate milk price that Fonterra pays its farmers was not originally regulated by the DIRA, although Fonterra adopted separate milk price and dividend payments from the 2009/10 season. The methodology adopted in the DIRA is largely in line with the methodology Fonterra applied prior to the 2012 DIRA Amendment. The Milk Price Regime also has a role to play in the factory gate market as it forms the basis for the price at which raw milk must be supplied under the Raw Milk Regulations. 114 Dairy Industry Restructuring Act 2001, s 150A. Contestability is provided for if any of the notional costs, revenues, or other assumptions taken into account in calculating the base milk price are practically feasible for an efficient processor. 2440192 ] 58 4.67.4 the monitoring of Fonterra’s Milk Price Manual and of Fonterra’s calculation of the base milk price by the Commission. 4.68 In order to achieve the subpart 5A purpose, Fonterra must set the base milk price consistent with certain principles such as using a portfolio of commodities that are likely to be the most profitable within a five-year period.115 However, there is no legal obligation in the current regulation for Fonterra to set a milk price consistent with the subpart 5A purpose. 4.69 Under subpart 5A, Fonterra is also able to pay a different milk price than that calculated under its manual, provided that it publically states the reasons why it has departed from the milk price recommended by Fonterra’s Milk Price Panel.116 Even when following the Milk Price Manual, Fonterra also has a degree of flexibility to change its approach from season to season, as the rules in the manual are not prescriptive. This could result in a higher or lower milk price. 4.70 However, the regulation does provide incentives for Fonterra to act consistently with the subpart 5A purpose in the following ways: 4.71 4.70.1 in our statutory reviews, we must state our views on whether Fonterra’s Milk Price Manual and base milk price calculation is consistent with the subpart 5A purpose;117 4.70.2 Fonterra is required to annually disclose its reasons papers outlining its view on the extent to which the Milk Price Manual and the base milk price calculation is consistent with the subpart 5A purpose.118 The base milk price regime also creates more transparency of information about how Fonterra’s farm gate milk price is set and gives IPs greater confidence that Fonterra’s milk price reflects market prices for commodities and efficient costs of collecting and processing milk. 119 115 Dairy Industry Restructuring Act 2001, s 150C. 116 Dairy Industry Restructuring Act 2001, s 150N. In the 2013/14 season, Fonterra exercised its right to pay a milk price different from a milk price calculated under the Milk Price Manual by paying a milk price that was 53 cents lower than the Manual calculated milk price. S 150G also allows Fonterra to amend or not amend the Manual, which is inconsistent with the milk price panel’s recommendation provided Fonterra publically states its reasons for doing so. 117 Dairy Industry Restructuring Act 2001, s 150I and s 150P. 118 Dairy Industry Restructuring Act 2001, s 150L and s 150T. 119 We note that Fonterra is likely to retain some transparency around price setting in the absence of the milk price regime. 2440192 59 4.72 The Commission’s monitoring of the base milk price calculation may also provide some disincentive for Fonterra to set the farm gate price of milk above competitive levels and so promote contestability in the farm gate market.120 4.73 IPs have generally favoured the increased transparency of the milk price provided by the milk price monitoring regime, but have voiced their concerns that there are insufficient constraints on how Fonterra sets its milk price.121 4.74 We consider that the increased transparency of information and the additional independent assurance provided by our reviews under the milk price regime may help reduce barriers to entry by IPs, by providing some disincentive for Fonterra to set the farm gate price of milk too high, and providing IPs and other stakeholders with a better understanding of and confidence in how Fonterra sets its milk price. The risk that Fonterra engages in strategic behaviour to discourage prospective entrants or expansion by rivals 4.75 Given its dominant market position, there is a risk that Fonterra could engage in strategic behaviour to discourage prospective entrants or expansion by rivals. 4.76 However, the DIRA Regulation has successfully lowered barriers to entry into and enabled expansion by IPs in the farm gate market. 4.77 Fonterra’s ability to use market power to engage in strategic behaviour to discourage prospective entrants or expansion by IPs is therefore largely constrained by the DIRA Regulation. 4.78 Fonterra’s ability and incentive to exercise market power against IPs in the absence of the DIRA Regulation is discussed in chapter 5, State of competition without the DIRA Regulation. Farm gate market—independent rivalry 4.79 As well as looking at the conditions faced by dairy processors entering or expanding in the farm gate markets, we assessed the current degree of independent rivalry in the farm gate markets. Independent rivalry refers to the extent to which Fonterra and IPs compete with each other to attract supply from farmers. 120 We note that Fonterra’s ability to flex the mix of the farm gate milk price and the dividends paid to farmers is arguably also moderated by farmer-shareholders being able to trade shares with other farmers and external shareholders, under Fonterra’s ‘trading among farmers’ clause. 121 For example, there is no legal obligation for Fonterra to consider or to make changes to the milk price setting based on our findings and the Commission cannot state its view on what the milk price should be. 2440192 60 4.80 Rivalry can constrain and lead to a reduction of market power. This is generally evidenced by lower levels of market concentration over time, entry and expansion by different firms in the markets and farmer switching: 4.80.1 changes in market concentration; 4.80.2 entry and expansion in the markets; and 4.80.3 the extent of farmer switching. Market concentration has changed over time, but Fonterra still has largest farm gate market share 4.81 The extent of competition for milk supply varies by region depending on whether or not Fonterra faces competition in its collection areas from rival IPs. 4.82 Nine IPs currently compete with Fonterra for milk supply across different geographic regions. 4.83 Fonterra principally competes for farmer supply with the IPs set out in table 4.1, which shows their market shares in different regions based on their milk collected for 2015. 2440192 61 Table 4.1 Farm gate market share of total collections, 2014/15 dairy season Northland Auckland / Waikato / Bay of Plenty Taranaki Canterbury Otago / Southland Fonterra [ [ ] [ ] [ [ Fresha Valley [ ] Open Country [ ] [ ] Tatua [ ] Miraka [ ] Green Valley [ ] ] ] [ ] Westland [ ] Synlait [ Oceania [ ] Danone Nutricia ] ] [ ] Source: Commission analysis based on information provided by Fonterra and market participants. 4.84 The market share of the largest firm in a particular market is an important indicator of market concentration and can also indicate possible market power. Relative market shares are also important, as a high market share when all other competitors have very low market shares is a stronger indicator of market power. The number of firms and the relative size of their shares of a particular market indicate the level of market concentration. 4.85 Table 4.1 shows that Fonterra dominates the regionals, except for the West Coast where Westland is the sole IP and is therefore excluded from our analysis. 4.86 Fonterra’s lowest market share is in the Canterbury region where it is still in excess of [ ]%. [ ] 4.87 Fonterra faces limited competition in all of these regions, with only a small number of competing IPs (between one and four) in the different regions, all of which are much smaller than Fonterra. 2440192 62 4.88 The market shares of IPs in the different regions range from [ ] There are no instances where Fonterra faces competition from more than one IP of any scale— no more than one IP has market shares exceeding 5% in any of the regions. 4.89 Table 4.2 shows that Fonterra also has the largest share of processing capacity in all regions. Table 4.2 Farm gate market share of total maximum processing capacity, 2014/15 dairy season Northland Fonterra [ ] Fresha Valley [ ] Auckland / Waikato / Bay of Plenty [ ] Open Country [ ] Tatua [ ] Miraka [ ] Green Valley [ ] Taranaki [ ] Canterbury [ Otago / Southland ] [ ] [ ] [ ] Westland [ ] Synlait [ ] Oceania [ ] Danone Nutricia [ ] Source: Commission analysis based on information provided by Fonterra and market participants. 4.90 122 Fonterra is a monopoly purchaser of raw milk in a substantial number of other regional markets. NERA’s report attached to Fonterra’s submission stated that around 73% of the milk it collects is in catchment areas where competitors also collect milk. This leaves Fonterra as a sole purchaser in the remaining 27% of its catchment areas.122 NERA Economic Consulting for Fonterra “Assessment of Competition in Raw Milk Markets and Costs and Benefits of the DIRA provisions" 17 August 2015, p.11. 2440192 63 Entry and expansion in the markets 4.91 In general, market power is more likely to exist if a firm has a persistently high market share. This is more likely where high barriers to entry and expansion impede competition. 4.92 However, our analysis shows that with the DIRA Regulation, there has been significant new entry, supporting the contestability of the farm gate market under regulation. Since 2010, three large IPs entered the market: Miraka, Oceania, and Danone Nutricia.123 Large IPs are now well-established in particular regions, and are competing to attract and retain supply of raw milk from farmers, including capturing a significant proportion of milk from new farm conversions and winning farmers from Fonterra. However, these IPs are still small relative to Fonterra, and unlike Fonterra typically only have one plant per region, while Fonterra often has multiple plants and sites per region. 4.93 These IPs have focused on entering or expanding in regions of highest milk growth and concentration of milk production.124 Most have entered and/or expanded into Canterbury (Synlait, Westland, Oceania), Southland (Open Country, Danone Nutricia), Taranaki (Open Country), and Waikato (Miraka, Open Country). 4.94 Table 4.3 shows the changes between 2010 and 2015 in the supply sources, volumes, and capacities of Tatua and Westland, and the processors that have entered the farm gate markets since 2001. 123 [ ] 124 Tatua "Consultation on substantive issues—review of the state of competition in the New Zealand dairy industry" 17 August 2015, para 2.4. 2440192 64 Table 4.3 Supply sources, volumes, capacity, and products, of IPs, 2010 and 2015125 Processing sites Entry date Processor 2010 2015 Supplying farms 2010 Number 2015 Number Volume processed 2010 2015 Million litres/year Capacity 2010 2015 Million litres/year Tatua 1919 1 1 112 [ ] 138 [ ] 200 [ ] Westland 1937 1 1 380 [ ] 479 [ ] 550 [ ] Open Country 2004 3 3 510 [ ] 740 [ ] 800 [ ] Synlait 2008 1 1 60 [ ] 290 [ ] 400 [ ] Miraka 2011 - 1 - [ ] - [ ] - [ ] Oceania 2014 - 1 - [ ] - [ ] - [ ] Danone Nutricia 2014 - 1 - [ ] - [ ] - [ ] Source: NERA –An assessment of the DIRA triggers (2010) p.15 and Commission analysis based on information provided by Fonterra and market participants. 4.95 While the volume of milk Fonterra collects continues to increase, Fonterra is growing at a slower rate than its competitors. Fonterra’s market shares are therefore gradually reducing as its share is eroded by IPs. 4.96 This trend may well continue as most of the large IPs are planning to expand their operations and to seek direct farmer supply to meet their increased demand for raw milk. 4.97 Table 4.4 sets out the expansion plans of IPs over the next six years. 125 NZDL was sold to Fonterra in 2012 and so is excluded from table 4.3. 2440192 65 Expansion plans of IPs seeking direct farmer supply126 Table 4.4 IP [ Future additional capacity usage ] Increase relative to 2015 [ [ 127 ] [ [ ] [ ] ] ] [ ] [ [ 129 ] [ 128 ] [ ] [ 130 ] [ ] ] [ ] [ 131 ] [ [ ] [ ] ] [ [ 133 ] 126 [ 132 ] ] This table lists the IPs that seek direct farmer supply rather than expansion through factory gate milk. It excludes those IPs that seek to expand by purchasing more milk under the 20% rule. 127 [ ] 128 [ ] 129 [ ] 130 [ ] 131 [ ] 132 [ ] 133 [ ] 2440192 66 IP [ [ Future additional capacity usage ] [ ] Increase relative to 2015 ] [ ] [ [ ] ] 134 Source: IPs’ submissions, information request responses and interviews. 4.98 In summary, over the next six years, IPs [ ] will be looking to add direct 135 farmer supply equivalent to [ ] farms. Using the average size of farm, this would equate to just over [ ]% of Fonterra’s approximately 10,600 shareholders.136 Together, this is just under [ ]% of Fonterra’s total current processing capacity. 4.99 We note, however, that Fonterra is also planning to expand, so the extent of any future changes in Fonterra’s market share is unclear. Extent of rivalry and farmer switching137 4.100 We assessed farmer switching—which describes the situation when a farmer shifts their supply of milk from one dairy processor to another—to help assess the extent of rivalry between Fonterra and IPs. However, we note that the level of competition may be understated by switching data if IPs and Fonterra are competing to retain existing suppliers because those that contemplate switching, but choose not to, will not be represented in the data. 4.101 The evidence we found on rivalry suggests that farmers have benefited from rivalry. [ ] stated that competition has meant farmers have greater choice and increased competition for service from dairy processors.138 We consider that 134 [ ] 135 [ ]. According to NERA’s estimate, there are 29 farms to every 50 million litres of milk on average. NERA Economic Consulting for Fonterra “Assessment of Competition in Raw Milk Markets and Costs and Benefits of the DIRA provisions" 17 August 2015, p.31. 136 [ 137 We cannot ascertain the extent of competition from simply looking at whether or not Fonterra faces competition in its collection areas from rival IPs. Even where there are rival IPs within a common catchment area, the degree of competition may be weak, or, alternatively, the market shares of those rivals may underrepresent their competitive impact. 138 [ 2440192 ] ] 67 effective competition is starting to emerge in the Canterbury, Southland, Taranaki and Waikato regions, while there is little or no competition in Northland, Wairarapa, or Hawke’s Bay regions. However, given the high level of concentration in all these markets, we do not consider the existing constraints from competition on their own would be sufficient to constrain Fonterra’s ability to exercise market power. 4.102 IPs are likely to compete more weakly with Fonterra for farmer suppliers the further their plants are from farms as this increases their transport costs, which in turn affects the price they can pay suppliers. 4.103 The competitive constraints on Fonterra will also be lessened where IPs do not have the capacity to compete and are small in comparison to Fonterra.139 In contrast, there are likely to be stronger competitive constraints on Fonterra where rival IPs are situated close to supplier farmers and have excess processing capacity. 4.104 Farmer switching is also affected by any costs of switching, despite the existence of the open entry and exit provisions. One such possible switching cost is the interestfree loan scheme Fonterra made available to its farmer suppliers on 1 September 2015.140 [ ] Figure 4.1 Analysis of switching at the farm gate, 2005/06–2015/16 [ ] Source: Information request by the Commerce Commission141 139 For example, according to Federated Farmers, Westland is not seeking new suppliers as it has enough milk to operate its Canterbury plant efficiently and seeking more suppliers could lead to a greater Fonterra presence damaging its home base; Federated Farmers "Review of the state of competition in the New Zealand dairy industry: Consultation paper—process and approach" 10 July 2015, paragraph 3.6.1. 140 Fonterra “Fonterra Farmers can now apply for co-operative support” (2015) (Viewed September 2015). 141 [ 2440192 ] 68 4.105 There has been an increase in the estimated competition ceases since the 2010/11 season, but conversions have until recently more than made up for Fonterra’s lost supply. Fonterra’s increase in ‘win backs’ from competitors in 2012/13 is likely to largely be accounted for by the farmer suppliers that accompanied Fonterra’s NZ Dairy Ltd purchase.142 4.106 While IPs have also grown their supply from conversions, most growth by IPs is due to suppliers switching from Fonterra.143 ]144 4.106.1 [ ]145 4.106.2 [ 4.106.3 [ ]146 4.106.4 [ ] 147 4.106.5 [ ]148 4.107 Corporate IPs sign farmer suppliers without those suppliers having to purchase shares in the company.149 NERA maintains that this is a key point of differentiation when competing for new conversions.150 [ ]151 Federated Farmers have noted that switching is 142 Rural News “Fonterra agrees to buy NZDL milk processing factory (2012) (Viewed 2 September 2015). 143 In general, we would expect entry to be more difficult if conversions declined. We would also expect Fonterra to be more concerned with switching by its existing suppliers if there were fewer conversions (or increases in milk production by existing farmers declined). As such, as conversions decline, there should be greater competition over existing suppliers. 144 [ ] 145 [ ] 146 [ 147 [ ] 148 [ ] 149 Westland and Tatua are also co-operatives that require their farmer suppliers to buy shares to back their production. 150 NERA Economic Consulting for Fonterra "Assessment of Competition in Raw Milk Markets and Costs and Benefits of the DIRA provisions" 17 August 2015, p.22. 151 [ 2440192 ] ] 69 attractive for farmers who wish to sell their Fonterra shares and realise their capital.152 4.108 The IPs have typically followed Fonterra’s lead in paying for milk at the farm gate and absorbing milk transport costs. Fonterra notes that processors that have been successful in contracting farms close to their plants often offer higher farm gate prices where their transport costs are less than Fonterra’s national average.153 4.109 The IPs sometimes pay a premium on the Fonterra price in order to attract farmers.154 Open Country also notes that IPs cannot attract supply from farmers without offering a price that is at or benchmarked against Fonterra’s milk price. 155 This suggests that price is a likely a key driver of switching by farmers. 4.110 Fonterra has a national pricing policy for farm gate milk; that is, it pays farmers the same price regardless of location.156 This may inhibit its ability to respond to regional competition for farmer supply since a price increase in one region would, according to the national pricing policy, result in a higher price in all regions. This should facilitate entry since the ‘cost’ of an increase in farm gate prices is lower for an IP than Fonterra.157 4.111 Competition by way of non-price incentives offered by Fonterra appears to be similarly non-region specific. National non-price incentives include financial assistance, interest-free loans, and access to support services such as food safety teams.158 These incentives are offered to all farmers, regardless of location. 4.112 As at 1 September 2015, Fonterra announced farmer shareholders could apply for an interest-free loan of 50 cents for each kilogram of share-backed milk solids 152 Federated Farmers “Review of the state of competition in the New Zealand dairy industry: Consultation paper—process and approach”, 10 July 2015, para 3.10- 3.11. 153 Fonterra "Review of the state of competition in the New Zealand dairy industry: Consultation paper— process and approach" 17 August 2015, paragraph 29.4. 154 IPs might be able to afford to pay a higher price because they have lower transport costs (cherry-picking) and/or because they have higher value product mix and/or are more efficient than Fonterra. 155 Open Country “Submission on the Commerce Commission’s Consultation Paper—Review of the State of Competition in the New Zealand Dairy Industry”, p.2. 156 [ ] We note that the milk price actually paid to each farmer and farmers in each region varies to reflect the fat/protein content of the milk. 157 We note that during 2008, Fonterra engaged in tactical pricing to retain supply, but it voluntarily abandoned this practice. 158 Fonterra “Supporting our Farmer Shareholders” (Viewed 4 October 2015). 2440192 70 produced from 1 June to 31 December 2015.159 While this scheme may not have been developed in response to competition for farmer suppliers, it has the potential to impact on that competition (in any region).160 4.113 The loan scheme may affect a farmer’s costs of switching to an IP. If a borrower leaves Fonterra, they are required to repay the entire loan immediately (as opposed to at their discretion, subject to the automatic repayment requirements), and the interest charged on the loan reverts to Fonterra’s usual rate on debit balances. That interest is currently [ ] a year not the 0% interest charged until at least 31 May 2017 (at which point the interest rate may increase but is still capped at a more favourable rate than Fonterra’s usual rate).161 4.114 The value of a 50 cent interest-free loan for a year (which would otherwise incur interest costs of [ ]) is [ ] per year per kgMS. However, its net impact on exit would likely be less as the payment would be offset to some extent by a reduction in dividend payment (since presumably profits would be higher in the absence of the interest rate subsidy). We note that uptake of this interest-free loan was 75% of farmer shareholders (7,800 farmers) by the end of September 2015. 162 4.115 Fonterra has also responded to regional competition with some non-price incentives in order to limit switching and compete for new dairy conversions. One such incentive is MyMilk, which allows farmers to supply milk to Fonterra for five years without needing to share-up.163 It is available in regions with strong 159 The loan is interest free until 31 May 2017. After that, interest may be charged at Fonterra’s cost of borrowing up to a maximum of the published wholesale inter-bank rate plus 0.5% per year. Automatic repayments begin when the total advance rate payments exceed $6.00 and will apply to season-to-date production. Fonterra “Fonterra Loan to Shared-Up Farmers – Response to Synlait Letter dated 12 August 2015” 31 August 2015. 160 In our draft report on the review of Fonterra’s 2015/16 Milk Price Manual (15 October 2015), we stated we would address the competitive aspects of the loan scheme in the report to the Minister on the state of competition in the New Zealand dairy industry. 161 The maximum interest rate is the published wholesale inter-bank rate plus 0.5% per year. Fonterra “Fonterra Loan to Shared-Up Farmers – Response to Synlait Letter dated 12 August 2015” 31 August 2015. 162 Radio New Zealand News “75% of Fonterra farmers take interest-free loan” (2015) (Viewed 3 October 2015) 163 NERA notes that Fonterra finds it difficult to attract new conversions because of its requirement to invest in shares. NERA Economic Consulting for Fonterra “Assessment of Competition in Raw Milk Markets and Costs and Benefits of the DIRA provisions" 17 August 2015, p.24. Fonterra has reacted by allowing farmers to ‘share up’ over time through the MyMilk scheme. The Fonterra constitution only allows 15% of total milks solids to be on a contract supply basis. NERA Economic Consulting for Fonterra "Assessment of Competition in Raw Milk Markets and Costs and Benefits of the DIRA provisions" 17 August 2015, p.24. 2440192 71 competitive pressures at the farm gate for conversions—Canterbury, Southland, and Otago. Farm gate market—Fonterra’s co-operative structure 4.116 Fonterra’s incentive to exercise market power at the farm gate is also limited by its co-operative structure. 4.117 As a co-operative, Fonterra has little incentive to depress the farm gate prices below competitive levels because this would not be in the interest of its shareholder farmers. 4.118 Due to its co-operative structure, Fonterra has also adopted a national pricing strategy, which may further limit incentive to depress farm gate prices. While Fonterra could, in theory, abandon the national pricing approach, any alternative pricing model is likely to face opposition from disadvantaged shareholder suppliers. Factory gate market—current state of competition 4.119 The factory gate markets are the markets in which dairy processors supply raw milk they have collected from farmers to other processors and some food and beverage manufacturers, including the DIRA milk by Fonterra to IPs. 4.120 In order to assess the current extent of competition in the factory gate markets we looked at: 4.120.1 factory gate markets and raw milk; 4.120.2 current customers for factory gate DIRA milk; 4.120.3 barriers to entry; 4.120.4 independent rivalry between Fonterra and IPs; and 4.120.5 constraints on Fonterra’s market power. 4.121 Fonterra is the dominant supplier in the factory gate markets. Fonterra is the only supplier of DIRA milk. The only other processors that supply non-DIRA milk are [ ]. Many of the large IPs that could potentially supply the factory gate market do not appear to want to supply small customers, and have indicated that they are unwilling to supply at the DIRA price. 4.122 Furthermore, even if such IPs were to enter the factory gate market Fonterra would remain the only seller of raw milk to IPs in a number of regions. 2440192 72 4.123 The factory gate markets are accordingly highly concentrated with limited participation by IPs. We therefore found that the regional factory gate markets are not currently characterised by effective competition with the DIRA Regulation. Overview of the factory gate market—raw milk 4.124 The factory gate market is the market in which Fonterra and (occasionally other processors) supply raw milk they have collected from farmers to other processors and some food and beverage manufacturers. 4.125 As factory gate raw milk is supplied directly from farms, not from the suppliers’ processing plants, a processor supplies raw milk to other processors when it already collects in that milk catchment zone. Factory gate raw milk supply is therefore similar to a collection service as this milk is not processed by the supplier. 4.126 The factory gate market is very small compared with the farm gate market. The total volume of milk supplied at the factory gate by Fonterra in the 2014/15 dairy season was [ ] litres compared with the total collected of 18.1 billion litres. This means that less than [ ]% of the milk Fonterra collected at the farm gate in the 2014/15 season was sold by Fonterra at the factory gate.164 4.127 Under the Raw Milk Regulations, if requested by a qualifying IP, Fonterra is required to supply set quantities of raw milk at a regulated price.165 The current Raw Milk Regulations require Fonterra to sell up to 50 million litres of raw milk per season to each qualifying IP subject to a total cap of 795 million litres per season. The volume of DIRA milk in 2014/15 was [ ]% of Fonterra’s total factory gate milk, but was less than [ ]% of the milk Fonterra collected at the farm gate. 4.128 There are separate factory gate markets for DIRA milk sold under the Raw Milk Regulations and non-DIRA milk sold outside of the Raw Milk Regulations. 4.129 Over [ ] of the factory gate milk that Fonterra supplied in the 2014/15 season was DIRA milk, with the balance being sold outside the Raw Milk Regulations.166 164 Fonterra information request response 11 June 2015. 165 Dairy Industry Restructuring (Raw Milk) Regulations 2012. 166 Fonterra supplied [ ] litres, or less than 3% of its factory gate milk, outside of the regulations. Fonterra “Information request by the Commerce Commission: Tranche 1 response to info request” 11 September 2015, para 12. 2440192 73 Factory gate market—current customers for DIRA milk 4.130 There are two distinct types of customers for factory gate milk: large IPs that also source their milk directly from farmers (‘own-source’), and IPs that do not (this includes Goodman Fielder). We discuss both types of buyer in more detail in the following paragraphs. 4.131 Table 4.5 summarises the supply of factory gate milk to IPs, excluding the large processors that also own-source milk. Table 4.5 IPs167 that purchase factory gate milk, 2014/15 Region IP Northland Goodman Fielder Alternative factory gate supply 168 [ ] Fresha Valley Dairy Goat Cooperative [ Emerald Foods Group Auckland/ Waikato/ Bay of Plenty Epicurean Dairy ] [ ] [ Factory gate milk (litres) [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] Own farmer supply Yes ] Green Valley Gopals Sweets & Snacks [ Yes ] Gisborne/ Hawke’s Bay/ Wairapapa/ Manawatu Waimata Cheese [ Goodman Fielder [ ] BioFarm Products [ Barrys Bay Cheese Goodman Fielder ] [ ] [ ] [ Yes ] [ ] [ ] [ Yes ] ] Canterbury Karikaas [ ] Serra Natural Foods [ ] [ ] Talbot Forest Cheese [ ] [ ] [ ] [ ] [ ] Whitestone Cheese Otago/ Southland Evansdale Cheese [ [ ] [ ] ] Tasman/ Marlborough 167 Excluding large IPs. 168 [ 2440192 Talleys [ ] ] 74 Source: Commerce Commission analysis based on information provided by Fonterra and market participants. Large IPs that buy DIRA milk to complement their own milk collection 4.132 DIRA milk is a support measure to facilitate entry into the farm gate markets and contestability and competition in downstream markets. 4.133 Until 2012, there was no restriction on IPs accessing DIRA milk irrespective of the volumes they sourced from farmers. 4.134 In 2012, the Raw Milk Regulations were amended to restrict the entitlement to three years where the independent processor’s own supply of raw milk in each of the three consecutive previous seasons was 30 million litres or more. 169 DIRA milk is therefore presently an interim support measure to overcome barriers to entry to the farm gate market for such IPs. 4.135 Most large IPs that buy DIRA milk to complement their own collection are subject to the sunset clause that will restrict their access to DIRA milk after 1 June 2016. These processors will be required under the sunset provisions to seek to fulfil all their milk requirements directly from farmers in the future. This group of processor customers mainly focuses on exports. 4.136 If a more robust factory gate market were to emerge, these IPs would be the most likely processors to enter the market to supply raw milk to other IPs. IPs that rely on DIRA milk for all or most of their milk requirement 4.137 The IPs that buy DIRA milk for all or most of their requirements typically produce dairy products such as cheeses, yoghurt and ice cream for the domestic dairy market. Few fully export. We expect these processors to continue to purchase DIRA milk under the DIRA Regulation. 4.138 These processors generally face significant barriers to collecting milk directly from farmers and typically prefer raw milk to other milk products (ie, pasteurised milk and milk powders). They therefore rely on DIRA milk if and to the extent they qualify for such milk, and factory gate milk at market prices otherwise. 169 Dairy Industry Restructuring (Raw Milk) Regulations 2012, s 6 (3). Under reg 6(3) Fonterra does not have an obligation to supply an independent processor (other than Goodman Fielder) with raw milk in relation to a season beginning on or after 1 June 2016 if the IP’s own supply of raw milk in each of the three consecutive previous seasons was 30 million litres or more. 2440192 75 4.139 The barriers that these processors face to collecting milk directly from farmers include: 4.139.1 the milk curve (the uneven seasonal supply of milk, high in spring, less in summer, and very little in autumn)—some IPs, typically smaller ones, cannot take milk along the milk curve as they require a constant volume (a ‘flatter’ milk supply) in order to use their processing capacity year round; 4.139.2 securing farmer supply—some new entrant IPs cannot secure sufficient farmer supply to be able to use their processing capacity efficiently. In addition, some smaller IPs are not large enough to take all of a farmer’s milk and, even if they are, farmers may be reluctant to supply smaller IPs because they fear that such processors would not have sufficient demand to collect milk daily; 4.139.3 own-supply costs—costs associated with own-supply (including vats, tankers and testing requirement) are not always viable for smaller IPs. 4.140 Figure 4.2 shows the volumes of DIRA milk purchased by IPs that source their milk from farmers and by those that do not (excluding Goodman Fielder) during the 2014/15 season. Figure 4.2 Volumes of DIRA milk purchased by IPs, 2014/15 season170 [ ] Source: Information request by the Commerce Commission.171 Goodman Fielder 4.141 Goodman Fielder supplies fresh milk and other dairy products to downstream markets in competition with Fonterra. It is the largest purchaser of DIRA milk. The total volume of milk supplied to Goodman Fielder at the factory gate by Fonterra in the 2014/15 dairy season was [ ]172 litres. 4.142 DIRA required Fonterra to divest shares in New Zealand Dairy Foods (NZDF). The NZDF business manufactured a number of well-known New Zealand consumer dairy brands and the sale made sure that Fonterra would not monopolise the downstream dairy market in New Zealand. NZDF eventually became the dairy division of Goodman Fielder. 170 [ 171 [ ] 172 [ ] 2440192 ] 76 4.142.1 As part of the negotiations, Fonterra and NZDF agreed commercial contracts allowing NZDF to purchase raw milk subject to the cap of 250 million litres allowed for under the DIRA Regulation. [ ]173 This contract is due to end in 2021.174 4.142.2 Goodman Fielder stated that it has only implemented minor incremental increases in milk purchases and production since 2011 and its growth has plateaued.175 In submissions to the Commission, Fonterra stated that it “considers the requirement to supply Goodman Fielder remains important for public confidence in downstream wholesale and retail markets”. Goodman Fielder’s DIRA milk purchases are shown in figure 4.3. Figure 4.3 Goodman Fielder purchases of DIRA milk, 2005/06–2014/15 [ ] Source: Information request by the Commerce Commission. 4.143 Goodman Fielder told us it will remain reliant on its entitlement to DIRA milk supply to compete downstream with Fonterra until competition develops in the factory gate market. It further claims that no IP, other than Fonterra, can guarantee year round supply of milk on the scale it requires, nor is an IP likely to develop this scale within a reasonable timeframe.176 As such, with the DIRA Regulation in place, Goodman Fielder is unlikely to consider any other source of supply in place of its entitlement to DIRA milk. Factory gate market—barriers to entry 4.144 Since any new entrant into the factory gate market would need to contract with farmers and invest in infrastructure for the collection of milk for delivery, the most obvious candidates to supply raw milk to the factory gate, in the absence of the Raw Milk Regulations, are other IPS—either current or future IPs. 173 [ 174 Goodman Fielder “Re: Commerce Commission Consultation Paper: Review of the state of competition in the New Zealand dairy industry” 10 July 2015, para 1.3(c) 175 Goodman Fielder "Review of the state of competition in the New Zealand dairy industry: Consultation paper—process and approach" 10 July 2015, para 2.1(a). 176 Goodman Fielder "Re: Commerce Commission Consultation Paper: Review of the state of competition in the New Zealand dairy industry." 10 July 2015, para 1.3. 2440192 ] 77 4.145 IPs that are large enough to source their raw milk from the farm gate for processing are unlikely to face significant barriers to supply at the factory gate in the absence of the Raw Milk Regulations. This is because supply would mainly entail transporting farm gate milk to a purchaser rather than the IP’s own processing facility. However, there may be additional transportation costs, as well as transactional costs, including those related to a change in business focus. 