CT IM196Mar22 · case IM196Mar22
1 COMPETITION TRIBUNAL OF SOUTH AFRICA Case No.:IM196Mar22 In the matter between: Corruseal Group Proprietary Limited First Applicant Neopak Holdings Proprietary Limited Second Applicant and The Competition Commission Respondent Mpact Operations Proprietary Limited First Intervenor APL Cartons Proprietary Limited Second Intervenor In re the intermediate merger between: Corruseal Group Proprietary Limited Primary Acquiring Firm And Neopak Holdings Proprietary Limited Primary Target Firm Panel: AW Wessels (Presiding Member) Mondo Mazwai (Tribunal Member) Liberty Mncube (Tribunal Member) Heard on: 07-11, 14, 16-18, 30 November 2022; 13-14 December 2022; and 02 February 2023 Order issued on: 22 February 2023 Reasons issued on: 14 July 2023 REASONS FOR DECISION 2 [1]After hearing an application for consideration of the proposed intermediate merger over 13 days, in terms of which Corruseal Group (Pty) Ltd (“Corruseal”) intends to purchase all the shares and shareholder loan claims in Neopak Holdings (Pty) Ltd (“Neopak”), the Competition Tribunal (“Tribunal”) on 22 February 2023 issued its decision to prohibit the merger. [2]Our reasons for prohibiting the proposed merger follow. Introduction [3]This matter is an application by Corruseal, the primary acquiring firm, and Neopak, the primary target firm, (collectively “the merging parties”) in terms of section 16(1)(a) of the Competition Act 89 of 1998, as amended (“the Act”), for consideration of an intermediate merger that was prohibited by the Competition Commission (“Commission”). [4]The merging parties seek an order by the Tribunal approving the proposed merger without conditions, alternatively an approval subject to conditions (which they tendered during the course of the hearing). The Transaction [5]In terms of the sale and purchase agreement, Corruseal intends to purchase all the shares and shareholder loan claims in Neopak. Post-merger, Corruseal will control Neopak. Ethos Fund IV (“Ethos”) (which currently holds a majority interest of 83.33%) in Neopak and NET (which currently holds 16.67%) are selling their entire respective interests in Neopak. The Rationale 3
[6]Corruseal submitted that the proposed merger is motivated by at least two objectives . First, Second, the acquisition of an The Neopak Rosslyn paper mill has two paper machines with a production width of each, while the Corruseal Enstra mill's two machines have a and a width, respectively. Re-allocating production between the mills w ill enable Corruseal to reduce waste. The reduction in waste will result in an increase in output.
[7]Neopak submitted that Ethos (as the seller), views the proposed merger as The Merging Parties’ Activities Corruseal
[8]Corruseal is a company incorporated in accordance with the laws of the Republic of South Africa. 1 1 Corruseal is ultimately controlled by and who are historically disadvantaged persons (“HDPs”) as defined in section 3(2) of the Act. 4
[9]It is a vertically integrated packaging firm whose activities include the manufacture and supply of recycled containerboard, corrugated board, and corrugated boxes. The manufacture of recycled container board by Corruseal is conducted through the Enstra Mill, which Corruseal acquired from SAPPI in 2015. Recycled container board is an input into the manufacture of corrugated board, which in turn, is used as an input in the manufacture of various packaging products.
[10]Corruseal owns five factories that produce "corrugated board" or "corrugated sheet". The corrugated sheet produced in the Corruseal Group’s factories is used by the Corruseal Group internally in the production of its own packaging products and is also sold to certain customers. Neopak
[11]Neopak is a company incorporated in accordance with the laws of South Africa.
[12]Neopak is an independent manufacturer and supplier of high-quality recycled containerboard. Neopak is not vertically integrated into downstream markets for the manufacturing of corrugated board and corrugated boxes, having in the past three to five years exited these markets.
[13]Neopak produces recycled containerboard paper through two divisions, namely – the Neopak recycling division, which collects wastepaper and supplies this to the manufacturing facility (paper mill) of Neopak in Rosslyn (Pretoria) where it is used as an input raw material; and the Rosslyn Mill, which makes use of wastepaper as a fibre source to produce recycled paper products. The primary recycled paper products produced by Neopak at the Rosslyn Mill are containerboard papers – these include liners, fluting, and a dual-purpose product 5 that can be used either as liner or fluting 2 . In addition, the Rosslyn Mill also produces a small volume of specialty recycled paper products (such as plasterboard, coreboard and fileboard).
[14]The proposed merger raises a product overlap between the activities of the merging parties with respect to the manufacture of containerboard.
[15]Further, the proposed merger also raises a vertical overlap between the activities of the merging parties as Neopak produces recycled containerboard paper which it supplies to Corruseal and which Corruseal uses to manufacture corrugated board (an input in the manufacturing process of various packaging products). Other Market Participants and Background to the Industry
[16]The first stage of the upstream level of the corrugated packaging value chain entails the collection and production of inputs. Producers of containerboard (also known as mills) source either wood pulp (to produce virgin fibre), recycled wastepaper (to produce recycled fibre) or a combination of both as essential raw inputs to produce containerboard. 3 Virgin fibre is obtained from wood pulp using a digester whereas recycled fibre is obtained from recycled wastepaper using a pulper. 4
[17]The market participants at this upstream level are Corruseal, Mpact (an intervenor in the proceedings), Golden Era, Select-a-Box, and Everest (including Boxlee and Pride Pak). These participants are fully integrated through the supply chain. SAPPI and Neopak are also active at this upstream level but they are not fully vertically integrated. 5 2 Corrugated sheets are generally made with two components: the liner and the fluting paper. Liner is typically found on the outside of a cardboard box, whereas the fluting is found in the centre of the box. 3 Merging Parties' Request for Consideration, para 2.2.3 at A111. 4 Transcript at page. 402. 5 FTI Expert Report, Figure 1 “Industry Value Chain” 6
[18]The next level entails the manufacture of containerboard from recycled or virgin fibre. Different kinds of containerboard can be produced from different types of pulp. The manufacturing process produces rolls of paper (referred to as containerboard) of various widths, grammages and quality. Depending on the input material, the containerboard may be classified as virgin containerboard, being made primarily from virgin fibre, or recycled containerboard, being made primarily from recycled fibre. The containerboard is then supplied downstream to firms called corrugators.
[19]The market participants at this level of the market are the fully integrated firms mentioned above, along with SAPPI and Neopak. 6
[20]At the next level, corrugators bind and glue the containerboard to form corrugated board, which is also known as sheet board. The corrugated board consists of 2 to 4 layers of liner (the flat outside layers) and up to 3 layers of fluting (the inner, corrugated paper situated between the liners). 7 This is illustrated as follows: 6 FTI Expert Report, Figure 1 “Industry Value Chain” 7 Merging Parties' Request for Consideration, para 2.2.5 at A111. See also the Transcript at page 64. 7 Figure 1: Cross -section of corrugated board Source: Merging parties Heads , para 12
[21]Participants at this level of the market are the vertically integrated firms and APL, Sunny Packs, and Houers. 8
[22]Corrugated board is then supplied to firms called converters who further process the corrugated board to produce corrugated packaging (such as boxes). The converters generally cut, glue and print on the corrugated boards to produce the final corrugated packaging to meet the end customers' specifications. 9 In most scenarios, corrugators will also have a converting line, enabling them to produce both corrugated board and the final corrugated packaging. Competitors at the converter level are the vertically integrated firms and Rotapak, Westrand Box, C -Pack, and Seyfert (and others making up a capacity . 8 FTI Expert Report, Figure 1 “Industry Value Chain” 9 Merging Parties' Request for Consideration, para 2.2.6 at A112. 8
[23]Containerboard is thus a key input in the value chain. Without it, corrugators would be unable to compete in their segment(s) of the market. Procedural background
[24]Prior to the hearing of the consideration application, the Tribunal heard two intervention applications, both of which were granted on a defined scope. The intervenors were Mpact Operations (Pty) Ltd (“Mpact”), a competitor of Corruseal (in the upstream market), and a customer of Neopak (in the downstream market); and APL Cartons (Pty) Ltd (“APL”), a customer of Neopak, Both intervenors ultimately submitted that the proposed merger should be prohibited. 10
[25]The consideration was duly heard by the Tribunal over a 13-day period from 07 November 2022 to 02 February 2023.
[26]Against this background it is necessary to expand on the role of the two intervenors in the industry, before dealing with their submissions in the appropriate sections of these reasons.
[27]The first intervenor is Mpact, a competitor to Corruseal and Neopak in the market for the supply of recycled containerboard, and a competitor to Corruseal in respect of its downstream corrugating activities. Mpact, is itself, vertically integrated. 10 APL was admitted as an intervenor by agreement between the parties (Tribunal Case Number.: IM196Mar22/INT086Aug22). Mpact’s intervention application was opposed by the merging parties and it was ultimately admitted as an intervenor by Tribunal Order (Tribunal Case Number.: IM196Mar22/INT017Apr22). Given that Mpact participated in these proceedings in accordance with the Tribunal Order and these reasons deal with the merits of the proposed transaction, in which Mpact participated, we find it superfluous to issue separate reasons for the Mpact intervention. 9
[28]In respect of the recycled containerboard market, Mpact owns and operates the Felixton mill located in KwaZulu-Natal. The Felixton mill supplies containerboard (liner and fluting), which is wholly produced from recycled fibre. It is also partially vertically integrated as it supplies recycled containerboard to Mpact’s corrugating plants, with the balance supplied predominantly to the merchant market.
[29]In respect of the production of corrugated packaging material, Mpact has nine corrugating plants, each with corrugator and converting facilities, producing corrugated board and boxes.
[30]The second intervenor is APL, a manufacturer of corrugated packaging products for end-consumers. APL is an “independent” corrugator in the downstream market in that it is not vertically integrated with a paper mill. APL is the largest independent corrugator in the market for corrugated paper packaging in South Africa. Its business entails the manufacture and supply of corrugated packaging products manufactured from recycled and virgin containerboard.
[31]At the hearing, the Commission called the following witnesses 11 : 31.1. Brian Seggar (“Mr Seggar”) of Everest Packaging Group (Pty) Ltd (“Everest”); 31.2. Graeme Wild (“Mr Wild”) of SAPPI Southern Africa Limited (“SAPPI”); and 31.3. Dr Thulani Mandiriza (“Dr Mandiriza”), a principal economist at the Commission and an economics expert.
[32]The merging parties called the following witnesses at the hearing 12 : 11 The three witnesses had all submitted witness statements prior to the hearing. 12 The three witnesses had all submitted witness statements prior to the hearing. 10 32.1. Mehul Mehta (“Mr Mehta”) of Corruseal; 32.2. Hamish Fraser (“Mr Fraser”) of Neopak; and 32.3. Prof Nicola Theron (“Prof Theron”) of FTI Consulting and an economics expert.
[33]Mpact called the following witnesses at the hearing 13 : 33.1. Hugh Michael Thompson (“Mr Thompson”); and 33.2. Patrick Smith (“Mr Smith”), of RBB Economics and an economics expert.
[34]APL called the following witnesses at the hearing 14 : 34.1. Frederik Christoffel Greeff (“Mr Greeff”); and 34.2. Johannes Wiehann Erasmus (“Mr Erasmus”)
[35]We used the ‘hot tub’ method of hearing evidence to hear the evidence of the economics experts. The Relevant Markets
[36]The purpose of market definition is to provide a framework for an analysis of the competitive effects of the merger. The relevant market (or markets) is the market within which the merger may give rise to a likely substantial prevention or lessening of competition (“SPLC”). The relevant market contains the most significant competitive alternatives available to customers of the merged companies, however, market definition is not an end in itself. Furthermore, the 13 The two witnesses had all submitted witness statements prior to the hearing. Christoff Botha (“Mr Botha”) who also had submitted a witness statement was not called at the hearing to submit his evidence. 14 The two witnesses had all submitted witness statements prior to the hearing. 11 boundaries of the market do not determine the outcome of an analysis of the competitive effects of the merger in a mechanistic way.
[37]In carrying out our assessment, we took note of the 2021 CMA Merger Guidelines which state that: Market definition involves identifying the most significant competitive alternatives available to customers of the merger firms and includes the sources of competition to the merger firms that are the immediate determinants of the effects of the merger. While market definition can be an important part of the overall merger assessment process, the CMA’s experience is that in most mergers, the evidence gathered as part of the competitive assessment, which will assess the potentially significant constraints on the merger firms’ behaviour, captures the competitive dynamics more fully than formal market definition. 15
[38]In practice, the analysis underpinning the identification of the market or markets and the assessment of the competitive effects of a merger may overlap, with many of the factors affecting market definition being relevant to the assessment of competitive effects and vice versa. Therefore, market definition and the assessment of competitive effects should not be viewed as distinct analyses. Our assessment of market definition below should therefore be read alongside our competitive assessment.
[39]In our assessment of the competitive effects of the proposed merger, we consider the following theories of harm: 15 UK CMA Merger Assessment Guidelines (2021) at para 9.2 12 39.1. vertical foreclosure of non-integrated corrugators from access to recycled containerboard; 39.2. horizontal unilateral effects in the supply of recycled containerboard; and 39.3. horizontal coordinated effects in the supply of recycled containerboard.
[40]As described above, the merging parties supply non-integrated corrugators recycled containerboard, which is a key input into the activities of corrugators. As such, we have considered market definition by reference to the operations of the merged entity’s supplies of recycled containerboard to its corrugator customers, and by reference to the downstream market in which Corruseal and its rivals are active.
