CT IM147Nov22 · case IM147Nov22
COMPETITION TRIBUNAL OF SOUTH AFRICA Case No.: IM147Nov22 In the matter between: AKZO NOBEL N.V. First Applicant KANSAI PLASCON AFRICA LTD Second Applicant KANSAI PLASCON EAST ARICA (PTY) LTD Third Applicant and THE COMPETITION COMMISSION OF SOUTH AFRICA Respondent In re the intermediate merger between: AKZO NOBEL N.V. Primary Acquiring Firm And KANSAI PLASCON AFRICA LTD KANSAI PLASCON EAST ARICA (PTY) LTD Primary Target Firms INTRODUCTION
[1]On 21 November 2023, the Competition Tribunal (“Tribunal”) prohibited the intermediate merger in terms of which AkzoNobel N.V. ("AkzoNobel") would acquire the majority of the issued ordinary shares of Kansai Plascon Africa Ltd ("KPAL"), and all the issued ordinary shares of Kansai Plascon East Africa (Pty) Ltd ("KPEA"), from Kansai Paint Co. Ltd ("Kansai Paint"). The reasons for the Tribunal’s decision are set out below. Panel: M Mazwai (Presiding Member) J Wilson (Tribunal Member) L Mncube (Tribunal Member) Heard on: 04-08, 11-12, 14-15 September 2023; and 07 November 2023 Order issued on: 21 November 2023 Reasons issued on: 18 June 2024 REASONS FOR DECISION 2 PROCEDURAL BACKGROUND
[2]On 8 August 2022, the merger parties notified the Competition Commission (“Commission”) of the proposed merger. In its investigation, the Commission found that the proposed merger raised unilateral effects and input foreclosure concerns, and that these concerns were not adequately addressed by the remedies proposed by the merger parties. Accordingly, on 1 November 2022, the Commission prohibited the proposed merger.
[3]On 15 November 2022, the merger parties filed a request for consideration with the Tribunal in terms of section 16(1)(a) of the Competition Act, 89 of 1998 (the “Act”), in which they requested the approval of the merger subject to the conditions they had proposed to the Commission during its merger investigation.
[4]The following factual witnesses gave evidence at the Tribunal hearing: 4.1. For the merger parties: 4.1.1. Mr Prejay Lalla, the Chief Executive Officer of KPAL; 4.1.2. Mr Johann Smidt, the Managing Director of AkzoNobel’s decorative coatings business in sub-Saharan Africa; 4.1.3. Mr Brad Carter, the Managing Director of BBS Mica Build (“BBS Mica”) in Ballito, KwaZulu-Natal; 4.1.4. Mr Herman Bornman, the Managing Director of the PITS Paint It Specialist group of retail paint stores (“PITS Paint”); and 4.1.5. Mr Vaughan Prost, the Chief Executive Officer of Promac Paints (Pty) Ltd (“Promac”). 4.2. For the Commission: 4.2.1. Mr Simon Stekhoven, the Managing Director of Duram (Pty) Ltd (“Duram”); 4.2.2. Mr Chirag Madhu, the Managing Director of Medal Paints (Pty) Ltd (“Medal”); 3 4.2.3. Mr Navin Mahabeer, the National Category Manager of the Build-It Group (“Build-It”); and 4.2.4. Mr Louis Greeff, the Chief Executive Officer of Elite Star Trading Africa (“EST”).
[5]In addition, the following witnesses were called as economic experts: 5.1. for the merger parties: Mr Patrick Smith, a partner at RBB Economics; and 5.2. for the Commission: Mr Yongama Njisane, a Principal Economist in the Economic Research Bureau of the Commission.
[6]The Tribunal also heard submissions from the following employee representatives: 6.1. Mr Walter Dlamini of the Chemical, Energy, Paper, Printing, Wood and Allied Workers Union (“CEPPWAWU”) on behalf of the employees of AkzoNobel; 6.2. Mr John Appolis of the General Industries Workers Union of South Africa (“GIWUSA”) on behalf of the employees of KPAL; and 6.3. Mr Alvin Varaden on behalf of the non-unionised employees of KPAL. THE PROPOSED TRANSACTION
[7]The proposed transaction involves the acquisition by AkzoNobel of 83.31% of the issued ordinary shares of KPAL, and all the issued ordinary shares of KPEA, from Kansai Paint. On completion of the proposed merger, AkzoNobel would exercise sole control over both KPAL and KPEA. THE MERGER PARTIES AND THEIR ACTIVITIES
[8]The primary acquiring firm is AkzoNobel. AkzoNobel is a public limited liability company listed on the Euronext Amsterdam Stock Exchange and is not controlled by any individual firm.