4.146 However, it is worth noting that DIRA milk has facilitated farm gate entry by IPs, and having entered the farm gate markets, the barriers to entering the factory gate in a future market are significantly reduced. Factory gate market—independent rivalry 4.147 Fonterra is the dominant supplier in the factory gate market. The other processors that supply this market are [ ]177 [ ] [ ]178 [ ]179 [ ]180 4.148 The market for non-DIRA milk is therefore very small and there is very little competition. Many of the large IPs that could potentially supply the factory gate market do not appear to want to supply small customers, and have indicated that they are unwilling to supply at the DIRA price. 4.149 Furthermore, even if such IPs were to enter the factory gate market Fonterra would remain the only seller of raw milk to IPs in a number of regions. 4.150 Given the limited participation by IPs in the factory gate markets there is currently little independent rivalry. 177 [ ] 178 [ ] 179 [ 180 [ 2440192 ] ] 78 Factory gate market—constraints on Fonterra’s market power 4.151 The Raw Milk Regulations constrain Fonterra’s market power in respect of DIRA milk supplied at the factory gate as the price is regulated. 4.152 However, when IPs purchase non-DIRA milk from Fonterra, they are charged a price for that milk that is [ ]181 higher than the DIRA price. 4.153 IPs have generally indicated that they would be willing to supply at the factory gate on a regular basis at prices similar to Fonterra’s non-DIRA price.182 This could suggest that Fonterra may be the price setter in the case of non-DIRA milk in the factory gate market or else that Fonterra and IPs have similar opportunity costs. 4.154 However, in some instances, the prices charged by Fonterra for non-DIRA milk [ ] 4.155 There are accordingly few constraints on Fonterra’s market power in the market for non-DIRA milk. Wholesale and retail supply of fresh processed milk—any new information on state of competition? Findings in the preliminary inquiry on whether to conduct an inquiry under Part 4 of the Commerce Act 4.156 Under the terms of reference for this report, we had to consider whether there is any new information that would alter the conclusions reached in relation to the wholesale and retail dairy markets, in our preliminary inquiry into domestic milk markets in 2011. 4.157 In addition to the farm gate and factory gate markets, the preliminary inquiry considered the following markets: 4.157.1 the market for the processing and wholesale supply of fresh pasteurised milk (referred to in this report as the market for the wholesale supply of fresh processed milk); and 181 Refer to Attachment E: Table E1 The difference between Fonterra’s average market price and the DIRA price, 2012/13 to 2014/15 182 [ ] 2440192 79 4.157.2 the market for the retail supply of grocery items including fresh processed milk (referred to in this report as the market for the retail supply of fresh processed milk). 4.158 Goods or services may only be regulated under Part 4 of the Commerce Act where they are supplied in a market where there is both “little or no competition” and “little or no likelihood of a substantial increase in competition”. In the 2011 preliminary inquiry, we concluded that the competition test of ‘little or no competition’ was not met in either of the above markets. 4.159 In relation to the market for the processing and wholesale supply of fresh pasteurised milk, we noted that rivalry from Goodman Fielder provided more than the threshold of ‘little or no competition’.183 4.160 We similarly considered that the competition test of ‘little or no competition’ was not met in the markets for the supply of grocery items, including fresh processed milk.184 4.161 There have been changes to the competitive dynamics in certain downstream domestic dairy markets since 2011 (such as the reduction in the number of wholesale suppliers of fresh processed milk (town milk suppliers) and the growth of gourmet yoghurt IPs). However, these changes do not change the conclusions in the 2011 report, since the competition test of ‘little or no competition’ would still not be met. Wholesale fresh processed milk 4.162 The information gathered as part of our review showed that there has been a trend in the exit of town milk suppliers. These suppliers are lower margin/lower valueadd and are therefore particularly exposed to farm gate milk price variability (which in turn is linked to the variability of export prices for commodities such as whole milk powder). 4.163 The wholesale market for the supply of fresh processed milk remains highly concentrated with Fonterra and Goodman Fielder still accounting for more than [ ]% of the market in both the North and South islands, although Fresha Valley also supplies town milk in the north of the North Island and Green Valley Dairies focuses on supplying premium town milk (organic, Lewis Road, etc.). 183 Goods or services may only be regulated under Part 4 of the Commerce Act if they are supplied in a market where - there is little or no competition (s 52G(1)(a)(i), and little or no likelihood of a substantial increase in competition (s 52G(1)(a)(ii). 184 Commerce Commission “Milk markets: Consideration of whether to initiate a Commerce Act Part 4 inquiry into milk prices” August 2011, paragraph 21. 2440192 80 4.164 In addition, according to Foodstuffs, only Fonterra and Goodman Fielder are able to compete to supply private label milk, which is the key driver of price pressure on other market participants in the retail market for fresh milk.185 4.165 We consider that the degree of competition still exceeds the “little or no competition” standard. Retail supply of fresh processed milk 4.166 There appears to have been little change in the market for the retail supply of fresh milk to domestic consumers since the preliminary inquiry: 4.166.1 milk is still supplied by supermarkets and retailers in the route trade (such as petrol stations and dairies); 4.166.2 the two main supermarket chains (Foodstuffs and Progressive Enterprises) continue to compete with each other across a basket of groceries that includes milk; and 4.166.3 there is still some competitive constraint from retailers in the route trade on the supermarkets’ behaviour in pricing milk. 4.167 185 We therefore consider that the degree of competition still exceeds the “little or no competition” standard. Foodstuffs “Dairy Competition Review Consultation Paper, 20 July 2015” 7 August 2015. 2440192 81 CHAPTER 5: State of competition without the DIRA Regulation Purpose of this chapter 5.1 This chapter presents our assessment of what competition would look like in the dairy industry in the absence of the DIRA Regulation. 5.2 We assessed whether Fonterra would be likely to: 5.2.1 exercise buyer side market power against farmer suppliers at the farm gate by decreasing (to below competitive levels186) the price Fonterra pays to the farmers for raw milk; 5.2.2 exercise market power against IPs at the factory gate by increasing the price Fonterra charges processors to buy raw milk; and 5.2.3 prevent IPs from effectively competing (generally referred to as foreclosure) by: 5.2.3.1 restricting IPs’ access to raw milk from farmers at the farm gate. This could be achieved by either increasing the farm gate price or locking farmers into longer term contracts; and/or 5.2.3.2 raising factory gate prices or restricting access to factory gate milk for IPs that sell to customers in domestic downstream dairy markets (such as milk and cream). Key findings on future competition in the farm gate market 5.3 Fonterra would have the ability but little incentive to exercise market power against farmers through lower prices for raw milk. 5.4 We do not think Fonterra would have the incentive to exercise its market power against farmers, as it would not be in the best interests of its farmer shareholders. 5.5 We do not think Fonterra would be likely to have the ability to offer new or returning farmers lower prices in many areas, and to the extent that it could, these prices would not be below competitive levels. 186 The competitive level is defined here as the price that would arise in a market characterised by a number of efficient suppliers. An exercise of buyer side power results in prices that are not high enough for farmers to cover their costs and so would result in a reduction in farmer output (either through farmer exit or retrenchment). 2440192 82 Key findings on future competition in the factory gate market 5.6 Without the DIRA Regulation, the price that Fonterra charges for DIRA milk would be likely to rise. 5.7 IPs with access to their own sources of raw milk would be likely to be less affected if there were no regulations, assuming access to farmers similar to that historically provided by the open entry and exit provisions. 5.8 The factory gate price may vary across regions and/or by buyer. 5.9 Factory gate price increases would be likely to have a negative impact on downstream market. 5.10 Buyers that could not afford a factory gate price that fully reflects all related costs (including opportunity cost) would be unable to compete in the market.187 5.11 Other buyers may be foreclosed if they cannot afford the price increase that reflects market power; however, any such foreclosing effect may be restricted by Fonterra’s (and other IPs that enter into the factory gate market) ability to price discriminate across price sensitive customers. The incentive to price discriminate in this way would depend on the availability of milk. Key findings on Fonterra’s incentive and ability to render competing IPs less competitively effective 5.12 At the farm gate, Fonterra would have limited incentive and little ability to render IPs less able to compete for farmers by increasing the farm gate price. What incentive Fonterra may have to increase the farm gate price in this way is tempered by Fonterra’s co-operative structure. 5.13 Fonterra’s ability to increase the farm gate price is largely in place now with the regulations, and so we do not consider the price would be much impacted, in the absence of the DIRA Regulation. 5.14 Fonterra is likely to have an incentive to render IPs less competitively effective by way of long-term farmer supply contracts in the absence of the open entry and exit provisions. Fonterra is not likely, however, to have the ability to do so in regard to well-established IPs, but could limit the development of new or newer ‘own-source’ IPs. 187 IPs that can only remain in the market at below competitive pricing would not be considered efficient. We consider in chapter 6, Our assessment of the efficiency costs and benefits of the DIRA Regulation, whether the DIRA price may have incentivised inefficient IP entry. 2440192 83 Fonterra’s buyer side market power in relation to farmers 5.15 We find that Fonterra would have the ability to lower farm gate prices to below competitive levels (this is referred to as buyer side market power) in the absence of the DIRA Regulation. Pricing below competitive levels means that prices would not be sufficiently high for all farmers to cover all their costs, and so would result in either farmer exit or retrenchment. 5.16 Fonterra would, however, have little or no incentive to use this to the detriment of shareholding farmers. Fonterra’s ability and incentive to exercise buyer side market power in relation to its existing farmer suppliers 5.17 The base milk price regulations may have some effect in protecting Fonterra’s farmer suppliers from Fonterra’s buyer side market power.188 If the regulations were removed, Fonterra would probably have some buyer side market power. The farm gate raw milk market is highly concentrated. Despite new IPs entering the dairy processing market and expanding their business, Fonterra remains the largest purchaser of raw milk at the farm gate. 5.18 Fonterra indicates that it does not face competition at the farm gate in 27% of its catchment areas.189 It is therefore the monopsony buyer190 of milk in these regions. It does, however, face some competition for farm gate milk in other regions. 5.19 If the DIRA Regulation was removed, Fonterra would likely have sufficient buyer side market power to reduce farm gate prices to a level that would reduce milk output (ie, reduce prices below competitive levels) to its farmers in at least some regions. We consider that Fonterra is unlikely to do so in the absence of regulation for the following reasons:191 5.19.1 Fonterra’s structure: Fonterra is a co-operative owned by its farmer suppliers, where non-supplier shareholders do not have voting rights. Lowering prices to below competitive levels is generally not in farmers’ 188 The regulations are primarily intended to provide assurance to both farmer suppliers and independent investors that an appropriate allocation of returns is made between the milk price and dividend. This also provides IPs with comfort that the milk price is set at an efficient level. 189 NERA Economic Consulting for Fonterra “Assessment of Competition in Raw Milk Markets and Costs and Benefits of the DIRA provisions" 17 August 2015, p.11. 190 ‘Monopsony buyer’ means that it is the only buyer. 191 This finding is consistent with the Commission’s findings in the Fonterra/NZDL merger. Fonterra Limited and New Zealand Dairies Limited (in receivership) [2012] NZCC 21. 2440192 84 (the supplier owners’) interest because it results in a cutback in farmer output due to an inability to cover costs.192 5.19.2 Independent oversight of the milk price: It is likely that some form of independent oversight of the base milk price rules would continue to occur. This would provide some assurance to Fonterra’s farmer and nonfarmer shareholders that the farm gate milk price is set at an efficient level.193 5.19.3 Profitability of a price decrease: The likelihood of such a price decrease being profitable is reduced by the fact that: 5.19.3.1 95% of Fonterra’s milk solids are exported.194 Fonterra is largely a price-taker195 in export markets. As such, any reduced downstream sales due to reduced farm gate purchases are unlikely to be accompanied by increased downstream prices; and 5.19.3.2 Fonterra’s buyer side market power varies by region. However, Fonterra has a national price for farm gate milk. If Fonterra lowered prices nationally, it would stand to lose farmer suppliers to competing IPs in at least some regions. This would further decrease the quantity of farm gate milk available to Fonterra. Fonterra’s ability and incentive to exercise buyer side market power against new and returning farmers 5.20 Fonterra is required under the DIRA Regulation to treat new shareholders the same as existing ones. However, if the regulations were removed, this may not be the case. We therefore considered whether Fonterra would have the incentive to lower prices below competitive levels to only new or returning farmers but found that it would have little ability to do so. 192 Where price here includes all concessions, including the dividend. 193 However, there may be less transparency over how the milk price is set. This could decrease the dairy industry’s confidence that the milk price is set at the competitive level. We acknowledge there may be times where independent oversight may not provide assurance that the milk price is set an efficient level. As outlined in chapter 4, State of competition under the DIRA Regulation, Fonterra are able to pay a milk price that is not consistent with its Manual. Fonterra’s board exercised its ability to pay a lower milk price for the 2013/14 season. 194 Fonterra NERA “ Review of Commerce Commission’s draft report into dairy competition” 4 December 2015, p.3. 195 We refer to a firm being a ‘price taker’ when its actions have no impact on the price of a good or service. 2440192 85 5.21 Farmers would only switch to Fonterra if the prices196 and supply terms they are offered are at or above that which they could get from competing IPs. 197 If a returning farmer has no options other than Fonterra, Fonterra would likely be better positioned to lower prices to that farmer. Fonterra is, however, unlikely to lower prices below competitive levels since that would either force the farmer to exit or decrease its output since it would no longer be able to cover its costs. 5.22 This is similarly the case for new conversions. Conversions are not likely to take place unless farmers have the expectation that they will have competitive terms from either Fonterra or an IP. No new entry would be induced where terms are less than competitive. Fonterra’s ability and incentive to exercise buyer side market power against contracted farmer suppliers 5.23 We also considered whether Fonterra could depress prices below the competitive level to contracted suppliers. The DIRA Regulation limits Fonterra’s ability to use contract suppliers but does not dictate its treatment of them.198 Therefore, removing the regulations is unlikely to change Fonterra’s pricing to contract farmers. 5.24 There is entry at the price Fonterra currently pays to contract farmers, which indicates that prices are at or above competitive levels.199 In fact, contract supply is seen as facilitating a pathway to sharing up.200 Given that shareholder capital is also Fonterra’s primary source of funding,201 Fonterra’s incentives to convert contract suppliers to shareholders would remain with and without the DIRA Regulation. 196 Where price here includes all concessions, including the dividend. 197 There may be scenarios where efficient farmers are excluded from supplying milk even though they are efficient. For example, farmers could switch from Fonterra to an IP. That processor, the only alternative milk purchaser in a region, could go out of business. If Fonterra in such a case does not accept the return of those efficient farmers (in order to strategically disincentivise farmer switching more broadly), then this might introduce inefficiencies into the market. [ ] 198 We also note that Fonterra’s constitution limits the proportion of contract farmers allowed to 15% [ ]. Fonterra Constitution, s3.22. 199 [ 200 [ ] 201 [ ] 2440192 ] 86 Fonterra’s impact on IPs without the DIRA Regulation 5.25 This section presents our assessment of Fonterra’s impact on IPs if the DIRA Regulation was no longer in place. 5.26 Could Fonterra exercise market power at the factory gate and profitability raise the factory gate milk price above competitive levels?202 This would be an exercise of Fonterra’s market power. 5.27 If Fonterra did exercise any market power it may have, would Fonterra have the incentive and the ability to seek to render IPs less competitively effective, in order for Fonterra to obtain additional market power either at the farm gate or in downstream domestic markets? This would be achieved by: 5.27.1 at the factory gate, further raising the factory gate price with the purpose of rendering IPs less competitively effective in downstream dairy markets, allowing Fonterra to ultimately raise prices in those markets; and 5.27.2 at the farm gate, raising the farm gate price above competitive levels or committing farmers to longer term exclusive contracts (or otherwise restricting their entry and exit from Fonterra) to prevent IPs from competing effectively for farmer suppliers and so ultimately reducing the price paid for raw milk. Summary of Fonterra’s impact on IPs without the DIRA Regulation 5.28 Without the DIRA Regulation, Fonterra would likely be able to exercise market power at the factory gate, raising prices above competitive levels. A likely indication of that price is the price at which non-DIRA milk is currently sold at the factory gate. 5.29 The effect of such a price increase would be that current buyers of that milk would buy less milk, leading to a decrease in their own output. This in turn would likely adversely impact the price, quantity and, potentially, the quality and variety of dairy products sold in domestic downstream markets.203 The extent of that adverse impact is unknown but would likely vary by product market with some markets not being as impacted as others. 5.30 Having exercised its market power at the factory gate, we consider that Fonterra would be limited in the extent to which it would be able to render its competitors 202 ‘Competitive levels’ is defined here as the price that would arise in a market characterised by a number of efficient suppliers. 203 The competitiveness of IPs that rely on DIRA milk to sell in export markets may also be reduced, but the overall exports are unlikely to be impacted since those sales are likely to be replaced by Fonterra sales. 2440192 87 less competitively effective either by further increasing factory gate prices or, at the farm gate, by changing either the price or non-price terms of contract. This is because: 5.30.1 At the factory gate, the benefit that Fonterra would gain from foreclosing exporters would likely be limited and, even if Fonterra were to face a benefit, it would have little ability to foreclose such IPs given their limited and decreasing reliance on factory gate milk. 5.30.2 Also, Fonterra’s ability to foreclose rivals that compete in the domestic market would also likely be limited since most have alternatives sources of milk.204 There are few domestic rivals that do not have alternative sources of supply, but Fonterra is unlikely to gain much benefit from their foreclosure. 5.30.3 At the farm gate, Fonterra’s current ability to foreclose IPs through farm gate price increases would not be impacted significantly by the removal of the DIRA Regulation. As such, the constraints that now limit foreclosure through farm gate price increases would largely continue to operate without the DIRA Regulation. 5.30.4 The impact of the removal of the open entry and exit provisions would likely be limited in the case of well-established IPs. There is a risk that less well-established IPs seeking to expand through increased farmer supply or new entrants to the farm gate market would be adversely impacted. Fonterra’s market power at the factory gate 5.31 To assess whether Fonterra has market power at the factory gate, we considered what would happen to the factory gate milk price without the DIRA Regulation. 5.32 In considering whether there is likely to be an exercise of market power, it is usually sufficient to determine whether prices are likely to increase. However, the nonregulated factory gate price may increase relative to the regulated DIRA milk price even without an exercise of market power. This is because the DIRA milk price might not reflect Fonterra’s costs of bringing this milk to market (where those costs include opportunity costs). 5.33 Therefore, before we assessed Fonterra’s market power, we considered whether the DIRA milk price is set at a level that covers the associated costs of producing it 204 The IPs that would form the basis for those alternative sources of milk include larger exporters. As such, this finding is in part linked to the finding that Fonterra would have limited incentive and ability to foreclose exporters. 2440192 88 (where those costs include opportunity cost). If not, then prices are likely to increase in the absence of the DIRA Regulation for this reason alone. Is the DIRA price for milk covering its costs? 5.34 In summary, we agree with NERA (for Fonterra) that there may be times when Fonterra does not recover all its costs, including the opportunity costs, associated with selling DIRA milk. 5.35 On the other hand, there may be times when the returns from selling DIRA milk more than recover those costs. 5.36 Fonterra sets the farm gate milk price to reflect the average efficient cost of milk. Since the farm gate price informs the DIRA milk price, the factory gate price should largely also reflect average opportunity cost. The extent to which it may not is beyond the scope of this report. 5.37 We accept, however, that without the regulations, there are scenarios under which the price of raw milk currently sold under the regulations may increase to reflect all of Fonterra’s costs, including its opportunity costs, or conversely, decrease. 5.38 Without the regulations, we would not expect Fonterra to sell milk at the factory gate to IPs if those sales were less profitable to Fonterra than if Fonterra instead processed those milk volumes itself. That is, without regulation we would expect Fonterra to at least seek to recover its opportunity cost on factory gate sales. 5.39 The DIRA Regulation requires Fonterra to supply DIRA milk for the whole of the current season at the most recent quarterly forecast of the farm gate milk price for that season plus the average transport costs Fonterra incurs in the collection of milk from farmers and the delivery of this milk to the purchaser IPs.205 5.40 The only additional cost included in the DIRA milk price, relative to the farm gate milk price, is the average cost of transporting the milk from the farm to processors. 5.41 There is potential for the DIRA price for milk to not reflect Fonterra’s opportunity cost or actual costs of supplying milk at the factory gate because it may not reflect: 5.41.1 205 the opportunity cost on DIRA milk sales as measured against Fonterra’s best alternative to DIRA sales; and Dairy Industry Restructuring (Raw Milk) Regulations 2012, ss 4(1), 19(2) and 19(4). Fonterra must offer the fixed quarterly pricing option to IPs that did not source any raw milk in the previous season from its own supply or who sourced raw milk in the previous season from own supply at a quantity less than 30 million litres. 2440192 89 5.41.2 any additional costs Fonterra may have as a result of supplying raw milk at the factory gate. 5.42 NERA, in its submission to our draft report, submitted that on a systematic basis, the DIRA price will be lower than Fonterra’s opportunity cost.206 5.43 We consider that on average, the DIRA milk price reflects Fonterra’s opportunity costs. This is because: 5.44 5.43.1 the DIRA milk price is based on the farm gate milk price, which in turn is based on a basket of commodity products (to fulfil the purpose of subpart 5A, Fonterra is required to calculate the milk price by using a basket of commodities that is likely to be the most profitable over a period that does not exceed five years207); and 5.43.2 the stream returns on commodity products sold in global markets are largely in ‘equilibrium’ in that they reflect the underlying fat and protein content.208 However, we acknowledge that there will be times when the DIRA price of milk may not reflect the cost, including the opportunity cost, of selling milk under the Raw Milk Regulations. Specifically in regard to opportunity cost, there may be times when the opportunity cost of selling raw milk may exceed the price Fonterra receives for selling DIRA milk; while at other times, the opportunity cost is likely to be less than the DIRA price of milk. This is because: 5.44.1 The DIRA price for milk is only updated quarterly but dairy product prices are volatile. They can show large changes within a season and also 206 NERA Economic Consulting for Fonterra “Review of Commerce Commission’s draft report into dairy competition” 4 December 2015, pp.12-13. NERA cite that at any point in time there will invariably be differences in relative stream returns for the reference commodity products used in the milk price calculation. NERA appears to assume that the production allocation in the milk price calculation is calculated as a weighted average for the entire year. However, the milk price calculation uses Fonterra’s actual monthly allocation of milk into the four milk price product streams made up of whole milk powder (WMP), butter milk powder (BMP), and anhydrous milk fat (AMF): (WMP/butter/BMP, WMP/AMF/BMP, SMP/butter/BMP, and SMP/AMF/BMP) to determine the hypothetical producer’s production for each reference commodity product for each month. This allows Fonterra to optimise its milk production in a monthly basis based on the relative stream returns for the reference commodity products. 207 Dairy Industry Restructuring Act, s 150C(2)(a). 208 We have noted in our 2015/16 base milk price calculation review that there can be instances where the relative prices for reference commodity products used in the milk price calculation and non-reference commodity products can get ‘out of sync’ for periods. For example, the global prices for WMP and SMP have recently been lower than cheese prices. However, market analyst views are that the trend would not continue in the long run. Commerce Commission “Final report: Review of Fonterra’s 2014/15 base milk price calculation” 15 September 2015, p.98. 2440192 90 potentially between updates of the forecast of the farm gate milk price. For example, the forecast of the farm gate milk price for the 2015/16 season for kilograms per milk solids (kgMS) opened at $5.25 per kgMS (May 2015), fell to $3.85 in August, and then partially recovered to $4.60 in September 2015.209 At any single point in time, the opportunity cost of selling raw milk to IPs may be greater than or less than the price forecast by Fonterra for all of that season. 5.44.2 Fonterra, like other processors, has capacity constraints—its ability to produce products enjoying high prices at any point in time may be constrained by the processing capacity it has available. For example, in times when commodity products like whole milk powder (WMP) or skim milk powder (SMP) make less profit than higher value products, Fonterra would be expected to ‘max-out’ its opportunities to produce those higher value products. However, Fonterra may not have the capacity to produce more of the higher value products, even if it did not have to sell milk to IPs under the regulations. 5.44.3 However, at times WMP and SMP may be significantly more profitable than making and selling other products.210 For example, in the 2013/14 season when raw milk volumes peaked and WMP and SMP prices were high and generating greater profits than other dairy products, Fonterra was constrained by its diversified range of processing plants from taking full advantage of the volumes and high WMP and SMP prices.211 That is, it had to produce larger volumes of lower margin product than was desirable. As a result, Fonterra had to adjust down the amount of money it could pay to its suppliers by 55 cents per kgMS. In that season, the sale of additional milk to IPs (at a price reflecting the assumed production of that milk into WMP and SMP) may have provided a higher return to Fonterra than Fonterra retaining the milk for processing into (at that time) lower return product. 5.44.4 The DIRA price of milk assumes raw milk is turned into WMP and SMP at costs that reflect the performance of a hypothetical efficient IP.212 If in 209 Fonterra “Farmgate Milk Prices” . 210 Dairy Industry Restructuring Act 2001, s 150C(2). In calculating the farm gate milk price, the portfolio of commodities must be determined having regard to the commodities that are likely to be the most profitable over a period not exceeding five years from the time when the portfolio is determined. Fonterra also produces products that are further processed and branded and that would not be considered commodities. 211 The volume of milk Fonterra was required to process was also greater than forecast. 212 We describe the ‘notional producer’ concept in: Commerce Commission “Review of Fonterra’s 2014/15 base milk price calculation: Final report” 15 September 2015, Attachment E. 2440192 91 reality Fonterra cannot match the assumed level of efficiency of that hypothetical processor, then selling raw milk at a price that reflects WMP and SMP selling prices and efficient processing and selling costs, but avoids actually incurring those costs, may be value-enhancing to Fonterra. 5.44.5 In the short term, Fonterra’s costs are fixed. Having to sell milk to IPs when that milk could have been processed through its own plants may reduce Fonterra’s ability to spread the fixed costs of its plants over greater volumes of milk product.213 5.44.6 However, in the longer term, growth in milk volumes available to Fonterra for processing (as they have been since it was established), would reduce any such above cost by offsetting the loss of milk volumes sold to other processors’ plants.214 5.44.7 Fonterra may incur some additional costs from supplying milk to IPs (eg, administration costs), but the supply of milk to other processors’ plants may also yield some efficiencies. For example, the ability to deliver milk to an IP’s site may allow Fonterra to reduce collection costs through shorter travel time for some of its tankers. 5.45 While the DIRA milk price, at any point, may be above or below opportunity cost, we consider that the flexibility that buyers of DIRA milk have to reduce their forecasted purchases under the current wording of the DIRA Regulation may impose a cost on Fonterra. 5.46 A consideration in assessing whether the DIRA price is at opportunity cost is the price that Fonterra charges Goodman Fielder. 5.47 [ ]215[ 213 We note too that volumes of milk sold under the regulations are small relative to the total quantity of milk processed by Fonterra, and that the need for processing capacity is determined by the amount of milk produced by cow herds in the spring, and inter-season variability in that level of production, rather than uncertainty over the amount of milk that will or will not be purchased by IPs under the regulations. Further, the volume of raw milk sold under those regulations has been falling (and will likely continue to fall as the sunset clauses impact in practice.) 214 We note also that the volume of milk sold to other processors under the regulations has been falling. [ ] 215 The volume purchased by Goodman Fielder is less than the 250 million litres it is otherwise entitled under the regulations. Goodman Fielder “Commerce Commission Consultation Paper: Review of the state of competition in the New Zealand dairy industry” 10 July 2015, para 3.7. [ ] 2440192 92 ] 5.48 We consider the factory gate price to Goodman Fielder without the regulations further in paragraphs 5.89–5.90. Does Fonterra have market power at the factory gate? 5.49 To assess whether Fonterra would have market power at the factory gate in the absence of DIRA Regulation, we examined: 5.49.1 Fonterra’s current market rate for non-DIRA milk sold at the factory gate to see whether it is likely to inform the price we would observe in the absence of the DIRA Regulation; 5.49.2 whether the competitive constraints Fonterra would face on its factory gate price would be likely to differ by region; and 5.49.3 whether the price would be likely to vary by customer. Fonterra’s current factory gate non-DIRA milk price 5.50 The raw milk that Fonterra sells outside the DIRA Regulation is considerably higher priced than sales made under the regulations. The difference between the price charged for DIRA milk and non-DIRA milk is [ ].216 The non-DIRA milk sales are often for types of milk that DIRA does not cover, such as the supply of organic milk, a flat supply profile, or volumes over and above the October maximum, for which a premium to the DIRA price would be consistent with competitive outcomes. 5.51 [ ]217 ]218 [ 216 Refer to Attachment E: Table E1 The difference between Fonterra’s average market price and the DIRA price, 2012/13 to 2014/15 217 [ ] 218 [ ] 2440192 93 5.52 [ ]219[ ]220 5.53 [ ]221 5.54 [ ]222[ ]223 5.55 The difference between the non-DIRA milk price and the DIRA milk price reflects, among other things, both the products on which the prices are based on ([ ]), the difference in historical prices and 224 actuals, [ ]. 5.56 To date, the Fonterra non-DIRA milk price has not had the effect of inducing entry (other than on an ad hoc basis) into any of the regional factory gate markets by IPs.225 This may be because the non-DIRA volumes are insufficient for processors to be interested.226 219 [ ] 220 [ ] 221 [ 222 [ 223 [ 224 The DIRA milk price is based on the forecasted farm gate price, which is in turn based on historical WMP and SMP product group prices. ] ] ] 225 [ ] 226 NERA contends that Fonterra is unlikely to exercise market power at the factory gate because the volumes sold at the factory gate are small relative to the total capacity of [own-source] IPs. As such, it would be easy for own-source IPs to undermine an exercise of market power by Fonterra by selling the milk themselves. NERA Economic Consulting for Fonterra "Assessment of Competition in Raw Milk Markets and 2440192 94 5.57 Based on our analysis of Fonterra’s current factory gate milk market price, its pricing methodology, and the fact that Fonterra currently faces little competition in the sale of non-DIRA milk, we consider that Fonterra’s current market price for non-DIRA milk likely reflects market power. The factory gate milk price may vary across regions 5.58 We conclude that without the regulations, Fonterra would be likely to price factory gate milk up to the price of buyers’ next best substitute.227 Whether this is likely to be a price that would otherwise induce entry (or, in the case of the case of those IPs that already supply at the factory gate, expansion) into the factory gate market is likely to vary by region. 