[41]Below we consider the definition of the relevant markets in which the effects of the proposed merger should be assessed. Product market definition
[42]The merging parties, the Commission and the intervenors agreed on the following: 42.1. Virgin containerboard has better strength characteristics than recycled containerboard; 16 42.2. Virgin containerboard is more moisture resistant than recycled containerboard and, therefore, can be used to make corrugated products that can be used in cold-chain environments; 17 16 Witness Statement of Mr Thompson, para 21; and Mr Wild, Transcript at pages 530 and 546. 17 Mr Mehta, Transcript at page 125 and Mr Wild, Transcript at page 530. 13 42.3. Virgin containerboard is generally used by corrugators to make packaging for agricultural applications where stronger boxes are n eeded to transport agricultural products; 18 42.4. Recycled containerboard, on a like gram for gram comparison, is generally cheaper than virgin containerboard and used for less demanding applications for shorter periods of time. Corrugators usually use recycled containerboard for applications in sectors such as fast-moving consumer goods. 19
[43]However, the merging parties submitted that these differences do not mean that virgin and recycled containerboard are not subst itutable for one another . They argued that lightweight virgin products are substitutable with heavier recycled containerboard products. 20 They further argued that corrugators could “mix and match” (including between virgin and recycled) the fluting and liners that make up the box t o achieve the desired strength characteristics at the lowest cost and hence “mixed ” conta inerboard is subst itutable with recycled containerboard. 21
[44]The merging parties submitted that the product market included recycled containerboard and “mixed products” (that is, Hi-Pact 22 , Ultraflex 23 , and DP 24 ). 18 Mr Mehta, Transcript at p ages. 213, 297, 298 and 299. 19 Transcript at p ages 474 and 477. 20 Mr Wild, Transcript at pages 596 – 598. 21 Me Wild, Transcript at pages 597 – 600. 22 Mpact's Hi-Pact product (" Hi-Pact") – consists of a minimum of virgin and between recycled material. 23 Sappi's Ultraflex product ("Ultraflex") – consists of virgin and recycled material. 24 Neopak's DP product (" DP") – this product is made of 100% recycled material. It does not include virgin containerboard. 14 Figure 2: FTI’s view of the product set (navy), expanded set based on testimony (cyan) Source: FTI Expert Presentation, slide 8 (adapted from the FTI Report ,Figure 2)
[45]The Commission and the intervenors disagreed with the merging parties on both arguments. The Commission submitted that the product market included all recycled containerboard products, including Ultraflex, but excluding DP and Hi- Pact. 15 Figure 3: Commission’s view of the product set (recycled excluding Neopak’s DP and Mpact’s Hi-Pact) Source: FTI’s Expert Presentation, slide 6 (adapted from the FTI Report, Figure 2)
[46]The intervenors (Mpact) submitted that the product market is limited to “100% recycled containerboard”. 16 Figure 4: RBB’s view of the product set (100% recycled) Source: FTI’s Expert Presentation, slide 7 (adapted from the FTI Report, Figure 2)
[47]The CMA merger guidelines state that: “…In many cases, especially those involving differentiated products, there is often no ‘bright line’ that can or should be drawn. Rather, it can be more helpful to describe the constraint posed by different categories of product or supplier as sitting on a continuum between ‘strong’ and ‘weak’. The CMA will generally not need to come to finely balanced judgements on what is ‘inside’ or ‘outside’ the market. Not every firm ‘in’ a market will be equal and the CMA will assess how closely two merger firms compete. The constraint posed by firms ‘outside’ the market will also be carefully considered.” 25 25 UK CMA Merger Assessment Guidelines (2021) at para 9.4 17
[48]This proposed merger relates to differentiated products. Our assessment of the potential competitive constraints posed by mixed-use and virgin containerboard on 100% recycled containerboard is set out in the competition assessment below. Overall, the evidence on Ultraflex is that SAPPI does not intend on supplying Ultraflex into the market. Mr Wild testified that: ‘[SAPPI’s] decision when we went into the market this year in July and August was to tell our customers that we would not contract any recycled products’. 26 The evidence on the Hi-Pact product is that this product is about more expensive than 100% recycled containerboard and Neopak’s DP product is about 10% more expensive than 100% recycled containerboard. 27 Furthermore, Neopak is the only firm that produces the DP product. This means that it is not a constraint to the merged entity’s ability to effect a small but significant increase in price.
[49]As part of our competition assessment, we consider whether mixed-use and virgin containerboard are competitive constraints on the supply of 100% recycled containerboard. We further consider the ability of local producers, of virgin materials, to switch into the production of 100% recycled containerboard. Overall, in our view the evidence indicates that the upstream relevant product market is not wider than the supply of 100% recycled containerboard.
[50]The merging parties, the Commission and the intervenors agreed on the downstream relevant product market as the supply of corrugated board. Geographic market definition
[51]The merging parties, the Commission and the intervenors agreed on a national market for the supply of corrugated board. 26 Mr Wild, Transcript at pages 560-561 This is because SAPPI has made investments into the production of virgin paper. While switching to manufacture recycled paper is technically possible, Mr Wild described it as ‘sub-optimal’. Transcript at page 641. 27 Dr Mandiriza, Transcript at page 1794. 18
[52]Regarding the upstream market for the supply of 100% recycled containerboard, the Commission and the intervenors submitted that imports are not a competitive constraint on the merged entity’s ability to effect a small but significant increase in price. The merging parties submitted that imports should be included in the relevant market.
[53]As part of our competition assessment, we consider whether imports are a competitive constraint on the national supply of 100% recycled containerboard. In our view, the evidence indicates that imports are unlikely to constitute a constraint on the merged entity. Conclusions on the relevant markets
[54]We assess the effects of the proposed merger in the following product markets on a nation-wide basis: 54.1. Supply of 100% recycled containerboard; and 54.2. Supply of corrugated board. Competition Assessment
[55]In this section we consider the competitive effects of the proposed transaction. The market relationship of the merging parties is both vertical and horizontal.
[56]The concern under a vertical theory of harm is that bringing together the merging parties creates or increases the ability and/or incentive of the merged entity to harm competition at one level of the supply chain through its behaviour at another level of the supply chain.
[57]The theories of harm raised by vertical mergers typically involve the merged firm harming the ability of its rivals to compete post-merger through foreclosure, for example by raising effective prices to its downstream rivals or reducing its supply (partial foreclosure), or by refusing to supply them completely (total foreclosure). 19 Such actions may harm the ability of the merged firm’s rivals to provide a competitive constraint into the future.
[58]The proposed transaction brings about a vertical overlap between the merging parties’ upstream supply of 100% recycled containerboard and Corruseal’s downstream corrugating activities.
[59]The theory of harm in this case arises from the combination of Corrus eal with Neopak. The theory of harm is that the proposed transaction might give rise to a significant shift in incentives as Neopak is not currently vertically integrated and therefore its incentives are not compromised by its downstream entanglements. The proposed transaction might give rise to a significant risk that non -integrated corrugators will be foreclosed from access to 100% recycled containerboard.
[60]Pre-merger, the contestable market for the supply of 100% recycled containerboard is supplied by Neopak (approximately tons), Corruseal (approximately tons) and Mpact (approximately tons ). Corruseal’s downstream operations require approximately tons per annum. Corruseal’s upstream operations, however, only produce approx imately tons per annum. Corruseal’s downstream operation is a net purchaser of about tons per annum. The Commission also pointed us to Exhibit 1, which indicates that Corruseal intends to . 28 28 Exhibit 1, Corruseal Board Meeting Minutes , dated 30 September 2021 , page 16 of Bundle F . 20 Source: RBB’s Expert Presentation, slide 7
[61]Post-merger, Neopak’s volumes will come under the control of Corruseal. A strategy of foreclosing (partial foreclosure or total foreclosure) non-integrated corrugators might be profitable for the merged entity overall if the profit gained is greater than the profit which it would lose from supplying non-integrated corrugators. 21 Figure 6: Market structure, post-merger Source: RBB’s Expert Presentation, slide 8
[62]Our approach to assessing the vertical theory of harm is to analyse: 62.1. Whether the merged entity would have the ability to carry out the foreclosure strategy; 62.2. Whether it would find it profitable to do so (i.e., the incentive), and 62.3. Whether the effect of any action by the merged firm would be sufficient to result in a likely SPLC.
[63]These conditions are cumulative: if we find that one condition is not met, we may not find it necessary to assess the other conditions. They may also overlap. For example, at the extreme end, with sufficient resources a firm is likely to be able to pursue almost any strategy, but if it is exceedingly costly to do, the firm is very unlikely to have the incentive to do so. 22
[64]Finally, we conclude on the resulting likely vertical competitive effects of the merger absent any countervailing factors.
[65]Following our assessment of the vertical issues arising from the merger, we examine the extent of horizontal competition between the merging parties. In doing so, we assess whether a loss of this competition could result in higher prices as a result of horizontal unilateral effects or horizontal coordinated effects. Vertical effects Ability to harm rivals
[66]As a starting point, we consider the following potential foreclosure mechanisms: 66.1. A refusal to supply 100% recycled containerboard strategy (total foreclosure); and 66.2. A strategy of increasing prices of 100% recycled containerboard or reducing the merged entity’s supply to downstream non-integrated corrugators (partial foreclosure).
[67]We assess the merged firm’s ability to foreclose rival non-integrated corrugators of access to inputs (100% recycled containerboard) in four steps: firstly, by setting out the merging parties’ views, secondly, by setting out those of the Commission and the intervenors, thirdly, by presenting our assessment, and finally by presenting our conclusions on ability. Merging parties’ views
[68]The merging parties submitted that for the merged entity to have the ability to foreclose rival non-integrated corrugators, it must have sufficient market power in the upstream 100% recycled containerboard paper market. The merging 23 parties submitted that t he Commission and intervenors overstate the merged entity's market shares. According to the merging parties, the merged entity will possess only a share of the market (excluding imports). The merging parties submitted that excluding internal sales of recycled containerboard from Neopak to Corruseal from the market share calculation, results in the merging parties having a post -merger market share of
[69]The merging parties also submitted that there exist numerous alternative sources of supply for 1 00% recycled containerboard to which third party customers could turn to if the merged entity ever sought to engage in a foreclosure strategy. These alternative sources of supply include suppliers of Corruseal downstream (in the event of Corruseal interna lising Neopak volumes, Corruseal would have to forego its current purchases from Mpact), competitors to Neopak, imports, the redirection of exports (by firms currently exporting such as Mpact) and the use of lightweight virgin products. The merging partie s submitted that these alternative sources exist and would be able and likely to expand were the merged entity to consider engaging in a foreclosure strategy.
[70]The merging parties submitted that while barriers to entry are indeed high if one wants to inves t at scale, it is unreasonable to assume that companies the size of Mpact and Sappi would not make the required investment and expand their operation if market conditions favoured them doing so. They submitted that small scale entry is possible and pointe d us to Select -a-box and Everest as examples of firms who have entered the market on a small scale in recent years. They submitted that upstream small -scale entry is open to other corrugators. The Commission and the interven or’s views
[71]In response, the Co mmission and the interven ors submitted that the merging parties have the ability to foreclose rival non -integrated corrugators because of the following factors: 24 71.1. The upstream market is characterised by high barriers to entry and that new entry or expansion in the short to medium term is unlikely; 71.2. Imports do not adequately (if at all) discipline the market conduct of domestic suppliers and have several significant disadvantages compared with domestic production; 71.3. Supply-side substitution by both Mpact and Sappi is unlikely to replace the production of 100% recycled containerboard that non -integrated corrugators receive from Neopak; 71.4. There is no evidence that increased domestic prices would cause exporters to re-direct their volumes back into the domestic market; and 71.5. Downstream non-integrated corrugators could not viably switch to substitute 100% recycled containerboard with mixed-use and virgin containerboard. Our assessment
[72]First, we consider whether the merged entity would have market power in the upstream market.
[73]The merging parties submitted in their heads of argument that: “In calculating market shares, both the Commission and FTI agreed, in principle, that third party sales should be used as the metric to determine market shares.” 29
[74]We agree that third party sales offer a better metric in determining market shares in this case. We therefore did not find Mr Smith’s calculation of market shares using capacities persuasive because capacity figures include internalised volumes. 29 Merging parties’ Heads at para 89. 25
[75]The Commission submitted that Neopak’s pre -merger share is approximately 30 The merger, if approved, will result in Corruseal attaining a post -merger market share of about with a n approximate market share accretion.
[76]In contrast, Prof Theron’s method of calculating market shares excluded Neopak volumes currently contracted to Corruseal resulting in the merged entity attaining a market share of approximately (and approximately excluding imports and exports) 31 . We are not persuaded that this is correct approach in this case. This is because the contractual arrangement for these sales is for a one-year duration. When that lapses, the volumes supplied by Neopak to Corruseal become contestable again to the non -integrated corrugators.
[77]Furthermore, Prof Theron was unable to provide a basis for her argument. The following exchange is in instructive: ADV WILSON SC: Can I then come onto another aspect of your evidence, which I understand is important in relation to market shares and this is the idea that you do not include sales from Neopak to Corruseal and as I understand it the explanati on you give for this is well, that’s not part of the counterfactual to the merger because there’s already a contract in place I effectively take those volumes out of consideration. And I wanted to ask you two things. The one is if there is any authoritativ e precedent for that kind of calculation… PROF THERON: So I think maybe just two steps back. So this was quite an important piece of contention between the Commission and the merging parties before we were briefed. So when we looked at this issue we were looking for authority to see what is the right thing 30 Commission ’s Expert Presentation, slide 13 (This is if one excludes exports). 31 FTI’s Expert Report, para163 -165. 26 to do and we couldn’t find anything. So maybe let me just answer your first question there. We couldn’t find anything to say that self -supply should be excluded…” 32
[78]Prof Theron conceded that the foundation of her market share assessment – which she herself termed “an important piece of contention between the Commission and the merging parties” – finds no support in economic theory nor in case precedent. 33
[79]Given the above , our view is that if Prof Theron’s calculations are corrected to include Neopak volumes currently contracted to Corrusea l, the merged entity is likely to enjoy a share of more than in the market for the supply of 100% recycled containerboard.
[80]We then consider ed whether the input is sufficiently important for non -integrated corrugators and found that non -integrated corrugato rs require 100% recycled containerboard in order to compete. For example, pre -merger, 78% of Neopak’s revenues are generated from non -integrated players. 34 Approximately of APL’s volumes are sourced from Neopak. 35 More than of Select -a-Box’s volumes and more than of Everest’s volumes come from Neopak. 36
[81]Below we set out our assessment in six steps: 81.1. First, we consider the natural experiment arising from Corruseal’s own response to a fire at its Enstra mill in January 2021; 32 Transcript at pages 2155 -2156 . 33 Prof Theron, Transcript at page 2156. 34 Commission’s Expert Presentation , slide 14 . 35 Commission’s Expert Presentation , slide 14 . 36 Commission’s Expert Presentation , slide 14 . 27 81.2. Second, we assess the ability of local suppliers currently exporting 100% recycled containerboard to redirect these exports into the local market to make up for any lost supply; 81.3. Third, we assess the ability of imports to credibly constrain the local market and make up for any foreclosed volumes; 81.4. Fourth, we assess the ability of corrugators to substitute the use of 100% recycled containerboard with mixed-use and virgin containerboard; 81.5. Fifth, we assess the ability of local producers, in particular of virgin materials, to switch substantially into the production of 100% recycled containerboard; and 81.6. Six, we assess the barriers to small-scale entry and expansion to compete with the merger entity. The natural experiment
[82]We consider relevant a natural experiment arising from Corruseal’s response to a fire at its Enstra mill in January 2021. Mr Mehta told us that, following the fire at the Enstra mill, Corruseal needed to make up its lost supply of 100% recycled containerboard in order to honour its contracts to its third-party customers. 37 We were interested in how Corruseal practically responded to this setback.