5.59 In Gisborne, Hawke’s Bay, and Nelson regions, there is unlikely to be an IP alternative to Fonterra. In Northland and Otago, it is less clear whether there is likely to be an IP alternative to Fonterra. In the remaining regions, [ ] are most likely to enter and [ ] are most likely to expand into the factory gate market if that price is sufficiently high. 5.60 Without the DIRA Regulation, we believe some factory gate customers would be unlikely to have an alternative supplier to Fonterra. In such situations, the constraint on Fonterra’s pricing would not be the possibility of entry or expansion, but rather the price that would compel those customers to switch to other inputs (such as direct farmer supply) or otherwise exit the market. Regions where IPs are more likely to enter the factory gate market 5.61 IPs that source milk directly from farmers are the most likely processors to enter or expand into the factory gate market (at the right price). 5.62 Most of these ‘own-source’ IPs are established in areas of high milk growth. Synlait, Oceania, and Westland are located in Canterbury. Danone Nutricia and Open Country are based in Southland. Miraka, Open Country,228 Green Valley, and Tatua are based in Waikato. Only Fresha Valley is based in an area of low milk growth— Northland. Costs and Benefits of the DIRA provisions" 17 August 2015, p.9. This, however, does not mesh with our view that the current market price for non-DIRA milk reflects market power and that this has not resulted in IPs competing the price down to competitive levels despite the fact that volumes of non-DIRA factory gate sales are small. Moreover, IPs have generally indicated that they are unwilling to supply small volumes and their willingness to supply at all will depend on price. 227 If Fonterra’s profit-maximising price increase, given a buyer’s elasticity of demand, is less than the price of the buyer’s next best substitute, Fonterra will only price up to that price. 228 Open Country also has a site in Taranaki. 2440192 95 5.63 While these IPs have indicated that they would not supply at the current DIRA milk price, some would consider supplying at a higher price.229 Regions where Fonterra is less likely to face a competitor at the factory gate 5.64 Factory gate milk is delivered by the IP directly from the farm to the factory gate customer. Fonterra admits there are no own-source IPs in approximately 30% of its catchment areas.230 5.65 Our analysis suggests that the catchment areas where Fonterra is the only ownsupply processor and where at least one IP purchases DIRA milk are Gisborne, Hawke’s Bay, and Nelson. Our information indicates three IPs collect factory gate milk from Fonterra across these areas and so would have no competitive processor alternative to Fonterra for factory gate milk. Gisborne 5.66 Waimata Cheese, in the Gisborne area, collects approximately 70%231 of its milk from its own farmers and therefore could potentially increase its own collection. Otherwise, the next closest own-supply IP to the Gisborne catchment area is Miraka in Waikato. 5.67 [ ]232 Hawke’s Bay 5.68 ]233[ [ ] Nelson 5.69 Talleys, located near Nelson, has no options for raw milk other than Fonterra. [ ]234 229 [ ] 230 Fonterra “Submission on review of the state of competition in the New Zealand dairy industry” 17 August 2015, para 9. 231 Waimata Cheese “source approximately 30% of their milk requirements from Fonterra”. Waimata Cheese “Review of the State of Competition in the NZ Dairy Market” July 2015, p.1. 232 [ ] 233 [ ] 234 [ 2440192 ] 96 5.70 [ ]235 Regions where Fonterra may face a competitive constraint at the factory gate 5.71 We identified Northland (with one IP) and Otago (with two IPs) as regions where it is uncertain whether there would likely be entry into the supply of factory gate milk. Northland 5.72 Fresha Valley in Northland is a small operation, collecting just [ ] million litres a year of milk from its own farmers. [ ]236 5.73 Goodman Fielder’s Puhoi plant is also located in Northland. Puhoi requires over [ ] million litres of milk a year. This volume would account for over [ ] of Fresha Valley’s milk. It is therefore unlikely that Fresha Valley would be able to provide an alternative to Fonterra for all of Goodman Fielder’s milk at the factory gate. Goodman Fielder has indicated that it is not interested in developing its own supply237 (Goodman Fielder’s options are discussed further at paragraphs 5.89– 5.90). Otago 5.74 Cadbury238 and Evansdale Cheese are both located in Otago. Two own-supply IPs, Danone Nutricia and Oceania, are close to the Cadbury and Evansdale Cheese plants, and therefore one or both of them could possibly supply these customers. [ ] 5.75 However, Fonterra has been the only source of milk for Evansdale Cheese since 1997. This might be because Evansdale Cheese sources small volumes of milk (just 1,500 litres per day).239 IPs may not be interested in supplying small volumes. 235 [ 236 [ 237 Goodman Fielder “Commerce Commission Consultation Paper: Review of the state of competition in the New Zealand dairy industry” 10 July 2015, para 3.1-3.2. 238 We refer to Mondelez New Zealand as Cadbury throughout our report. 239 Evansdale Cheese “About Us” (Viewed on 17 February 2016). 2440192 ] ] 97 5.76 Fonterra’s costs of collecting raw milk and delivering it an IP’s factory gate may be lower than that for other IPs. In a given region, this may be because Fonterra has a larger catchment area around its plant resulting in farmer suppliers being relatively closer to the factory gate of a customer IP. Fonterra also potential has greater economies in scale and scope in its collection costs than other IPs.240 5.77 Danone Nutricia and Oceania’s transportation costs may be relatively higher than Fonterra’s because IPs typically have smaller catchment areas than those of Fonterra. So, while their plants may be close to these customers, they might not have a farmer supplier as close to these factory gate customers as Fonterra has. 5.78 Cadbury, meanwhile, seeks much larger quantities (over [ ] million litres in the 2014/15 season) and therefore it may be more attractive for factory gate suppliers in the absence of regulations. Other IPs have discussed supplying Cadbury, but Cadbury can get milk from Fonterra so there is no driver for Cadbury to look elsewhere for milk.241 Price increase by type of buyer 5.79 Prices may also vary by type of buyer. In particular, it will likely vary by a buyer’s ability to access or increase their own supply of milk cost-effectively. That ability will in turn depend in part on whether, without open entry and exit provisions, Fonterra would be able to limit such access. This is more fully considered in the discussion below in relation to foreclosure of IPs at paragraph 5.97. For purposes of this section, we assume that access to farmers would remain as under the regulations. 5.80 Factory gate prices by buyers will likely differ across: 5.80.1 large IPs that are subject to the sunset provisions; 5.80.2 IPs that source milk from farmers but are not subject to the sunset provisions; and 5.80.3 IPs for whom own-supply is not an option. 240 For example, in the Otago region Danone Nutricia and Oceania’s transportation costs may be relatively higher than Fonterra’s because IPs typically have smaller catchment areas than those of Fonterra. So, while their plants may be close to these customers, they might not have a farmer supplier as close to these factory gate customers as Fonterra has. 241 [ 2440192 ] 98 5.81 Since Fonterra might be able to set different prices for different factory gate customers, we consider each group in turn. We also comment on Goodman Fielder’s claims that it would have no alternative to Fonterra.242 5.82 The price increase buyers would be likely to face will also likely differ based on their price sensitivity. Buyers who face vigorously competitive downstream markets or are otherwise low-margin businesses are less well placed to sustain a factory gate milk price increase. Large IPs subject to the sunset provisions 5.83 IPs that will be subject to the sunset provisions, and so will no longer have access to DIRA milk by mid-2016, are already seeking to increase their own-supply of milk directly from farmers. Therefore, these IPs are unlikely to rely on factory gate milk except while they transition to exclusively supplying themselves or if they happen to fall short in that supply. Own-supply IPs not subject to the sunset provisions 5.84 Own-supply IPs not facing the sunset provisions are largely located in regions with higher milk growth.243 Similarly, entrants that would seek their own supply from farmers are likely to establish plants in these regions. These IPs are therefore likely to be positioned to increase their own-supply, without regulations. 5.85 [ ]244 Also, He Run, which had planned to enter into Waikato, did not plan on taking DIRA milk at its planned entry in 2016. This was because enough farmers had indicated that they would switch from Fonterra.245 IPs for whom own-supply is not an option 5.86 For many small IPs, own-supply is not an option owing to the volumes and consistent uptake that farmers require. Sourcing milk from farmers under the 20% rule is also impractical for many IPs, given the requirement to maintain separate vats. 5.87 For these IPs, not only are competitive alternatives reduced as a result of not being able to take milk directly from farmers, there is also some uncertainty as to 242 The alternative of using powered milk instead of sourcing raw milk is not considered. [ ] 243 The main exception is [ 244 [ 245 Farmers Weekly “New Dairy Plant Goes Organic” (2015) http://farmersweekly.co.nz/article/new-dairyplant-goes-organic?p=7 viewed 16 February 2016. He Run has subsequently decided not to go ahead with its planned entry due to disruptions in the Chinese capital markets. 2440192 ] ] 99 whether they would be supplied even if a factory gate market were to emerge, given their small volume purchases (at least in certain regions). 5.88 IPs have told us they are not interested in supplying small volumes. However, Fonterra currently supplies small volumes to small IPs at the market rate. It is not clear why, at the right price, IPs that enter or expand in the factory gate market would not be willing to do the same. That said, IPs for whom own-supply is not an option are more likely to have only Fonterra as their factory gate supply option. Goodman Fielder 5.89 Goodman Fielder, based in Auckland, Canterbury, and Manawatu, told us that no processor other than Fonterra can guarantee year round supply of milk on the scale it requires, and nor is any processor likely to develop this scale within a reasonable timeframe.246 5.90 However, it is not clear why an own-supply IP would not supply factory gate milk, if that price was high enough. This includes supplying Goodman Fielder. It is possible there would be a transition period before all Goodman Fielder’s supply requirements could be met by such suppliers. If Goodman Fielder purchased even a portion of its milk volumes from alternative suppliers, it might constrain Fonterra’s pricing to some degree.247 We note that Goodman Fielder is [ ].248 Price sensitivity by classes of buyer 5.91 The price increase faced by buyers is also likely to differ based on their price sensitivity. 5.92 Many dairy exporters are, in large part, price takers. This limits their ability to passthrough any cost increase since such a cost increase would not be industry-wide (given that the export market is international). The competitiveness of commodity export market also suggests that, after having paid a factory gate price that reflects opportunity cost, exporters are likely to have low margins. The consequent sensitivity to factory gate price increases (meaning that a small price increase is 246 Goodman Fielder purchases some raw milk from Westland on an ad-hoc basis when Westland has excess supply. This amounts to approximately 8% of the total volume of raw milk Goodman Fielder procures. Goodman Fielder "Review of the state of competition in the New Zealand dairy industry: Consultation paper – process and approach" 10 July 2015, paragraph 2.4. 247 Although there may be efficiencies associated with having a single supply contract for a plant it is clear that Goodman Fielder has sourced milk from other IPs in the past [ ]. 248 [ 2440192 ] 100 likely to result in a large decrease in demand) would limit Fonterra’s profit maximising price to such buyers. 5.93 The IPs Auscow Dairies, Envictus Dairies, and the Dairy Goat Co-operative do not own-supply and are focused mainly on export markets. 5.94 Small IPs that do not collect their own milk are in a similar situation. They are likely to be quite price sensitive because they tend to be relatively high cost and they face competitive downstream markets. In particular, since these IPs sell various downstream dairy products (eg, ice cream or cheese), the raw milk price increase they would be able to profitably sustain will depend in part on whether their competitors also face a similar cost increase. If they do not, an IP is less able to fully pass-through a cost increase and so less able to sustain such an increase in the first place. This will in turn reduce its willingness to pay and so the price that Fonterra can profitably charge it. In this way, Fonterra’s price increase to an IP with no alternatives may nonetheless be disciplined by the factory gate milk price increase to an IP with alternatives. 5.95 If Fonterra can effectively price discriminate across customer types, the factory gate price increase that different buyers would face in the absence of regulations may differ based on their differing ability to sustain a price increase. Should Fonterra have milk after having sold milk to those buyers with the highest willingness to pay, it would be profitable for Fonterra to sell that milk to the remaining buyers at a lower price, as long as that price was at least as great as its opportunity cost. 5.96 The implications of this type of pricing is that those buyers with few or no alternatives to Fonterra and with little ability to sustain a price increase, would not necessarily face a market foreclosing price increase but rather a price increase that they could just afford. Fonterra’s incentive and ability to render IPs less competitively effective 5.97 This section presents our assessment of whether Fonterra would be likely to have the incentive and ability to render rival IPs less competitively effective in order to ultimately further increase Fonterra’s profits by dampening competition at the farm gate or by reducing competition in downstream dairy markets. 5.98 We considered the following strategies that could be available to Fonterra: 5.98.1 2440192 restricting IPs’ access to farmers at the farm gate by: 5.98.1.1 increasing the farm gate price; and 5.98.1.2 locking farmers into long-term contracts, or otherwise limiting or preventing exit or re-entry; 101 5.98.2 restricting access to milk at the factory gate, where this includes further increasing its price. 5.99 It is of note that this conduct would not likely be profit maximising for Fonterra in the short term to act in a way to render IPs less competitively effective. This is because it would require Fonterra, in the case of farmers, to increase the farm gate price to levels that are likely higher than that which can be justified based on downstream prices or provide concessions to farmers for taking on longer term contracts. In the case of the factory gate, the price would be above profit maximising levels and would thus forego profitable sales. Fonterra would only engage in such conduct if in the long-term it allowed Fonterra to retain milk supply it was otherwise at risk of losing, or it allowed Fonterra to gain additional, compensatory market power in downstream markets. 5.100 Whether Fonterra has the incentive and ability to render its rivals less competitively effective is considered in the context of the DIRA Regulation no longer being in place. This means we assume Fonterra would have already raised its factory gate prices to the point where it covers its opportunity cost plus a margin to reflect its market power. It also means we assume the open entry and exit rules are no longer in place. At the farm gate, Fonterra’s current ability to foreclose IPs through farm gate price increases would not be impacted significantly by the removal of DIRA. As such, the constraints that now limit foreclosure through farm gate price increases would largely continue to operate without regulation. The impact of the removal of the open entry and exit provisions would likely be limited in regard to wellestablished IPs. There is a risk that less well-established IPs seeking to expand through increased farmer supply or new entrants to the farm gate market would be adversely impacted. 5.101 At the factory gate, Fonterra would not likely have a significant incentive to foreclose exporters, and it would not be able to foreclose those domestic downstream buyers that would have alternatives sources of raw milk (either by sourcing milk directly from farmers or from IPs that enter the factory gate market as a result of higher prices). Some IPs would not have such alternatives but there is likely an insufficient number for Fonterra to gain much benefit from their foreclosure. Does Fonterra have the incentive and ability to render IPs less competitive effective in order to soften competition at the farm gate? 5.102 We assessed whether, in the absence of the DIRA Regulation, Fonterra would have an incentive to render IPs that source milk directly from farmers less competitively effective, in order to reduce competition for access to farmer suppliers. Having found the possibility of such an incentive, we further considered whether Fonterra would have the ability to act in this way. 2440192 102 Incentive to foreclose ‘own-source’ IPs from the farm gate market 5.103 IPs provide competition to Fonterra for farmers. Fonterra principally competes for farmer supply with the IPs set out in table 4.3 in chapter 4, State of competition under the DIRA Regulation. The biggest of these are Open Country and Synlait. [ ]249 5.104 Fonterra is somewhat constrained by its national pricing policy in the vigour with which it can compete with ‘own-source’ IPs to maintain and grow its farmer suppliers. However, Fonterra has reacted to farm gate competition by improving the range and quality of services if offers and easing farmers’ conditions of sharing up. 5.105 Given the sizes of these various ‘own-source’ IPs in certain geographic areas, Fonterra may have an incentive to foreclose processors that compete at the farm gate, to soften the competition for farmer suppliers and so ultimately be able pay farmer suppliers less (either by way of lower farm gate prices or a lower dividend). This incentive is, however, tempered by Fonterra’s co-operative structure. This structure seeks to assure that farmers obtain a return that reflects the market value of the milk they produce. 5.106 If competing IPs are willing to pay more, or provide other favourable concessions, this presumably reflects the value of that milk and so Fonterra too should be willing to pay more250 to reflect that underlying value, rather than foreclose competitors and so potentially deny farmers the benefit of that value. 5.107 While Fonterra would not likely wish to see its farmers rendered worse off, it does likely have an incentive to maintain and grow its co-operative base, and so, in the absence of the open entry and exit provisions, seek to lock farmers into long-term contracts or otherwise restrict their entry and exit. 5.108 Unlike in the case of a price that does not fully reflect the value of their product, farmers may be less inclined to view such contracts as being outside their best interests. [ ]251 249 Refer to Attachment D, Table D1: Farm gate market shares (processing capacity and milk collection), by region for the 2014/15 season 250 The price paid should not be considered in isolation but rather in aggregate with the dividend and any services and concessions farmers may receive. 251 [ 2440192 ] 103 5.109 Given that the co-operative structure might not fully address an incentive to foreclose at the farm gate, we go on to consider Fonterra’s ability to engage in such foreclosure. Fonterra’s ability to foreclose ‘own-source’ IPs from the farm gate market 5.110 We considered two mechanisms through which Fonterra might be able to foreclose its upstream rivals, in the absence of the regulations: 5.110.1 without the Milk Price Monitoring regime, Fonterra might be able to raise the price of farm gate milk so that IPs would no longer be able to afford farm gate milk;252 and 5.110.2 without the open entry and exit provisions, Fonterra might seek longer term contracts with farmers and so make them unavailable to rival processors. The effect of the current regulations is that Fonterra cannot lock-in its farmer for one more than a year since shareholding farmers are free to exit and re-enter Fonterra at will.253 Fonterra would have little ability to render IPs less competitive effective by raising the farm gate price 5.111 We consider that there would not be significant changes to Fonterra’s milk price setting if the DIRA Regulation is removed and that Fonterra would retain most aspects of the current milk price setting process, including its Milk Price Manual and milk price governance arrangements. 5.112 Fonterra’s ability to foreclose rivals by increasing the farm gate milk price is largely the same with and without regulation. That is, even with the DIRA Regulation, Fonterra could pay a higher price than that which results under the Milk Price Manual. 252 Rather than seeking to raise farm gate prices generally, Fonterra may seek to raise prices to only those farmers who would otherwise exit Fonterra to sell their milk to an IP. Such farmers would essentially become contract suppliers. Fonterra could potentially pay such farmers a higher price than the price it pays its shareholder farmers so as to deny these farmers to its rival processors. Any potential for such targeted pricing is not affected by the regulations. The limit on the number of contract suppliers Fonterra may retain is, rather, contained in Fonterra’s constitution. We do not, thus, consider the possibility of such targeted pricing. We note, however, our understanding that Fonterra generally pays its contract suppliers a lower price than its farmer shareholders. 253 Subject to the 33% rule, whereby 33% or more of milk solids produced within a 160km radius of any point in New Zealand must be available to be supplied under contract with IPs, Fonterra can lock-in contract suppliers for longer term contracts. It is possible that Fonterra’s interest free loan scheme may be raising the costs of a farmer switching to an IP. Such a switching cost is present both with and without the regulations. 2440192 104 5.113 However, Fonterra is constrained from raising the farm gate milk price to above competitive levels because of the importance to Fonterra of its TAF scheme and a need to protect the divergent interests of its supplier shareholders,254 outside investors, and contract suppliers. 5.114 A number of parties seek information and transparency over the farm gate milk price (including farmers, TAF investors, and analysts). In the absence of DIRA Regulation, we consider that Fonterra would have a continued incentive to provide milk price information to such parties and to ensure there is effective monitoring of that price. Without this, the attractiveness of investing in TAF units would be likely to be reduced. 5.115 We also consider that Fonterra would be financially constrained in its ability to fund a milk price that is at a level that would render IPs less competitively effective whilst requiring funding for its capital programme255 and maintaining a sustainable business.256 5.116 We consider that the most significant change to the milk price regime without regulation would be the removal of the Commission’s milk price monitoring. Without the milk price monitoring regime, Fonterra would technically have greater than its current level of discretion to raise the price of farm gate milk as there would no longer be a review that has a focus on the practical feasibility of the milk price calculation (the contestability dimension of the subpart 5A purpose). 5.117 However, we do not consider that Fonterra would, without the regime, make significant changes to how the milk price is set.257 5.118 Our annual review of Fonterra’s Milk Price Manual and milk price calculation provides the benefit of greater transparency and has resulted in Fonterra disclosing more information on how its model for calculating the milk price works. 5.119 We consider that the level of transparency created by the milk price monitoring regime would create a standard of transparency which Fonterra would find it hard to remove in the absence of regulation. 254 This includes supplier shareholders who own ‘dry’ shares above their ‘wet’ production shares. 255 The main source of Fonterra’s capital is from regular retentions of distributable profits. 256 Miraka considers that the milk price is Fonterra’s primary measure of performance and a ‘profit imperative’ is not sufficiently strong or important to prevent Fonterra inflating the milk price. Miraka “Submission to the Commerce Commission: Review of the state of competition in the New Zealand Dairy industry, Draft report” 4 December 2015. We consider that Fonterra is likely limited in its incentive and ability to increase the milk price as outlined in paras 5.105 and 5.113 – 5.115. 257 Fonterra’s milk price setting through its Milk Price Manual and governance of the milk price is currently codified in its constitution. 2440192 105 5.120 Fonterra has also adjusted aspects of its approach in response to our findings. However, the overall changes to the milk price as a result of our monitoring have been relatively small. As a result, we do not think removing the regulation would result in a significant increase in the farm gate milk price. Fonterra would have limited ability to render IPs less competitively effective through the use of longer term contracts 5.121 If the open entry and exit provisions are no longer critical to own-source IP entry and growth, Fonterra increasing the length of its contracts with farmers would not have an adverse effect on these processors since they would nonetheless be able to cost-effectively attract and maintain farmers. Whether this is likely to happen will depend on the type of IP affected: whether the IP is a new entrant or, if already in the market, how well-established it is in regard to its reputation and financial stability. 5.122 The majority of IPs raised concerns that Fonterra would lock-in farmers to impede entry in the first place and to preclude growing processors from acquiring new sources of milk supply.258 5.123 They are concerned that, in face of the choice of longer term contracts with Fonterra or staying with an IP (even a well-established one), farmers, who tend to be conservative in their business decisions, would switch to Fonterra. That is, conservative famers see Fonterra as a ‘sure thing’ given its size and co-operative structure. Moreover, should sufficient switching away from IPs occur, this could weaken the processors as ongoing businesses and in turn further drive farmers back to Fonterra. We have been cautioned that this negative feedback loop could result in a ‘mass exodus’ of farmers from IPs to Fonterra.259 5.124 We consider Fonterra could limit the development of new or newer ‘own-source’ IPs, but a number of factors mitigate Fonterra’s ability to adversely impact established IPs: 258 Danone Nutricia “Consultation on substantive issues – review of competition in the dairy industry” 17 August 2015; Westland “Submission on the dairy competition review consultation paper 12 June 2015” 10 July 2015, para 35. [ ]; Talleys “Re: Commerce Commission review of the state of competition in the New Zealand dairy industry” 22 June 2015, para 6. 259 [ ] Miraka “Consultation Paper—process and approach: Review of the state of competition in the New Zealand dairy industry” 10 July 2015. 2440192 106 5.124.1 IPs tend to have longer contracts with farmers than Fonterra, which could stem switching.260 5.124.2 the processors typically pay a premium for milk at the farm gate and most do not require sharing up, potentially making them more appealing for certain farmers; and 5.124.3 certain own-source IPs, such as Westland261 and Tatua,262 are wellestablished and so are at lower risk of losing farmers were Fonterra to change the terms of its farmer contracts. Even after the liquidation of NZDL in 2012, farmers continued to switch away from Fonterra to Oceania, Synlait, and Westland.263 5.125 An additional consideration is whether a longer term supply contract is not anticompetitive, despite any foreclosing effect it may have, but instead is an efficient way to reduce the risk of uncertain supply. Security of supply can give a processor the confidence to invest in risky processing assets. In the absence of the open entry and exit provisions (which currently enable farmers to enter Fonterra at will), farmers may seek longer term contracts, either with Fonterra or another IP, prior to undertaking the risk of investing in new or additional capacity. This may be efficient in that it precludes investment in high cost output for which there is no demand. We discuss further in chapter 6, our assessment of the efficiency costs and benefits of the DIRA Regulation. 5.126 Multi-year contracts are standard among own-source IPs, running from [ ] Incentive and ability to render IPs less competitively effective in order to enhance market power downstream 5.127 This section examines Fonterra’s incentive and ability to foreclose IPs to enhance its market power downstream. We separately consider this for IPs that principally export and for those that sell in downstream domestic markets. 260 [ ] 261 Fonterra has not found it profitable to sign farmers from the Westland region. 262 Tatua is a closed co-operative that predates Fonterra. [ 263 ] Although this occurred at a time when farmers were assured they could switch back to Fonterra, it still indicates that not all farmers are so risk averse that they are unwilling to switch, particularly in regard to certain IPs. 2440192 107 Incentive to foreclose IPs from the export market 5.128 It is not clear that Fonterra would obtain much benefit from rendering rival exporters less competitively effective by raising the price of factory gate milk to them (or foreclosing their access to such milk altogether) as long as Fonterra was otherwise selling factory gate milk at a price that at least covered its opportunity cost. This is because the export market is generally understood to be a competitive, international market in which Fonterra, while being a large player, does not have significant market power. In this situation, Fonterra would have little or nothing to gain from foreclosure. 5.129 Fonterra’s key competitors in export markets are global players and so IP exit in New Zealand would not significantly change the competitive dynamic in those markets.264 5.130 [ 5.131 [ ]265 ]266 5.132 To the extent that Fonterra and other New Zealand exporters are seeking to develop competing branded products, the incentive to foreclose could be increased. Figure 5.1 New Zealand exporter’s WMP prices in China, 2014/15 season [ ] Source: Information request by the Commerce Commission267 5.133 [ ] (see figure 5.1), if Fonterra is recovering its opportunity cost in factory gate sales, we do not consider it to have a strong additional incentive to foreclose domestically based exporters through further increases of the factory gate price. To the extent that this incentive is stronger, we consider that the following discussion in regard to Fonterra’s ability to foreclose own-source IPs applies. This ability is limited. 264 Fonterra’s Europe and US Roadshow, June 2015. 265 [ 266 [ 267 [ 2440192 ] ] ] 108 Incentive to render IPs less competitive effective in domestic downstream markets 5.134 Fonterra would be likely to have an incentive to render IPs less competitively effective in domestic downstream markets were the DIRA Regulation not in place. 5.135 We assessed Fonterra’s incentive to foreclose IPs that participate in the domestic downstream market by further raising the factory gate market price.268 We also assessed this incentive in the context of the 20% rule. That is, whether without the regulations, Fonterra would prevent IPs that rely on this rule from accessing Fonterra’s farmers. IPs that rely on this rule participate almost entirely in the domestic market. 5.136 Fonterra would have an incentive to foreclose downstream competitors if they have or are likely to have a significant effect on the competitiveness and the dynamism of domestic dairy markets. We consider they have affected certain dairy markets and that there is the possibility for increased effect, and so, as noted above, Fonterra would likely have an incentive to foreclose in the absence of the regulations. 5.137 The domestic downstream dairy markets are almost entirely supplied by New Zealand dairy processors. Fresh dairy products are difficult to transport given their relatively short shelf lives so domestic dairy markets face few, if any, import constraints—although product is transported around the country. Hard cheeses and very small quantities of artisan cheeses are sometimes imported and tend to be premium products. 5.138 The strength of competition that Fonterra faces in domestic downstream markets from IPs varies by product market; however, Goodman Fielder is its most significant competitor in all the main markets. Private label brands also have significant market share in certain product markets but all private label products are produced by either Fonterra or Goodman Fielder, and so are not considered further in our assessment of foreclosure incentives outside our consideration of any incentive to foreclose Goodman Fielder. 5.139 The main domestic dairy processors after Goodman Fielder are Green Valley 269, Fresha Valley, and Tatua.270 268 Fonterra could also attempt to foreclose downstream IPs by reducing their access to farmer supply by way of change to the open entry and exit rules. As we addressed this previously, we do not consider it any further. 269 Some of Green Valley’s products are manufactured for third parties such as Lewis Road Creamery, which produces fresh milk, flavoured milk, cream, and butter. Lewis Road Creamery “We take animal welfare seriously” . [ 2440192 109 5.140 IPs (apart from Goodman Fielder) typically specialise in cheese, yoghurt, or ice cream. Table 5.1 shows the collective market share of IPs, other than Goodman Fielder, is particularly significant in cream cheese and dips ([ ]), followed by yoghurt and chilled desserts ([ ]). Their collective market shares range are [ ] in cheese, butter, and margarine, and [ ] in UHT flavoured milk. The only product categories where their collective market shares are less than [ ] are UHT white milk ([ ]) and fresh milk and cream ([ ]). Market share in main domestic product categories, August 2015271 Table 5.1 Fresh milk and cream UHT flavoured milk UHT white milk Fonterra [ ] [ ] [ Goodman Fielder [ ] [ ] [ ] Private 272 label [ Other [ ] ] [ [ ] ] Cream cheese and dips Yoghurt and chilled desserts [ [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ] ] [ ] ] [ ] [ ] [ ] Butter and margarine Cheese Ice cream [ ] Source: Information request by the Commerce Commission.273 5.141 We understand that this competitive dynamic has been quite stable for a while. 274 5.142 We consider the particular competitive role played by Goodman Fielder and smaller cheese, yoghurt and ice cream makers in the following sections. ] 270 [ ] 271 Note: Figures are rounded, so do not sum to 100%. 272 Where private label brands are produced by Fonterra or Goodman Fielder. Private-label products or services are typically those manufactured or provided by one company for offer under another company's brand. 273 [ 274 [ 2440192 ] ] 110 Goodman Fielder 5.143 [ ]275 5.144 [ ]276 [ ] Cheesemakers 5.145 IPs indicated they generally cannot compete against Fonterra in bulk products such as 1kg cheeses, as Fonterra and Goodman Fielder have lower cost bases.277 Instead, these IPs tend to differentiate their products by producing premium products. 278 5.146 This, however, does not mean that Fonterra is not in competition with such cheesemakers. There is some price competition279 and they exert pressure on Fonterra’s quality.280 As shown in table 5.1, smaller cheesemakers have collectively in this way gained about [ ] market share as at August 2015. Yoghurt and chilled dessert makers 5.147 As shown in table 5.1, producers other than Fonterra and Goodman Fielder collectively account for almost [ ] of the yoghurt and chilled dessert product category. Lion,281 which on its own accounted for about [ ] market share as at August 2015, is one of these IPs. There has been steady growth by the remaining players, gaining market share from [ ] in early 2012 to [ ] in August 2015. 5.148 As in the case of cheese, the growth of these players appears to have had placed pressure on both price and quality.282 275 [ 276 [ 277 Grinning Gecko Cheese “Submission on the Dairy Competition Review Consultation Paper, 20 July 2015” 14 August 2015. 278 [ ] ] ] 279 [ ] 280 [ ] 281 Lion, registered as ‘Lion – Beer, Spirits & Wine NZ’, sell Yoplait Yoghurt. 282 [ 2440192 ] 111 Ice cream makers 5.149 As shown in table 5.1, Fonterra accounts for [ ] the ice cream category, which IPs have made in-roads into. Based on Fonterra’s analysis of market shares at Foodstuffs’ South Island outlets, ‘other manufacturers’ (excluding Goodman Fielder) account for about [ ] of market share [ ].283 Conclusion on incentive to foreclose in domestic downstream markets 5.150 Fonterra remains the most significant player in domestic downstream markets, particularly for fresh milk and cream (typically supplied indirectly through private label toll manufacturing). However, smaller IPs have made significant in-roads in certain product categories. There appears to have been some consequent price pressure on Fonterra (although this may be limited since most of the smaller IPs products are premium products), as well as a quality pressure. 5.151 We consider that Fonterra would have an incentive to foreclose smaller IPs in event of no regulations. Whether this incentive extends to Goodman Fielder is less clear given its observed ability to negotiate a discount off the DIRA price for its milk supply. Ability to render IPs less competitively effective in domestic downstream markets 5.152 Given our view that Fonterra would have some incentive to foreclose downstream domestic competitors, we considered whether Fonterra would be able to foreclose. 5.153 This outcome appears likely only for the IPs that do not have competitive alternatives to Fonterra either because they are not in a good position to selfsupply or no other IP is geographically positioned to supply milk to them at the factory gate. There does not appear to be sufficient of the latter for Fonterra to consider a strategy of exclusion. Fonterra would have little to gain in the downstream market from their exclusion. 5.154 In reaching this conclusion we assessed the alternatives available to IPs that buy factory gate milk from Fonterra and IPs that rely on the 20% rule. IPs that use Fonterra factory gate milk 5.155 283 Table 5.2 indicates by geographic area whether downstream competing IPs that use Fonterra factory gate milk are likely to have a competitive alternative should Fonterra further raise factory gate milk prices above levels that only reflect its opportunity cost and market power. The identified alternatives assume that Fonterra has not otherwise successfully foreclosed own-supply IPs or precluded [ 2440192 ] 112 new access to farmer suppliers by effectively locking in farmer suppliers to Fonterra. 5.156 We identified 14 domestically focused IPs that use Fonterra factory gate milk. Of these, it appears that six may have alternatives to Fonterra, including by way of increasing their sourcing from farmers. These include Goodman Fielder’s Manawatu and Canterbury plants. In terms of the possible cost effectiveness of these alternatives, it seems likely that these processors would be willing to supply at Fonterra’s current market rate and that price may be an indication of the factory gate price that would induce entry or expansion in that market. 5.157 The remaining nine IPs (including Goodman Fielders’ Puhoi plant in Northland) that do not appear to have alternatives currently account for approximately [ ] of total DIRA milk purchases in the 2014/15 season. They principally sell specialty cheeses, gourmet yoghurts and ice cream. 2440192 113 Table 5.2 Region Northland IPs that rely on factory gate milk, 2014/15 IP Alternative factory gate supply Goodman Fielder [ Factory gate milk (litres) 284 ] Fresha Valley Auckland/ Waikato/ Bay of Plenty ] Own farmer supply Compete against Fonterra Products Yes Domestic dairy Yes Domestic dairy [ ] [ ] [ ] Yes Infant milk formula export Yes Dairy Goat Cooperative [ Emerald Foods Group [ ] [ ] Yes Domestic ice cream Epicurean Dairy [ [ ] Yes Domestic yoghurt [ ] Yes Domestic dairy [ ] Yes Domestic yoghurt Yes Domestic specialty cheese Yes Domestic dairy Yes Domestic yoghurt Yes Domestic specialty cheese Yes Domestic dairy Yes Domestic specialty cheese ] Green Valley Gopals Sweets & Snacks [ Yes ] Gisborne/ Hawke’s Bay/ Wairarapa/ Manawatu Canterbury Waimata Cheese [ Goodman Fielder [ BioFarm Products [ Barrys Bay Cheese [ Goodman Fielder [ ] [ ] ] ] ] [ Yes ] [ ] [ ] [ Yes ] ] Karikaas 284 [ 2440192 [ ] [ ] ] 114 Otago/ Southland Tasman/ Marlborough Serra Natural Foods [ ] [ Talbot Forest Cheese [ ] [ Whitestone Cheese [ Evansdale Cheese [ Talleys [ ] ] ] ] [ ] Yes Domestic yoghurt Yes Domestic specialty cheese Yes Domestic specialty cheese [ ] Yes Domestic specialty cheese [ ] Yes Domestic ice cream ] Source: Commerce Commission analysis based on information provided by Fonterra and IPs. IPs that rely on the 20% rule 5.158 We found smaller cheesemakers tend to purchase milk under the 20% rule. Without access to milk under the 20% rule, these cheesemakers would face higher barriers to entry and expansion, and so otherwise competitively effective entry might not occur. Further, some existing cheesemakers may be foreclosed as they do not have access to milk at the factory gate and are not large enough to take all of a small farm’s milk. 5.159 About half the Specialty Cheesemakers Association members use the 20% rule.285 They typically source milk from a single farm as three-quarters of them take less than 100,000 litres per season. These companies are unlikely to have another costeffective source of milk. Fonterra generally restricts its volume of delivery, per delivery, to above 3,000 litres. 5.160 Based on our view that IPs that enter the factory gate market are expected to do so on terms similar to Fonterra, we expect these IPs may also restrict deliveries to above 3,000 litres (although, whether an IP could sell at the factory gate costeffectively would likely be affected by the location of buyers relative to the factory gate seller’s collection routes). If the IPs restrict deliveries to above 3,000 litres, the IPs that rely on the 20% rule would still not likely gain access to raw milk, even if new processors entered the factory gate market. 285 Thirty-seven members of NZSCA are currently making cheese for sale. Of these, eight companies do not use cow’s milk, three use cow’s milk exclusively from their own animals, and three companies are very large (Goodman Fielder, Open Country, and Fonterra) and so do not find the 20% rule useful. Of the remaining 23 cheese makers, 16 members currently take milk from farms under the 20 % rule. Another four companies are planning to do so. 2440192 115 5.161 Cheesemakers prefer to know where their milk comes from, and so tend to source milk from farms close to them. Being able to trace the milk to that farm is important. 5.162 Further, IPs that purchase milk directly from farmers under the 20% rule are concerned that Fonterra would foreclose access to those farmers. A representative of the Specialty Cheesemakers Association, Karikaas, noted that generally and historically dairy companies and co-operatives do not like, and therefore contractually prohibit, a farmer from dual supply. The Specialty Cheesemakers Association understands that there may be a drive to do away with 20% rule, which would be a concern for these processors.286 Two cheesemakers claim that their businesses would not survive without access to this milk.287 5.163 Self-supply also appears to be an unlikely alternative to most cheesemakers as this would require significant investment in farms and they take a very small proportion of a farm’s milk. 5.164 We consider it likely that Fonterra would be able to foreclose IPs that rely on the 20% rule. However, while Fonterra may generally have an incentive to foreclose downstream competitors, the amount of competition provided by those cheesemakers that rely strictly on the 20% rule may be so small that Fonterra’s incentive would be much reduced. Conclusion on incentive and ability to foreclose in domestic downstream markets 5.165 Our analysis suggests that while Fonterra may have an incentive to foreclose downstream competitors, its ability to do so would be limited because many such IPs would likely have alternatives at the factory gate or be able to increase their own supply. 5.166 This presumes that the withdrawal of the open entry and exit provisions would have limited impact on those IPs that are well placed to supply at the factory gate, as well as limited impact on an IPs’ ability to increase their own supply. In the absence of the open entry and exit provisions, Fonterra may have the ability to adversely impact new and less well-established IPs. 5.167 Processors that rely on the 20% rule, plus some factory gate users, would not be likely to have an alternative to Fonterra. Fonterra would probably be able to foreclose these processors. However, given the small market share such processors collectively account for, Fonterra may not get a benefit from their foreclosure. 286 [ 287 Over the Moon “Submission on the Dairy Competition Review Consultation Paper, 20 July 2015” 13 July 2015. Grinning Gecko “Submission on the Dairy Competition Review Consultation Paper, 20 July 2015” 14 August 2015. 2440192 ] 116 CHAPTER 6: Our assessment of the efficiency costs and benefits of the DIRA Regulation Purpose of this chapter 6.1 This chapter outlines our assessment of the efficiency costs and benefits of the DIRA Regulation and our conclusion that competition in the dairy industry is insufficient.288 Due to this and our consideration of the asymmetric risks of deregulation we do not recommend removing the DIRA Regulation at this time. 6.2 To reach this view, we assessed whether the relevant markets would be more efficient with or without the DIRA Regulation.289 This involved assessing: 6.3 6.4 6.2.1 the efficiency costs and benefits that accrue from the DIRA Regulation; and 6.2.2 the risk to efficiency associated with removing the DIRA Regulation too soon versus removing it too late. In carrying out our assessment, as required by the terms of reference, we considered: 6.3.1 efficiency costs and benefits with the current provisions of subparts 5 and 5A in place, and the current provisions of the Raw Milk Regulations; and 6.3.2 efficiency costs and benefits if the current provisions of subparts 5 and 5A and/or the Raw Milk Regulations were completely removed. We did not consider the efficiency effect of refining the DIRA Regulation. However, in chapter 7, Pathways to deregulation, we discuss refinements and amendments to the DIRA Regulation which may improve efficiency. Key findings on the sufficiency of competition 6.5 We do not consider competition to be sufficient to ensure the efficient and contestable operation of the relevant markets in the absence of the DIRA Regulation. In particular:290 6.5.1 Subpart 5 should remain in place: It is not clear that efficiency could be improved by removing the open entry and exit provisions. We consider 288 As required by the terms of reference. 289 In our framework, we consider efficiency to be the overriding goal of the DIRA. Chapter 3, Process and framework for evaluating and reporting on the state of competition, contains more detail on this. 290 Chapter 5, State of competition without the DIRA Regulation, outlines our view of what competition would look like without DIRA. 2440192 117 these have lowered the barriers to entry and expansion to the farm gate market, without imposing significant costs on Fonterra;291 6.5.2 Subpart 5A should remain in place: Under the milk price regime, the Commission provides additional oversight and review of Fonterra’s setting of the farm gate price. While this involves some costs, the additional scrutiny provided by the Commission provides stakeholders with assurance that the price is set at the efficient level; and 6.5.3 The Raw Milk Regulations should remain in place:292 We think that there is a risk that efficiency could be reduced by removing the Raw Milk Regulations at this time.293 6.5.4 However, we consider the Raw Milk Regulations may hamper the development of competition within the factory gate market. We consider how to mitigate this in chapter 7, Pathways to deregulation. 6.5.5 Costs and benefits of the DIRA Regulation indicate balance but risks are asymmetric: Our assessment of the potential efficiency costs and benefits of the DIRA Regulation indicates that they are in a similar order of magnitude. However, there are asymmetric risks attached to the timing of when we remove DIRA. 6.5.6 Full deregulation at this stage may put at risk some of the competition gains achieved to date. Fonterra would likely have significant market power in both the farm gate and factory gate markets in the absence of DIRA Regulation.294 Immediate removal of DIRA Regulation may allow Fonterra to exercise its market power to the disadvantage of some farmers, IPs, and consumers in downstream dairy markets. 6.5.7 The immediate removal of access to DIRA milk may also result in significant disruption to downstream dairy markets, because Fonterra may be able to significantly increase its price for raw milk, and it would take time for IPs to obtain alternative supplies. This could damage long-term efficiency, which offers potentially significant future benefits. We consider that the impact of this risk outweighs any impact of the risk that the DIRA Regulation may remain in place too long. Due to this, we recommend 291 The Open entry and exit Provisions (including the 20% Rule) and the milk price regime. 292 Though, as discussed in chapter 7, Pathways to deregulation, we recommend adopting a transition pathway to reduce inefficiencies in the current regulations. 293 That is, the requirement to supply IPs factory gate milk at the DIRA Milk price. 294 Due to the resulting market power that Fonterra’s would have. See Chapter 5, State of competition without the DIRA Regulation, for more detail. 2440192 118 keeping the DIRA regulation in place at this time. However we discuss potential amendments to the regulations which could help facilitate a pathway to deregulation in chapter 7. 6.6 In reaching this view we considered points and issues raised in response to our draft report. The efficiency costs and benefits that accrue from the DIRA Regulation 6.7 We have considered the various potential efficiency costs and benefits that may accrue from DIRA Regulation. In our analysis, the efficiencies from the DIRA Regulation are considered to result from restraining Fonterra from being able to fully exercise its market power. This section draws upon the state of competition without regulation and considers the likely impacts on Fonterra, IPs and farmers to gauge the extent of the efficiencies which may accrue. 6.8 Table 6.1 summarises our key findings in relation to the efficiency costs and benefits of each form of regulation. 2440192 119 Table 6.1 Summary of the efficiency costs and benefits of the DIRA Regulation Regulation Cost Benefits Subpart 5: open entry and exit and related provisions Relatively low cost. No material impact on: Acts as a safeguard helping IPs enter and expand in the farm gate market. This improves the ability of these IPs to compete in the factory gate market. Subpart 5A: milk price regime Raw Milk Regulations  excess capacity;  crowding out of higher value investment;  asset stranding; and  inefficient conversions. Keep the regulations in place at this time. The 20% rule has helped niche food processers access raw milk and compete in domestic downstream markets. Costs to Fonterra of the Commission monitoring of the Milk Price Manual and calculation are approximately 295 $[ ] million a year. Some of these monitoring costs are likely to remain even if subpart 5A is removed. Additional benefits from the Commission monitoring role (more flexibly of scope). Moderate direct cost in the region of $6 million a year for Fonterra due to excess capacity. Prevents a factory gate price increase which would be detrimental to domestic downstream markets. May be hindering the development of the factory gate market for non-DIRA milk. Recommendation Keep the regulations in place at this time. There is widespread industry support of the milk price monitoring regime. We estimate this to be in the region of $3.5 million to over $13 million a year. Keep the regulations in place at this time but we recommend adopting a transition pathway to mitigate inefficiencies. Source: Commerce Commission analysis based on information provided by Fonterra and market participants. Efficiency costs and benefits of subpart 5, open entry and exit and related provisions 6.9 Subpart 5 relates to the open entry and exit and related provisions (including the 20% rule). These provisions help IPs and food processors to obtain their own supply of raw milk direct from farmers. 6.10 We see benefit to keeping the DIRA Regulation in place as a safeguard to ensure that IPs can continue to access farm gate milk. This helps IPs enter and expand in the farm gate market. It also improves the ability of these IPs to compete in the 295 This only includes the cost of the Commission’s monitoring—it does not include the cost to Fonterra of maintaining the milk price panel. Commerce Commission meeting with Fonterra Milk Price Panel, 7 December 2015. 2440192 120 factory gate market (including smaller and niche producers who access milk through the 20% rule) and ultimately domestic downstream markets. Own-supply IPs could potentially supply a factory gate market for non-DIRA milk should one develop. Retention of the open entry and exit provisions could assist IPs that take DIRA milk to develop their own supply if the DIRA entitlements were to be reduced. 6.11 6.12 We have considered the various potential costs associated with these benefits, including: 6.11.1 the need for Fonterra to hold additional excess capacity; 6.11.2 crowding out of Fonterra’s investment in higher value plant; 6.11.3 the increased risk of the stranding of Fonterra’s assets; and 6.11.4 inefficient dairy conversions. However, we have not found the open entry and exit provisions to have a significant or material impact on these costs. As such, we are of the view that the provisions are relatively low cost. We provide more detail on the costs considered below. Excess capacity 6.13 The open entry and exit provisions mean that Fonterra must accept milk from farmers who apply to join the co-operative.296 However, we do not consider the provisions to have materially contributed to the excess capacity Fonterra holds. 6.14 In its submission on our draft report, Fonterra submitted that the open entry and exit provisions require it to hold excess capacity to manage uncertain milk volumes. It submits that this has a similar impact [ ] as the Raw Milk 297 Regulations. However, our assessment found that there is little evidence to support this. Further to this, Fonterra’s non-price response to regional competition appears to contradict it. 6.15 We do not consider that the open entry and exit provisions’ have materially contributed to the milk volume uncertainty Fonterra manages because: 6.15.1 We consider that Fonterra’s milk growth strategy and the associated uncertainty with that trajectory of milk volumes exceeds that created by the open entry and exit provisions; and 296 To mitigate this risk, Fonterra can issue a capacity constraint notice for one year, [ 297 The impact of the Raw Milk Regulations is discussed in paragraphs 6.46-6.64. 2440192 ]. 121 6.15.2 The information Fonterra provided, shows that Fonterra explicitly takes into account the uncertainty caused by the Raw Milk Regulations when considering capacity requirements. However, uncertainty caused by the open entry and exit provisions is not explicitly accounted for and is instead included in the general forecast of uncertainty. 6.16 Fonterra’s growth strategy shows that it wants to maintain and grow the volume of milk that it collects, and to do so quickly.298 This can be evidenced by the introduction of MyMilk in response to regional competition. MyMilk assists Fonterra in maintaining and growing its milk supply in more competitive regions by allowing farmers to supply milk for five years without having to share-up. 6.17 We note that it is difficult to identify the difference between investment in plant that Fonterra makes as part of its overall growth strategy and investment made due to the open entry and exit provisions.299 However, we think that Fonterra’s growth strategy and response to regional competition demonstrate its desire to increase milk volumes. Given these factors, we think that the open entry and exit provisions do not create material additional volume uncertainty. 6.18 In attachment F, Evidence that informed our conclusions on excess capacity and resetting market share thresholds, we further discuss milk volume uncertainty and Fonterra’s investment in capacity which demonstrates Fonterra’s strategy [ ]. Investment in excess capacity has 300 costs as well as benefits. Crowding out investment in higher value plant 6.19 In its submission on our draft report, Fonterra submitted that its obligation to accept supply limits its ability to phase supply growth to match its medium term investment programme. Fonterra argues that this obligation makes it prudent to invest in low cost, low-value capacity to a greater extent than is optimal, thereby using up scarce capital. 298 Fonterra’s growth strategy includes an objective to “grow milk production volumes to protect our place in a growing market” and another to “execute our strategy at speed”. Detail on Fonterra’s strategy is available here: . 299 Fonterra’s desire to maintain and grow milk volumes can further be evidenced by the introduction of MyMilk in response to regional competition. MyMilk assists Fonterra in maintaining and growing its milk supply in more competitive regions by allowing farmers to supply milk for five years without having to share up. 300 We note that in the 2013/24 season capacity constraints resulted in a material decrease in Fonterra’s potential earnings. In particular, “Fonterra’s asset footprint impeded its ability to fully respond [to changing international prices] by switching production to milk powders”. Fonterra “Farmgate Milk Price Statement for the Season Ended on 31 May 2014” p.11, 23 September 2014. 2440192 122 6.20 However, the number of new farm conversions and increased supply is likely driven by the farm gate milk price, which is dependent on changes in global commodity prices. Regardless of the DIRA Regulation, Fonterra is likely to increase capacity to meet global demand. 6.21 On balance, we do not consider the open entry and exit provisions to be material barriers to investment in value-add capacity. There are various factors that are likely to have a larger impact on Fonterra’s decisions to invest in low value capacity over higher value capacity than the DIRA Regulation. These include: 6.21.1 Fonterra’s co-operative structure: this imposes some natural capital constraints, likely to impact investment decisions. Its structure also means that Fonterra is committed to processing all of the milk produced by its farmer shareholders. This may constrain Fonterra’s high value investment more so than the obligation to accept increased supply imposed by the DIRA Regulation does; 6.21.2 Fonterra’s strategy: IPs such as Westland and Tatua have invested in value-added product lines. However it is not clear that Fonterra has adopted the same strategy on an equivalent proportional scale. Fonterra has a strategic focus on growing volume. In the absence of the DIRA Regulation, it appears likely that Fonterra would remain geared more toward commodity capacity growth to meet growing milk volumes; and 6.21.3 regional competition: Fonterra has noted that farmers switching to it from an IP are likely to be in competitive regions and regions with milk growth. In these areas, Fonterra has responded to regional competition through non-price incentives such as MyMilk.301 This suggests that Fonterra is driven to limit switching away from it and to compete for new conversions. The impact of switching on investment decisions may therefore be limited. Asset stranding risk 6.22 Asset stranding risk exists because the open entry provisions oblige Fonterra to accept milk from new applicants that want to supply it while open exit provisions mean Fonterra cannot secure supply on restrictive terms. Fonterra must have the capacity to accept this supply, or if it does not, it must build capacity.302 However, farmer suppliers can then leave Fonterra. This creates the risk that the asset will be left underutilised. 301 MyMilk is available in regions with strong competitive pressures at the farm gate for conversions— Canterbury, Southland, and Otago. 302 Fonterra can issue a capacity constraint notice for one year, [ 2440192 ]. 123 6.23 Asset stranding risk is an industry-wide risk faced by Fonterra, IPs, and farmers. For example, IPs often use longer term contracts to secure supply and farmer suppliers also do so to secure a market for their raw milk.303 This mitigates asset stranding risk for both parties, but under the open entry and exit provisions Fonterra is restricted in its ability to agree longer term contracts with farmers. We also note that an allowance for asset stranding risk is made in the way the farm gate milk price is set.304 We are only concerned with any incremental asset stranding risk for Fonterra created by the open entry and exit provisions. 6.24 We agree in principle that the DIRA Regulation may create incremental asset stranding risk, however we have not seen evidence to suggest that this is currently a material risk and consider there to be a number of mitigating factors. 6.25 Milk growth is likely to have partially mitigated asset stranding risk in the past for both Fonterra and IPs. New Zealand has experienced milk growth of 6% compound annual growth rate (CAGR) over the last 5 years to the 2014/15 season. 6.26 Future projected milk pool growth is uncertain but it is lower than in recent history. However, we expect the total milk pool to grow at around [ ]% CAGR in the six years to the 2020/21 season, consistent with Fonterra’s own ‘medium’ growth. We therefore think milk growth may still assist in mitigating Fonterra’s future asset stranding risk. 6.27 For relatively small farms we do not think the open entry and exit provisions contribute materially to incremental asset stranding risk. This is because volumes lost due to smaller suppliers leaving Fonterra could likely be mitigated by volumes gained through suppliers joining Fonterra and growth in supply from existing farmers. 6.28 However, in the case of cornerstone suppliers the incremental asset stranding risk created by the open entry and exit provisions may be more material.305 A cornerstone supplier is a large single entity for which significant capacity investments must be made to accommodate its supply. 303 Commerce Commission “Draft Report: Review of the State of Competition in the New Zealand Dairy Industry” 6 November 2015 Para 5.98–5.101. 304 The weighted average cost of capital (WACC) used to calculate Fonterra’s farm gate milk price includes an allowance of asset stranding risk. For Fonterra’s notional producer this is calculated at approximately [ ] for the 2014/15 season. 305 In response to our draft report, NERA (on behalf of Fonterra) submitted the [ ] case study to demonstrate the materiality of asset stranding risk in the case of a cornerstone supplier. NERA “Review of Commerce Commission’s draft report into dairy competition” 4 December 2015, pp.11–12. 2440192 124 6.29 We think that cornerstone supplier asset stranding risk could be mitigated through a contract between the supplier and Fonterra.306 There may be reasons why each party may have incentives to enter such a contract: 6.29.1 Fonterra may have the incentive to enter a longer term contract as doing so mitigates the risk created by free exit that would exist if a new conversion shared up immediately; and 6.29.2 new conversions may have the incentive to take up multiple year contracts to supply, with the ability to share-up later. Doing so reduces their initial capital requirements and guarantees a buyer for their milk.307 6.30 We also note that in a previous submission NERA (on behalf of Fonterra)308 acknowledged that under the 33% rule “Fonterra has quite a lot scope to contract suppliers, for investment certainty.”309 6.31 In addition to this, Fonterra has some options (which come at a cost) for mitigating asset stranding risk, which may help reduce the materiality of any incremental risk caused by the open entry and exit provision. These include: 6.32 6.31.1 Flexibility in where it processes raw milk. In the past, Fonterra has transported milk across areas of New Zealand (at a cost) to realise option value in its product mix; and 6.31.2 [ [ ]310 ]311 In chapter 7, Pathways to deregulation, we outline a potential transition pathway option that may help to mitigate the incremental asset stranding risk. Inefficient conversions 6.33 While a possibility, we have not found any evidence to suggest that the open entry and exit provisions have incentivised inefficient conversions. Inefficient entry would 306 Such a contract is allowed in certain circumstance under the DIRA Regulations due to the 33% rule. 307 The alternative to supplying Fonterra would be supplying an IP who would require the farmer to enter a multi-year contract in any case. 308 NERA “Assessment of competition in Raw Milk Markets and Costs and Benefits of the DIRA Provisions” 17 August 2015, section 5.3.3. Free Exit, p.43. 309 Currently subject to the 15% limit on contract supply in Fonterra’s constitution. 310 Tactical pricing may be used in order to entice suppliers to switch from IPs to Fonterra. 311 [ 2440192 ] 125 occur if, despite the exception clauses provided for in the DIRA,312 Fonterra took on dairy conversions it otherwise would not. 6.34 Inefficiencies could arise from Fonterra incurring collection costs (eg, expanding capacity, transport costs) that exceed the benefit of taking the milk.313 This may occur, for example if a dairy conversion occurred in an inefficient location. If this occurred there may be further inefficiency caused by the land being put to suboptimal use. 6.35 However, the evidence we have on collection costs suggests that these are not material to the overall price of raw milk. Further to this, [ ]314 this could be outweighed by other economies of scale generated by the greater volume of raw milk collection.315, 316 6.36 Fonterra forecasts [ ]317 This lessens concerns that the DIRA Regulation may result in land being put to sub-optimal use. 6.37 Inefficient conversions could also occur if the price farmers receive for raw milk is too high. This price is determined as part of the milk price regime, however it does not appear likely that deregulation would lead to a materially lower price (see chapter 5, State of competition without the DIRA Regulation). Consequently, we do not consider such inefficiency if any due to the DIRA Regulation. 6.38 Further to this, Fonterra has informed us [ ]318 312 Fonterra may reject an application from a new entrant if it does not meet minimal supply requirements or on the grounds of transportation costs. Dairy Industry Restructuring Act 2001, ss 94–96. 313 These conversions are likely neutral in terms of making the farm gate market more efficient and contestable. 314 [ ] 315 Synlait “Submission on the Commerce Commission’s 2014/15 Base Milk Price Calculation Review Draft Report” 1 September 2015, paragraph 28(iii). 316 For example, Fonterra will enjoy economies of scale generated by it larger size. This includes fixed costs being spread across larger milk volumes—in 2013/14 the notional producer’s fixed administrative costs of $251 million equated to 16 cents per kgMS. However, a 25% reduction in volume would have increased this cost to 21cents per kgMS. 317 [ 318 [ 2440192 ] ] 126 Efficiency costs and benefits of subpart 5A, milk price regime 6.39 Subpart 5A relates to the milk price regime. These provisions have had some influence on how Fonterra sets its base milk price and also improve the transparency of the milk price setting process. This is beneficial to farmers, IPs and investors. We do not think that removal of the regime would necessarily result in significant changes to how or the level at which the milk price is set. 6.40 The cost to Fonterra of the Commission reviewing the Milk Price Manual and calculation is approximately $[ ] million a year. We think that if subpart 5A was removed, Fonterra would need to retain an independent review process to assure supplier shareholders and non-supplier investors that the milk pay-out is appropriately divided between dividend and a supplier payment. 6.41 If Fonterra seeks an independent audit of its process similar to that undertaken by the Commission, it would incur costs similar to those incurred with subpart 5A in place. Alternatively, it may rely on the processes it already has in place for auditing its milk price setting. In the latter case, there would be a cost saving, but there would also be a loss of the benefits provided by the Commission’s monitoring. 6.42 Provided Fonterra seeks an independent audit of its process, it would incur costs which may be similar to those it incurs with the DIRA Regulation. We therefore do not consider the costs associated with the milk price regime monitoring in our balancing exercise. 6.43 Overall we consider that any cost savings are likely to be small. We therefore do not consider the costs associated with the milk price regime monitoring in our balancing exercise. 6.44 We see some additional benefits from the Commission’s role in monitoring the milk price regime. Our independence and scope of review likely provides greater assurance to stakeholders than Fonterra’s Milk Price Panel or another independent party such as an external auditor. Our independent review on behalf of stakeholders complements the governance oversight of the Milk Price Panel and the financial assurance provided by Fonterra's external auditor.319 6.45 We also note that there is widespread support of the milk price monitoring regime. This indicates that there is value in the milk price regime, including the Commission’s role in the milk price monitoring. 319 such as calling for improved transparency of information from Fonterra 2440192 127 Efficiency costs and benefits of Raw Milk Regulations 6.46 In our view, the 2012 amendments to the Raw Milk Regulations successfully mitigated material inefficiencies created by the Raw Milk Regulations.320 Nonetheless, the current regulations either require or restrain Fonterra’s behaviour in ways that may result in some inefficiency. 