[83]Mr Smith’s analysis, as shown below, examined the available evidence and found that Corruseal did not make up the shortfall through imports, diverted exports, or demand or supply-side substitution. 38 Mr Smith illustrated that the fire reduced Corruseal’s internal supply of recycled containerboard, and 37 Mr Mehta, Transcript at page 100. 38 Mr Mehta at Transcript at page 393: “ADV WILSON SC: Okay, and can I also understand that there is no indication of any significant increase in imports by you in order to accommodate those, that lost production. Is that correct? MR MEHTA: No, we didn’t need to import. We had the additional supply from both Mpact and Neopak.” See also Mr Smith, Transcript, at page 1867. 28 Corruseal made up that shortfall by increasing its purchases from Neopak and Mpact. Figure 7: Demand substitution, the SNAP test – Corruseal Enstra Fire Source: RBB’s Expert Presentation, slide 23
[84]In our view, this evidence shows that when Corruseal needed to secure additional volumes to make up for its lost output on PM3, it did so through the only other two local producers of recycled containerboard – namely Mpact and Neopak. Mr Mehta also confirmed that Corruseal approached Mpact and Neopak and “renegotiated larger supply agreements from both these suppliers”. 39 39 Mr Mehta, Transcript, at page 101. 29
[85]Our view is that this evidence clearly illustrates that when Corruseal itself was faced with a significant shortage of 100% recycled containerboard, it did not turn to imports, diverted exports, or demand or supply-side substitution. The re-direction of exports
[86]We consider the merging parties’ submission that there were up to 65 000 tonnes of exports that could be re-directed into South Africa, 40 the majority of which emanated from Mpact. Mr Mehta was of the view that this would constrain the ability of the merged entity to engage in input foreclosure, because non- integrated corrugators could rely on the re-diversion of exports back into South Africa to make up any lost volume. Mr Mehta further argued that a producer of containerboard would “always want to fill your mill with local supply” and would only export “opportunistically” for “very short periods of time where the market is at a peak”. 41
[87]We note that Prof Theron conceded in her testimony that exports are only between 4.4% to 5% of the market. 42 Our view is that even assuming that exports could be re-directed, the volumes are very small and unlikely to constitute a competitive constraint.
[88]We consider relevant the evidence suggesting that the bulk of these exports (on average “are not suitable for most domestic corrugator customers”. 43 This is because the narrow width or “deckle” of the containerboard which is exported (being approximately 1,8m) is unsuitable for corrugators in South Africa (which operate corrugating machines at width of 2.4m and above. 44 We note that 40 Mr Mehta, Transcript, at pages 109-110. 41 Mr Mehta, Transcript, at page 110, 42 Prof Theron, Transcript, at page 1893. 43 Witness Statement of Mr Thompson, at para 78. 44 Commission’s Expert Report, at para 150. 30 Corruseal similarly exports narrower width containerboard. 45 Mr Mehta conceded that if corruga tors were to utilise narrower widths (of 1.8m) this would result in a reduction in throughput of %, and to make up the lost capacity would come at an additional cost. 46
[89]We further consider relevant the evidence that Mpact exports to a long -standing customer in the SADC region. 47 In our view, Mpact does not have the incentive to redirect export volumes to the domestic market to make up any shortfall of supply of 100% recycled containerboard at competitive prices to assist downstream rivals.
[90]Based on the evidence we conclude that the redirection of export volumes is unlikely to constitute a substantial constraint on the ability of the merged entity to engage in an anticompetitive foreclosure strategy. The ability of imports to credibly constrain the local market
[91]The merging parties submitted that c orrugators can import recycled containerboard with “relative ease” and that imports are a viable alternative for 45 Commission’s Expert Report, Table 11. 46 Mr Mehta, Transcript at pages 479-480. 47 Witness Statement of Mr Thompson, at para 7 9, and Mr Thompson, Transcript page 1364 -5: “on average of my exports are not suitable for the domestic corrugator market … 31 local corrugators facing a shortage of domestically produced 100% recycled containerboard. 48
[92]However, we note that the bulk of the evidence pointed overwhelmingly to the contrary. On average, on an annualised basis, imports of recycled containerboard to South Africa account for between 7000 to 10 000 tonnes, and in 2021 they were 8 000 tonnes. 49 Prof Theron’s own evidence was that “imports constituted approximately 4.4% of Corruseal’s corrugators’ total purchases, including third party purchases and internal sales of recycled containerboard paper between 2016 to 2021.” 50 These actual tonnages fall far below the 15% benchmark identified by UK Office of Fair Trading and adopted by the Tribunal in Langeberg Foods International and Ashton Canning Company (Pty) Ltd 51 as constituting a credible constraint. In Ashton, the Tribunal explained that: “significant levels of import penetration, at least 15% of domestic sales, should be required before one can regard import competition as capable of exerting decisive influence on price levels.” 52
[93]In our view, the economic evidence illustrates that imports are not a significant constraint even in the face of a significant price rise. For example, when Mpact increased its prices by 11% and 24%, as indicated in the analysis below by Mr Smith, the levels of imports did not increase. Furthermore, following the significant price rises in recycled containerboard in 2021 and 2022, there was no significant increase in imports – even post the October 2022 negotiation of new contracts. 53 48 Witness statement of Mr Mehta, at paras 22.6 and 31. 49 Mr Mehta, Transcript, at page 202. 50 FTI Expert Report, footnote 195. 51 Tiger Brands Ltd / Ashton Canning Company (Pty) Ltd / Newco and Langeberg Foods International Ashton Canning Company (Pty) Ltd (46/LM/May05) [2005] ZACT 82 (23 November 2005) (“Ashton Canning”). 52 Ashton Canning at para 28. 53 Mr Smith, Transcript, page 1815 32 Figure 8: Monthly imports of recycled containerboard, January 2018 – October 2022 Source: RBB’s Expert Presentation, slide 13
[94]We interpret this evidence as illustrating that that imports are unlikely to substantially constrain the ability of the merged entity to engage in a foreclosure strategy.
[95]We also considered the merging parties’ argument captured on slide 27 of Prof Theron’s expert presentation (shown below) comparing the landed price of imported fluting with the price on offer by Neopak (fluting prices across eight different dates between January and November 2022). 33 Table 1: Fluting import prices offered to Corruseal 54
[96]The slide reveals the following: 96.1. On every single one of the eight dates reflected, imports were more expensive than domestic prices; 96.2. On three of the eight dates, imports were almost double the domestic price; and 96.3. On only one of the eight dates were the prices comparable. And even then, the import price exceeded the local price.
[97]We interpret this evidence to mean that, even on the evidence of the merging parties’ expert, imports are consistently more expensive than domestic prices, and in many instances almost double the domestic price. It is common cause that for extended periods, imports are not viable. It is true that imports will sometimes be priced competitively – for instance in November 2022 – but that is cold comfort in circumstances where imports were significantly more expensive for the preceding 10 months.
[98]Furthermore, Mr Mehta’s evidence was that the “global market is very volatile”. 55 Mr Mehta emphasised that one would “get different pricing from different regions 54 FTI’s Expert Presentation, slide 27 55 Mr Mehta, Transcript, at page 62. 34 constantly”, 56 and that Corruseal views imports as 7
[99]Mr Mehta’s conce ded that a purchaser is required to wait longer for imported paper than for domestically produced containerboard. 58
[100]Mr Mehta’s conce ded that it would not necessarily be easy for a new player to open a credit facility with imports, 59 and that existing partici pants in the market like APL would not be able to rely solely on imports. 60
[101]Mr Thompson testified that : 61 “So, imports do come in at the fringes. Many of the corrugators or the customers that we deal with do have levels of import, but there are – to build your business on imports make no sense. I heard the Everest example where all starts aligned. However, you know it was very quickly turned around to indicate that there are really good reasons for ensuring that you have local supply in terms of having security of offtake, working capital management, flexibility in the product offering and I – you’re not exposed to the logistics around the supply chain. So, there are – imports do play a role in this market. Imports as a percentage of the total market are very small and I’d be surprised if we’d find players that would be building their business model on imports ag ain.” 56 Mr Mehta, Transcript, at page 86. 57 Mr Mehta, Transcript, at page 86. 58 Mr Mehta, Transcript, at page 211. 59 Mr Mehta, Transcript, at page 224. 60 Mr Mehta, Transcript, at page 465. 61 Mr Thompson, Transcript, at page 1359. 35
[102]In short, Mr Thompson’s evidence was that imports only come in “at the fringes”, 62 and that they make up a very small percentage of the total market. Furthermore, he would be surprised if players would be able to build a business model on imports in the current market. 63
[103]Mr Greeff’s evidence was that imports are not a viable source of supply for APL because they are unreliable, there is limited supply, one is dependent on ad hoc spot pricing, and there is no long-term commitment. 64 Mr Greeff’s further explained that if imports were readily available, everyone would be importing, “It would be manna from heaven but it is not, unfortunately”. 65
[104]Mr Seggar’s explained Everest’s use of imports in the following manner: 66 “We found ourselves (a) because of our growth rate using a purely import model our working capital was under pressure. So, your payment terms typically on imports is bill of lading or cash with order ,which if you convert that into working capital cycle it’s probably 60 days sharper than what’s 10 offered locally. So, in order to supply your customers you’d have to carry stock on your floor, plus extent terms to them from a debtor’s perspective. So, it created quite a big working cap squeeze. So, we knew eventually we had to move some volume to local supply from a working cap perspective. We also knew this sort of gravy train couldn’t last forever. When we started up it was, you know it was very easy for us to be nimble and have a high-risk approach in terms of procurement, but it is your main raw material. You don’t have a business without it. So, our view was to shift a 62 Mr Thompson, Transcript, at page 1359. 63 Mr Thompson, Transcript, at page 1359. 64 Mr Greeff, Transcript, page at 1528. 65 Mr Greeff, Transcript, page at 1529. 66 Mr Seggar, Transcript, pages at 1130-1131. 36 significant portion to local supply as we started steadying the business because it’s easy to take big gains and big wins when you’re a sta rt- up and be fairly nimble. But, as you have this base load and customers that now rely on you, you have to forego potential opportunistic margins that you may attract on the import model and localise to more of a stable supply.”
[105]In brief, w hile the merg ing parties sought to place emphasis on the fact that Everest started out its business in 2019 on the basis of imports, we were persuaded by Mr Seggar ’s testimony that while Everest had used imports to its advantage when starting out, it is not possible to operate a viable business with scale by relying on imports. Mr Seggar testified in this regard that it had become necessary for Everest to ‘localise’ its purchasing to ensure ‘stable supply’. 67
[106]Everest currently secures around ocal paper and imported recycled papers, primarily from the UAE. In terms of local production: (1) Neopak supplies% of Everest’s recycled containerboard; and (2) Corruseal supplies of Everest’s recycled con tainerboard. 68 The volumes imported by Everest are spot imports. There are no contracts. 69
[107]Overall, we conclude that although import prices can occasionally fall below domestic prices – and one can then engage in a short -term opportunistic import – on average and over an extended period imports are more expensive than domestic supply. Imports are also complicated by factors such as: (i) rising prices and high fre ight charges, (ii) long lead times 70 , (iii) high working capital 67 Mr Seggar, Transcript, at page s 1130 -1131. 68 Mr Seggar, Transcript, at page 1133. 69 Mr Seggar, Transcript at page 1194. 70 Mr Seggar, Transcript , at pages 1130 and 1206 . 37 requirements 71 , (iv) depreciation of the Rand against major currencies 72 , and (v) the inability of new entrants to secure credit lines with foreign suppliers 73 .
[108]Taking into account these factors, we consider it more likely than not that imports are unlikely to constitute a constraint on the ability of the merged entity to engage in a foreclosure strategy in respect of the supply of 100% recycled containerboard to rival non-integrated corrugators that require security of supply. The ability of corrugators to substitute the use of 100% recycled containerboard with mixed-use and virgin containerboard
[109]We considered the argument raised by merging parties that the ability of the merged entity to engage in input foreclosure would be constrained by customers of 100% recycled containerboard who would switch to procuring alternative materials, in particular virgin containerboard (the merging parties’ chain of substitution argument) 74 . Prof Theron argued that there are light weight virgin containerboard products that are very likely to be substitutable with heavier recycled containerboard as their performance metrics and prices were comparable with one another. In other words, one can cost effectively replace a higher grammage recycled product with a lower grammage virgin product.
[110]However, we are not persuaded by this argument. This is because theoretical arguments need to be supported by actual evidence. Mr Mehta acknowledged in his evidence that the concept of a chain of substitution was developed by the merging parties’ expert, 75 and, when he was asked whether he could point to 71 Mr Seggar, Transcript, at page 1130. 72 Transcript, at page 205. 73 Mr Mehta, Transcript, at pages 222 – 224. 74 Mr Mehta, Transcript, at pages 136-137. 75 Mr Mehta, Transcript, at page 395. 38 any particular studies or documentation in the record to support the existence of the “so-called chain of substitution that [Corruseal’s] economic expert refers to”, Mr Mehta responded: “No, I haven’t brought out anything specifically on substitute[ion]”. 76
[111]Mr Mehta acknowledged that it is not commercially viable to substitute Ultra Flute and UltraFlute Plus for recycled paper 77 because it is significantly more expensive. Mr Mehta accepted the same constraint for Kraft Liner. 78
[112]Mr Fraser conceded that, although he believed that DP could theoretically compete with UltraFlute, “it’s quite tricky because those board plates have to be approved by the Citrus Growers Association” and DP was not certified for use in export agricultural cartons. Mr Fraser acknowledged that this comparison was “one of those sorts of blue sky can we actually go and compete with our recycled product in this virgin market”. 79
[113]Mr Seggar told us the following: 80 “MR SEGGAR: I think the principle of substitution, I mean I don’t know to what extent has it been laboured on, but if I can give a short version. The principle of substitution relies on the fact that you can take high performing paper at a lower grammage to substitute heavier low performing paper, right. So, in order to substitute recycled paper that has a hundred and forty gram – 140 gsm you could substitute it with virgin, for example, at a hundred and twenty gram. And the reason for that are twofold. One, it has to tick the performance box. So, do the 76 Mr Mehta, Transcript, at page 400. 77 Mr Mehta, Transcript, at page 243. 78 Mr Mehta, Transcript, at page 244. 79 Mr Fraser, Transcript, at pages 1014-1015. 80 Mr Seggar, Transcript, at pages 1140- 1142. 39 less grams perform equally as well as the heavier paper because it’s superior in terms of its makeup? So, it has to tick that box. And the second box it has to tick is the commercial viability. So, the price per square metre has to be competitive. Within the context of South Africa, my view is it is limited what you can substitute on two factors. One, we don’t have significant – we don’t have a range of grammage supplied into the market by Sappi that’s at the grammage level to substitute for the significant volume that you’re using in recycled paper. So, 80% of your volume that you use in the recycled to supply the industrial market sits between a hundred gram and a hundred and forty gram. In order to get the commercials to work on a price per square metre level you’d need lower than that. So, my question is is does Sappi produce virgin paper lower than a hundred gram? Would it? and the answer is no. So, you wouldn’t be able to substitute. The next sort of limiting factor on substituting virgin in for recycled is the availability of it. So, Sappi is extremely constrained in terms of capacity and what they supply to the market. So, all players, I hope and as I understand it, are on allocation. So even if you require significantly more volume from Sappi on virgin products they probably won’t be able to give it to you.”
[114]Mr Seggar also told us that Everest has an innovation team that investigates possible substitute products. In cross-examination by Corruseal’s counsel in relation to the potential substitution of recycled containerboard for lightweight virgin containerboard, Mr Seggar explained that there are certain substitutions that Everest makes in relation to short cold chains (i.e. meat cartons from abattoirs to retail, or short application shipping of bananas, with a shipping time of two of three weeks). 81 However, he explained that this is complicated by 81 Mr Seggar, Transcript, at page 1285. 40 moisture absorption rates, particularly in the agricultural markets. 82 Mr Seggar explained that on “deep sea exports, i.e. citrus or table grapes going to Europe and China you couldn’t do these substitutions” 83 . Moreover, substitution is not possible at the heavier end of the recycled grammage spectrum. Mr Seggar was not aware of any substitutions happening from 160g upwards. 84
[115]Mr Greeff’s analysis indicated that substitution was simply not feasible from a cost perspective. 85 Mr Greeff told us that while there may be some theoretical supply -side substitution bet ween the lighter grammage of UltraFlute Plus 112 for 120g recycled, at higher grammages of recycled than this, substitution of the sort the merging parties were contending for simply did not happen. 86 In addition, the Commission found that Ultraflute 112 and 120 are produced in limited volumes and so the substitution opportunities aren’t easily available in this scenario . 87
[116]Furthermore, Mr Greeff emphasised the context in which the hypothetical substitutions would occur in. If one were to look at, for example, 155g fluting for 120g UltraFlute plus, Mr Greeff explained this was an outlier because 155g fluting is too expensive to use in a carton in any circumstances. APL does not use this grade fluting in any of its board grade configurations. These outliers, Mr Greeff stated, do not reflect the common use of cartons and the common specifications in the market. Mr Greeff explained: 88 82 Mr Seggar, Transcript, at page 1282 . 83 Mr Seggar, Transcript, p1286 84 Mr Seggar, Transcript, at pages 1285-1286. 85 Mr Greeff, Transcript, at pages 1533 -1535. 86 Mr Greeff, Transcript, at page 1607. 87 Commission’s Expert Presentation, slide 7. Mr Wild also states “ I know the 112s are very limited. In fact at one point this year we were considering not making it at all, but we reversed that decision. But there is – ja, I don’t know the absolute number, but it is in the – it is the smaller proportion (of Sappi’s total production )” See Transcript, at page 645. 88 Mr Greeff, Transcript, at page 1612. 41 “MR GREEFF: The problem with that is that you must look at what generally happens. You won’t be able to sell a lot of those cartons into the market space because there’s not an application for it, what I’m trying to get to. (Indistinct) are lower grammages. If the argument is can you replace that by virgin, then that’s a proper argument, but to talk about an outlier on the top end, it’s true, you can substitute it by virgin, but it doesn’t prove anything related to the commercial application of substituting recycled paper by virgin paper.”
[117]Mr Thompson evidence was that demand-side substitution is done “only to a certain degree”. This is because “Recycled paper has a specific place in the market. Virgin paper has a specific in – place in the market. There is a level of overlap and through reconfiguring of cartons, et cetera, it can happen, but it doesn’t happen to a significant degree.” 89
[118]Dr Mandiriza’s assessment of the chain of substitution argument was that: “Despite Corruseal being in the market for that long there was no shred of evidence about these claims, chain of substitution that the economists claimed. And the only linkage that had to come through was through some of the factual witnesses, but at the time of writing the report it was very clear that it was a theoretical discussion and the evidence on record clearly shows that none of the parties deal with those kinds of things in that particular manner.” 90
[119]Mr Smith reinforced this view by pointing out that: 89 Mr Thompson, Transcript, at page 1356. 90 Dr Mandiriza, Transcript, at page 1796. 42 “when we actually look at the economic evidence and we look at what the factual witnesses have said we don’t find any substantial demand substitution that again would constrain the ability of the merged entity to engage in input foreclosure in regard to its 100 % recycled containerboard”. 91
[120]In the context of merger assessment, economic evidence can be useful in assisting the Tribunal to determine the issues, if it clearly articulates the application of economic reasoning to the facts of a case and not as abstract theory. Unfortunately, the chain of substitution argument in this case was used to produce smoke rather than light.
[121]Based on the above we find that the merged parties’ contention that demand- side substitution would constrain input foreclosure by the merged entity in respect of the supply of 100% recycled containerboard is not supported by the evidence. The ability of local producers, of virgin materials, to switch into the production of 100% recycled containerboard
[122]We considered the merging parties’ submission that supply-side substitution will constrain the ability of the merged entity to engage in input foreclosure in respect of 100% recycled containerboard. Mr Mehta told us that “the same paper machine can use different forms of fibre, either virgin fibre or virgin pulp fed into the machine or recycled fibre” and thus produce either virgin or recycled containerboard. 92 Mr Mehta also told us that, from a technical perspective Sappi Ngodwana could easily switch into the production of 100% recycled 91 Mr Smith, Transcript, at pages 1817-1818. 92 Mr Mehta, Transcript, at page 122. 43 containerboard at no cost and Mpact’s mill at Piet Retief could also switch into producing greater volumes of 100% recycled containerboard. 93
[123]We note that for supply-side substitution to be a competitive constraint, it must be shown that it is not only technically possible, but also that it is economically feasible and commercially likely.
[124]We were not persuaded by Mr Mehta because of the contradictory evidence of both Sappi and Mpact – the two firms that Mr Mehta told us could engage in supply-side substitution – indicated that such substitution is highly unlikely and would not constrain input foreclosure by the merged entity in respect of 100% recycled containerboard.
[125]Mr Thompson indicated that the demand for virgin products is growing faster than the demand for recycled products, at close to 5% per annum. 94
[126]Mr Wild told us that switching products on a machine is significantly more complicated than suggested by the merging parties, particularly in relation to the extraction of water, and the impact of the process on the efficiency of the machine 95 . Mr Wild explained: “So, you can change products on a paper machine, but if you move outside of the range of product that it was designed to make, you increasingly become less and less efficient and it makes it more and more expensive to make any product that’s outside of that scope, either in lost time or production on the machine or you have to start adding 93 Mr Mehta, Transcript, at page 401. 94 Mr Thompson, Transcript, at page 1335. 95 Mr Wild, Transcript, at page 539. 44 perhaps expensive additives to enable the machine to run closer to its design speed or capacity.” 96
[127]Mr Thompson also testified that there is a shortage of semi-chem fluting worldwide 97 and Mr Wild stated that while they are technically capable of producing recycled test liner at both Tugela and Ngodwana, their cost position relative to their competitors is substantially higher. Sappi’s costs to produce recycled containerboard are around higher than its competitors from a cost point of view. 98 Mr Wild further emphasised: “So while you can make products that don’t cover all of their variable and fixed costs for some period in time, once you make that proportion of products too big then you put the profitability of the entire asset base at risk.” 99
[128]Mr Thompson furthermore indicated that it would not be viable to run Piet Retief as a recycled mill. This is because Piet Retief (like the Sappi Ngodwana mill) has significant capital infrastructure and capital investments that support the production of virgin paper, including a chipping plant, digesters and the like. 100 Mr Thompson made it clear that Mpact’s firm strategy is to migrate Piet Retief to producing 101 102 Mr Thompson explained it to us in the following manner: “MR THOMPSON: So, yes, so Piet Retief Mill has a significant capital investment structure. It’s got a digest – it’s got a chipping plant, it’s got 96 Mr Wild, Transcript, at pages 533-534. 97 Mr Thompson, Transcript, at page 1335 98 Mr Wild, Transcript, at page 668. 99 Mr Wild, Transcript, at page 669. 100 Mr Thompson, Transcript at page 1357. 101 Mr Thompson, Transcript at page 1358. 102 Mr Thompson, Transcript at page 1414. 45 digesters. It’s got the paper machine and various ancillary bits of equipment. This structure has – comes with a significant overhead cost. Piet Retief is not able to run efficiently if it were to basically reduce the virgin content of its paper. So, quite simply Piet Retief cannot run at 100% recycled fibre input to produce a one hundred percent recycled fibre output that would be able to compete effectively in the market. The quality of what it would produce without pulp in it would not enable it to be competitive in the market to basically perform at the right level with the right quality specs, to the extent that we were trying to migrate Piet Retief to produce only a Hi-Pact liner and no semi-chem, again the economics of what would take place at the mill would basically say, yes, theoretically it could happen. You would produce a certain amount of Hi- Pact, but economically we would actually end up shutting the mill. And I think Mr Wild made that similar representation vis-à-vis Tugela. You take away from your primary product to make a product that you’re selling for a less economic value it just economically doesn’t give you the returns that you require, given the overhead structure that exists at a mill like Piet Retief or Tugela where you got significant pulping capacity that has to run, given the constraints that exist within the paper mill. So, let me just reemphasise, we would have to run the pulping capacity to be able to produce a sheet that could compete in the market and as soon as you do that and you then produce a sheet that has less virgin in it and you sell it at a lower price and you go the entire overhead cost that you’re busy recovering, and you’re selling your – you’re displacing a valuable product to make recycled products, it’s economic suicide. You would end up shutting the mill.” 103 103 Mr Thompson, Transcript, at page 1358. 46
[129]Mr Mehta conceded that it was necessary to have regard to the question whether it would make economic sense for Mpact or Sappi to engage in supply- side substitution, having regard to the relative value of the alternative output products. 104 We furthermore note Mr Mehta’s confirmation that a firm could not switch production from a product that is already contractually committed. 105
[130]Given the above, we conclude that while some degree of supply side substitution might be technically possible, there are significant technical and economic limitations to this type of switching. We consider it more likely than not that supply side substitution is unlikely to constitute a constraint on the ability of the merged entity to engage in input foreclosure in respect of the supply of 100% recycled containerboard to non-integrated corrugators. The barriers to small-scale entry and expansion to compete with the merger entity
[131]The merging parties submitted that Select-a-Box has invested in a small mill with a capacity of tons, and that Everest procures suppliers from Union Papers’ Eswatini mill (which has a similar capacity). 106
[132]We considered whether buying a smaller, older second-hand machine is feasible, however Mr Wild indicated that the cost of moving, stripping and reinstalling a second-hand machine is more than the cost of buying a machine. 107 Furthermore, given a scenario of switching products on a machine (for example, a machine that is designed to make virgin paper and its capability to switch to transform to make recycled paper) would be possible, however moving outside of the ranges of products the machine is initially designed to 104 Mr Mehta, Transcript, at page 404. 105 Mr Mehta, Transcript, at pages 516- 517. 106 FTI Expert Report, at para 238. 107 Mr Wild, Transcript, at page 564. 47 make, becomes increasingly more expensive or less efficient to run. 108 Furthermore, there are many other costs associated with a second-hand machine to turn it into an operational plant. Mr Greeff put it in the following manner: MR GREEFF: Well, we have got a recent case study in the form of Select-A-Box that was discussed and some of the feedback you received via the Chair. The problem with that is that you may start with a smaller capital investment buying second-hand equipment. There’s a reason why it’s second-handed obviously, because it’s old, antiquated, and redundant. So, then he’s upgraded it, by the time you’ve upgraded it you are approaching R 100 million in any circumstance but then you don’t have your waste [the input for recycled paper], you don’t have your pulp, you don’t have your effluent plant, you don’t have your water, you don’t have your energy. There’s all that into the equation. Also ongoing maintenance going forward, your ongoing upgrades going forward, so before you wipe out your eyes they [are] 250 or R300 million. So, to try and explain to anybody out there that you can arrive at a second-handed mill, buying it for 50 or R60 million, maybe able to buy the scrap that way, ja but to convert scrap into a proper mill soon, soon very quickly ended you with at least a R300 million investment with a mill with limited potential, limited capacity and a limited broad grade mix. How do you compete on such a mill? I would like to say if that was so possible and so easy many people would’ve done it already.” 109
[133]Mr Mehta conceded that, other than Everest and a corrugator called “Capstan”, there has been no new entry in the last five years. 110 108 Mr Wild, Transcript, at pages 532- 534. 109 Mr Greeff, Transcript, at pages 1536-1537. 110 Mr Mehta, Transcript, at pages 448-449. 48
[134]Overall, we were not persuaded that small scale entry removed the ability of the merged entity to harm downstream rivals. Conclusion on ability to foreclose
[135]Considering the six factors discussed above that are particularly relevant to our assessment of the merged entity’s ability to engage in foreclosure strategies, and after careful consideration of the views of the Commission, the merging parties and the intervenors, we conclude that there are a range of mechanisms through which the merged entity would have the ability to harm non-integrated corrugators. These include a total refusal to supply or reducing its supply or increasing the prices of 100% recycled containerboard.
[136]We were not persuaded by the merging parties’ arguments that there were features of the market which would prevent such mechanisms intended to harm non-integrated corrugators downstream from being implemented. We therefore conclude that the merged entity would have the ability to harm non-integrated corrugators downstream.