6.47 Our review found that the main role the DIRA Regulation plays, in particular access to DIRA milk through the Raw Milk Regulations, is to prevent Fonterra exercising its market power to increase prices in the factory gate market price in the absence of effective competition. This is beneficial as it: 6.48 6.49 320 6.47.1 has aided the development of IPs that may compete in the factory gate market. The benefits of this should feed through to domestic downstream markets; and 6.47.2 has aided the entry and expansion of IPs that directly serve the downstream domestic markets by providing access to raw milk at the DIRA milk price. We also considered the various potential efficiency costs that arise from the requirement on Fonterra to supply raw milk, including: 6.48.1 excess capacity: Fonterra must hold excess capacity to manage the milk volume risk created by DIRA milk. We found this resulted in a direct cost to Fonterra in the region of $6 million per annum; 6.48.2 recovery of opportunity cost: There may be times that the DIRA milk price under or over recovers Fonterra’s opportunity costs. We have no reason to believe that this has caused any material cost to efficiency; and 6.48.3 hampering the development of the factory gate market for non-DIRA milk: Access to DIRA milk may be hampering the development of the nonDIRA milk factory gate market. This includes preventing IPs from competing to supply Goodman Fielder. We provide more detail on our assessment of the efficiency benefits and costs of the Raw Milk Regulations below. For example, the amendments to the Raw Milk Regulations address specific efficiency concerns with Fonterra investing in excess capacity. Broadly the ‘October rule’ addressed concerns on taking DIRA milk along the milk curve, the sunset clauses mitigated dependence on the DIRA Regulation by large ownsource IPs. 2440192 128 Restraining a factory gate price increase 6.50 To estimate the benefits of DIRA milk, we estimated the potential impact of Fonterra using its position of market power in the factory gate market to increase price were it to be deregulated now. 321 This involved attempting to isolate the impact of the volumes of milk that might be priced out of the market.322 We assumed that the efficiency cost caused by a factory gate price increase (ie, the cost of the lost volume) represented the efficiency benefits of the Raw Milk Regulations. 6.51 Our estimate gives us an indication of the potential order of magnitude of the impact of the Raw Milk Regulations. It is very sensitive to the underlying assumptions. Therefore we have been cautious in interpreting our estimate and have used it to inform rather than determine the outcome of our assessment. 6.52 We gained a view of the order of magnitude using the following assumptions: 6.52.1 We estimated the size of this potential market using factory gate volumes in 2014/15.323 We then added in volumes of milk Fonterra supplied for the domestic market within Fonterra. These Fonterra volumes are incorporated because a market-wide cost increase is likely to increase all prices in those downstream markets including Fonterra prices.324 6.52.2 We used the [ ] as an estimate of the likely price increase. This figure can fluctuate significantly; therefore, we used a 25% mark-up as an approximation.325 321 In economics terms, this is the deadweight loss. In estimating it, we assumed a horizontal supply curve. This means that the estimated deadweight loss is likely to be underestimated. 322 That is, the deadweight loss. 323 We have the volumes of milk supplied by Fonterra from which we excluded the larger IPs: Synlait, Miraka and Oceania who primarily serve international markets. [ ] 324 We estimated the volumes internal to Fonterra from data supplied by Fonterra on milk delivered to their Fonterra Brand plants and Kapiti Fine Foods. This is only an approximation given some of these sales may be for export. We do not believe this will be material and note the overall shares of milk volumes for Fonterra broadly approximates their downstream market share. [ ] 325 The estimates are affected not only by the estimated mark-up but also by the base DIRA price. In 2013/14 this was $8.40 whilst in 2014/15 it was $4.40. We have examined the difference between the prices achieved under the DIRA Raw Milk Regulations and those outside DIRA Regulation, these range between [ ] with the majority being in the region of [ ]. We recognise there are other factors that can affect this price difference, such as type of milk being supplied and the terms under which they are supplied. [ ] 2440192 129 6.52.3 To isolate the efficiency impact, we took a range of potential changes in volume that might result from the increase in the factory gate milk price.326 6.53 Based on a 25% price increase, we found that the efficiency benefits accruing from DIRA preventing a factory gate price increase (or put another way, the efficiency cost of removing DIRA) can range from $3.5 million to over $13 million a year. 327 6.54 However, this estimate is very sensitive to its underlying assumptions (including the DIRA milk price, which can vary significantly). As such, it is a very rough guide to the efficiency benefits of DIRA. Other factors that could significantly affect this estimate include: 6.54.1 Opportunity costs: The analysis assumes that the DIRA milk price includes all opportunity costs of the on-sale of raw milk. As discussed in chapter 5, State of competition without the DIRA Regulation, we consider on average that the DIRA price reflect opportunity cost but there will be times where it may not. 6.54.2 Uncertainty around the magnitude of the price increase: 328 Our estimates are based on a price mark-up of 25%. [ ] It is possible that this may under or overestimate the actual mark-up that would occur. 6.54.2.1 Underestimate: If other large IPs did not prove to be an effective constraint, including if they have higher opportunity costs than Fonterra, then this may lead to higher prices than currently observed in the non-regulated factory gate. 326 Technically this is the price elasticity of demand. We have used a range of between -0.5 to -1.0. As this is a derived demand from retail markets for dairy products, we do not expect this to be highly elastic at this level. However, again the results are highly sensitive to these assumptions. We note that while a lower elasticity implies a lower volume impact, it is also implies a higher price rise from the exercise of market power. For simplicity we have assumed perfectly elastic supply, this likely underestimates the deadweight loss estimates. 327 We have also estimated the associated consumer welfare lost through a factory gate price increase (which can be interpreted as the consumer welfare gained due to the DIRA Regulation). We estimate this to be in the range of $51.9 million to over $92.4 million. It is subject to the same sensitive assumptions as the estimated deadweight loss. The consumer welfare estimates did not form part of our balancing of efficiency costs and benefits because DIRA is concerned with overall efficiency rather than how the benefits of it are distributed. 328 The price mark-up has a non-linear impact on the estimated deadweight loss. If we doubled the price mark-up, the estimated deadweight loss increases to the range of $14 to $50 million. 2440192 130 6.54.2.2 6.55 6.54.3 A further source of uncertainty as to the size of the price increase if DIRA was removed is the extent to which Fonterra is able to price discriminate across regions. If Fonterra was able to price discriminate regionally or between IPs this would reduce the impact on efficiency. 6.54.4 The assumed elasticity of demand: We have assumed that the demand is relatively inelastic. This means that we do not think that demand would change much relative to the change in price.329 6.54.5 However, our assumed elasticity of demand is derived from downstream dairy markets. Actual elasticity of demand for raw milk may vary from our assumed range. The more elastic the demand, the higher the efficiency cost at any assumed price rise. 6.54.6 Static versus dynamic efficiencies: This estimate is based on ‘static’ efficiency benefits of the Raw Milk Regulations. This means it does not capture the full efficiencies and benefits that competitive rivalry can bring over time. 6.54.7 Typically, more weight is attached to dynamic efficiencies as they bring important benefits that often outweigh static measures of efficiency—for example, more efficient investment. 6.54.8 The dynamic efficiency benefits of the Raw Milk Regulations could be significant. In chapter 5, State of competition without the DIRA Regulation, we noted that a price increase could particularly impact on smaller producers who cumulatively represent a material competitive constraint on Fonterra. Such benefits, while very difficult to estimate, can be significant over the long run. Ultimately a factory gate price increase is likely to be passed through in downstream markets. We can therefore cross check the materiality of our estimates by considering the impact on these markets. Two conflicting factors are important when doing so: 6.55.1 329 Overestimate: If other IPs prove to be a more effective constraint than assumed, then this may lead to lower prices than currently observed through increased competition at the non-regulated factory gate. The cost of raw milk only makes up a proportion of the costs of the downstream products. Therefore, any price increase is diluted. We This gives greater scope to exercise market power through increased price. 2440192 131 previously estimated that raw milk contributes to a quarter of the retail price of a 2 litre bottle of processed milk.330 This means that a 25% increase in the price of raw milk at the factory gate is likely translate into a smaller increase (up to 6.25%) on the retail price. 331 6.55.2 6.56 Demand in downstream markets is likely to be more sensitive to price increases than demand in the intermediate markets (ie, the factory gate market). This means that the volume impact of a price increase, and hence the economic efficiency impacts, are likely to be larger.332 We limited our analysis to the impact on the domestic markets. We did not consider the impact on international markets because while we think there may be some limited incentive for Fonterra to foreclose domestic IPs that supply the export market we think that the ability to do so is likely to be similar with or without DIRA. Excess capacity 6.57 The Raw Milk Regulations require Fonterra to supply DIRA milk to IPs under specified terms and at a specified price. The terms in these regulations mean that Fonterra bears the risk of IPs not taking DIRA milk according to the allowed tolerances on IPs’ forecasts.333 This means that Fonterra may have to invest in capacity in excess of what it would otherwise have. It may also have to bear additional transaction costs to manage the milk volume uncertainty.334 6.58 We estimated that this volume uncertainty has a direct cost in the region of $6 million a year on Fonterra in terms of the excess capacity Fonterra must hold to manage it. However, we note that this direct cost to Fonterra is not the same as the cost to economic efficiency. 330 Commerce Commission “Consideration of whether to initiate a Commerce Act Part 4 inquiry into milk prices” August 2011, paragraph 43. 331 This will depend on what proportion of the cost increased is passed on to consumers and what is absorbed by the company. 332 If we are correctly measuring the deadweight loss impacts in the immediately impacted market (the factory gate), this should capture the overall costs. 333 The DIRA Regulations require IPs taking raw milk under the Raw Milk Regulations to provide forecasts of their requirements to Fonterra (ss 9–11, DIRA). For milk other than winter milk an IP must provide a quantity forecast at least three months before the delivery date and must provide a forecast again, one week before the delivery date. This second forecast may vary by up to 40% from the first forecast. Likewise, for winter milk a forecast must be provided to Fonterra 18 months before the delivery date and must again provide a forecast one week before the delivery date. This second forecast may vary by up to 40% from the first forecast. 334 For example, Fonterra may keep a drier on when it otherwise could have been off, or it builds capacity earlier. 2440192 132 6.59 There is a great deal of uncertainty in forecasting future milk volumes. This uncertainty would exist with or without DIRA. It is difficult to isolate the effect of DIRA milk from other drivers of milk volume uncertainty. However, without DIRA there are likely to be better incentives to manage and price the risk appropriately. This means that the cost to economic efficiency of this uncertainty would be borne by the industry participant best placed to manage it. As such, we consider the $6 million that we have estimated to be the upper bound on the excess capacity costs due to the requirement that Fonterra supplies DIRA milk. 6.60 We provide further detail on milk volume uncertainty and excess capacity, including how we estimated the direct cost to Fonterra in attachment F, Evidence that informed our conclusions on excess capacity and resetting market share thresholds. Recovery of opportunity costs 6.61 The DIRA Regulation protects against Fonterra’s prices reflecting market power. However, as noted in chapter 5, State of competition without the DIRA Regulation, there may be times when Fonterra might not recover all of its costs associated with selling DIRA milk, which could also result in inefficiency. This includes times it may not recover all of its opportunity cost. 6.62 If this occurs regularly, it could cause: 6.63 6.62.1 inefficient entry by IPs reliant on access to DIRA milk at a price that does not fully incorporate opportunity cost;335 and 6.62.2 inefficient site choice upon entry by IPs that might not take into account factory gate market conditions for non-DIRA milk. This could result in supply costs which are higher than at an efficient location and could impede a functioning factory gate market from emerging. However while we consider that the DIRA price is likely to cover Fonterra’s average opportunity costs most of the time, we have not undertaken a review of how often the DIRA price may or may not cover opportunity costs under the current DIRA Regulation. However, we have no reason to believe that any inefficiencies created are material. Access to DIRA milk may be hampering the development of the factory gate market 6.64 335 Access to DIRA milk may be hindering or preventing an effective factory gate market for non-DIRA milk from developing. This would effectively prolong the DIRA There is also inefficiency in that Fonterra could make more profit without the DIRA Regulation without exercising market power. 2440192 133 Regulation. DIRA milk may hinder the development of the factory gate market for non-DIRA milk in the following ways: 336 6.64.1 Prices that do not reflect the costs of supplying DIRA milk: IPs have indicated they would not supply the small volumes some factory gate customers require. However, this may relate to the current factory gate price, which at times may not fully reflect all cost, including opportunity costs. 6.64.2 Reduced incentive to seek supply elsewhere: The benefits IPs receive from being able to access DIRA milk (eg, the DIRA milk price, the tolerance limits and guaranteed supply) may reduce incentives for customers of factory gate milk to seek supply outside the DIRA. This in turn can affect the viability of a factory gate market. 6.64.3 Preventing own-source IPs from competing to supply Goodman Fielder: Fonterra is required to supply Goodman Fielder a maximum of 250 million litres of raw milk per season.336 Under the Raw Milk Regulations, Fonterra must either negotiate a price for this supply, or, if a price cannot be agreed,337 Fonterra is required to charge the default milk price.338 6.64.4 The Raw Milk Regulations may prevent own-source IPs from competing to supply some or all of Goodman Fielder requirements at the factory gate. This could hamper the development of the factory gate market without the DIRA Regulation. It may also discourage Goodman Fielder from seeking its own supply. As the largest single customer, if Goodman Fielder were to regularly seek at least some of its volumes outside the DIRA Regulation, this may help encourage the development of this market. Dairy Industry Restructuring (Raw Milk) Regulations 2012, s 7. In addition, supply to Goodman Fielder is also subject to monthly limits, supply for August, September, November, December, January, February, March, April, or May, is limited to 110% of the amount taken by Goodman Fielder in October. 337 [ ] 338 Dairy Industry Restructuring (Raw Milk) Regulations 2012, s 20. The default milk price means the farm gate milk price for that season plus: (a) the reasonable cost of transporting the raw milk to the independent processor; and (b) for winter milk, the additional cost of winter milk in the island in which the milk is supplied; and (c) for organic milk, the reasonable additional costs to new co-op of procuring and supplying the organic milk. 2440192 134 Asymmetric risk to removing DIRA 6.65 We found that the potential efficiency benefits of the DIRA Regulation339 are in a similar order of magnitude of the potential efficiency costs.340 However, we think that there is asymmetric risk attached to removing DIRA—the detrimental impact of removing DIRA too early outweighs the detrimental impact of keeping DIRA too long. 6.66 We consider the asymmetric risk of removing regulation too soon or too late in chapter 7, Pathways to deregulation. 6.67 Table 6.2 summarises the risks of removing DIRA too soon and the risks of removing DIRA too late. 339 In our analysis, we assume that the efficiencies of regulation result from restraining Fonterra from being able to fully exercise market power. 340 This analysis is based on very sensitive assumptions. 2440192 135 Table 6.2: Comparing the risks of removing DIRA too early and too late Risks of removing DIRA too early  Full deregulation at this stage may put at risk some of the competition gains achieved to date by allowing Fonterra to exercise its market power to the disadvantage of some farmers and IPs.  Consumers in downstream dairy markets could also be disadvantaged through Fonterra exercising market power because Fonterra may be able to significantly increase its price for raw milk and it would take time for IPs to obtain alternative supplies.  This could harm long-term efficiency, which offers potentially significant future benefits. Risks of removing DIRA too late  Over-reliance on access regulation when no longer needed can weaken incentives for IPs to enter factory gate market, dampening competition.  Fonterra may have to invest in excess capacity to manage uncertainty caused by the volume of milk IPs take under the Raw Milk Regulations. This may be at the expense of more long-term profitable investments.  Inefficient entry of IPs, although we have not found any evidence that this has been happening thus far. Source: Commerce Commission Analysis on risk of early removal of DIRA. 6.68 Due to this asymmetric risk and the efficiency costs and benefits of DIRA Regulation, we do not think that competition is sufficient to ensure the efficient and contestable operation of the relevant dairy markets if the DIRA Regulation was removed. It is not clear that removing subparts 5 and 5A341 would improve efficiency. Our analysis suggests that, since there is not an effective non-DIRA factory gate market, efficiency could be reduced by the wholesale removal the Raw Milk Regulations. 6.69 However, we note that there are potential costs to efficiency in the current DIRA Regulation that need to be addressed. We discuss various options for potential amendments to the regulations with a view to mitigating these costs in chapter 7, Pathways to deregulation. 341 Subpart 5 relates to the open entry and exit provisions (including the 20% rule) and subpart 5A relates to the base milk price disclosure and monitoring rules. 2440192 136 CHAPTER 7: Pathways to deregulation Purpose of this chapter 7.1 This chapter gives our analysis of options the Minister could consider as transition pathways to deregulation, and outlines the approach that could be taken to reset the market share thresholds, including the potential to introduce new or alternative expiry triggers. 7.2 In this chapter, we outline: 7.2.1 our overall recommendations; 7.2.2 why we have identified transition pathways to deregulation; 7.2.3 our recommendations for resetting the market share thresholds; 7.2.4 our recommendation to reset the time limit provision; 7.2.5 our recommendation that additional or alternative expiry triggers are not necessary; and 7.2.6 our identified options, and of those the options we recommend, for transition pathways to deregulation. Our recommendations 7.3 So far as the available information allows, we recommend the Minister should consider: 7.3.1 resetting the market share thresholds in the North and South islands to 30%; 7.3.2 resetting the time limited provisions of the DIRA to the 2021/22 season; 7.3.3 not including any additional or alternative expiry triggers to the DIRA; and 7.3.4 as a transition pathway to deregulation, exploring amendments to the Raw Milk Regulations (including the Goodman Fielder provisions) in order to facilitate better functioning factory gate markets. 7.4 We consider the main potential concern from full deregulation now is Fonterra increasing the price of raw milk at the factory gate to a level above the competitive price. We are primarily concerned about the end impact on domestic retail markets where manufacturers are dependent on access to raw milk at the factory gate. 7.5 We consider that the domestic dairy markets could benefit from a better functioning factory gate market. We recommend that the Minister consider options to facilitate the development of a functioning factory gate market as a transition 2440192 137 pathway to deregulation. A functioning factory gate market would reduce the risks associated with further deregulation in the future. 7.6 We have drawn upon our investigation of the sufficiency of competition to identify options for transition pathways to deregulation, but we have not evaluated any of the options in detail. 7.7 Detailed policy issues and choices on how to best give effect to the efficiency objectives of the DIRA are outside the scope of this report. Why we identified transition pathways to deregulation 7.8 The DIRA Regulation was designed to facilitate efficient entry by IPs and facilitate efficient growth and expansion. 7.9 It was envisaged that eventually, it would be possible to remove the DIRA Regulation, as IPs became able to effectively compete with Fonterra in farm and factory gate markets (as well as downstream domestic dairy markets). 7.10 Our review has concluded that the current state of competition is insufficient in one or more of the relevant dairy markets, and therefore we have considered: 7.11 7.10.1 whether either or both of the market share thresholds specified in section 147 of DIRA should be reset (and if so, to what level); 7.10.2 whether any other expiry triggers should be introduced in place of, as well as, the market share thresholds; 7.10.3 what options, if any, there are for a transition pathway to deregulation; and 7.10.4 whether any of the options (or sets of options) for a transition pathway to deregulation should be pursued. Under the DIRA, automatic deregulation cannot occur before an expiry threshold or a provided date is triggered (and after the resulting competition review is reported to the Minister). The efficiency costs and benefits of the DIRA Regulation evolve as competition develops 7.12 Economically inefficient outcomes can occur both by retaining the DIRA Regulation too long, or removing it too soon. Our views on the potential error costs associated with each of these outcomes may help inform views on the expiry thresholds. 7.13 Removing the DIRA Regulation too soon could harm efficiency and structural competition gains achieved through regulation. For example, Fonterra may raise raw milk prices in the factory gate above competitive levels to earn higher profits. This could result in a: 2440192 138 7.14 7.15 7.13.1 loss of innovation in the domestic markets, including the evolution of new and higher value products, through insufficient diversity and competition in the production and marketing of New Zealand dairy products; and 7.13.2 reduced level of contestability and competition in the farm, factory and retail markets. Keeping the DIRA Regulation in place for too long could risk other types of inefficiency, such as: 7.14.1 inefficient incentives that discourage IPs from entering the factory gate market thereby dampening competition. For example, the Raw Milk Regulations may crowd out the development of a functioning factory gate market; 7.14.2 the cost to Fonterra of having to invest in excess capacity to take on unexpected milk volumes from either returning farmers or variances in regulated milk deliveries (at the expense of more long-term profitable investments);342 and 7.14.3 inefficient entry by IPs that may not be economic without the regulations. We note that in the absence of the DIRA Regulation, Fonterra would remain subject to the provisions in the Commerce Act 1986, which restrict anti-competitive behaviour (see chapter 5, State of competition without the DIRA Regulation).343 We recommend resetting the market share thresholds of the DIRA 7.16 This section discusses our recommendation on how the market share thresholds in the North and South islands should be reset. 7.17 Our recommendation is to reset the market share thresholds to 30% in the North and South islands. 7.18 We consider that triggering of either of these market share thresholds should continue to result in a competition review of the dairy industry.344 342 Although that risk is in turn offset by other factors such as Fonterra’s large network providing processing flexibility, its ability to temporarily refuse to take new supply through the use of capacity constraint notices, its ability to flex its product set depending on the volume of milk it is receiving (including, for example, producing non-standardised product). 343 In particular, section 36 of the Commerce Act 1986 makes it illegal for any business with a substantial degree of market power to take advantage of that power to deter or prevent rival businesses from competing effectively. Section 27 of the Commerce Act makes it illegal to enter into or give effect to an agreement that has the purpose or effect of substantially lessening competition. 2440192 139 7.19 We took into account the following points, which we discuss in this chapter: 7.19.1 market share thresholds are simple but imperfect proxies for competition. However, the market share thresholds trigger a competition review rather than automatic deregulation; 7.19.2 an effective market share threshold should reflect the asymmetric risks of deregulation; 7.19.3 our current review suggests the current market share thresholds are too low; 7.19.4 submissions on our draft report and wider contextual factors such as Fonterra’s co-operative structure; and 7.19.5 historic observation and judgement to project future market shares. Market share thresholds are simple but imperfect proxies of competition 7.20 We consider that the best expiry trigger is one that is simple and well understood. 7.21 Market share measures do not necessarily provide a fully meaningful representation of the level of competition in the industry as these thresholds are not, for example, based on the relevant geographic markets, nor do they indicate the level of market concentration in the relevant markets.345 7.22 However, the current expiry triggers trigger a competition review, not automatic deregulation as was the case before the last review in 2011, which in our view addresses this limitation of the market share thresholds.346 The market share thresholds can trigger a competition review before the time limit 7.23 We consider that the next competition review should not be triggered too early. It is appropriate for any transition pathway amendments made to the DIRA Regulation to be given time to improve the sufficiency of competition before we reassess the markets. 344 Submitters support a review process, rather than a trigger leading to full deregulation. For example: Tatua “Submission on review of the state of competition in the New Zealand Dairy Industry” July 2015, para 6.5. Open Country “Submission on consultation paper” 10 July 2015, p.4. Miraka “Submission on consultation paper” 10 July 2015, para 4.1.4. 345 The level of market concentration depends on the market shares and the number of firms active in the relevant markets. 346 But, where a trigger has been met deregulation will occur in the relevant island(s) unless enactments are passed to reset the market share thresholds and/or extend the DIRA Regulation. 2440192 140 7.24 If there is unanticipated industrial change before the time limit provision is met the market share thresholds can act to trigger a competition review. While our concerns are focused on the factory gate market we consider the farm gate market share thresholds will act as an adequate proxy to trigger the next competition review.347 7.25 The market share thresholds and time limit triggers should provide sufficient regulatory predictability for existing IPs, future new entrants in dairy processing and, potentially, new dairy conversions such that they can benefit from the DIRA Regulation for a period of time, and enter or expand accordingly. An appropriate market share threshold should consider asymmetric risk 7.26 On balance, we consider that if the risks associated with removing the DIRA Regulation too early eventuate this would likely have a greater and more detrimental impact than removing the DIRA Regulation too late. 7.27 At this time, we consider that the potential adverse consequences of regulating too long (the ongoing efficiency costs) are smaller than the consequences of removing the regulations prematurely.348 7.28 In chapter 6, Our assessment of the efficiency costs and benefits of the DIRA Regulation, we discussed the asymmetric risks of deregulating too early. In particular we were concerned with the possible loss of the downstream competitive gains that the DIRA Regulation has facilitated while competition has been insufficient in the farm and factory gate dairy markets. 7.29 We also note that a staged approach to deregulation could minimise the disruption to domestic markets in the transition to deregulation. Our review of competition in dairy markets has informed our view on the market share thresholds 7.30 The market share thresholds are an average over the different geographic farm gate markets in each island. For example, the Canterbury region has a relatively competitive farm gate market, with IPs holding approximately [ ] market share in the 2014/15 season. 347 In this context, we agree with NERA's observation that the demand side of the farm-gate market contributes to the supply side of the factory-gate market. NERA "Review of the Commerce Commission's draft report into dairy competition" 4 December 2015, p.1. 348 This view was informed by the small size of the domestic market relative to Fonterra’s international activities. We view the DIRA Regulation as relatively low cost. The competition analysis suggests that in the absence of regulation the efficient and contestable operation of the farm and factory gate markets could not be ensured. Commerce Commission “Draft Report—Review of the state of competition in the NZ Dairy Industry” 6 November 2015, para 6.85–6.91. 2440192 141 7.31 In chapter 5, State of competition without the DIRA Regulation, we reported our finding that the ability to exercise market power may be more muted in certain regions, such as Canterbury. However, there is nonetheless likely some ability to exercise market power in Canterbury even with an IP market share of [ ]. 7.32 Therefore, we consider that setting the market share thresholds closer to 25% would likely be too low to reasonably expect a finding of sufficient competition in the relevant dairy markets, based on our findings.349 The next increment in share is 30% and is our starting point for considering the appropriate trigger for another competition review. We considered submitters' views on the market share thresholds 7.33 We noted the views of submissions on our draft recommendation to reset the market share thresholds to 30%, in particular: 7.33.1 Danone Nutricia considers that resetting the market share thresholds to 30% is too low and risks undertaking another competition review prematurely.350 It notes that this threshold would be viewed as low in foreign contexts.351 It also considers that the competitive issues present at the 20% thresholds are still likely to be present at a 30% threshold;352 7.33.2 Westland submitted its support for resetting the market share thresholds to 30% for both islands. It does not recommend additional expiry triggers, other than the time limit provision;353 and 7.33.3 Fonterra submitted that 30% market share thresholds are inappropriately high. It considers this threshold is too high given the speed of industrial change and the risk of leaving costs and inefficiencies in place too long— particularly if any review process takes up to two years to be completed and implemented.354 349 For example, in Canterbury which is characterised by a relatively competitive farm-gate market the independent processors’ collective market share in 2014/15 was approximately [ ]. 350 Danone Nutricia “Submission on the Dairy Competition Review Draft Report” 4 December 2015, paragraph 1.2(c). 351 Danone Nutricia “Submission on the Dairy Competition Review Draft Report” 4 December 2015, paragraph 4.2. 352 Danone Nutricia “Submission on the Dairy Competition Review Draft Report” 4 December 2015, paragraph 4.2. 353 Westland “Submission on the Dairy Competition Review Draft Report” 4 December 2015, paragraph 5–6. 354 Fonterra “Submission on Draft Report: Review of the state of Competition in the New Zealand Dairy Industry” 4 December 2015, paragraph 14. 2440192 142 7.34 Federated Farmers questioned the need to have a market share threshold at all.355 It stated it was unable to offer an alternative threshold, but challenges whether the market share thresholds are needed given they trigger a competition review. 356 7.35 Open Country Dairy submitted that the 30% market share threshold should trigger a competition review only, and not require future legislative change to prevent automatic deregulation.357 Open Country Dairy considers this amendment would be sensible as “everyone agrees a review needs to be undertaken to determine whether at any point the New Zealand dairy industry is ready for deregulation”.358 We considered wider contextual factors in making our recommendation on the market share thresholds 7.36 We have had regard to broader factors that may be informative to resetting the market share thresholds, including: 7.36.1 Fonterra’s co-operative structure (which, all else being equal, may support lower market share thresholds). The market share thresholds should not be reset ‘too high’ as while Fonterra is likely to dominate at the farm gate in some regions (eg, Northland) for the foreseeable future its co-operative structure might also mitigate some of the adverse efficiency effects of its market power; 7.36.2 competing IPs are operating at or near capacity (which, all else being equal, may support higher market share thresholds). For example, a market share of 30% may overstate actual contestability as some IPs are at capacity and not able to increase farmer supply. This may be a feature of the landscape as an IP that is underutilising its plant would likely bear additional costs; 7.36.3 expected growth of the industry (which, all else being equal, may support lower market share thresholds). This reveals the dynamism in the market including the size of the IPs that can provide a competitive constraint on Fonterra; and 355 Federated Famers “Submission on the Dairy Competition Review Draft Report” 4 December 2015, paragraph 1.3. 356 Federated Famers “Submission on the Dairy Competition Review Draft Report” 4 December 2015, paragraph 3.6. 357 Open Country Dairy “Submission on the Commerce Commission’s Draft Report—Review of the State of Competition in the New Zealand Dairy Industry” December 2015, p.1–2. 358 Open Country Dairy “Submission on the Commerce Commission’s Draft Report—Review of the State of Competition in the New Zealand Dairy Industry” December 2015, p.2. 2440192 143 7.36.4 overseas use of market share thresholds (which, all else being equal, may support higher market share thresholds). Though there is no precise threshold, competition authorities worldwide generally consider a large market share is required to establish significant market power. A 30% market share threshold (with Fonterra holding 70%) would be viewed as low in foreign contexts.359 We considered historic trends when making our recommendation on the market share thresholds 7.37 In attachment F, Evidence that informed our conclusions on excess capacity and resetting market share thresholds, we explain how we have projected IP farm gate market shares in the North and South islands. 7.38 Our projections suggest that in 2021/22 neither the North or the South island IPs will collectively have a market share of more than 30%. Our recommendation to reset the market share thresholds to 30% for the North and South islands is conservative (ie unlikely to be triggered) before the proposed time threshold (2021/22) is reached. 7.39 We consider that conservative lower market share thresholds are appropriate for triggering a competition review. We consider that in terms of selecting a specific market share threshold: 7.40 359 7.39.1 relying on market share thresholds alone to trigger deregulation carries a risk of regulating for ‘too long’ if we wait for imperfect triggers to be met (though the costs of this occurring are likely to be relatively low). The risk of regulating for ‘too long’ is also low if the triggers are not set too high or a time limit is retained; 7.39.2 triggers set very low that result in a more frequent review and potential DIRA amendment may deter investment through a perceived uncertainty in future regulations; and 7.39.3 therefore, we consider that a market share threshold trigger should be adopted at the lower end of a range that is most likely to be indicative of sufficient competition. We note that there is significant uncertainty inherent in projecting market outcomes years in advance. In particular developments in the factory gate market We note in passing a 2008 European Commission merger decision required the merged company to effectively reduce its market share of farm gate raw milk, ensuring 30–40% farm gate raw milk is available to IPs. Commission of the European Communities [17/12/2008] Case No COMP/M.5046 – Friesland Foods / Campina. 2440192 144 could bring forward the time at which an inquiry could be triggered. That, in turn, would also reflect increased competition, which would suggest an inquiry is appropriate. 7.41 We consider the market share threshold is likely to be met first in the South Island. On this we note: 7.41.1 future analysis might find that deregulation should apply only to the dairy market in the South Island, leaving the DIRA Regulation in place in the North Island; and 7.41.2 regional deregulation within the South Island might be difficult to administer and result in perverse factory gate outcomes.360 A time limit provision is appropriate to accompany the market share thresholds 7.42 This section discusses our recommendation to reset the DIRA Regulation time limit provisions. 7.43 Our recommendation is that the time limit provisions of the DIRA Regulation be reset to the 2021/22 season.361 We consider this is appropriate given the inherent uncertainty as to the future development of the markets. 7.44 The time limit provision requires the Minister to request a report on the state of competition by a certain date in the event that the market share thresholds have not been triggered.362 The time limit provisions will likely trigger the next competition review 7.45 With our recommended time limit provision, our recommended market share thresholds could be viewed as a simple mechanism to bring forward a competition review in the event sufficient competition emerges more quickly than anticipated by the market share forecasts.363 360 For example, DIRA factory gate- dependent IPs may choose geographically inefficient sites if there is regional deregulation within an island. This could potentially create further barriers to a functioning factory gate market with the complete absence of the DIRA Regulation. 361 Dairy Industry Restructuring Act, s 148A. 362 If the Minister does not promote a new enactment to reset the market share thresholds and/or extend the DIRA Regulation then it will be revoked, that is, the Minister must respond to the report because if he does not, the DIRA will be revoked in the island(s) where the threshold has been met. 363 Westland notes continued activity in irrigation in the South Island and significant variation in levels of competition across the farm gate and factory gate markets as an example of future potential industrial change justifying a time limit provision set to the 2020/21 season. We note that the market share 2440192 145 7.46 As discussed in chapter 5, State of competition without the DIRA Regulation, our main concern is with the functioning of the factory gate market. Since it will take some time for any changes to the DIRA Regulation to be implemented and influence the markets it is important to allow sufficient time between one review ending and another beginning.364 7.47 We are doubtful that sufficient competition will develop without first triggering the 30% market share thresholds. However, the time limit provision triggering a competition review will allow an assessment of competition and the impact of any transition pathways to deregulation. It will act to mitigate the risk of sufficient competition developing without first triggering the 30% market share thresholds. We considered submitters' views on the time limit provision 7.48 We consider that setting the time limit provision earlier than the 2021/22 season may prematurely trigger the next competition review. 7.49 The 2021/22 season is an appropriate time limit provision because of the time it may take for adopted transition pathways to influence factory gate market outcomes, the potential for market share thresholds to trigger a review earlier than the 2021/22 season, and the asymmetric risks of deregulation. 7.50 Submitters supported the use of a time limit provision, but generally considered that setting the time limit provision to the 2021/22 season was too distant. In particular: 7.50.1 Federated Farmers supports the use of a time limit provision, though it considers the 2021/22 season to be too distant. It recommends setting the next competition review to 1 July 2020 and every five years thereafter;365 7.50.2 Westland supports a time limit provision. However, given the potential for sufficient change to occur in the industry, it prefers that a four-year-time limit is adopted (ie, the 2020/21 season);366 thresholds could trigger an earlier competition review if these potential industrial changes are material. Westland “Submission on the Dairy Competition Review Draft Report” 4 December 2015, paragraph 14.1. 364 Fonterra submits that, assuming the next competition review is triggered by the 2021/22 time limit provision, it is unlikely that any changes to the DIRA Regulation will be effected before the 2023/24 season. It considers this to be overly conservative. We note that there must be sufficient time before the next review to allow the markets to develop and the adopted transition pathways to take effect; otherwise it is likely that the next review will come to similar conclusions as this review. The effect of setting a tight time limit provision may be that the DIRA Regulation persists longer than necessary. See Fonterra “Submission on Draft Report: Review of the state of Competition in the New Zealand Dairy Industry” 4 December 2015, paragraph 55.2. 365 Federated Famers “Submission on the Dairy Competition Review Draft Report” 4 December 2015, paragraph 3.5. 2440192 146 7.50.3 Fonterra considers setting the time limit provision to 2021/22 to be inappropriate given the speed of change in the industry and risk of leaving inefficiencies and costs in place for too long;367 and 7.50.4 Tatua generally supports our recommended time limit provision but noted that it would have preferred a longer term review mechanism be implemented.368 We do not recommend additional or alternative expiry triggers 7.51 This section discusses our recommendation on whether there are additional expiry triggers that should be provided for, either in addition to or in place of the market share thresholds. 7.52 Our recommendation is to not introduce any expiry thresholds in place of or in addition to the market share thresholds. Alternative or additional triggers should only be adopted if they add value to the DIRA Regulations 7.53 Additional or alternative expiry triggers to the market share thresholds should capture new or more information to assist in justifying a competition review or, with a necessarily higher standard, automatic deregulation. 7.54 Given that a competition review is triggered by the expiry triggers, we consider an effective expiry trigger should be robust against small changes in the market; be linked to our concerns in the industry; and be simple, transparent, and well understood.369 We considered options for additional or alternative expiry triggers 7.55 We have not identified suitable additional or alternative expiry triggers to the market share thresholds that provide better information than the farm gate market share thresholds. 7.56 While different measures have relative advantages, there is no perfect measure, and they would likely introduce additional complexity with limited timing benefits in triggering a competition review or automatic deregulation. 366 Westland “Submission on the Dairy Competition Review Draft Report” 4 December 2015, paragraph 5. 367 Fonterra “Submission on Draft Report: Review of the state of Competition in the New Zealand Dairy Industry” 4 December 2015, paragraph 14. 368 Tatua "Review of the state of competition in the New Zealand dairy industry—draft report" 4 December 2015, paragraph 6.1. 369 However, if a market share threshold has been met deregulation will occur unless enactments are passed to reset the market share thresholds and/or extend the DIRA Regulation. 2440192 147 7.57 7.58 Measures which we considered as expiry triggers might instead be more appropriate in informing a future competition review. This includes measures of: 7.57.1 industry concentration;370 or 7.57.2 industry dynamics.371 An expiry trigger should be based on a geographic area that we consider could be deregulated independently of the rest of New Zealand. In addition to the North and South islands, we considered triggers covering: 7.58.1 New Zealand;372 and 7.58.2 sub-island regions.373 Factory gate market share thresholds measures may not be workable as an expiry trigger 7.59 Our main concerns with achieving sufficient competition are focused on the development of a functioning factory gate market. We considered adopting factory gate market share thresholds in place of the farm gate market share thresholds. 7.60 On balance, we consider the farm gate market share threshold is likely to be a more effective expiry trigger than a factory gate market share threshold. The incremental benefits of adopting a factory gate market share threshold do not outweigh the drawbacks. 7.61 We assessed the concept of factory gate market share thresholds relative to farm gate market share thresholds. We concluded that factory gate market share thresholds are: 7.61.1 less robust. There is less information on market shares in the factory gate. It would be practically difficult to set an appropriate threshold; 7.61.2 better linked to our concerns. Factory gate market shares are directly linked to our concerns. However, the demand-side of the farm gate market contributes the supply-side of the factory gate market;374 370 This could use established measures, such as Herfindahl indices, or policy specific such as ‘two IPs with 20% market share’. 371 This could include measures such as the proportion of new supply going to Fonterra, proportion of capacity owned by Fonterra, Fonterra’s farmer churn, and rates of change in other measures over a period of time (eg, percentage change in market share over three years). 372 An appropriate national trigger could be used as a sufficient condition for automatic deregulation. 373 Deregulating regional areas of sufficient competition may be an appropriate transition pathway to deregulation. We discuss regional deregulation in paragraphs 7.70 – 7.80 below. 2440192 148 7.61.3 not transparent or well understood. There is limited information on market share at the factory gate. This would make it difficult for market participants to monitor and predict when another review would be triggered. Options for transition pathways to deregulation 7.62 This section discusses our options and recommendations for transitional pathways to deregulation. 7.63 We have identified opportunities to better promote the efficiency purpose of the DIRA. Our recommendations for transition pathways that should be pursued are: 7.64 7.65 7.63.1 a staged approach to transition pathways to deregulation is appropriate at this time; and 7.63.2 the Minister consider potential amendments to the Raw Milk Regulations (including the Goodman Fielder provisions) that have the potential to incentivise additional demand and supply in the factory gate market outside of DIRA milk; We concluded in chapters 5 and 6 that: 7.64.1 the open entry and exit provisions may still be important for entry and expansion by own-source IPs.375 These regulations impose modest costs; 7.64.2 the existing own-source IPs will be the most likely potential future suppliers in a deregulated factory gate market; and 7.64.3 the Raw Milk Regulations are the most appropriate area of focus for a transition pathway to deregulation. The Raw Milk Regulations may be distorting the development of a factory gate market. They result in Fonterra having to invest in some additional capacity. We consider that transition pathways to deregulation should: 7.65.1 not encourage a dependency on the regulations. Transition pathways should allow for efficient market participants to stand on their own two 374 For an IP to be able to supply other IPs at the factory gate it must have raw milk that it can sell on. This is most likely to have been procured at the farm gate. Therefore, a well-functioning farm gate market is indicative of the potential to supply at the factory gate. 375 The sunset clauses on DIRA milk may make the Raw Milk Regulations less of a critical factor for new entry and expansion by own-supply IPs. However, the three-year clause likely addresses a key entry concern of filling capacity just after entry into the market. 2440192 149 feet, facilitating a future sufficiency of competition in the relevant dairy markets; and 7.65.2 maintain confidence in the direction of the regulatory regime. The regulations are temporary and the deregulation process should allow a well signalled, simple glide path to market participants. We recommend a staged approach to transition pathways to deregulation 7.66 We recommend a staged approach as the most appropriate means of achieving transition pathways to deregulation. This approach is the same approach taken in the 2012 amendments to DIRA which in our view successfully removed material inefficiencies created by the DIRA Regulation.376 We discuss the 2012 Raw Milk amendments in attachment C. 7.67 If competition was sufficient, we would recommend the DIRA Regulation be removed.377 As we have concluded that competition is insufficient, we consider what options there are for transitional pathways to deregulation.378 Support for our recommendation of a staged approach to deregulation 7.68 Our approach to staged deregulation focuses on recommending amendments to DIRA Regulation that could better promote the efficiency purpose of DIRA and may facilitate future deregulation. 7.69 Westland supports our recommendation of a staged approach to deregulation. In its submission, Westland highlighted the importance of a transition pathway to be introduced when resetting the expiry triggers. In particular, Westland considers:379 …that to reset the thresholds right now without also introducing a transition pathway creates a degree of dependence on the Regulations by some IPs and with it the potential for inefficient farm gate and factory gate markets. 7.70 As competition in the relevant dairy markets strengthens and becomes sufficient, policy makers could consider a staged repeal of subparts of DIRA as the most 376 In particular, the sunset provisions added to the Raw Milk Regulations: if an IP collects more than 30 million litres of raw milk from its own farmer suppliers for the past three years it no longer has access to regulated milk; adopting a fixed quarterly price rather than an annual price with wash-up price for regulated milk; and the quantity of raw milk an IP has access to following the milk curve. 377 See chapter 3, Process and framework for evaluating and reporting on the state of competition. 378 See chapter 6, Our assessment of the efficiency costs and benefits of the DIRA Regulation. 379 Westland “Submission on the Dairy Competition Review Draft Report” 4 December 2015, paragraph 18. 2440192 150 appropriate transitional pathway. At this point, we have not identified any subparts to repeal in entirety.380 7.71 A staged approach to repealing subparts might involve first repealing the regulation that contributes least to efficiency and contestability. 7.72 We consider that a predetermined pathway to complete deregulation is inappropriate at this time. The order of a future staged repeal, or areas of amendment, to the DIRA Regulation should depend on the state of future competition. 7.73 We have recommended amendments to the Raw Milk Regulations in order to facilitate a functioning factory gate market as a transition pathway to deregulation. 7.74 If a functioning factory gate market emerges, then there may be an opportunity to consider full deregulation. If a functioning factory gate market does not emerge then this transition pathway to deregulation would need to be re-assessed. Some submissions do not support deregulation 7.75 We note Federated Farmers statement that “[it] does not support the Commerce Commission’s push for deregulation.”381 Instead it considers that the DIRA should be changed in order to allow the industry to work in an innovative and efficient manner over a period of decades.382 Federated Farmers disagrees that the DIRA Regulation is temporary. 7.76 We consider that our recommended transition pathways will result in improved efficiency in the domestic dairy markers. We considered regional deregulation as staged deregulation 7.77 We considered whether the Raw Milk Regulations and the open entry and exit provisions could be removed in particular regions. 7.78 We recognise the practicalities of regional deregulation might be too problematic to successfully implement. For example, oversight and enforcement of region specific Raw Milk Regulations might be difficult to administer and costly for Fonterra and/or the enforcement agency. 380 Fonterra agrees with the Commission that a further review will be necessary in order to “sensibly implement a staged approach to deregulation.” Fonterra “Submission on Draft Report: Review of the state of Competition in the New Zealand Dairy Industry” 4 December 2015, paragraph 53. 381 Federated Famers “Submission on the Dairy Competition Review Draft Report” 4 December 2015, paragraph 3.1. 382 Federated Famers “Submission on the Dairy Competition Review Draft Report” 4 December 2015, paragraph 1.1.1. 2440192 151 7.79 7.80 Repealing the Raw Milk Regulations in particular regions may result in sub-optimal factory gate outcomes overall later in time. For example: 7.79.1 it is likely that a well-functioning farm gate market is indicative of the potential to supply at the factory gate. The regions where the DIRA Regulation would most likely be repealed are also areas of strong milk growth, and entry and expansion of own supply IPs (eg, Canterbury); 7.79.2 removing the Raw Milk Regulations in these areas, but not others in the same island, could incentivise IPs dependent on DIRA milk to locate in regions further away from the areas of milk growth and self-supply IPs. This may be more likely if access to DIRA milk is more favourable than contracting with own-source IPs at the factory gate;383 and 7.79.3 in eventual absence of the DIRA Regulation in both the North and South islands, the IPs dependent on DIRA milk may be inefficiently located, or not make up the critical mass required to incentivise supply at the factory gate by own-source IPs. The benefits of regional deregulation include: 7.80.1 removing unnecessary regulation in areas of sufficient competition where and when it develops;384 and 7.80.2 providing an opportunity for experimentation in transition pathways to deregulation before adopting the preferred pathway nationwide. For example, discovering more about how a functioning factory gate market would develop in the absence of DIRA Regulation may be prudent. Reduce the dependency of IPs in the factory gate market on the Raw Milk Regulations 7.81 An option for a transition pathway to deregulation is to reduce the dependency of factory gate-dependent IPs on the Raw Milk Regulations. 7.82 The analysis in chapter 5, State of competition without the DIRA Regulation, suggests that the Raw Milk Regulations in their current form may at times be setting the DIRA milk price below competitive levels which may be impeding the 383 This might be particularly true if the DIRA price of raw milk at the factory gate is below the market price that might emerge in the absence of regulation. This could be limited by allowing a price increase on DIRA milk and signaling the temporary nature of the DIRA Regulation. 384 At this time, we have not identified a region of New Zealand where competition might be considered sufficient. The Canterbury region, for example, has an IP market share of approximately [ ] in the 2014/15 season. 2440192 152 development of a functioning factory gate market.385 However, the volatility of commodity prices makes it hard to be definitive about whether the DIRA price is properly reflecting opportunity costs. 7.83 Currently, there is uncertainty on the ability of IPs that compete in downstream markets to efficiently source their full factory gate raw milk requirements to meet domestic demand in the absence of DIRA Regulation. It is this domestic market, for fresh processed milk, yoghurt, and cheese where the DIRA Regulation’s main competitive benefits terms of in price and quality are potentially felt. 7.84 To reduce the risk of deregulation, a viable source of supply other than directly contracting with farmers may be needed. A functioning factory gate market outside of DIRA would achieve this. Therefore, it is important to consider how refining the regulations may aid the development of such a market. 7.85 The regional options for factory gate suppliers vary given the location of the large own-source IPs. These large own-source IPs compete in the international markets and are best placed to contract directly with farmers and potentially supply the factory gate market as competitors to Fonterra in the absence of DIRA Regulation.386 Efficient demand and supply should be facilitated in the factory gate markets 7.86 7.87 Currently, we have not observed a functioning factory gate market outside DIRA Regulation. We consider this may reflect: 7.86.1 the DIRA entitlements affecting the demand for raw milk outside the DIRA Regulation. For example, the Goodman Fielder provisions that allow for its total raw milk requirements, and the forecast tolerances allowed to other IPs; and 7.86.2 the DIRA price for raw milk at the factory gate might not cover Fonterra’s opportunity cost at some points in time. This may affect the incentives of own-source IPs to supply factory gate milk if this price is also below their opportunity cost. This combination of incentives created by the Raw Milk Regulations may act to prevent a functioning factory gate markets from developing. On this we note Fonterra’s submission that “[t]he lack of apparent interest in the factory gate 385 On the other hand we note the Raw Milk Regulations have facilitated entry by IPs that are now potential entrants into a future factory gate market. 386 [ 2440192 ] 153 market on the part of particularly larger exporting IPs is largely attributable to insufficient incentives rather than any other identifiable factor”.387 Options to facilitate the development of functioning factory gate markets 7.88 We considered potential options to facilitate the development of a functioning factory gate market, as a transition pathway to deregulation. A gradual removal of the Raw Milk Regulations would assist in mitigating the risks associated with deregulation. 7.89 It may take some time for a functioning factory gate market to develop. The small scale of the factory gate market outside the DIRA milk may mask other factors that could also hinder its development. A gradual shift may also mitigate transitional costs to the industry. The DIRA raw milk entitlements could be amended or allowed to expire 7.90 An option is to amend the DIRA raw milk entitlements, or, alternatively, let the entitlements expire. Signalling the expiry of the raw milk entitlements now may cause IPs dependent on DIRA milk to explore their alternative options now and hence create greater demand for raw milk outside the DIRA Regulation.388 The following options that might facilitate this could be explored: 7.90.1 Tighten the forecast tolerances allowed for in the Raw Milk Regulations. The extent to which these provisions provide a free option to IPs may crowd out the development of a functioning factory gate market (as well as imposing milk volume risk and potentially milk coordination costs on Fonterra).389 7.90.2 Allow Fonterra to apply a take or pay approach to DIRA milk sales. 7.90.3 Investigate further whether Fonterra is recovering its opportunity cost through the DIRA price. 387 Fonterra “Submission on draft report: Review of the state of competition in the New Zealand dairy industry” 4 December 2015, para 36. 388 We note Federated Farmers submission that some of options for transition pathways to deregulation “could work but [it] consider[s] that these still give those IPs that have no interest in adding value to the farm gate or factory gate markets the incentive to stay on [regulated] milk.” Federated Famers “Submission on the Dairy Competition Review Draft Report” 4 December 2015, para 3.12.12. 389 Fonterra has also provided two examples of situations where DIRA milk forecasted by IPs has not ended up being taken: [ ] 2440192 154 7.90.4 7.90.5 Learn from the success of the sunset clauses introduced for larger ownsupply IPs in 2012 (and which appear to have worked well to reduce dependence on DIRA milk)390 and: 7.90.4.1 extend sunset clauses to IPs that collect some of their ownsupply even if under the current cap. This may create incentives for factory gate-dependent IPs to demand raw milk from a factory gate market outside DIRA Regulation; 7.90.4.2 introduce sunset clauses for factory gate-dependent IPs that do not have their own-supply.391 Providing for these IPs to consider their factory gate alternatives and make necessary adjustments to their business (eg, location, whether to accept own-supply) may require longer sunset clauses. There appear to be diminishing marginal benefits from further IP entry in some regions and at least the potential for existing IPs to expand into other regions. In this context an option could be to remove DIRA entitlements, particularly for IPs that are largely serving the competitive international markets. Allow trading of DIRA raw milk entitlements 7.91 7.92 An option is to consider whether DIRA raw milk entitlements should be allowed to be traded: 7.91.1 In particular, being able to trade the DIRA raw milk entitlements may act as a signalling mechanism about the ‘correctness’ of the DIRA price while allowing for the raw milk volumes to flow to the most valued ‘use’; and 7.91.2 We consider that potential trading of DIRA raw milk entitlements would need to be regional in scope. Regional trading could assist to mitigate potentially inefficient volume risk imposed on Fonterra. However, a number of practical issues would need to be addressed to implement traded entitlements. A related option is to grandfather the rights392 of IPs currently taking DIRA milk under the Raw Milk Regulations based on previous volume but: 390 For details, see 2012 Raw Milk Regulations Amendments, Attachment C. 391 Federated Farmers supports this option and suggests a three-year sunset clause. Federated Famers “Submission on the Dairy Competition Review Draft Report” 4 December 2015, para 3.12.8–3.12.10. 392 That is, when new regulation is in place, allow the old rule to apply in particular circumstances. 2440192 155 7.92.1 in particular, the previous volumes taken under the Raw Milk Regulations could be allowed for at the DIRA milk price (and possibly tradeable). Additional volumes would be obtained at a market price in the factory gate market; 7.92.2 new entrants would have to obtain raw milk in the unregulated factory gate market; and 7.92.3 it would mean that incumbent IPs would face an efficient price for any additional milk they demanded, while minimising possible disruption to the downstream markets. But their access to DIRA milk could potentially give the incumbent IPs a material advantage in downstream domestic markets. Create an auction system for DIRA milk 7.93 An option is to set up an auction system for DIRA milk as an alternative means of allowing IPs to purchase raw milk provided the price discovered is attractive to them.393 7.94 An auction system could help improve the functioning of a factory gate market through aiding price discovery, improving efficiency by ensuring buyers pay market prices, and assisting smaller IPs access the market. 7.95 However, we note the practical considerations that would have to be taken into account in determining the design and feasibility of an auction. Participation and volumes,394 a platform,395 safeguards,396 and choice of products397 are examples of issues to be taken into account. 7.96 Also, we have not carried out detailed assessment of the merits and practicalities of an auction. The effectiveness of a factory gate market auction system as a transition pathway may depend on the extent to which the development of it is industry led versus mandated through legislation. 393 Federated Farmer's suggested an auction system which can allow price discovery of factory gate milk and the value a purchasing IP can add. It submits that an auction could help support the IPs that may experience difficulty in a full transition to the factory gate market. Federated Farmers “Submission on the Dairy Competition Review Draft Report” 4 December 2015, paragraphs 3.12.11–3.12.14. 394 Fonterra (and potentially other larger IPs) may have to be obligated to sell minimum volumes in the auction. Additional rules may also be required to ensure that smaller IPs can access the small volumes that they require. 395 A platform would be required for the auction that was accessible to smaller IPs. 396 There may need to be some safeguards, for example buy-side rules, to prevent those obligated to supply the auction from placing matching bids on the other side. 397 For example, the auction could cover a futures market and/or a near-term market. 2440192 156 Reduce dependency of Goodman Fielder on the Raw Milk Regulations 7.97 An option for a transition pathway to deregulation is to reduce the dependency of Goodman Fielder on the Raw Milk Regulations.398 7.98 Fonterra is required under the Raw Milk Regulations to provide Goodman Fielder with up to 250 million litres of regulated raw milk per season. Fonterra and Goodman Fielder signed a commercial contract agreeing the terms of purchase of raw milk subject to the cap of 250 million litres. This contract is due to end in 2021.399 7.99 In submissions to the Commission Fonterra stated that it “considers the requirement to supply Goodman Fielder remains important for public confidence in downstream wholesale and retail markets.”400 7.100 We consider that scaling down Goodman Fielder's regulatory access to raw milk may contribute toward a functioning factory gate market and a future sufficiency of competition. Scaling down Goodman Fielder's entitlement may facilitate efficient demand and supply in the factory gate market 7.101 An option for a transition pathway to deregulating the Goodman Fielder raw milk provisions could include: 7.101.1 reducing Goodman Fielder's access allowance to its most recent volumes, or somewhat below those. This may provide sufficient incentive for Goodman Fielder to consider its alternative options for own-supply and 398 The creation of DIRA required that Fonterra divest shares in NZDF (New Zealand Dairy Foods) The NZDF business manufactured a number of well-known New Zealand consumer dairy brands and the sale made sure that Fonterra would not monopolise the downstream dairy market in New Zealand. NZDF is now the dairy division of Goodman Fielder New Zealand Limited. 399 The original contract was with NZDL (now part of Goodman Fielder New Zealand Limited). Limit to supply for Goodman Fielder can be found in the Dairy Industry Restructuring (Raw Milk) Regulations 2012 s1(7) Details of contract end date were sourced from Goodman Fielder “Commerce Commission Consultation Paper: Review of the state of competition in the New Zealand dairy industry” 10 July 2015, para 3.1 . We understand the contract has the following features: [ ] 400 Fonterra "Review of the state of competition in the New Zealand dairy industry: Consultation paper— process and approach" 17 August 2015, para 17. 2440192 157 factory gate supply outside the Raw Milk Regulations.401 This option likely mitigates potential undue risk on the downstream domestic markets; and 7.101.2 regional deregulation may be an appropriate option for Goodman Fielder. At least two of its plants (ie, Canterbury and Manawatu) out of its three plants (ie, Northland) appear to have factory gate alternatives to Fonterra.402 Reducing rather than eliminating its entitlement. This approach would afford Goodman Fielder protection in those areas where Fonterra is the only option. A good transition time and certainty will likely be important for Goodman Fielder to adjust its source and terms of supply.403 7.102 We note that Goodman Fielder is not currently taking its full 250 million litres entitlement of regulated milk.404 In the 2014/15 season, Goodman Fielder took [ ]405 million litres of raw milk under the Raw Milk Regulations. 7.103 An advantage of our options for transition pathways to deregulating the Goodman Fielder raw milk provisions may be in spurring a factory gate market outside of the DIRA earlier on account of Goodman Fielder’s large volume requirements.406 7.104 A disadvantage of our options for transition pathways to deregulating the Goodman Fielder raw milk provisions could be the risk that Fonterra will exercise market power [ ].407 However, with only a scaling back of Goodman Fielder's raw milk entitlements the impact of this risk on downstream markets could be mitigated. 7.105 A further potential disadvantage of deregulating Goodman Fielder’s raw milk entitlements is the risk of Fonterra attempting to foreclose Goodman Fielder in 401 Goodman Fielder’s contracted supply of raw milk from Fonterra ends in 2021. Goodman Fielder “Commerce Commission Consultation Paper: Review of the state of competition in the New Zealand dairy industry” 10 July 2015, para 3.1. 402 We note that in the 2014/15 season Goodman Fielder processes approximately [ ] of its raw milk in Canterbury, [ ] in Manawatu, and [ ] in Northland, respectively. Goodman Fielder has potential options for factory gate raw milk in 1) Canterbury: Westland, Synlait and Oceania, 2) Manawatu: Miraka and Open Country. 403 We note the likely increase in factory gate and downstream prices. 404 Goodman Fielder “Commerce Commission Consultation Paper: Review of the state of competition in the New Zealand dairy industry” 10 July 2015, para 3.7. 405 Fonterra “Information request by the Commerce Commission: Final milk volumes 14/15 season.” 22 September 2015, Annex 1A Q6.2. 406 We note a perception by at least one submitter that Goodman Fielder has limited incentive to seek factory gate supply outside of DIRA Regulation. Trever Bain “DIRA and The Commerce Commission on Fairness Within The Dairy Industry” January 2016, para 4. 407 We discussed this in chapter 5, State of competition without the DIRA Regulation. 