[137]In line with our approach to assessing vertical mergers, below we go on to consider the merged entity’s incentives. Incentive to foreclose
[138]We assess the merged firm’s incentive to foreclose downstream rivals in four steps: firstly, by setting out the merging parties’ views; secondly, by setting out those of the Commission and the intervenors; thirdly, by presenting our own assessment, and finally by presenting our conclusions on incentives. Merging parties’ views 49
[139]The merging parties submitted that three factors militate against a finding that the merged entity will have the incentive to foreclose. These are: 139.1. The quantitative analysis does not show an incentive to foreclose. The scenarios sketched in Mr Smith’s analysis are premised on an unsubstantiated factual matrix in which the merged entity increases its price of recycled containerboard by 10% and loses only 30% of its demand. 139.2. A risk diversification strategy is an important consideration when evaluating the merged entity’s incentive to foreclose. A diverse customer base for upstream paper mills allows a firm to limit its exposure to the risks of its vertically integrated downstream operations. Similarly, having a diverse range of suppliers means that downstream corrugators are less exposed upstream; and 139.3. There are factors which indicate Corruseal would opt not to expand corrugation capacity through internalisation of containerboard supply.
[140]The merging parties submitted that the merged entity would have no incentive to engage in a foreclosure strategy because it would be unprofitable for it to do so. They argued that any input foreclosure strategy would impose a substantial cost on Corruseal's operations in terms of profitability and efficiency, and risk profile.
[141]The merging parties submitted that the market behaviour of firms currently active in the market is to mitigate risk through supplier and customer diversification. For example, Mpact supplies third party corrugators and exports containerboard on an ongoing basis. Further, they argued that Corruseal has itself chosen to prioritise supplying customers over itself in market shocks (like that following the fire at the Enstra mill) and equally subscribes to the risk mitigation strategy of diversification over internal consolidation. 50
[142]The merging parties submitted that Corruseal is investing in a The project is expecte d to increase capacity of PM6 from tons to tons tons). 111 The merging parties also submitted that the mere fact that Corruseal is in the process of increasing upstream capacity runs counter to the idea that there is an incentive to foreclose downstream corrugators. The Commission and the interven or’s views
[143]In response, the Commission and the interven ors submitted that the merging parties have an incentive to foreclose downstream non -integrated corrugators because a foreclosure strategy is likely to be profitable (relying on the evidence of Mr Smith in which he posited three scenarios in which the price of recycled paper from the Rosslyn mill would be increased by 10% resulting in a loss of demand of 30%) and be in line with the stated strategy of the merged entity.
[144]The Commission and the interven ors submitt ed that the stated business strategy of the merged entity is that: 112 “the acquisition of a second paper mill will Fire and natural disasters are key risks to paper mills. For example, Corruseal experienced a fire at the Enstra mill in January 2021. The fire permanently damaged the PM3 machine in the mill, which produced around of paper per year. 111 Witness Statement of M r Mehta , at para 7.6. 112 Commission ’s Expert Report at para 241 . 51 Our assessment
[145]Below we set out our assessment of the merged firm’s in centive to foreclose rival non -integrated corrugators in three steps: 145.1. First, we analyse the benefits to the merged firm of foreclosure; 145.2. Second, we analyse the costs to the merged firm of foreclosure, separately both for partial and total foreclosure; and 145.3. Third, we discuss the quantitative analysis of foreclosure . The benefits of foreclosure
[146]We noted that pre -merger, the contestable market for FY2021 was as follows: 146.1. Approximately tons supplied by Neopak; 146.2. Approximately tons supplied by Corruseal; and 146.3. Approximately tons supplied by Mpact.
[147]Pre-merger, therefore, non -integrated corrugators can compete for a total of about tons. Corruseal downstream currently sources externally about .
[148]In contrast, post -merger, there is a benefit to the merged entity arising from a strateg y which internalises most of Neopak’s tons, with only excess 52 volumes offered to non -integrated corrugators ). The effect of this benefit is that volumes that were available to non -integrated corrugators will be reduced by Corruseal’s ne eds.
[149]Pre-merger Neopak is the only non -integrated supplier of 100% recycled containerboard and has no downstream corrugating activities, and therefore is unlikely to have the benefit from a foreclosure strategy. However, post -merger, when Neopak (as the only non -integrated upstream producer of 100% recycled containerboard) is owned by Corruseal (the largest downstream corrugator), its incentives would change. Post-merger, there would be a direct and obvious route by which the merged entity could benefit from internalising volumes.
[150]The Commission pointed us to evidence regarding Corruseal’s past conduct as it withheld supply following its acquisition of the Enstra Mill and allocated about % of its volumes to its downstream business. 113
[151]We attach weight to the merging parties’ internal documents suggesting that the merger will result in Corruseal consuming a large portion of Neopak’s volumes. In the words of Ethos: ” 114
[152]Similarly, in the Board meeting minutes of 24 August 2021 115 Corruseal speaks of the need to “ invest in capacity ” and to make up the “ PM3 Shortfall ”. 116 It is further recorded in the minutes that “ based on [the PM3 Shortfall] ” the Neopak transaction needs to be pursued. 117 113 Dr Mandiriza, Transcript , at page 1952. 114 Neopak Exit Investment Committee Paper, Bundle D , page 436 3, bullet point 3. 115 Exhibit 1 116 Mr Mehta, Transcript, at page. 266. 117 Mr Mehta, Transcript, at page. 272. 53
[153]We further note that Corruseal’s business model has historically centred around self-supply. A Neopak document dated July 2019, and confirmed as accurate by Mr Fraser, recorded that “Corruseal and Golden Era produce exclusively for internal consumption.” 118 We recognise that this position remained largely unchanged two years later. A Neopak document dated August 2020, and another one dated July 2021, recorded that “Corruseal and New Era produce nearly exclusively for internal consumption.” 119
[154]We note that Corruseal generally applies that self -supply model to those firms with which it merges. Mr Mehta was asked about the acquisitions Corruseal had made during the period 2017 to 2021. Mr Mehta listed the acquisitions and confirmed that in each instance, Corruseal had acquired the firms in question and then internalised sales volumes. 120
[155]The very rationale of the merger is, in Corruseal’s words, 121 We note that, implies that the merger will deprive non - integrated corrugators downstream of supply they would otherwise have been able to contest for when the market is constrained.
[156]We considered the merging parties submissio n that Corruseal is in the process of increasing upstream capacity and there is no benefit to foreclosing downstream corrugators. The merging parties submitted that the increase would comprise of: which would come from Corruseal upgrading PM6 and tons from PM3. 118 Neopak Rosslyn strategic overview FY20 -22 Presentation (July 2019), Bundle D, page 3243 , bullet point 4; and Mr Fraser, Transcript , at page 8 74. 119 Mr Fraser, Transcript , at page 8 75. 120 Mr Mehta, Transcript , at page s 178 – 182. 121 Witness Statement of Mr Mehta, para 17 . See also merging parties’ Joint Competitiveness Report, Bundle B, p ages B55 7 ;Mr Mehta, Transcript, at pages 93-94. 54
[157]However, we note that the evidence was that Corruseal was conducting a feasibility study for the In Mr Mehta’s words, Corruseal had requested a supplier in Germany to commence with “ an engineering study to get the process going.” 123 Mr Mehta further confirmed that it would be before the project would even be commissioned. 124 Our view is that in the circumstances there is simply no evidence that the project will be commissioned, let alone start.
[158]We are aware that currently . We note that Mr Mehta dealt with in his witness statement at paragraph 7.7 where he stated: 125 We were not persuaded by the merging parties’ arg uments regarding . As set out above, we reached this view on the basis that if . Furthermore, Mr Mehta had to acknowledge in cross -examination that the . 126
[159]The idea behind the analysis of incentives to foreclose is to predict the merged entity’s behaviour. Sometimes it may be possible to understand the incentives to foreclose directly from past conduct, business strategy and merger rationale. The above anal ysis of the merger firms’ internal documents clearly show s that 122 Mr Mehta, Transcript , at page 286. 123 Mr Mehta, Transcript , at page 78. 124 Mr Mehta, Transcript , at page 78. 125 Witness Statement of Mr Mehta, at para 7.7. 126 Mr Mehta, Transcript, at page 358 55 it would be strategically beneficial to foreclose downstream rivals. From this conclusion, it may not be nece ssary to try to infer the behaviour of merger firms from their financial incentives, however we consider the financial incentives in the quantitative analysis below as a cross -check.
[160]Based on t he above assessment, our view is that these benefits of foreclosure are likely to be substantial and would emerge relativel y quick ly. The costs of foreclosure
[161]We consider the costs of potential partial and total foreclosure strategies separately.
[162]On the costs of partial foreclosure (increasing prices of 100% recycled containerboard or reducing the merged entity’s supply to downstream non - integrated corrugators by first internalising), we were not persuaded that the merging parties could face substantial costs in terms of lost revenues if they pursued this strategy. Our view is that, if there are to be any costs at all, the magnitude of these costs is likely to be small because on a partial foreclosure strategy , excess volumes would be made available to non -integrated corrugators.
[163]We reach this view on the basis that Corruseal’s upstream operations, only produce approximately tons per annum and Corruseal downstream is a net purchaser of about tons per annum.
[164]We then considered the merging parties’ second point, that there are several alternatives to the merged entity in respect of 100% recycled containerboard. Our view is that these potential alt ernatives are unlikely to emerge in practice and that barriers to entry are high (see above our assessment on the ability of the merged entity to foreclose downstream rivals, in particular, the six issues considered in the assessment of ability to foreclos e). 56
[165]Our conclusion is therefore that the merged firm would likely experience only limited costs as a result of a partial foreclosure strategy.
[166]On the cost of total foreclosure, our view is that there are some opportunity costs to a strategy of total for eclosure because Neopak supplies approximately tons while Corruseal downstream is a net purchaser of about tons per annum.
[167]Total foreclosure of existing non -integrated corrugators customers will likely have an impact on the merged entity’s profitability. It would directly result in the loss of the revenues currently earned by Neopak from the non -integrated corrugators custome rs that the merged entity forecloses. However, there is a possibility that the merged entity could export the excess volumes. Recall, Mr Mehta’s evidence was that : “Exports would be there if you couldn’t get enough local sales and you would want to export the marginal difference. In the export market you would opportunistically make money higher than the local price, but only for very short periods of time where the market is at a peak .” 127
[168]For these reasons our view is that the merged firm would i ncur some costs if it was to engage in total foreclosure (the opportunity costs of not selling volumes that are in excess of its need) and, therefore, such a strategy was less likely than partial foreclosure. However, for the avoidance of doubt, we consid er that a total foreclosure strategy remains at the merged entity’s disposal in the scenario in which the merged entity exported volumes excess of its need. 127 Mr Mehta, Transcript, at page 110 . 57 Quantitative analysis
[169]Mr Smith submitted a quantitative analysis of the likely gains and losses to the merged firm of a partial foreclosure strategy directed at downstream non - integrated corrugators.
[170]Mr Smith postulated the following three simple, conservative scenarios of foreclosure, all of which would be profitable for the merged entity (i.e., there would be an incentive for the merged entity to engage in this behaviour). The starting point for the analysis, Mr Smith explained, is that, post -transaction, the merged entity will make around tonnes more of 100% recycled containerboard than it consumed in that year. 128
[171]The scenarios begin with an illustrative example, namely that if Neopak, as part of the merged entity, increased prices of recycled containerboard by only 10% or R1000, and if they were to lose, for example, around 30% or tonnes of demand, 129 then there would be at least three scenarios under which Corruseal would regain any initial loss of profits: 171.1. The first scenario is for Corruseal as the downstream corrugator to simply absorb that lost demand of approximately tonnes. Even with no increase in downstream pricing this would turn a R8 million “na ive upstream loss” into a R59 million profit for the merged entity. Mr Smith 128 RBB Expert Presentation , slide s 7 and 8. 129 Mr Smith, Transcript, at page 1940, in which he explains the basis for the assumptions made. Mr Smith, Transcript, at page 1940. Mr Mr Smith explained that these assumptions (30% margin) and the fact that a 10% price increase may result in a 30% drop in demand were checked against the available gross margin data and are broadly consistent (Mr Smith, Transcript, p age 1940 ). Furthermore, Mr Smith explained that Prof Theron’s suggestion of an elasticity of minus one is “simply impossible” as a matter of economic theory. Mr Smith also demonstrated that on Prof Theron’s elasticity of minus one, the incentives are even greater: there would be virtually no upstream loss and hence a massive incentive to engage in anticompetitive input foreclosure (Mr Smith, Transcript , at pages 1978 -1980). 58 told us that given the history of Corruseal’s aggressive acquisition and expansion, this seems to be the most likely outcome. 171.2. In the second scenario, the non-integrated corrugators could pass through just half of the increase in costs (with no downstream diversion). This would result in an additional R83 million profits. 171.3. In the third scenario, if the non-integrated corrugators passed through just over 5% of this R1,000 increase in costs (no downstream diversion), the profit to Corruseal in the downstream market would entirely outweigh the “naïve” upstream loss. 130
[172]Mr Smith submitted that, the potential for upstream loss is quite limited; however, the opportunity for downstream gain is very much more significant. 131
[173]We note that the quantitative analysis undertaken by Mr Smith was a high-level one that considered a number of indicative partial foreclosure scenarios in order to assess the broad magnitude of the potential costs and benefits of partial foreclosure and served as a cross-check on our qualitative assessment.
[174]We observe that for the partial foreclosure scenarios we considered, the results of Mr Smith’s analysis were consistent with those of our qualitative assessment; namely that the benefits of partial foreclosure are likely to be substantially greater than the costs. As a result, we found that, to the extent it is possible to place any weight on a quantitative analysis of incentives, this provides further support for the conclusions of our qualitative assessment.
[175]We did not accept the merging parties’ submission that the quantitative analysis was built on unsubstantiated assumptions and therefore did not properly assess the merged firm’s incentive to foreclose. Our view was that Mr Smith sufficiently 130 RBB Expert Presentation, at slide 35. 131 Mr Smith, Transcript, at page 1944. 59 explained his assumptions as arising from economic theory. Mr Smith explained that these assumptions (30% margin) and the fact that a 10% price increase may result in a 30% drop in demand were checked against the available gross margin data and are broadly consistent. 132 Furthermore, Mr Smith explained why Prof Theron’s alternative suggestion of using elasticity of minus one is “simply impossible” as a matter of economic theory. Mr Smith also demonstrated that on Prof Theron’s elasticity of minus one approach, the incentives are even greater: there would be virtually no upstream loss and hence a massive incentive to engage in anticompetitive input foreclosure. 133
[176]In summary we note the quantitative analysis of the likely gains and losses to the merged firm of a partial foreclosure strategy and our view is that all the scenarios considered in the quantitative assessment support our qualitative assessment. Conclusions on the incentive to foreclose
[177]In order to assess whether the merged firm would have the incentive to foreclose its rival non-integrated corrugators, we assessed the benefits and costs it would face from implementing these foreclosure strategies.