2440192 158 order to increase market power in downstream markets against its largest domestic rival. In chapter 5, State of competition without the DIRA Regulation, we considered this as unlikely to occur. At this time, we do not suggest repealing Goodman Fielder's entitlements entirely 7.106 In our draft report we outlined an option for a transition pathway to deregulation that included the repeal of Goodman Fielder's regulated access to raw milk. 408 7.107 At this time, we do not consider that repealing Goodman Fielder's entire entitlement is an appropriate option for a transition pathway to deregulation.409 Rather, a more gradual scaling down Goodman Fielder's allowances may facilitate the development of a functioning factory gate market as a transition pathway to deregulation. 7.108 Goodman Fielder submitted that increased factory gate competition should be facilitated. It considers that its regulated supply of 250 million litres of raw milk should continue until a viable alternative source of year round milk supply develops.410 In particular, it stated: 7.108.1 there are no alternative suppliers that can meet its requirements and that this is unlikely to change before it renegotiates its contract with Fonterra in 2021; 7.108.2 its purchases from alternative suppliers such as Westland or Synlait are “typically negligible (if any)”.411 It considers that “it is clear from past experience that IPs will only supply [it] in the event that they have collected raw milk in excess of their own requirements”;412 and 408 Commerce Commission "Review of the state of competition in the New Zealand dairy industry draft report" 6 November 2015, para 7.77.1. 409 Notwithstanding the fact that Goodman Fielder’s contracted supply of raw milk from Fonterra ends in 2021, which would likely offer some time for Goodman Fielder to assess its own-supply and factory gate options outside the DIRA Regulation. 410 Goodman Fielder “Cross-submission on the Dairy Competition Review Draft Report” 18 December 2015, para 1.3. 411 Goodman Fielder “Cross-submission on the Dairy Competition Review Draft Report” 18 December 2015, para 3.1(a). 412 Goodman Fielder “Cross-submission on the Dairy Competition Review Draft Report” 18 December 2015, para 3.2. 2440192 159 7.108.3 removing or reducing its entitlement to raw milk “would have a severe impact on Goodman Fielder’s business and would impact on its ability to compete against Fonterra downstream”.413 Our conclusions on options for Goodman Fielder’s entitlements 7.109 We consider that a scaling back of Goodman Fielder's entitlement to raw milk under the DIRA may facilitate the development of a functioning factory gate market. This will allow Goodman Fielder to source most of its requirements through DIRA with only a portion of its supply being non-DIRA milk. 7.110 Market dynamics will then determine who will supply, and the price and volumes, of any non-DIRA milk Goodman Fielder purchases. It may be that Fonterra still supplies these volumes, but at a different price. If this price is above the competitive price then this could induce entry from other IPs best placed to supply the factory gate market. 7.111 We consider that while Goodman Fielder can rely on DIRA milk for all of its needs it has little incentive to seek supply from outside of the Raw Milk Regulations. 7.112 In its submission, Goodman Fielder noted “[ ]”.414 This suggests there are IPs interested in supplying Goodman Fielder at the factory gate. 7.113 [ ]415 The open entry and exit provisions 7.114 The open entry and exit provisions of the DIRA Regulation ease farmer switching and as such potentially facilitate the enhancement of competition in both the farm gate and factory gate markets.416 413 Goodman Fielder “Cross-submission on the Dairy Competition Review Draft Report” 18 December 2015, para 3.4. 414 Goodman Fielder “Cross-submission on the Dairy Competition Review Draft Report - Confidential Version” 18 December 2015, p.2-3, paragraph 3.3. 415 [ 416 We note the general support provided by Mr. Willans and Mr. Borst behind our draft conclusions on the Open Entry and Exit provisions. Richard Willans and Robert Borst “Cross submission on the review of the state of competition in the New Zealand Dairy Industry Draft Report dated 6 November 2015” 22 January 2016. 2440192 ] 160 7.115 We note that the open entry and exit provisions may make it easier for current IPs that are wholly dependent on DIRA milk to attract farmer suppliers to own-source their raw milk. This may be an important consideration if the Raw Milk Regulations are amended to reduce DIRA entitlements. Entry to the farm gate market by IPs obtaining their own supply may further facilitate the development of a functioning factory gate market. 7.116 We found little evidence that the open entry and exit provisions contribute materially to costs on Fonterra or to inefficiency more generally in the domestic dairy markets. 7.117 In our draft report we requested further evidence on the efficiency grounds for relaxing these provisions, in particular in relation to open entry.417 We have considered the efficiency impacts of the open entry and exit provisions in chapter 6, Balancing efficiencies and inefficiencies of regulations. Open entry for large new conversions could contribute to asset stranding risk 7.118 In chapter 6, our assessment of the efficiency costs and benefits of the DIRA Regulation, we considered the asset stranding risks created by the open entry and exit provisions. 7.119 We questioned the materiality of the asset stranding risk caused by the open entry and exit provisions of the DIRA Regulation and conclude that at this time this stranding risk is likely to be more of a risk in theory than in practice. 7.120 The strong milk growth in New Zealand has reduced the risk to Fonterra of asset stranding as farmers exit the co-operative. We have received little evidence that the scale of asset stranding risk is a material concern, and we note that Fonterra’s internal documents provided to us have not identified this as a significant risk. New Zealand milk growth is forecast to be somewhat weaker over the foreseeable future, which may make asset stranding somewhat more likely. But we consider that even with the lower forecast growth that is expected, the risk of stranding of Fonterra’s investments will remain low. Options that may mitigate asset stranding risk 7.121 417 We considered whether it would be possible to mitigate asset stranding risk during the transition, without significantly undermining the competition benefits arising from the open entry and exit provisions as outlined in chapter 5, Concerns about deregulation. Commerce Commission "Review on the state of competition in the New Zealand dairy industry draft report" 6 November 2015, paragraph 7.80. 2440192 161 7.122 An option would be to allow Fonterra discretion in accepting new farm conversions as shareholding suppliers.418 The provisions that allow exit and re-entry to Fonterra facilitate efficient switching, but open entry for new farm conversions does not facilitate competition (IPs and Fonterra can already compete for new conversions). In particular: 7.122.1 The open entry and exit provisions as applied to new farm conversions likely provide fewer efficiency benefits at the farm gate but contribute to Fonterra’s asset stranding risk because Fonterra needs to invest without having any security of ongoing supply. The stranding risk may be significant for very large conversions. 7.122.2 Allowing Fonterra the ability to sign large conversions to longer term contracts may mitigate asset stranding risk through protecting Fonterra’s investments, but will not prevent existing large shareholding suppliers switching to or from Fonterra. 7.122.3 For existing suppliers (of all characteristics) the efficiency benefits of facilitating efficient switching need to be weighed against the efficiency costs (such as asset stranding risk). Our analysis suggests that at this time the open entry and exit provisions generally provide net efficiency benefits and should be retained. 7.123 Fonterra supports relaxation of the open entry requirements for new conversions. We note that relaxing the open entry provisions as proposed by Fonterra may raise practical issues as to what constitutes a new farm conversion or new entrant.419 Considering amendments to the transport costs exceptions to open entry may contribute toward mitigating any potential inefficiency caused by the open entry and exit provisions.420 Amend open entry and exit provisions for large IPs 7.124 Another option is to amend the open entry and exit provisions for large suppliers in order to reduce asset stranding risk. In particular: 418 While Fonterra may have discretion to refuse a new conversion for the alternative of contracted supply (the new conversions have not yet sunk capital in converting and therefore farmers’ assets would not be stranded by a change in the rules) once a new conversion signs with a competing IP it would then have open entry rights into Fonterra under DIRA (ie, it is no longer a new conversion). 419 Fonterra "Assessment of Competition in Raw Milk Markets and Costs and Benefits of the DIRA provisions, Public Version" 17 August 2015, paragraphs 53–62. And Fonterra “Submission on draft report: Review of the state of competition in the New Zealand dairy industry” 4 December 2015, paragraph 51. 420 Dairy Industry Restructuring Act 2001, s 95. 2440192 162 7.124.1 Given Fonterra’s strategy of milk growth,421 asset stranding risk is likely to be created by cornerstone suppliers leaving Fonterra. The open entry and exit provisions could provide for longer notice periods for larger suppliers only. 7.124.2 We consider that Fonterra may be able to amend its constitution and change the standard terms and conditions to differentiate between suppliers based on volumes. Fonterra could likely achieve this without breaching the 33% rule.422 7.125 We note that there may be difficulty in defining a large supplier. Possible opportunity to amend the 20% rule and reduce regulatory burden 7.126 The 20% rule is used by many small IPs, particularly artesian cheese makers. It is uncertain how effective this rule is going to be in the future; as Fonterra submitted it was not aware the 20% rule was being used so widely.423 7.127 We consider the 20% rule is potentially important in creating a collective constraint on Fonterra in downstream domestic markets. 7.128 We further consider that the 20% rule is still important for small IPs to access milk. It provides an important pathway for them to enter the market and to grow.424 Until a functioning factory gate market has developed, we consider that the 20% rule should be maintained, potentially with some amendments. 7.129 The 20% rule requires the use of separate vats. The significant cost of installing a second vat, may act as a barrier to supply under the 20% rule. The Minister might consider whether this rule could be amended to reduce this regulatory burden. 421 See Fonterra’s strategy: https://www.fonterra.com/global/en/about/our+strategy. 422 Dairy Industry Restructuring Act 2001, s 107. The highest percentage of total supply Fonterra has contracted in an IP’s collection zone is [ ] percent, based on Fonterra’s May 2015 analysis for the 2015/16 season. NERA on behalf of Fonterra "Assessment of Competition in Raw Milk Markets and Costs and Benefits of the DIRA provisions, Public Version" 17 August 2015, p.6. 423 Fonterra "Cross-submission on review of the state of competition in the New Zealand dairy industry" 31 August 2015, para 31. 424 Karikaas considers that the 20% rule is an important pathway for cheesemakers to become established and grow. Karikaas "Submission on the dairy competition review substantive issues” 17 August 2015, page 2. Grinning Gecko and Over the Moon Dairy submit they would not be able to access any milk without the 20% rule. Grinning Gecko “Submission on the dairy competition review paper 20 July 2015” 17 August 2015 and Over the Moon Dairy “Submission on the dairy competition review consultation paper, 20 July 2015” 17 August 2015. 2440192 163 7.130 Amending the 20% rule would have to balance any legitimate concern Fonterra had about the IPs in accessing their vats and any food safety concerns. These concerns could be considered alongside the benefit of downstream competition. 7.131 We have not obtained sufficient information on the practicalities of the 20% rule to outline any other potential improvements apart from highlighting it as an area of future potential work. 7.132 We note that in the absence of DIRA Regulation, Fonterra's Board would have discretion without amending the constitution to set the conditions under which its suppliers can split their supply of raw milk. The base milk price disclosure and monitoring rules 7.133 It is likely that some form of independent oversight of the base milk price setting arrangements would occur in the absence of DIRA Regulation, in order to provide assurance to Fonterra’s farmer shareholders and non-supplier investors that the farm gate milk price is set at an efficient level. 7.134 It appears the Commission base milk price reviews are valued by industry stakeholders425 and those parties would be less comfortable with a process controlled by Fonterra.426 7.135 We have received a number of submissions asking for the milk price setting process to be changed or parts it off handed over to the Commission427 and for the efficiency standard employed in the Milk Price Manual to be changed, and that the price is too high.428 We note that these suggested changes to the milk price setting regulations are outside the scope of this report. 425 Open Country Dairy “Review of the state of competition in the New Zealand dairy industry: Consultation paper—process and approach" 10 July 2015, p.4. 426 Tatua "Consultation on substantive issues—review of the state of competition in the New Zealand dairy industry" 17 August 2015, paragraphs 3.12–3.15. 427 For example: Castalia Strategic Advisors for Open Country Dairy "Review of the State of Competition in the New Zealand Dairy Industry: Comments following submissions on substantive issues" 31 August 2015, section 3. 428 For example: see Tatua "Consultation on substantive issues—review of the state of competition in the New Zealand dairy industry" 17 August 2015, para 3.18, and Miraka "Review of the state of competition in the New Zealand dairy industry: Consultation paper—process and approach" 17 August 2015, para 2.1– 2.20.1. 2440192 164 Attachment A: Terms of reference for a report on the state of competition in the New Zealand dairy industry Purpose of the terms of reference A1 A2 The Dairy Industry Restructuring Act 2001 (DIRA) requires the Minister of Primary Industries, in consultation with the Minister of Commerce, to request a report on the state of competition in the New Zealand dairy industry: A1.1 when market share thresholds are met; or A1.2 in the event that they have not been met by 1 June 2015, as soon as practicable after that date. The report must provide an assessment of the state of competition in the dairy industry and advise the Minister as to whether the market share thresholds should be reset, and options for a pathway to deregulation (if any). Context A3 The creation of Fonterra resulted in a lessening of competition in the New Zealand dairy industry. As a result, DIRA imposed regulatory requirements on Fonterra to promote contestability and efficiency in New Zealand dairy markets. Specifically, s4(f) of DIRA stipulates the purpose of the DIRA regulatory requirements is to “promote the efficient operation of dairy markets in New Zealand by regulating the activities of Fonterra to ensure New Zealand markets for dairy goods and services are contestable”. A4 The DIRA regulatory requirements are aimed at the following two dairy markets: A5 A4.1 The market for milk at the farm gate, which is a market for the purchase of raw milk from farmers. Dairy processors compete with each other to purchase raw milk from farmers and this rivalry is beneficial to dairy farmers. A4.2 The factory gate milk market, which is a wholesale market where dairy processors trade raw milk among themselves prior to processing. There is demand at the wholesale level for unprocessed raw milk, from: A4.2.1 dairy processors who are in the early stages of their operations looking to supplement their own inputs of raw milk to improve the capacity utilisation of their processing plants; and A4.2.2 food and beverage manufacturers, which require unprocessed raw milk as an input for their manufacturing processes. To ensure the contestability of the farm gate milk market, DIRA contains ‘open entry and exit’ provisions, as well as the farm gate milk price monitoring regime (as set out in subparts 5 and 5A of Part 2 of DIRA). The requirement for Fonterra to on-sell a 2440192 165 small proportion of its raw milk to other dairy processors (as specified in the Dairy Industry Restructuring (Raw Milk) Regulations of 2012 [referred to in this report as ‘Raw Milk Regulations’] further promotes the contestability of the farm gate milk market by providing newly established dairy processors with temporary access to wholesale raw milk supply. The Raw Milk Regulations also provide access to wholesale raw milk for food and beverage manufacturers that choose not to have access to their own farm gate milk supply. A6 Critically, the DIRA regulatory requirements were designed to be transitional in nature. The regulatory regime is designed to expire if and when the New Zealand dairy markets are considered to be sufficiently competitive, so that industry-specific regulation is no longer required over and above the normal application of general competition law. A7 The process to expire these requirements begins when IPs have collected 20% or more of milk solids in either the North or the South Island of New Zealand in a dairy season. If this market share threshold is not met by 1 June 2015, the DIRA sets out a process to ascertain the state of competition in the New Zealand dairy industry (pursuant to section 148A). A8 The first step in the process requires a report to be produced on the state of competition in the New Zealand dairy industry. This document sets out the requirements for that report. Objectives and scope A9 The provider of the report (the provider) must ascertain: A9.1 the state of competition that exists in the New Zealand dairy industry; and A9.2 if the provider determines that the state of competition in the New Zealand dairy industry or any specified New Zealand dairy market or markets is insufficient,— (A) whether either or both of the market share thresholds specified in section 147 should be reset; and (B) the options for a transition pathway to deregulation (if any) and whether a particular option or a set of options (if any) should be pursued. A10 In addressing the objectives above, the report should address the following questions: A10.1 What is the state of competition that exists in the relevant New Zealand dairy markets, in particular, in the farm gate and factory gate markets? In relation to the wholesale and retail dairy markets, is there any new information that would alter the conclusions reached in the Commerce Commission’s preliminary inquiry into domestic milk markets in 2011 under Part 4 of the Commerce Act 1986? A10.2 Is the current state of competition in the relevant New Zealand dairy markets sufficient to ensure the efficient and contestable operation of these 2440192 166 markets in the absence of the provisions of subparts 5 and 5A of Part 2 of the DIRA and/or the provisions of the Raw Milk Regulations? In particular, would Fonterra (given its market shares, nationwide collection and processing networks, and ownership and capital structure arrangements) be likely to have both the ability and the incentives to exercise market power against competitors, suppliers or purchasers in one or more of the relevant New Zealand dairy markets (in the North and/or the South island), in the absence of the provisions of subparts 5 and 5A of Part 2 of the DIRA and/or the provisions of the Raw Milk Regulations? A10.3 If the research determines that the current state of competition in the relevant New Zealand dairy markets is insufficient in one or more of the relevant New Zealand dairy markets, so far as the available information allows the provider of the report to reach a conclusion: A10.3.1 Should either or both of the market share thresholds specified in s 147 of the DIRA be reset (and, if so, to what new level(s))? A10.3.2 Are there other expiry triggers that should be provided for, either in addition to or in place of the market share thresholds? A10.3.3 What options, if any, are there for a transition pathway to deregulation? A10.3.4 Are there any options (or sets of options) for a transition pathway to deregulation that should be pursued? Stakeholder consultation A11 The provider of the report must consult at least once with the following on each of Questions 1-3 above before providing its final report: A11.1 New Zealand dairy farmers through the following representative groups; A12 A11.1.1 DairyNZ A11.1.2 Federated Farmers A11.1.3 Fonterra; and A11.1.4 IPs. The method of consultation may include: A12.1 Providing a draft report for written submissions; A12.2 Facilitating a workshop, following which participants may provide written submissions. 2440192 167 Attachment B: Summary of the key provisions of the DIRA Regulation Regulations under subpart 5 of Part 2 that assist farmer switching and access to raw milk by IPs at the farm gate Section 73 Subject to a limited number of exceptions, Fonterra must accept an application by a new entrant to become a shareholding farmer and must also accept applications by shareholding farmers to increase volumes of milk 429, 430 supplied. Section 97 A shareholding farmer who wants to cease or reduce their supply of milk as a shareholding farmer to Fonterra may give a notice of withdrawal. Section 106 Fonterra may not discriminate in the terms of supply between its suppliers (new entrants and shareholding farmers), and the terms of supply that apply to a new entrant must be the same as the terms that apply to a shareholding farmer in the same circumstance. Fonterra must also not treat a shareholding farmer who exercises an entitlement under subpart 5 of the DIRA any less favourably than a shareholding farmer who does not do so. Section 109 A shareholding farmer who withdraws totally from Fonterra may require Fonterra to sell a milk vat situated on the withdrawing shareholding farmer’s farm to the shareholding farmer or an IP. 429 Section 5 of the DIRA defines a new entrant as “a dairy farmer who is not a shareholding farmer who applies to become a shareholding farmer under section 73”. 430 Section 94 of the DIRA provides that Fonterra may reject an application by a new entrant or a shareholding farmer if the supply of milk solids obtainable from milk to be supplied by the applicant in a season is less than 10 000 kilograms. Section 95 of the DIRA further provides that Fonterra may reject an application by a new entrant if the cost of transporting the milk of the new entrant exceeds the highest cost of transporting another shareholding farmer’s milk. 2440192 168 Section 107 Fonterra may only offer new entrants and shareholding farmers contracts for milk supply for more than one season if it ensures that, at all times 33% or more of the milk solids produced within a 160km radius of any point in New Zealand is: supplied under contracts with IPs; or supplied under contracts with Fonterra that expire or may be terminated by the supplier at the end of the current season without penalty to the supplier; and on expiry or termination, end all the supplier’s obligations to supply milk to Fonterra. Section 108 Fonterra shareholding farmers are entitled to allocate up to 20% of their weekly production to IPs throughout the season; this may not be a higher percentage of their weekly production than their average weekly allocation to 431 IPs in October. 431 A shareholding farmer who exercises this entitlement must give Fonterra 20 working days’ notice of the arrangements for the collection of this milk and may require storage in separate milk vats. 2440192 169 The base milk price regime under subpart 5A of Part 2 which promotes contestability and efficiency at the farm gate Sections 150A to 150C Fonterra’s base milk price setting regime sets out the purpose of subpart 5A, which is to promote the setting of a base milk price that provides an incentive for Fonterra to operate efficiently while providing for contestability in 432 the market for the purchase of milk from farmers (section 150A). Fonterra can choose to use certain assumptions in setting the base milk price (section 150B). Fonterra must set the base milk price in a way that is consistent with certain principles (section 150C). Sections 150D to 150G and 150N Fonterra must establish a Milk Price Panel (panel) and maintain a Milk Price Manual (manual). Fonterra must establish a panel, where the majority of its members and the chair are independent, and set the publicly available terms of reference for the panel (sections 150D and 150E). Fonterra must maintain a manual that sets out how the milk price is calculated, and make it public, including any amendments (section 150F). For each season, the panel must: supervise the calculation of the base milk price; advise Fonterra on the application of the manual; and recommend the base milk price to Fonterra (section 150D). If Fonterra does not accept a recommendation of the panel, or amends the manual without a recommendation, it must make its reasons for doing so publically available (section 150G). If Fonterra does not accept the recommendation of the panel when setting the base milk price or sets the base milk price without a recommendation by the panel, it must make its reasons for doing so publically available 433 (section 150N). Sections 150H to 150M The Commission must review the manual and report on the extent to which the manual is consistent with the subpart 5A purpose, namely “to promote the setting of a base milk price that provides an incentive to new co-op to operate efficiently while providing for contestability in the market for the purchase of milk from farmers.” Sections 150O to 150U The Commission must review Fonterra’s base milk price calculation and report on the extent to which the assumptions adopted and the inputs and process used in the milk price calculation are consistent with the 432 Under section 150A(2) of the DIRA, the setting of a base milk price provides for contestability if any notional costs, revenues or other assumptions taken into account in calculating the base milk price are practically feasible for an efficient processor. 433 The effect of these provisions is limited as Fonterra is free to deviate from the milk price produced by the calculation under the manual. 2440192 170 subpart 5A purpose, namely “to promote the setting of a base milk price that provides an incentive to new co-op 434 to operate efficiently while providing for contestability in the market for the purchase of milk from farmers.” Raw Milk Regulations that promote access to raw milk by IPs at the factory gate 2012 Raw Milk Regulations Fonterra is required to:  supply IPs (including Goodman Fielder) with up to 795 million litres per season (regulation 4 and 5(1));  supply each IP (other than Goodman Fielder) with up to 50 million litres per season, subject to both 435 maximum monthly limits and, for months other than October, 110% of the amount taken in October (regulation 6(1));  supply Goodman Fielder with up to 250 million litres per season, subject to a monthly limit in August, September, November, December, January, February, March, April, or May, of 110% of the amount taken in October (regulation 7(1));  offer to supply the raw milk required by the IP (other than Goodman Fielder) for the whole of the current season at the most recent forecast of the farm gate milk price for that season (fixed quarterly price) subject 436 to that IP having own supplied less than 30 million litres in the previous season (regulation 19); and  supply an IP (other than Goodman Fielder) whose own supply of raw milk exceeded 30 million litres in the 437 previous season at the default milk price (regulation 20). Under regulation 6(3) Fonterra does not have an obligation to supply an IP (other than Goodman Fielder) in a season beginning on or after 1 June 2016, if the IP’s own supply of raw milk in each of the three consecutive previous seasons was 30 million litres or more. Fonterra may also require that a contract for supply includes terms that are reasonable having regard to industry practice before the commencement of the regulations, but this may not include a “take or pay requirement” (regulation 21). 434 The DIRA limits the extent of our review as section 150P(3) states that we must not state the amount of the base milk price according to our own calculations. We are further not required to calculate the costs of an IP. 435 The maximum monthly limits are: August, 3 million litres; September, 6 million litres; October, 7 million litres; November, 7 million litres; December, 6 million litres; January, 6 million litres; February, 5 million litres; March, 4 million litres; April, 4 million litres; May, 2 million litres. 436 Under regulation 19(4) the most recent forecast of the farm gate milk price means the forecast published under regulation 24(1)(c) plus the reasonable cost of transporting the raw milk to the IP, plus an additional amount for winter milk (milk supplied in June or July) and/or organic milk. Regulation 3(1) defines own supply as “raw milk collected from dairy farmers by or on behalf of an independent processor”. 437 Under regulation 20(2) the IP and Fonterra may also agree on the price for the supply of the raw milk. Under regulation 20(1) an IP whose own supply of raw milk was less than 30 million litres in the previous season may also choose to obtain raw milk at the default price rather than at the fixed quarterly price. Under regulation 20(4) the default milk price means the farm gate milk price for that season plus the reasonable cost of transporting the raw milk to the independent processor, plus an additional amount for winter milk and/or organic milk. 2440192 171 Attachment C: 2012 Raw Milk Regulations amendments Purpose of this attachment C1 The purpose of this attachment is to: C1.1 outline what has happened in the dairy markets as a result of the MAF 2012 amendments to the Raw Milk Regulations; and C1.2 comment on the how the amendments have performed. The 2012 DIRA amendments focused on the Raw Milk Regulations C2 Table C1 summarises the amendments to the Raw Milk Regulations and the reasons for amendment. C3 Overall, we consider these amendments have improved efficiency through reducing dependence on the regulations, improvements in administering the pricing of regulated milk, and ensuring IPs face the reality of the New Zealand milk curve when making investment and production decisions under the regulations. Table C1: The 2012 DIRA amendments to the Raw Milk Regulations Regulation as at 2001 Amendment to the regulation Reason for amendment All IPs, including those with their own established farmer supply, would have access to regulated milk for as long the regulations are in place. A processor would be ineligible for DIRA milk if it collected a minimum of 30 million litres of milk for three consecutive seasons. They are considered to have their own farmer supply. To create stronger incentives for IPs to grow the amount of their own farmer supply sooner. 600 million litre cap on DIRA milk supplied by Fonterra. The maximum quantity of raw milk that Fonterra must supply to all IPs is 795 million litres per season, which is equivalent to 5% of the total raw milk collected by Fonterra To provide greater flexibility, allowing the total regulated milk quantity to move in line with the total quantity of milk Fonterra collects. Ensure that access to regulated milk is targeted to IPs seeking an entrance pathway into the farm gate market and those with minimal or none of their own farmer supply, rather than established processors with their own-supply. The 5% gap is based on the average quantity of milk Fonterra received over the previous three seasons and would be reviewed at a minimum every three seasons to ensure that it continues to represent approximately 5% of Fonterra’s total milk supply. Goodman Fielder entitled to up to 250 million litres per season (with seasonal restrictions on winter milk and does not include the above cap). October rule: A processor can only 2440192 Monthly maximum volume limits to reflect the seasonal supply curve and Conditions more closely reflective of those in a competitive farm gate market. 172 take up to 110% of what they took in October. retention the October rule. Does not disadvantage processors who need to be able to flatten the curve (such as town milk suppliers) as the quantities generally fall below the monthly limits. To mitigate risk that a processor would use the regulations purely to access milk in the shoulder months and not any other month. Price for DIRA milk was Fonterra’s farm gate milk price + 10 cents per kgMS. Final farm gate milk price for the season (default milk price). The DIRA price was amended to ‘Fonterra’s farm gate milk price’ plus reasonable costs of transport, winter milk and organic milk with the introduction of the monthly quantity limits (implementation of introducing monthly limits to reflect the seasonal supply curve and retaining the October rule). The additional 10 cents is no longer required as the extra cost was to Fonterra providing IPs with the option of ‘flattening’ the profile of the regulated milk. Introduction of quarterly pricing as an option for DIRA milk pricing. Fonterra required to offer to sell regulated milk at Fonterra’s most recent quarterly farm gate milk price forecast. This option is only available to those processors without their own significant milk supply. To provide more price certainty to those IPs that need it the most. With the introduction of monthly limits, IPs that take up large quantities can no longer flatten its profile. What has happened since the restriction and monthly limit amendments C4 DIRA milk sales have declined since 2011/12 (when the three consecutive season sunset clause amendment was implemented). IPs have anticipated that they will no longer have access to DIRA milk and so have sought to increase their own farmer supply.438 C5 [ ]439 C6 ]440 [ C6.1 [ C6.2 [ C6.3 [ ] ]; ]; 438 As outlined in chapter 4, State of competition under the DIRA Regulation. 439 [ 440 [ 2440192 ] ] 173 C6.4 [ ]; and C6.5 [ ]. C7 The amendments (particularly the restriction on who can access DIRA milk and monthly limits) have resulted in IPs increasing their own farmer supply and decreasing Fonterra’s farm gate market share (which increases competition in the farm gate market). C8 By retaining the October rule, Fonterra may not incur additional costs for providing raw milk to processors who take DIRA milk in the shoulder months and not during the peak. C9 [ ]441 C10 441 [ 442 [ Because the fixed quarterly pricing is only available to processors that have less than 30 million litres of their own-supply, larger IPs442 have purchased DIRA milk using the default price. 2440192 ] ] 174 Attachment D: Market structure Purpose of this attachment D1 This attachment provides information about the current structure of the dairy markets and the structural changes that have occurred since Fonterra was established and MAF’s 2010 review of the DIRA Regulation expiry thresholds. D2 We identified three broad trends in respect of market structure that are relevant to competition in the farm gate and factory gate markets: D3 443 D2.1 milk production has been growing and we expect this to grow further, although at a slower rate; D2.2 we understand that Fonterra is growing at a slower rate than rival IPs but has still claimed over two thirds of all new milk growth; and D2.3 although there has been significant entry and expansion of IPs, both since Fonterra was established and since MAF’s 2010 DIRA Regulation review,443 the markets are still highly concentrated and Fonterra remains a monopsony purchaser (it is the only buyer) of raw milk at the farm gate, and a monopoly seller of raw milk at the factory gate in many regions. We present the evidence behind each of these trends in the context of the following wider developments: D3.1 milk production and growth; D3.2 sources of Fonterra’s milk; D3.3 market shares and processing capacities of Fonterra and IPs; D3.4 IP growth and entry over time; D3.5 other co-operatives; D3.6 other IPs; and D3.7 own-source IPs subject to the sunset clause. MAF “Extension of the Pro-Competition Provisions of the Dairy Industry Restructuring Act 2001: Regulatory Impact Statement” July 2010. 