[178]In summary, our qualitative assessment of the benefits to the merged firm of foreclosure indicated that this would allow it to obtain a number of benefits.
[179]The costs for partial foreclosure would be limited. This is because the merged entity is unlikely to lose substantial revenues because market participants are dependent on it, and alternatives are weak and not practical. In contrast, the merged firm would face some opportunity costs from a total foreclosure strategy 132 Mr Smith, Transcript, at page 1940. 133 Mr Smith, Transcript, at pages 1978-1980. 60 from a loss of revenues arising from exce ss volumes, but these were not substantial because the merged entity could, in alternative export these excess volumes.
[180]We noted that the quantitative assessment also broadly confirmed our qualitative assessment of the merged entity’s incentives.
[181]In conc lusion, on the basis of the above, the merged entity is likely to have the incentive to foreclose its rival non -integrated corrugators. Effects of foreclosure
[182]In light of our findings above that the merged firm would have both the ability and incentive to engage in a foreclosure strategy, we assess whether foreclosure of rival non -integrated corrugators would likely result in harm to competition.
[183]We do not accept the merging parties’ submission that there is no evidence to suggest that a foreclosure strategy would bring about harm in the downstream market. The merging parties submitted that is because Neopak supplies 7 customers other than Corrusea l. Four of these are exclusively narrow -width customers (and approximately 50% of the fifth customer’s purchases are narrow widths) which are very unlikely to be foreclosed, given Mpact’s significant exports of these widths. The merging parties also subm itted that these customers are, in any event, small operators. The merging parties further submitted that of the remaining two customers (APL and Everest), and there is nothing to suggest that it would exit the m arket if foreclosed and Everest has shown itself to be quite capable of meeting shortfalls of supply through import channels.
[184]In an asset review dated 13 October 2020, Ethos stated that Neopak was a 61 . 134 Ethos acknowledged that ” . 135 Mr Fraser acknowledged that: “It is the preference of smaller convertors to keep an independent paper mill as you’ve seen by like APL intervening. I don’t disagree that that’s their preference.” 136
[185]Our view is that the proposed merger would bring about a substantial change to the market structure. It would remove Neopak, the only independent or “outside option” for rival non -integrated corrugators. Our view it that the proposed transaction is a verti cal merger that eliminates the only remaining viable long -term source of supply for rival non -integrated corrugators. In other words, the fact that rival non -integrated corrugators will be harmed by the merger necessarily means that it will result in harm to competition.
[186]It is appropriate for us to carefully consider the views of the merged entity’s customers (i.e. , the rival non -integrated corrugators ) because we are mainly assessing vertical effects, and therefore the possibility that the competitiven ess of these rivals, as customers of merged entity, could be directly harmed by the merged firm. As regards APL, Mr Greeff told us that if the merger were to be approved: 137 134 Ethos Presentation titled “Neopak Asset Review” dated 13 October 2020, Bundle D, p ages D4374 - 76; and Mr Fraser, Transcript, at page 847. 135 Mr Fraser, Transcript, at page 848. 136 Mr Fraser, Transcript at page 851. 137 Mr Greeff , Transcript, at page 1546 . 62 ”
[187]Our view is that the exit of existing rival non-integrated corrugators will lead to a significant loss of competition in the market that is ultimately to the detriment of consumers.
[188]In addition, we do not accept the merging parties’ submission that rival non- integrated corrugators are “in any event, small operators” and therefore we should not be concerned.
[189]In assessing the effects of the merger on competition, we found that without the merger, Neopak plays a critical role in enabling competition between the non- integrated corrugators. This reliance on Neopak, may enable the merged entity to influence competition through a number of mechanisms which, if implemented, would likely affect competition in the short and long-term.
[190]Considering all of the evidence in the round, we conclude that post-merger Corruseal’s ownership of Neopak would be used to disadvantage rival non- integrated corrugators and/or favour Corruseal. This would result in an immediate loss of rivalry with a longer-term effect on competition, including a loss of head-to-head competition between the merged entity and its rival non- integrated corrugators. Conclusion on effects of foreclosure
[191]Overall, and subject to our assessment of countervailing factors (through merger efficiencies or through the entry and/or expansion of third parties in 63 reaction to the effects of a merger) below, we conclude that a foreclosure strategy would result in a li kely SPLC . 138 Efficiencies
[192]While mergers can harm competition, they can also give rise to efficiencies. Efficiencies arising from the merger may enhance rivalry, with the result that the merger does not give rise to a likely SPLC .
[193]To form a view that the claimed efficiencies will enhance rivalry so that the merger does not result in a likely SPLC , our approach was to consider the following: 193.1. The efficiencies must be timely, likely and sufficient to prevent a likely SPLC from arising; and 193.2. The efficiencies must be merger specific, i.e., a direct consequence of the merger, judged relative to what would happen without it.
[194]The merging parties claimed that post -merger, the y, through efficient allocation of orders between the Enstra and Ros slyn mills, are able to realise an extra tons of recycled containerboard 139 as a result of optimising the deckle on the Corruseal and Neopak machines.
[195]The Commission pointed us to the fact that of those tons were going to be passed on to just a single customer, Everest. This would leave about tons for the rest of the market. 138 We are of the view that the conditions of competition under the counterfactual would not be materially different from the pre -merger conditions of competition. The merging parties, the Commission and th e interv enors did not suggest that in the counterfactual , conditions of competition would be materially different to the pre -merger situation . 139 Mr Mehta, Transcript at p age 95-97. Witness Statement of Mr Mehta , at para 18.3 and Annexures MM2 to MM4 at pages 160-162 of Bundle A . 64
[196]In our assessment, we note that Neopak’s contestable volumes tons) will reduce by the amount that Corruseal downstream currently sources tons). Therefore, the contestable volumes will reduce from tons to tons as a result of a foreclosure strategy.
[197]We noted that this figure must then be supplemented by the claimed efficiencies of tons and include Corruseal’s third party sal es of tons. 140 In total, if the parties were to pursue a foreclose strategy, the merged entity will sell approximately tons to third parties. This contrasts with the combined pre -merger third party sales of tons tons from Neop ak and tons from Corrusea l). 141
[198]Mpact’s sales into the non -integrated corrugators market would likely remain unchanged at tons. Therefore, total sales into the non -integrated corrugators market would reduce from pre -merger volumes of tons to approximately tons.
[199]Mr Mehta acknowledged that efficiences on their own are not the basis for pursing the transaction. 142
[200]Based on the assessment above, our view is that even with these claimed efficiency volumes, the proposed mer ger would result in a likely SPLC .
[201]The merging parties submitted that in addition to the efficiency volumes, the merger will result in increased lead times and improved quality of paper due to Neopak's expertise 143 and that common customers of Neopak and Corruseal 140 Commission’s Expert Report, Table 14 . 141 Commission’s Merger Report, at paras 129 to 131. 142 Mr Mehta Transcript at page 344 . 143 Witness Statement of Mr Mehta at para 18 ; Mr Mehta, Transcript , at page 94. Neopak Rosslyn Mill Strategic Overview FY21+ at page 3342 of Bundle D . 65 will also benefit from combining their respective volumes to qualify for larger rebates. 144 The merging parties also submitted that the transaction will result in a sharing of costs and will lead to larger scale research and development operations, all to the benefit of the customer and end consumers. 145 However, we noted that the merging parties did not quantify the benefits of these additional claimed efficiencies nor set out whether these would enhance rivalry. Conclusion on efficiencies
[202]In conclusion, on the basis of the above, our view is that in the absence of any evidence from the merging parties on the significance of these claimed efficiencies, we do not consider that efficiencies would mitigate the anticompetitive effects of the merger. Barriers to entry and expansion
[203]Having found that as a result of vertical effects the proposed transaction would likely result in a SPLC, we considered whether market entry or expansion might prevent a likely SPLC.
[204]In assessing whether market entry or expansion might prevent a likely SPLC, our approach is to consider whether such entry or expansion would be: (a) timely; (b) likely; and (c) sufficient.
[205]We note that the merging parties conceded that barriers to entry in the upstream supply of 100% recycled containerboard are indeed high if one wants to invest at scale. The Commission pointed us to a Neopak internal document dated 144 Prof Theron, Transcript, at page 2065. 145 FTI Expert Report, at para 334. 66 July 2019 that records: “Market has relative protection from new entrants due to high CAPEX barrier to entry ~R1bn.” 146
[206]We note that the capital requirements for the establishment of a recycled paper mill of sufficient scale are very significant. SAPPI has estimated that a mill with a capacity of between 60,000 and 70,000 tonnes would cost in the order o ; New Era has estimated that a facility of 100,000 tonnes (slightly less than that of the Rosslyn mill) would be in the order of ; and Mpact has estimated that to replicate its Felixton mill would cost approximately R billion. 147
[207]Our view is that entry at scale would be very costly, difficult to achieve and not timely 148 thereby resulting in high barriers to entry and expansion.
[208]We also considered small scale entry facilitated through second-hand machinery, but as discussed above we find that it is not likely that non- integrated corrugators could acquire “smaller” facilities to effectively compete. Conclusion on barriers to entry
[209]Based on our assessment above, it is our conclusion that barriers to entry and expansion in the upstream supply of 100% recycled containerboard are substantial. Entry and/or expansion would not be timely, likely, and sufficient to mitigate the likely SPLC. 146 Neopak Rosslyn strategic overview FY20-22 Presentation (July 2019), Bundle D p3243 and Transcript page 875 – 877. 147 Commission’s Expert Report, Table 18. 148 SAPPI estimates that it will take approximately 6-12 months to compete in the market. Mpact estimates it would take 2-3 years for an entrant to enter the market and New Era estimates 2-5 years for a new entrant to enter and compete. 67 Conclusion on vertical effects
[210]In our competitive assessment, we considered vertical theories of harm: we considered the merged entity’s ability and incentives to foreclose rival non- integrated corrugators, and the potential effects on competition of a foreclosure strategy.
[211]Based on an assessment of the vertical theories of harm, and taking into account the likely effects overall, we conclude that the merger between Corruseal and Neopak is likely to result in a SPLC as a result of vertical effects. Horizontal effects
[212]We considered whether the proposed transaction would likely result in a SPLC as a result of horizontal coordinated effects or horizontal unilateral effects.
[213]On horizontal unilateral effects, the Commission’s case was that the merged entity will acquire the ability to profitably raise prices and / or degrade the quality of containerboard supplied to corrugators. We considered many of the issues highlighted in this theory of harm as part of our vertical assessment.
[214]On horizontal coordinated effects, the Commission submitted that the 100% recycled containerboard market is highly concentrated, and any merger will lead to even higher concentration. In addition, the Commission submitted this market is highly transparent and the merger will increase the transparency. According to the Commission this level of transparency has in the past led to a history of collusion in the market for the manufacture and supply of corrugated packaging. Our view is that this theory of harm lacks the detail required to properly assess whether horizontal coordinated effects are likely as a result of the proposed transaction. 68 Conclusion on horizontal effects
[215]Overall, we make no conclusion on the whether the proposed transaction might result in a likely SPLC as a result of horizontal unilateral effects. Remedies
[216]Having concluded that the proposed merger will likely result in a SPLC as a result of vertical effects, we considered whether remedies could cure the SPLC.
[217]The merging parties tendered various iterations of proposed behavioural conditions, the first during the Commission’s investigation, which was supply condition for a limited duration of two years. The Commission however found that the tendered behavioural conditions do not address the fundamental and permanent structural shift in the recycled containerboard paper market in South Africa and therefore the SPLC brought about by the proposed transaction. The Commission also noted that third parties raised significant concerns regarding the sufficiency of pricing and supply conditions in remedying the concerns arising from the merger.
[218]After the Commission's prohibition of the proposed transaction, the merging parties tendered a further set of behavioural remedies in an attempt to respond to the Commission’s identified competition concerns. Ultimately, the merging parties requested that the Tribunal unconditionally approve the proposed transaction, alternatively, approve the proposed transaction in line with a set of behavioural conditions tendered by them. We consider the ultimate set of conditions proposed by the merging parties. 69 Merging parties ’ tendered conditions
[219]These conditions tendered have two components (i) a volume component; and (ii) a price component. Volume component
[220]The merged firm commits to supply its recycled containerboard customers with recycled containerboard for a period of five years i.e., from October 2022 to September 2027. Furthermore, the merged firm commits to supply additional volume s to customers , from the baseline of the October 2022 volume supplied per customer - the so -called "synergy upside" volumes, being the alleged efficiency gains of tons, will be made available post-merger . This additional volume will be capped per cus tomer, as explained below.
[221]The merging parties submitted that the proposed volume remedy will be implemented, in practical terms, as follows: 221.1. During each yearly contract negotiation, a customer will be able to oscillate between its baseline volume - the O ctober 2022 volume - and the greater of (i) 20% additional volume; or (ii) 1 000 tons additional volume. 149 221.2. Should a customer drop below the baseline volume, the merged firm will make its best endeavours to supply the baseline volume during the following c ontract year provided that such volume has not already been re - allocated during the normal course of business. 149 Customers will be ab le to increase their volume allocation by this synergy upside on a pro rata basis within three months after the implementation date of the proposed transaction. In order to realise the synergy upside, the merged firm will determine which mill will service the relevant orders. 70 221.3. Any force majeure event on the part of a customer will not affect their volume allocation for the next contract year and any force majeure event on the part of the merged firm will lead to a proportionate reduction of volumes (of both external corrugators and internal Corruseal customers). Price component
[222]The tendered price condition takes the form of a "price cap" or a maximum selling price that the merged entity will charge to recycled containerboard customers post-merger – for a specified, limited period.
[223]In terms of its duration, the tendered price cap will endure for a four-year period from 1 October 2022 to 30 September 2026. From 1 October 2026, the price cap falls away and the price will be negotiated.
[224]The merging parties submitted that the price remedy will in practical terms be implemented as follows: 224.1. The October 2022 prices will be used as a baseline and will be adjusted yearly by making use of a cost driver formula. 224.2. The merging parties identified various cost driver components and attached a specific weight to each. According to the merging parties, the cost driver components are reflective of a proportional split of the actual cost drivers of the Neopak business. They submitted that a similar model has been used before by Neopak in long term contracts (i.e., longer than a year) concluded with APL and Corruseal. merging parties 224.3. The price cap will increase or decrease in line with the cost drivers and the metrics linked to the cost drivers.
[225]The merging parties argued that the cost drivers are linked to objective metrics, such as actual invoices received, the Producer Price Index (“PPI”) and the 71 Consumer Price Index (“CPI”). We n ote that the merging parties in the tendered pricing conditions al located a weight to a basket of costs labelled ‘Labour, Overheads, Depreciation, Capex and Other’ which approach and weight were queried by the customers that testified, as discussed bel ow.
[226]The merging parties furthermore included a non -discrimination clause in the price condition tendered.
[227]They further contended that the large majority of the respective customers are comfortable that the proposed conditions remedy any concerns which may exist post-merger. However, they did not call any customers to testify. Commission and intervenor’s submissions
[228]The Commission argued that, based on the evidence, the general view of customers was that they are opposed to the proposed merger, even if approved subject to the proposed behavioural conditions. It argued that the merging parties were unable to find a si ngle witness willing to support the tendered ‘generic’ behavioural remedy offered to the market. It submitted that the general view expressed by customers was that that the tendered behavioural remedy simply does not serve as a proper substitute for an in dependent Neopak negotiating with independent downstream firms. In the words of Mr Seggar, the benefit of having Neopak independent is that it assists in “ keeping integrated players honest ”. 150
[229]The Commission advanced several reasons why the proposed behavioural conditions do not address the competition concerns resulting from the proposed transaction. 150 Mr Seggar, Transcript , at page 1145. 72
[230]First, it does not address the structural problem brought about by the proposed transaction, i .e., the post -merger removal of the only independent supplier to customers. Greeff testified, “ The merger has got major structural negative or downside impacts on the industry .” 151 The Commission argued that structural competition concerns require structura l remedies, and that they are not cured by behavioural remedies. “ … as long as you speak … about the remedy you don’t address the structural downside effect of the merger. You should start there. The remedies doesn’t sort out anything .” 152
[231]Second, both the volume and price commitments are of limited duration, whereas the adverse effect of the merger is a long -term structural alteration of the market. As indicated above, the tendered pricing component of the tendered conditions is for a four -year period and the volume commitment is for a five-year period. Once that four -year period is over, the merging parties will be at liberty to charge supracompetitive prices and once the five -year period is over, the merged entity will be free to withhold supply from it s downstream rivals. Furthermore, customers do not have a viable alternative source of supply to switch to.
[232]Third, it is common cause that the baseline price is determined with reference to prices as of October 2022 . 153 The Commission contended that the b aseline off which the suggested price cap is to be calculated, is unreasonably high and, in effect, the currently high prices will form the basis of all future prices in terms of the proposed remedy.
[233]Fourth, the factual witnesses raised concerns regarding the weight given by the merging parties in the proposed pricing conditions to certain cost drivers. The Commission noted that the tendered conditions attribute as much as of the 151 Mr Greeff, Transcript , at page 154 1. 152 Mr Greeff, Transcript , at page 1546. 153 Transcript , at pages 42, 43 and Mr Fraser, Transcript, 1094. 73 post-merger price to a “basket” of the merged entity’s costs, i.e., Labour, Overheads, Depreciation, Capex and Other. Mr Greeff testified that in the pre - existing contract between Neopak and APL that figure is only . He was therefore concerned that the merging p arties increase d that figure in the proposed conditions to . 154 Mr Greeff also noted that part of that weight was the merging parties ’ profit margins that would in terms of the proposed conditions be “ very much open to manipulation”. 155 His evidence was that the lack of transparency flowing from a globular weight, read with the items included in the basket, raised concerns: “ Now in overheads you can hide anything, from profit margin to anything which we don’t have any control about …” 156 and “So we have to question mark why all of a sudden is there such a weight to something which you really can’t measure out there or control out there .” 157
[234]Fifth, the Commission argued that on the evidence, the merging parties offered different behavioural remedies to different customers and that creates further competition problems. This is because if corrugator X (in this case Everest according to the Commission and Intervenors) are offered substantially better terms than corrugator Y (other customer(s) of the merging parties ), then corrugator Y is arbitrarily prejudiced and that distorts actual competition in the marketplace. A competition regulator cannot endorse merger conditions that treat customers dissimilarly, thereby creating new competition problem s, argued the Commission. As it transpired during the hearing, Corruseal offered Everest a more favourable deal shortly before the commence ment of the hearing, which deal would become effect ive if the merger is approved. 154 Mr Greeff, Transcript , at page 1545. 155 Mr Greeff, Transcript , at page 1545. 156 Mr Greeff, Transcript , at page 1647. 157 Mr Greeff, Transcript , at page 1647. 74
[235]Naturally, the Commission was highly concerned about and critical of what was referred to as a “side deal”, being struck between the merging parties and Everest. In the course of proceedings, the Commission alleged that such customer influencing, specifically of a Commission witness, has adverse implications for competition in the market. We share the Commission’s and Intervenors’ concerns in this regard, and we deal with this aspect below.
[236]Similarly, the Intervenors raised various concerns with the proposed remedy, that overlap with the abovementioned issues raised by the Commission. They ultimately argued that the tendered behavioural conditions do not address the competition concerns and therefore that the proposed merger ought to be prohibited. Our assessment of the proposed conditions
[237]It is trite that a suitable remedy is one that is likely to avoid a SPLC that would otherwise occur, relative to the counterfactual.
[238]In Imerys the Competition Appeal Court (“CAC”) clarified that where the Tribunal determines that the merger is likely to cause an SPLC the issue of an appropriate remedy is one of discretion by the Tribunal: “Where … the Tribunal is asked to approve the merger with conditions rather than prohibit it, the choice of remedies is in the nature of a discretion.” 158 (emphasis added)
[239]The CAC held further “I reject the proposition that the Commission bears the burden of proving that the proposed conditions will not adequately address the likely SPLC. The Tribunal has the power to prohibit the merger if it is not satisfied that the conditions will adequately remedy the likely SPLC. And 158 Imerys South Africa (Pty) Ltd and Another v Competition Commission (147/CAC/Oct16, IM013May15) [2017] ZACAC 1 (2 March 2017) at para 40. 75 regardless of where the onus lies in respect of proposed conditions (if it is accurate to speak of onus at all), I do not think that the Tribunal is obliged to approve a merger just because it finds it more probable than not that the conditions will neutralise the likely SPLC.” 159 (emphasis added)
[240]The CAC went on to point out “One should bear in mind, in this regard, the real problem in such cases will not necessarily be competing views as to the probable future state of the market but an inability to make reliable predictions at all. I think it is permissible for the Tribunal to reason thus: ‘The merger will likely give rise to an SPLC. Although the proposed conditions are more likely than not to remedy the likely SPLC, there is a reasonable possibility that they will fail to do so. Therefore we prohibit the merger.’” 160 (Emphasis added)
[241]The CAC also explained how the Tribunal could exercise its discretion over remedies: “In exercising its discretion, the Tribunal could be expected to take into account, on the one hand, the precise likelihood and extent of the SPLC; and, on the other, the precise extent of the risk that the conditions will fail to remedy the likely SPLC. The public interest may also enter into the balancing exercise, particularly the public importance of the markets which would be directly or indirectly prejudiced if the conditions failed to remedy the likely SPLC.
[242]The CAC has further held that the risk of inadequate relief i.e., uncertainty as to the preservation of competition, should not be borne by consumers. 161
[243]It is with the above guidance of the CAC that we assess the merging parties’ tendered behavioural conditions. 159 Imerys at para 40. 160 Imerys at para 40. 161 Imerys at para 42. 76
[244]We note that in the context of mergers which are likely to lead to a SPLC, it is well-established that structural remedies are preferable to behavioural conditions to address the SPLC. 162 International competition jurisdictions in their respective merger guidelines deal with the inappropriateness of behavioural conditions as a means to address a structural competition problem.
[245]The UK CMA explains: 163 “[Structural remedies] are normally preferable to measures that seek to regulate the ongoing behaviour of the merging parties (so-called behavioural remedies, such as price caps, supply commitments or restrictions on use of long term contracts). Behavioural remedies are unlikely to deal with an SPLC and its adverse effects as comprehensively as structural remedies and may result in distortions when compared with a competitive market outcome.” The UK Merger Remedies Guidelines further state at paragraph 2.11: “Particular types of behavioural remedy such as … supply commitments control or restrict the outcomes of business processes. These aim to control the adverse effects expected from a merger rather than addressing the source of the SPLC. This type of remedy may not only be complex to implement and monitor but may also create significant market distortions.” (emphasis added)
[246]The European Commission adopts a similar approach. It states “[c]ommitments relating to the future behaviour of the merged entity may be acceptable only exceptionally in very specific circumstances”. 164 (emphasis added)
[247]The United States Department of Justice expresses its views on behavioural (conduct) remedies as follows: 165 “Conduct remedies substitute central decision making for the free market. They may restrain potentially procompetitive 162 See, for example, the UK CMA’s Merger Remedies guidance (December 2018) at para 3.5(a). 163 Ibid. 164 European Commission’s notice on remedies (22 October 2008, Official Journal of the European Union),at para 17. 165 US Department of Justice Merger Remedies Manual, at page 4. 77 behavior, prevent a firm from responding efficiently to changing market conditions, and require the merged firm to ignore the profit-maximizing incentives inherent in its integrated structure. Moreover, the longer a conduct remedy is in effect, the less likely it will be well-tailored to remedy the competitive harm in light of changing market conditions. Conduct remedies typically are difficult to craft and enforce. For these reasons, conduct remedies are inappropriate except in very narrow circumstances” and “‘Regulating conduct is inadequate to remedy persistent harm from a loss in competition.” (emphasis added)
[248]The Australian ACCC in its Merger Guidelines states that it has a strong preference for structural undertakings — that is, undertakings to divest part of the merged firm to address competition concerns since structural undertakings provide an enduring remedy with relatively low monitoring and compliance costs. It further states that behavioural remedies are rarely appropriate on their own to address competition concerns. 166
[249]It is common cause that customers raised concerns regarding the proposed transaction from both a security of supply (i.e. volume) and post-merger (increase in) price perspective. Mr Mehta explained in his evidence that Corruseal has tendered the conditions in an attempt to address the “fear, and … comments coming out of the Commission and certain key customers …”. 167
[250]Customers in their evidence rejected the merging parties’ proposed behavioural remedy. Both Mr Greeff of APL and Mr Seggar of Everest testified that they do not regard the tendered remedy as an acceptable condition to cure the competition concerns. Mr Greeff furthermore dealt with the likely adverse impact should the proposed transaction be approved subject to the proposed 166 Australian Competition and Consumer Commission’s (ACCC’s) November 2017 Merger Guidelines, Appendix 3: Undertakings, at paras 11 and 12. 167 Mr Mehta, Transcript, at page 107. 78 conditions. 168
[251]We record that Mr Seggar of Everest at the time of his witness statement held the following view about behavioural conditions: “ Everest is of the view that the adverse impact of the proposed transaction on the structure of the packaging industry in South Africa cannot be addressed by the imposition or behavioural remedies related to security of supply. Such conditions however constructed are notoriously difficult to enforce, largely incapable of providing timely relief to firms, which may be negatively impacted by transgressions and are temporary at best .” 169 (emphasis added).
[252]However, as indicated, he was offered a more favourable deal by the mergin g parties shortly before the hearing . During his testimony, the fact that he had a favourable deal came to light.
[253]On 14 November 2022, Mr Seggar for the first time, informed the Commission that he was not available to testify as scheduled because he had travel plans to Mozam bique with staff and involving customers. 170 Subsequently, to the Commission’s surprise, the merging parties indicated that Mr Seggar was available to testify. It transpired that Mr Mehta had called Mr Seggar on 14 November 2022 to find out why he was not testifying . In Mr Mehta’s words: “ I was quite surprised because I had spoken to him and he had not mentioned any travel or the fact that he wasn’t available. In fact if you go back to the correspondence of the 10th and possibly on the 11th there was a discussion that he was going back – he was going to be on the witness stand on that 168 Mr Greeff, Transcript , at page 1546. 169 Witness St atement of Mr Seggar, at para 59. 170 Transcript, at pages 949 -953. 79 Monday. So I was just surprised that he wasn’t.” 171 Thus, the merging parties had been discussing the Commission’s witness’ upcoming testimony with him just days before he was scheduled to take the stand. Following that approach by the merging parties, Mr Seggar underwent an about turn and was suddenly again available to testify as required. 172
[254]Ultimately, Mr Seggar testified that he would be reluctant to agree to the conditions tendered “So, I think my testimony was as follows, that if I was offered purely the generic remedy I would be reluctant to agree to it based on the issues I raised yesterday. What motivated us to agree on a remedy was the material differences that we’ve highlighted.” 173 He refers in the latter sentence to the bespoke remedy that was offered by the merging parties to Everest should the proposed transaction be approved (discussed below).
[255]In terms of its effects, the proposed merger will bring about a permanent structural change in the market, and the anti-competitive threat it poses is likewise one of indefinite duration. The behavioural conditions offered by the merging parties do not address the competition concerns associated with the proposed merger because, as behavioural conditions, based on a volume commitment (for a limited period of time) and a price cap (also for a limited period of time), they do not address the permanent structural change brought about by the proposed merger in the market and therefore do not address the SPLC in the market over the long-term.
[256]There is no basis on which, on the evidence, we can conclude that after five years (or even a longer period of behavioural conditions should that be considered), the competition concerns with the proposed merger will dissipate. 171 Mr Mehta, Transcript, at page 1728. 172 Transcript, at pages 1054-56 and email from the Commission to the Tribunal dated 14 November 2022 regarding Mr Seggar’s availability. 173 Mr Seggar, Transcript, at page 1306. 80 Mr Smith in our view accurately summarised why in this context the proposed behavioural conditions do not address the competition concerns resulting from the prosed merger and are not an appropriate remedy: 174 “… it appears unlikely that new entry, or some other substantial change in market structure will occur in the next two or five years. Barriers to new entry in the production of 100% recycled containerboard appear to be substantial, and accordingly new entry seems unlikely. If there is not any new entry nor any other substantial pro-competitive improvement in market structure, then it is difficult to see how a two-year 175 pricing commitment would remove the likely effects of the merger on non- integrated corrugator customers. Similarly in regard to the five-year volume commitment, if there is not any new entry nor any other substantial pro-competitive improvement in market structure, then it is difficult to see how such a five-year pricing commitment would remove the likely effects of the merger on non- integrated corrugator customers.”