2440192 175 Milk production and growth D4 As of December 2014, New Zealand is the ninth largest milk producer globally, but is the world’s largest exporter of dairy products.444 D5 Milk production in New Zealand is reflective of the seasonal pattern of pasture growth. D6 Typically, the shoulder periods occur in September to October and December to March with peak production occurring in late October to late November. D7 Figure D1 shows milk production for the last four seasons.445 Figure D1: Milk production, 2011/12–2014/15 (million kg) 300 250 200 150 2011/12 2012/13 100 2013/14 2014/15 50 0 Source: Dairy Companies Association of New Zealand. Milk growth has occurred through increased production and conversions D8 Milk volumes have been trending upwards. This upwards trend is due to increased farm productivity due to more efficient land use and feed being used to supplement pasture growth.446 It is also due to an upward trend in farm conversions. Figure D2 shows the increase in production since Fonterra was created in 2001. D9 Total milk production has increased by around 12% since the 2011/12 season. 444 DairyNZ “QuickStats about dairying—New Zealand” December 2014 (Viewed 4 November 2015). 445 The New Zealand dairy season runs from 1 June to 31 May. 446 DairyNZ “New Zealand Dairy Statistics 2013-2014” < http://www.dairynz.co.nz/publications/dairyindustry/new-zealand-dairy-statistics-2013-14/> (Viewed 4 November 2015). 2440192 176 Figure D2: Milk processed nationally and by island, 2001/02–2013/14 (million kg) Source: New Zealand Dairy Statistics 2013-2014 LIC and Dairy NZ. D10 Milk growth has been much higher in the South Island than in the North Island, see figures D3 and D4. Figure D3: Milk processed in the South Island (million kg) Source: New Zealand Dairy Statistics 2013–2014 LIC and Dairy NZ. 2440192 177 Figure D4: Milk processed in the North Island (million kg) Source: New Zealand Dairy Statistics 2013–2014 LIC and Dairy NZ. D11 Most of the growth in the South Island has occurred in Canterbury and Southland and has been driven by both dairy conversions and farm productivity growth. D12 Most of the growth in the North Island has taken place in the Waikato region. This growth has largely been driven by higher milk output per cow rather than by dairy conversions. Expectations for milk growth in the future D13 There is an expectation that milk volumes will grow overall over the next 10 years, but that the rate of growth is likely to slow. D14 In the short term, milk production may slow owing to low milk prices as farmers cut back on supplementary feed and decrease the size of their herds.447 The Ministry for Primary Industries expects milk solids production to fall in the 2015/16 season as lower prices lead to cutbacks in supplementary feeding.448 Over the longer term, 447 More than one million cows were expected to be culled in the 2014/15 season owing to the low milk solids pay-out. New Zealand Herald (2005) “Record cattle cull after low payout” (viewed 15 September 2015). 448 Ministry for Primary Industries “Situation and Outlook for Primary Industries” June 2015, p.20. 2440192 178 Fonterra expects milk growth to continue. MPI expects the national herd size to continue growing to 2019.449 D15 There is also potential for further conversions in the South Canterbury/North Otago region because irrigation schemes are being developed. The Central Plains Water Enhancement Scheme is a large-scale proposal to supply water for irrigation to an area of 60,000 hectares on the Canterbury Plains due for completion in 2019. Construction started on the scheme in 2014; phase 1 will distribute water to 120 farms or about 20,000 hectares.450 These irrigation schemes are expected to drive conversions. Fonterra’s growth over time D16 Fonterra collects and processes a high percentage of the raw milk produced in New Zealand. D17 Fonterra is the world’s second largest milk processor and remains the largest processor and milk collector in New Zealand. Fonterra collected [ ] of the raw milk in New Zealand in the 2014/15 season. This amounted to approximately [ ].451 D18 Fonterra processes the milk it collects at 33 sites across New Zealand.452 Figure D5 shows Fonterra’s processing sites and the different products they produce. 449 Ministry for Primary Industries “Situation and Outlook for Primary Industries” June 2015, p.20. 450 For example, http://www.3news.co.nz/nznews/canterbury-irrigation-project-making-progress2014110418#ixzz3mQOLy7rg. 451 [ 452 Fonterra “Fonterra in New Zealand: Sites” (viewed February 2016). 2440192 ] 179 Figure D5: Fonterra’s New Zealand processing sites [] Source: Fonterra presentation to Commerce Commission.453 D19 Fonterra’s peak volume processing capacity has increased over the last three seasons. Fonterra processed just over [ ] at its peak in the 2014/15 454 season. Figure D6 shows Fonterra’s peak volume processing capacity over the last three seasons. Figure D6: Fonterra’s peak volume supply, 2012/13–2014/15 Source: Fonterra “Sustainable Value Generation–New Zealand–drive efficiency and optionality”.455 D20 Fonterra expects to add [ [ ]:456 ] of processing capacity between D20.1 2015 saw the completion of a [ addition of [ ] at Pahiatua, and the ]. 453 [ 454 [ 455 Fonterra – Europe and US Roadshow June 2015 “Sustainable Value Generation: New Zealand – drive efficiency and optionality” < https://www2.fonterra.com/files/docs/europe-and-us-roadshow-june2015.pdf> p.13. 456 [ 2440192 ] ] ] 180 D20.2 [ ] D20.3 Fonterra has plans [ ]. Sources of Fonterra’s milk D21 Fonterra is a co-operative, owned by its shareholder suppliers. Almost all Fonterra’s farmer suppliers are shareholders. D22 Fonterra requires its shareholder suppliers to invest in Fonterra by purchasing shares that reflect the volumes of milk that they supplied (equal to the kgMS supplied to Fonterra). These shares are known as ‘wet shares’. We understand these requirements have softened so that farmers now have three years to adjust their shares if they increase output. D23 Fonterra’s constitution allows 15% of total milks solids supplied by farmer shareholders to be on a contract supply basis.457 Only about [ ] of Fonterra’s milk is presently sourced from contracted suppliers.458 Many of these contracted suppliers are in the process of sharing up to Fonterra. Only [ ] of the milk that Fonterra collects has no share backing.459 D24 In 2014, Fonterra introduced MyMilk as an alternative to investing in the cooperative. This scheme is open to farms in the Canterbury, Otago, and Southland regions that are not currently Fonterra shareholders. MyMilk enables farmers who do not want to invest in Fonterra to supply Fonterra for up to five seasons without immediately becoming a Fonterra shareholder. D25 In 2012, Fonterra acquired the dairy processing assets of NZDL after it was placed in receivership. As part of the merger, farmers who supplied New Zealand Dairies were required by Fonterra to take up a Growth Contract.460 457 NERA Economic Consulting for Fonterra "Assessment of Competition in Raw Milk Markets and Costs and Benefits of the DIRA provisions" 17 August 2015, p.24. 458 [ 459 Fonterra sourced [ season and only [ [ 460 Under a Growth Contract, suppliers must purchase an initial 1,000 shares followed by the purchase of sufficient shares to cover one third of their total milk supply at the start of the fifth, sixth and seventh milking seasons. [ ] Fonterra Limited and New Zealand Dairies Limited (in receivership) [2012] NZCC 21. 2440192 ] ] kgMS from shareholding farmers (not all fully share backed yet) in the 2014/15 ] kgMS from suppliers that are not shareholders. ] 181 D26 The proportion of Fonterra’s contract milk supply has increased over time as it has relaxed the requirements to share-up in order to make it easier for farmers to join the co-operative. D27 Fonterra suppliers are also able to invest in Fonterra by purchasing shares that are unrelated to production. These shares are known as ‘dry shares’. D28 In 2012, Fonterra introduced the Trading Among Farmers scheme (TAF). TAF provides farmers who supply milk to Fonterra more flexibility in their investment in Fonterra. It enables farmer shareholders to trade shares among themselves and also gives non-Fonterra suppliers an opportunity to invest in Fonterra through Fonterra’s shareholder fund. Market shares and processing capacities of Fonterra and IPs D29 Fonterra competes at the farm gate for the purchase of raw milk against ‘ownsource’ IPs. D30 While the volume of milk Fonterra collects has increased since 2001, Fonterra is growing at a slower rate than its competitors. Fonterra’s national market share of milk collected has declined since 2001 from around 96% to 85%.461 Figure D7 shows the steady decline in Fonterra’s market share and the corresponding increase in the market share of IPs nationally. Figure D7: Fonterra and IP market shares of milk collected, 2001/02–2014/15 [ ]Source: Commission analysis based on information provided by Fonterra and market participants. D31 Fonterra’s share of the milk collected has declined from around [ ]% to [ ]% in the North Island, and from around [ ]% to [ ]% in the South Island since 2001. Figure D10 shows the decline in Fonterra’s market share of milk collected and the increase in the market share of IPs in the North and South islands. Figure D8: Percentage of IP milk collection North and South islands, 2001/02–2014/15 North Island South Island [ ][ ] Source: Commission analysis based on information provided by Fonterra and market participants. 461 Our analysis based on information provided by Fonterra and market participants. 2440192 182 D32 Figure D8 shows IPs collect proportionately more milk in the South Island than in the North Island. This is because there are a number of large IPs with farmer suppliers in the South Island and because Fonterra does not collect milk from the West Coast. D33 Fonterra’s market share varies within the North and South islands. Table D1 sets out the market shares of Fonterra and the IPs across different regions in New Zealand based on the 2014/15 dairy season. 2440192 183 Table D1: Farm gate market shares (processing capacity and milk collection), by region for the 2014/15 season Northland Auckland/ Waikato/ Bay of Plenty Taranaki Canterbury Otago/ Southland Capacity Milk collection Capacity Milk collection Capacity Milk collection Capacity Milk collection Capacity Milk collection Fonterra [ [ [ [ [ ] [ ] [ [ [ [ Fresha Valley [ ] [ ] [ ] ] ] ] ] ] ] ] ] [ ] Open Country [ ] [ ] Tatua [ ] [ ] Miraka [ ] [ ] Green Valley [ ] [ ] Westland [ ] [ ] Synlait [ ] [ Oceania [ ] [ ] [ ] [ ] [ ] [ ] ] Danone Nutricia Source: Commerce Commission analysis based on information provided by Fonterra and market participants. IP growth and entry over time D34 IP growth and entry since the formation of Fonterra has had various forms: expansion by other co-operatives, and entry and expansion by IPs that, at least to some extent, source milk directly from farmers, and IPs that rely on DIRA milk. Other co-operatives D35 The two dairy co-operatives to remain independent when Fonterra was established in October 2001 are Westland on the West Coast and Tatua in Waikato.462 D36 Westland and Tatua have both grown their businesses by increasingly investing in value-added product lines. Westland has also expanded its milk collection. It now has capacity of [ ] of Fonterra’s total processing capacity.463 [ 462 [ 463 Commission analysis based on information provided by Fonterra and market participants. 2440192 ] 184 ]464 Tatua has [ ]% of Fonterra’s total processing capacity.465 Other IPs D37 A number of IPs have entered the New Zealand dairy market since the establishment of Fonterra. In 2006, the two co-operatives, Westland and Tatua, accounted for 94% of milk supply outside of Fonterra, but by the 2015 season, this was under 30%. 466 D38 Currently [ ] IPs collect more than 30 million litres per season directly from their own farmers and so are subject to the DIRA sunset clause.467 Under the sunset clause in the Raw Milk Regulations, an IP that collects more than 30 million litres from their own farmers for three consecutive seasons will no longer qualify for DIRA milk.468 This leaves [ ] large-scale IPs that do not collect 30 million litres directly from their own farmers and so are not subject to this clause.469 D39 Some small IPs obtain all or part of their milk directly from farmers. The precise total volume of milk collected by small IPs from farmers is unknown but [ ] which is [ ]% of total raw milk collected. Own-source IPs subject to the sunset clause D40 Historically, own-source IPs focused investment into processing commodity products such as WMP. As they have matured, some IPs have invested in more value-add products such as UHT milk and IMF. D41 Open Country was established in 2008 when Dairy Trust Limited acquired Open Country Cheese. It is majority owned by Affco New Zealand Limited, which is owned by Talley’s Group Limited. It is currently the second largest milk processor and manufactures milk powders, milk proteins, milk fats, and cheese. It has manufacturing plants in Waharoa, Whanganui, and Awarua. It has the capacity to process [ ]470. 464 [ 465 Commission analysis based on information provided by Fonterra and market participants. 466 Commission analysis based on information provided by Fonterra and market participants. See also Infometrics (2015) “Why is Fonterra losing grip of market share in the dairy industry?” . 467 [ ] ] 468 [ ] 469 [ 470 [ 2440192 ] ] 185 D42 Synlait was established in 2008 and is owned by Synlait Limited and Bright Dairy of China. It is located in Canterbury and manufactures milk powders and infant formula. Since entering the market, it has become [ ]471 D43 Miraka was established in 2011 and is owned by a group of Māori trusts and corporations.472 Its plant in Mokai produces milk powders and UHT products for export. It has a [ ]473 capacity WMP dryer plant and two UHT lines with combined capacity of 60 million litres.474 D44 Oceania was established in 2014 and is a wholly-owned subsidiary of Inner Mongolia Yili Industrial Group. Oceania produces milk powder for export to produce infant formula in China. Its processing plant is located in Glenavy, Southern Canterbury. It began production with a single 10 tonne dryer operating a commodity powder dryer475 and [ ].476 D45 Danone Nutricia was established in 2014 and is owned by Danone, which is a global dairy food corporation based in Paris. It manufactures infant formula from its facilities in Auckland and manufactures the milk powder needed for infant formula in Balclutha.477 [ ].478 D46 Green Valley, based in the Waikato, [ ].479 D47 Fresha Valley is a smaller processor [ ].480 471 [ 472 Wairarapa Moana Incorporation, Tuaropaki Trust, Waipapa 8 Trust, Hauhungaroa Partnership, Tauhara Moana Trust and Huiarau Farms. 473 [ 474 Miraka (2016) “Our Products: UHT” < http://www.miraka.co.nz/> (viewed January 2016). 475 Foodprocessing-technology “Oceania Dairy’s Glenavy Milk Processing Factory, South Canterbury, New Zealand” . 476 [ 477 Danone Nutricia “Announcements” (2016) . 478 [ 479 [ 480 [ 2440192 ] ] ] ] ] ] 186 D48 Yashili, based in Pokeno, south of Auckland 481 481 2440192 187 Attachment E: Market definition Purpose of this attachment E1 This attachment describes the tool we used to determine the relevant markets for this review, in the sections: E1.1 the ‘hypothetical monopolist’ or SSNIP test;482 E1.2 the farm gate markets for raw milk; and E1.3 the factory gate markets for raw milk. Hypothetical monopolist, or SSNIP test E2 The SSNIP test is the most common tool for assessing the scope of markets E3 To apply this test, it is necessary to determine whether a hypothetical monopolist could profitably impose a SSNIP. At the heart of this test is substitutability on both the demand-side (will customers switch?) and supply-side (will businesses switch production?). The principal focus is usually on the demand-side. E4 The test starts with the narrowest possible market. If imposing a SSNIP would be profitable, then this is the relevant market. If it is not profitable, then the market is widened and the test re-applied, until it is passed. E5 The following explains these questions with reference to the geographic dimension of market definition: E6 482 E5.1 on the demand-side, can a business in a chosen geographic area increase prices without consumers switching to a nearby supplier of a substitute good or service in sufficient numbers so as to render the price increase unprofitable? E5.2 on the supply-side, can the business increase prices without attracting supply from other firms outside the chosen geographic area in sufficient quantities so as to render the price increase unprofitable? If the answer to these two questions is ‘yes’, then it is likely that a hypothetical monopolist in the relevant area could profitably impose a SSNIP. The relevant A small but significant and non-transitory increase in price. For the purpose of determining relevant markets, the Commission will generally consider a SSNIP to involve a five percent increase in price for a period of one year, refer: Commerce Commission “Merger and Acquisition Guidelines” July 2013, p.15. For assessing market definition in non-merger cases (eg, allegations of anti-competitive behaviour), the relevant benchmark is the hypothetical competitive price rather than the prevailing price. Since it is usually unclear what the competitive price would be, the SSNIP test is primarily used in anti-competitive practice investigations as a conceptual framework for considering the various substitution possibilities, rather than as an empirical framework. 2440192 188 geographic market therefore is likely to be the area tested. If the answer to one or both of these questions is ‘no’, then the area would be increased in size and the test re-applied. Farm gate markets for raw milk Product dimension E7 Raw milk is traded at the farm gate between suppliers (farmers) and customers (processors). Consistent with previous decisions, the Commission considers that there is limited differentiation in raw milk supplied by farmers and that there are no close substitutes for raw milk.483 E8 While we recognise that some differentiation exists within the raw milk market,484 this would not make any difference to our analysis and we therefore define the relevant product dimension of the market as the supply of farm gate raw milk by farmers to dairy processors. Geographic dimension E9 Raw milk collection or ‘catchment’ areas tend to be localised. This is because raw milk is a perishable product and needs to be collected from farms daily and cannot be stored for any significant period in its raw form. There is a high cost of transporting milk relative to its value owing to the proportion of water that is extracted from raw milk during processing.485 Therefore much of the transport costs are spent on hauling water. E10 The geographic scope of the market is determined by what options farmers have to sell their raw milk and what options processors have to purchase raw milk. These options differ regionally. Some farmers may be able to earn a premium on their milk because processors are competing for supply from farmers in that region. E11 Fonterra faces impediments to fully competing in these regional markets. These impediments include its national pricing policy and potential disincentives to price above the milk price as set out in the Milk Price Manual. E12 Under national pricing, the cost of competing for a farmer is not just the price increase to that one farmer, but an equivalent increase to all farmers. However, Fonterra has introduced innovations at the farm gate in the form of the MyMilk contracts in order to counter regional competition in the high milk growth areas of 483 Fonterra Limited and New Zealand Dairies Limited (in receivership) [2012] NZCC 21. 484 For example, organic milk, which constitutes a very small proportion of the milk collected in New Zealand. 485 [ ] 2440192 189 Canterbury, Southland, and Otago.486 This suggests that Fonterra is responding to some extent to regional competition at the farm gate. E13 Consistent with our analysis in previous cases, we therefore consider that there are regional markets for farm gate milk.487 Factory gate market for raw milk Product dimension E14 The factory gate market for raw milk is a wholesale market where processors purchase raw milk from each other. Geographic dimension E15 As factory gate raw milk is supplied directly from farms, not from the suppliers’ processing plants, a processor supplies raw milk to other processors when it already collects in that milk catchment zone. E16 Therefore, factory gate raw milk supply is akin to a distribution service as this milk is not processed by the supplier. E17 The geographic scope of the market is determined by what options selling processors have to sell their raw milk and purchasing processors have to buy raw milk. As with the farm gate market, the factory gate market for raw milk is localised because it is a highly perishable product with high relative transport costs. E18 As these options differ regionally, we consider that there are regional markets for factory gate milk.488 We note that the adoption of narrower regional market definitions for purpose of our analysis differs from the North and South island markets we adopted in our 2011 preliminary inquiry under Part 4 of the Commerce Act 1986.489 Milk supplied under the Raw Milk Regulations versus outside of the Raw Milk Regulations E19 Under the Raw Milk Regulations, Fonterra is required to supply up to 50 million litres of raw milk to each IP, capped at a total of 795 million litres in aggregate per season.490 Fonterra is also required to supply Goodman Fielder with up to 250 486 NERA maintains that Fonterra finds it difficult to attract new conversions because of its requirement to invest in shares. It has responded by allowing farmers to “share up” over time through the MyMilk scheme. NERA Economic Consulting for Fonterra "Assessment of Competition in Raw Milk Markets and Costs and Benefits of the DIRA provisions" 17 August 2015. 487 Fonterra Limited and New Zealand Dairies Limited (in receivership) [2012] NZCC 21. 488 Fonterra Limited and New Zealand Dairies Limited (in receivership) [2012] NZCC 21. 489 Commerce Commission “Milk markets: Consideration of whether to initiate a Commerce Act Part 4 inquiry into milk prices” August 2011. 490 Dairy Industry Restructuring (Raw Milk) Regulations 2012, subpart 1(5)(1), subpart 1(6)(1). 2440192 190 million litres of raw milk per season under the Raw Milk Regulations.491 We refer to this milk as ‘DIRA milk’. E20 [ ] of the factory gate milk that Fonterra supplied in the 2014/15 season was DIRA milk.492 E21 The price at which Fonterra supplies DIRA milk is regulated. The DIRA milk price equals Fonterra’s forecast farm gate milk price plus average transport costs. 493 E22 IPs have a choice of: E22.1 a default price which is the forecast milk price every quarter with a wash-up at the end of each quarter; and E22.2 a fixed quarterly price at the forecast price without a wash-up mechanism. ]494 E23 [ E24 However, some IPs do not qualify for DIRA milk for all or part of their factory gate requirements.495 Such IPs are charged Fonterra’s ‘market’ prices [ ].496 E25 Fonterra’s average market prices are [ ] than the DIRA price. These price differences over the last three seasons are summarised in table E1. 491 Dairy Industry Restructuring (Raw Milk Regulations 2012, subpart 1(7). 492 [ ] 493 Dairy Industry Restructuring (Raw Milk) Regulations 2012 Subpart 3(20)(4)(a) provides that IPs pay ‘reasonable’ cost of transport to the processor. 494 [ 495 For example, they may only require supply on some days; they may not obtain supply in October; they may have failed to comply with the required notice provisions; or they may have exceeded their DIRA milk quota. ] 496 [ ] 2440192 191 Table E1: Season The difference between Fonterra’s average market price and the DIRA price, 2012/13 to 2014/15 Quarter Fixed DIRA milk price Final DIRA milk price Non-DIRA price charged by Fonterra Difference between the average market price and the fixed DIRA price 1 - [ ] 2 - [ ] 3 - [ ] 4 - [ ] 1 $7.00 [ ] [ ] 2 $7.80 [ ] [ ] 3 $8.30 [ ] [ ] 4 $8.65 [ ] [ ] 1 $7.00 [ ] [ ] 2 $6.00 [ ] [ ] 3 $4.70 [ ] [ ] 4 $4.70 [ ] [ ] Difference between the average market price and the final DIRA price 2012/13 2013/14 $5.94 $8.40 [ ] [ ] 2014/15 $4.40 [ ] 497 Source: The Commission’s own table based on data provided by Fonterra. E26 Although IPs do not sell much factory gate milk and, for the most part, do not do so on terms similar to DIRA milk sales, we have been told that [ ]498 E27 Because of the price differential between the prices of DIRA milk and market milk, DIRA milk customers are unlikely to switch to market milk if the price of DIRA milk were to increase by a small but substantial amount. E28 The supply of DIRA milk is therefore considered a separate market to the milk supplied by dairy processors to each other outside the Raw Milk Regulations. 497 [ ] 498 [ ] 2440192 192 Milk supplied at the factory gate versus milk supplied at the farm gate E29 We have previously considered whether the regulatory arrangements had caused the farm gate and factory gate milk markets to merge, in that acquiring milk directly from farmers might be a close substitute for acquiring milk from a dairy processor. 499 E30 In our August 2011 preliminary inquiry, we maintained the separate market definitions.500 E31 Consistent with our approach in the above matters, we consider that maintaining the separate market definitions for farm gate milk and factory gate milk is also appropriate for purposes of the current analysis. E32 IPs that do not have their own sources of milk are unlikely to switch to purchasing milk directly from farmers based on a SSNIP on the DIRA price because of the cost and difficulties involved. E33 Switching from factory gate to farm gate milk likely involves significant costs. These costs include: E33.1 managing the seasonality of milk supply; E33.2 collection and transport of milk to the factory; E33.3 managing multiple suppliers; and E33.4 the risks associated with both obligations to accept supply and supply disruptions. E34 New entrant and small IPs may also find it hard to attract farmers if they do not already have established reputations and/or some production and sales arrangements via supply from the factory gate markets. E35 IPs that source their own milk and are subject to the sunset clause will be seeking to fill their full milk requirements from farmers in the future. As such, they collect a portion of their milk under the DIRA as an interim measure. A small increase in the DIRA milk price is not likely to have much impact on the speed at which switching to own-source supply takes place. E36 For IPs that source some of their own milk in addition to buying DIRA milk, a small increase in the DIRA price might prompt a search for increased farmer supply, but many of these IPs are also likely to face the same barriers faced by IPs that do not currently collect from farmers. 499 Fonterra Co-operative Group Limited and Kapiti Fine Foods Limited and United Milk Limited [2006] NZCC 574, paragraph 129. 500 Commerce Commission “Milk markets: Consideration of whether to initiate a Commerce Act Part 4 inquiry into milk prices” August 2011, Appendix 3, para 18–19. 2440192 193 Attachment F: Evidence that informed our conclusions on excess capacity and resetting market share thresholds Purpose of this attachment F1 The purpose of this attachment is to provide detail on the evidence that informed out conclusion on the need to hold excess capacity as a result of the DIRA Regulation and our recommendation on resetting the market share threshold. In particular it: F1.1 recaps our findings from Chapter 6 on the efficiency cost of the DIRA Regulations in terms of Fonterra’s need to hold excess capacity; F1.2 expands on our discussion from Chapter 6 on milk volume uncertainty, Fonterra’s investment in capacity and the impact of the DIRA Regulations on Fonterra’s need to hold excess capacity; F1.3 outlines how we estimated the efficiency cost of excess capacity due to the Raw Milk Regulations included in Chapter 6 ; and F1.4 sets out evidence that supports our recommendation in Chapter 7 to reset the IP market thresholds for the North and South islands to 30%. Efficiency cost of DIRA Regulation in terms of Fonterra’s need to hold excess capacity F2 In chapter 6, we found that the open entry and exit provisions had no material impact on excess capacity that Fonterra must hold. However we found that the Raw Milk Regulations do have an impact. We estimated a modest direct cost to Fonterra from maintaining this additional capacity in the region of $6 million per year. F3 We note however that the direct costs to Fonterra of $6 million are not the same as the cost to economic efficiency from maintaining this capacity. This is because, even without the regulations, due to milk volume uncertainty, such excess capacity may exist but the costs would potentially be incurred by another industry participant who may have better incentives to manage this risk and price it appropriately. As such we consider $6 million as an upper bound on the costs. Milk volume uncertainty, Fonterra’s investment in capacity and impact of the DIRA on excess capacity Forecasting milk volumes and planning capacity is complex and uncertain F4 There is significant uncertainty in future milk volumes and it appears that the regulations are not the primary driver of Fonterra’s capacity investment. It is difficult to isolate the effect of DIRA Regulation from other drivers of milk volume uncertainty although it is likely to have some impact. 2440192 194 F5 Fonterra’s planning and capacity investments are complex. [ [ ] ]501 F6 Forecasting future milk volumes is important for making optimal investment decisions. This is particularly challenging given the range of variables and uncertainties that influence the value of different products and volume of milk supply. Fonterra notes that [ ]502 F7 [ ]503 F8 The challenge with uncertain milk volumes is that it is difficult to react quickly to the need for new capacity, as it can take as long to set up a new processing plant (a ‘greenfield investment’) as it would to expand an existing plant ([ ]) because expansion often involves duplicating plant. [ ]504 F9 Figure F1 illustrates Fonterra’s milk volume forecast accuracy over the past few years. 505 501 [ ] 502 [ ] 503 [ ] 504 [ ] 505 [ ] 2440192 195 Figure F1: Fonterra’s Milk Volume Forecast Performance [] Source: Information request by the Commerce Commission.506 F10 [ ] 507 This competition also complicates Fonterra’s forecasting of capacity requirements. F11 Overall, DIRA Regulation appears to be only one of several factors driving uncertainty in required capacity and is not the main driver. The evidence before us does not suggest DIRA Regulation has a large impact F12 [ ]508 F12.1 [ ] F12.2 [ ] F12.3 Figure F2: [ ] Fonterra’s capacity forecasts [ ] Source: Information request by the Commerce Commission509. F13 [ ].510 This points to the DIRA Regulation 506 [ ] 507 [ ] 508 [ ] 509 [ ] 2440192 196 being a small component of overall capacity risk as these measures can mitigate, but not totally avoid, costs. It is [ F14 ] in the absence of DIRA Regulation [ ]511 ]512 [ F15 ]513[ [ ]514 We note that the other IPs that source their own milk are generally at or near capacity. F16 ]515 Fonterra has calculated that [ F16.1 [ ]; and F16.2 [ ]. Figure F3: Fonterra’s capacity buffers [ ] Source: Information request by Commerce Commission516 F17 [ ]: F17.1 [ ]517 and 510 [ ] 511 [ 512 [ 513 [ 514 [ 515 [ 516 [ 517 [ 2440192 ] ] ] ] ] ] 197 F17.2 F18 ]518 [ F17.2.1 [ F17.2.2 [ F17.2.3 [ ]; ]; and ]. [ ]. F19 Figure F3 [ ]. F20 [ ]. DIRA Regulation contributes toward general uncertainty which Fonterra must manage F21 NERA contends that [ ]519 F22 NERA argues [ F23 [ ]. ]520 F24 [ ]. F25 However, if Fonterra had more discretion on open entry and exit and the Raw Milk Regulations we consider this could have some effect on Fonterra’s investments. ] 518 [ ] 519 [ ] 520 [ ] 2440192 198 Estimating the efficiency cost of excess capacity due to the Raw Milk Regulations F26 While around [ ] of Fonterra’s milk collection is allocated as DIRA milk, [ ] Based on this information, we can estimate the costs of DIRA Regulation on capacity investment are likely to be in the order of $6 million per year, but may be somewhat lower in the future as the sunset clause continues to have an impact.521 F27 This cost does not necessarily represent a cost inefficiency imposed on the industry. Uncertainty of volume implies a cost whether or not DIRA Regulation requires Fonterra to bear this cost. It appears more likely that purchasers may have better incentives to manage that risk if they faced the costs of capacity risk. The Fonterra forecast evidence suggests [ ]. F28 This could occur if suppliers to the factory gate market priced milk to reflect agreed tolerances to trade off the managing and bearing of this risk. F29 Overall, on the evidence outlined in this attachment, the costs associated with the additional capacity generated by the regulations appears modest and below $6 million. Evidence to support our recommended market share threshold expiry trigger Historic observation and judgement on the future to project market shares Historic observation F30 As a starting point for estimating how the market shares might evolve, we used the past evolution of market shares. F31 In the last five years to the 2014/15 season we observe that:522 F31.1 the total New Zealand milk pool grew at around 6% CAGR; F31.2 IP milk collection grew at around [ ]% CAGR; and F31.3 Fonterra milk collections grew at around [ ]% CAGR. Projected milk growth 521 [ 522 Commission analysis of Dairy NZ data and provided by IPs and Fonterra. There was a high level of milk growth observed in the past 5-10 years, [ ]. Fonterra has a range from low to high [ ]. We have used a combination of low and medium growth to simulate a scenario where the growth continues, but at a slower rate than in the recent past and where IPs are growing faster than Fonterra. 2440192 ] 199 F32 We also looked at forward projections of milk pool growth based on assumptions used by Fonterra and reference to other projections.523 F33 We used a higher growth rate for the total New Zealand milk pool to maintain the observation that IP milk collections will grow faster than Fonterra, albeit from a much lower base. In particular: F33.1 the total milk pool is assumed to grow at around [ ] CAGR in the 6 years to 2020/21 season.524, 525 This growth is based on Fonterra’s own ‘medium’ growth scenario; F33.2 IP milk collections grow around [ ] CAGR;526 and F33.3 Fonterra collections grow at less than [ ] CAGR in the six years to the 2020/21 season. [ ]527 Projected IP market shares in the North and South islands F34 Figures F4 and F5 show how IP market share at the farm gate might evolve over the next six years. Figure F4: North Island projected market share of Fonterra and the IPs, 2015/16 – 2021/22 [ ] Source: Commission analysis. Data provided by DairyNZ and Fonterra. 523 Ministry for Primary Industries “Situation and Outlook for Primary Industries 2015” p.21, forecasts milk solids growth of 1.7% CAGR 2015 to 2019. 524 [ 525 Six years chosen to proxy any new legislation coming into effect during the 2016/17 season with a fiveyear competition review trigger. 526 The choice of different growth scenarios is to simulate a situation where IPs are able to capture a higher proportion of the milk pool in the future. This scenario [ ]. ] 527 [ ] 2440192 200 Figure F5: South Island projected market share of Fonterra and the IPs, 2015/16 – 2021/22 [ ] Source: Commission analysis. Data provided by DairyNZ and Fonterra. F35 Our projections suggest that the North Island IP market share could reach [ ]%, and that of the South Island [ ]% by the 2021/22 season, respectively. F36 The North Island IP market share growth may continue to lag behind that of the South Island. This is consistent with the experience of farm conversions in the Canterbury, Otago and Southland regions. F37 The North Island will likely remain the largest milk pool, but based on our projections, the split between the North and the South islands could decline from 75:25528 in 2001/02 to almost [ ]529 in 2021/22. 528 Dairy NZ “New Zealand Dairy Statistics 2013-2014 LIC and Dairy NZ” . 529 [ 2440192 ]