[257]In the circumstances, any behavioural conditions, if considered, would have to be of an indefinite duration. However, indefinite behavioural conditions are inappropriate and impractical to impose, and would be unreasonably burdensome on the competition authorities to monitor and enforce. We further note that the CAC in Imerys held “If the merger is conditionally approved and the conditions turn out to be inadequate to neutralise the SPLC, the harm cannot be reversed. If, on the other hand, the merger is prohibited and with the 174 RBB Economics Expert Report, at paras 206-208. 175 The same applies to a behavioural price cap condition with a five-year duration as ultimately tendered by the merging parties. 81 passing of time it becomes clear that the merger will no longer give rise to SPLC, the transaction can be renewed.” 176
[258]Above we highlighted various jurisdictions’ merger guidelines warning that behavioural remedies are difficult to craft, complex to implement and monitor and to adequately enforce. They may also create significant market distortions. This matter is a case in point. The enforcement difficulties with the tendered conditions are demonstrated inter alia by the design flaws pointed out by the customers that testified.
[259]The evidence has borne out that the price cap proposal will be an ineffective remedy since the proposed base price utilised, is set when prices are high. Recall that the merging parties’ tendered pricing condition selects the price of October 2022 as the base price. When asked why this (high) baseline price of October 2022 had been selected by the merging parties in their proposed remedy, Mr Fraser was not able to give a satisfactory explanation, other than to say “okay, this is like the get-go. From here, every future year I’m going to take this price and now increase it by the movement and drivers.” 177 Mr Fraser conceded “yes, the paper pricing at the moment is high” but contended that the remedy would act as a “cap” and that customers would still have the option to attempt to negotiate a lower price. 178 For obvious reasons, we do not regard the option of post-merger price negotiating as a remedy at all, since it is not responsive to the structural change to the market brought about by the proposed transaction and therefore the competition concerns associated with the proposed merger. 179 176 Imerys at para 41. 177 Mr Fraser, Transcript, at page 1094. 178 Mr Fraser, Transcript, at page 1096. 179 See evidence of Mr Fraser, Transcript, at pages 1096 and 1099. 82
[260]Furthermore, the evidence pointed to inherent design issues or flaws with the tendered behavioural conditions that would make them open to post -merger manipulati on and disputes, and impossible for the Commission to adequately monitor and enforce. For example, in relation to costing, as pointed out above, the proposed pricing condition allocates a weight of to a basket of costs that include labour, overheads, depreciation, CAPEX and other. Mr Greeff raised concerns regarding this costing structure and the inclusion of the “basket” in the proposed remedy, warning that scope exists for the merged entity in terms of the proposed conditions to manipulate prici ng post-merger . 180 He explained: “Now in overheads you can hide anything, from profit margin to anything which we don’t have any control about and my team tried to reconcile that to what they believe should be reality and they keep missing numbers in that equation, anything varying between depending on how you look at it. So we have to question mark why all of a sudden is there such a weight to something which you really can’t measure out there or control out there. I’ll give you an example. The weight towards baled waste for instance, the first item on that formula, that’s something which we have direct knowledge about because we see our waste with the plant. So you’re your own referee as far as the correctness of that number. Overheads there ar e no referee because you have no insight in what’s hiding in those very quantum jumped weight. That’s our concern. So as soon as that happens we say well, what’s going on here, why is that and it could not have been explained to us so far”. 181 Mr Seggar rai sed similar concerns and explained that this approach of the merging parties in the tendered conditions makes him very uncomfortable: “… why I’d be reluctant to settle on face value 180 Mr Gree ff, Transcript , at page 1545. 181 Mr Greeff, Transcript , at page 1647. 83 is too much of the price is sitting in another column of . Typically on a cost push model you’d have a lot more granularity around your drivers harbouring of it in one bucket is – would make me very uncomfortable .” 182
[261]Mr Greeff went on to say that, based on past practice and experience, another weigh t in the proposed conditions, i.e., that given to baled waste, raises concerns: “… they had to reduce something else. They’d prefer to reduce the most obvious one. Obvious I mean it’s the one that you could control best by knowing what it’s all about is ba led waste. That has declined in ratio. And we argued well baled waste is such a strong component of manufacturing recycled paper that it should carry a higher weight, So that being so obvious, so why do you change the obvious, the common number which everybody knows about, why do you reduce that in weight and replace it with something which we’d have no insight into?”. 183
[262]The above illustrates that the tendered conditions would be difficult for the Commission to adequately monitor and enforce, would likely lead to disputes and furthermore would be ineffective.
[263]Although the parties in their Heads indicated that they are willing to further refine the Labour and Other Cost Drivers, this was not done and , in our view, would not cure the overall inadequacy of the tendered conditions to address the structural change brought about by the proposed merger. 182 Mr Seggar, Transcript , at page s 1114-1115 . 183 Mr Greeff, Transcript , at page 1648. 84
[264]Even if it were possible to fix all the design issues / flaw aspects of the tendered remedy, it remains conceptually deficient for the reasons given above since it remains a behavioural volume and price cap commitment that does not address the long-term negative effects of the proposed merger on competition. Furthermore, the proposed conditions are of limited duration. They also raise significant concerns from a monitoring and enforcement perspective.
[265]As indicated above, the CAC has confirmed that the Tribunal has discretion over remedies and may conclude that it prohibits a merger because there is a reasonable possibility that the proposed conditions will fail to remedy the likely SPLC. We conclude that that is the case here.
[266]We next deal with the bespoke remedy that the merging parties offered to Everest, a Commission factual witness, prior to the hearing and during the hearing, and the circumstances surrounding that. Preferential terms offered to Everest Commission’s arguments
[267]The Commission argued that it is unheard of for a litigant (in this case the merging parties) to approach an opponent’s witness, in an attempt to convince that witness to urgently make himself/herself available to testify. The Commission further argued that the reason for the merging parties’ insistence that Mr Seggar testify soon became apparent: they had concluded a preferential “side deal” with Everest as a customer - should the merger be approved by the Tribunal.
[268]The Commission argued that presumably, the merging parties’ strategy was to offer Everest such a sweet deal that it could no longer testify objectively as to the competitive harm of the merger in the downstream market. The Commission was of the view that the merging parties’ deal offered to Mr Seggar influenced his testimony and that his evidence was contradictory at times. In his witness 85 statement he was categorical about the harmful effects of the merger, and he also confirmed in his testimony that the conditions put up by the merging parties (as opposed to the bespoke remedy offered to Everest) would not be acceptable to Everest. 184 At other times, however, he appeared to suggest that the merger was not anti-competitive. The Commission argued that his inconsistent evidence is understandable since he was clearly conflicted, in multiple respects, due to the merging parties’ influence. Merging parties’ arguments
[269]The merging parties in relation to the “bespoke” remedy offered to Everest, contended that customers are perfectly entitled to negotiate better terms than those outlined in the ‘generic’ remedy, as Everest did. Customers are entitled to exercise their countervailing power to extract the best possible deal and there is nothing untoward about engaging with customers on the remedy and to provide such customers with a pro-competitive deal should they seek to negotiate one. Our assessment
[270]We see the offer of a better deal to one customer, who is a Commission witness, in a very serious light since it potentially influences a witness in an improper way. Moreover, it potentially subverts any assessment of the competition effects of the proposed transaction. It creates a risk that the Tribunal will not be presented with an accurate picture of how the merger will affect the relevant market. merging parties 184 Mr Seggar, Transcript, at pages 1291, 1305 and 1306. 86
[271]We note that Mr Mehta (on behalf of t he merging parties ) 185 who testified prior to Mr Seggar, did not disclose in his evidence to the Tribunal that Corruseal was offering a bespoke and better deal to an opponent to the merger, Everest. This only emerged when Mr Seggar gave evidence before the Tribunal. 186 Mr Mehta was also not able to provide a satisfactory explanation of his failure to disclose to the Tribunal this bespoke deal with Everest. Mr Fraser (also on behalf of the merging parties ) also failed to mention that the merging parties were offering Everest a bespoke remedy. When Mr Fraser testified, the merging parties and Everest had reached an in -principle agreement on the specific terms of the bespoke agreement. 187
[272]We consider next if Everest, based on the factual evidence, was offered a better deal by the merging parties than that offered to other customers in terms of the proposed behavioural conditions, also referred to during the proceedings as the ‘generic’ reme dy.
[273]To recap, in terms of additional volumes to be offered to customers post- merger , the tendered ‘generic’ conditions involve the merging parties offering customers an additional 20% or 1 000 tons post -merger, whichever is the greater, for a specified per iod. 188 Everest’s pre -merger volumes purchased 189 are tons per year. 190 A 20% increase on that figure (which is what Everest would have access to under the tendered generic conditions) is a total of tons per year. 191 However, the deal offered by the merging parties to Everest if the deal was approved was for a total supply volume of tons 185 Mr Mehta was the first factual witness to testify. He testified from 7 to 9 November 2022. One of the topics traversed by Mr Mehta was the merging parties ’ proposed remedy and, more specifically, how the remedy would address the competition concern s identified by the Commission and Intervenors. 186 Mr Seggar, Transcript inter alia pages 1289 and 1290. 187 Transcript , at page 1185 et seq. 188 The general market remedy is Exhibit 14. 189 From both Corruseal and Neopak. 190 Mr Seggar, Transcript , at page 1166. 191 Mr Seggar, Transcript , at page 1166. 87 per year. 192 This means that the volume offer to Everest represents an increase of approximately % from Everest’s pre -merger volumes and an increase of approximately % above the ‘generic’ conditions on offer to other c ustomers in the market.
[274]Apart from that volume benefit, the bespoke Everest deal also created greater certainty of supply from Everest’s perspective because if the merged entity failed to supply Everest with its tons per annum, it would pay a penal ty. This was termed a “make or pay” clause; if the merged entity did not make Everest’s volumes, a penalty would accrue to it. 193 Everest’s offered deal also included a “waste swap agreement”.
[275]Given the abovementioned significant volume differences between the bespoke offer to Everest versus the ‘generic’ volume remedy offered to the rest of customers, we conclude that the volume offer that was made to Everest is substantially better than that tendered for other customers in a market where security of suppl y (i.e., access to volume) is a major competitive factor. We therefore find that the bespoke deal would artificially skew the market in favour of Everest. The implications of this will be further dealt with below.
[276]We note that the bespoke deal offered to Everest was contingent upon the merger being approved. And so, despite being called as the Commission’s witness in opposition to the merger, Mr Seggar now had a vested interest in the merger being approved. This situation potentially could taint the evide nce of Mr Seggar as a witness and therefore the Tribunal’s ability to assess the evidence.
[277]From the evidence it appears that the bespoke remedy was agreed with Everest in the month preceding the merger hearing, because Everest was not satisfied 192 Mr Seggar, Transcript , at page 1166. 193 Mr Seggar, Transcript , at pages 1168 -1169. 88 with the generic remedy offered by the merging parties to customers. It seems clear that the bespoke remedy was offered to Mr Seggar in an attempt to ensure that he would not object to an approval of the merger since the same offer was not made to other customers in terms of the proposed conditions. 194 Furthermore, neither Mr Mehta nor Mr Fraser disclosed this in their evidence. The merging parties were clearly under a duty to disclose the full details of the deal with Everest to the Tribunal (and to the other parties in the merger hearing), especially given the merging parties’ knowledge that the Commission was intending to call Mr Seggar as a factual witness and given their request that the Tribunal approve the merger subject to the ‘generic’ remedy (that had been rejected by Everest).
[278]Bespoke behavioural remedies that significantly favour one party, in this case Everest, could have other negative consequences for competition. Mr Greeff objected to the merging parties offering different remedies to different customers in the market as he stated: “No, this is unfair. It should not be. The merger has got major structural negative or downside impacts on the industry … So, whatever is offered should be offered to all customers in total, transparent, open, no strings attached, nothing behind the scenes.” 195 We concur with this. If a special deal is offered to one or just certain customers to get a merger approved but not to others, then instead of curing the competition harm the remedies can exacerbate competition problems. In this case, the bespoke remedy would result in Everest – one of the largest corrugators – securing a major volume increase post-merger relative to the rest of the industry and that artificially preferences Everest at the expense of other firms, which in turn distorts the competitive landscape in the corrugator market. What is even more concerning to us is the fact that the better security of supply was offered to a Commission factual witness prior to his testimony and that fact was not 194 Mr Seggar, Transcript inter alia pages 1107 and 1180 to 1191. 195 Transcript pages 1541-1542. 89 disclosed by the merging parties’ factual witnesses during their testimony. By definition, remedies must address the competition problem, they may not be (mis)used to secure preferential testimony from a witness who stands to benefit disproportionately from a bespoke offering.
[279]After the above came to light the merging parties indicated, ultimately, that Corruseal would be comfortable, should the Tribunal deem it appropriate, to extend the waste-swop arrangement and the additional volumes (subject to the take-or-pay arrangement 196 ) with other customers, should they want to exceed their synergy upside volume. However, for the reasons articulated above, this belated tender does not cure the fact that the proposed conditions are in many other respects inadequate and not responsive to the structural change brought about by the proposed merger. Conclusion For all the above reasons, we conclude that the merging parties’ proposed behavioural conditions do not address the competition concerns associated with the proposed merger. Therefore, the proposed merger is prohibited. 14 July 2023 Prof Liberty Mncube Date Concurring: Ms Mondo Mazwai and Mr Andreas Wessels 196 We note that APL indicated that it would be opposed to such a condition, see Transcript, at pages 1542, 1632 –1634. 90 Tribunal case managers: Kameel Pancham and Leila Raffee For the merging parties: Adv Mike van der Nest SC and Adv Gavin Marriot instructed by Cliffe Dekker Hofmeyr and Webber Wentzel For the Commission: Adv Tembeka Ngcukaitobi SC, Adv Yanela Ntloko, and Adv Shannon Quinn For Mpact: Adv Jerome Wilson SC and Adv Sarah Pudifin- Jones instructed by Nortons Inc. For APL: Adv Luke Kelly instructed by Maserumule Attorneys
Read from the court's own list. Judgments are official texts; verify against the court's file before you rely on this page. Firms on AttorneyOS get judgments like this one against the matters they touch, in the Monday edition.