[9]AkzoNobel is a Dutch multinational group that manufactures and sells decorative and industrial coatings around the world. Within Africa, AkzoNobel has manufacturing plants in South Africa, Botswana, Zambia, and Mauritius. AkzoNobel has three manufacturing 4 plants in South Africa – two located in Gauteng (in Alberton and Vanderbijlpark) and one in KwaZulu-Natal (in Umbogintwini).
[10]In South Africa, AkzoNobel controls the following firms: AkzoNobel Powder Coatings South Africa (Pty) Ltd; AkzoNobel South Africa (Pty) Ltd; ICI Dulux (Pty) Ltd; and PJA (South Africa) (Pty) Ltd.
[11]The primary target firms are: (i) KPAL, a company incorporated under the laws of South Africa; and (ii) KPEA, a company incorporated under the laws of Mauritius. KPAL and KPEA are both subsidiaries of Kansai Paint, a Japanese paint manufacturing conglomerate with operations around the world.
[12]KPAL manufactures decorative and industrial coatings in South Africa, Zimbabwe, Zambia, and Malawi. KPAL has four manufacturing plants in South Africa – two located in Gauteng (Krugersdorp and Clayville), one in KwaZulu-Natal (Mobeni), and one in the Eastern Cape (Gqeberha). KPAL also owns and operates ICC Kansai Colourants (Pty) Ltd (“ICC”), a firm that manufactures colourants that are used in the manufacture of ready-mix paints and for in-store tinting.
[13]KPEA mainly operates in the Eastern Africa region, with manufacturing plants in Tanzania, Kenya and Uganda. KPEA does not have any operations in South Africa. RATIONALE FOR THE TRANSACTION AkzoNobel
[14]Mr Smidt explained that the primary rationale for the proposed merger from AkzoNobel’s perspective is to improve its distribution network and footprint in sub-Saharan Africa, and particularly Eastern Africa, where AkzoNobel has a limited presence. Mr Smidt stated that KPEA has more established routes to market in these countries, which AkzoNobel would be able to utilise post-merger to better serve consumers in the region. 1
[15]Mr Smidt testified that the proposed merger would also give rise to manufacturing, distribution and cross-selling synergies in Southern Africa. For example, in South Africa, the proposed merger would enable AkzoNobel to better allocate its decorative coatings 1 Smidt witness statement, Trial Bundle, Part A, pp 252-253, paras 8-9; Hearing Transcript at pp 570-571. 5 operations between AkzoNobel’s old, relatively inefficient and capacity-constrained plant in Umbogintwini and KPAL’s plants. 2 Kansai Paint
[16]Mr Lalla testified that, from Kansai Paint’s perspective, the merger would enable it to exit its decorative coatings operations in Africa in accordance with Kansai Paint’s global strategy to focus on industrial and automotive coatings. Mr Lalla noted that the proposed transaction was negotiated between AkzoNobel and Kansai Paint 3 THE COMMISSION’S FINDINGS ON THE COMPETITION AND PUBLIC INTEREST EFFECTS OF THE MERGER
[17]In its reasons for prohibiting the proposed merger, 4 the Commission found that the proposed merger would give rise to a horizontal overlap in the manufacture and supply of coatings used in various segments. The Commission also found that the merger had a vertical dimension because the colourants produced by ICC are a significant input in the manufacture of decorative coatings, and are currently supplied to downstream competitors of the merger parties.
[18]The Commission found that the merger would give the merged entity significant market shares in the following candidate markets: 18.1. a broad national market for the manufacture and supply of decorative coating products; 18.2. national markets for the manufacture and supply of decorative coating products in separate product tiers (premium, mid-tier and economy) and customer segments (retail and specified trade customers); and 18.3. a national market for the manufacture and supply of colourants and tints.
[19]In its investigation, the Commission considered two theories of harm, namely: (i) unilateral effects in decorative coatings (the removal of a close and effective competitor), 2 Smidt witness statement, Trial Bundle, Part A, pp 252-253, para 9. 3 Hearing Transcript, Lalla, pp 288-299. 4 Reasons for decision, Trial Bundle, Part A, pp 2-195. 6 and (ii) foreclosure effects relating to colourants as inputs in the manufacturing of decorative coatings products.
[20]In its unilateral effects assessment, the Commission found that AkzoNobel and KPAL, with their respective Dulux and Plascon brands, are close competitors in all the candidate decorative coatings markets referred to above, and particularly in the premium and specified trade segments. The Commission found further that barriers to entry are high, and that there is limited countervailing power, in these markets, especially because of the strength of the Dulux and Plascon brands. Based on the above findings, the Commission concluded that the merger would result in a substantial lessening of competition in the relevant South African decorative coatings markets.
[21]In its input foreclosure assessment, the Commission found that the merged entity would have both the ability and incentive to foreclose its downstream competitors from access to colourants given its high market share in the colourants market, and that this would have an anti-competitive effect in the downstream market.
[22]The Commission found that the anti-competitive effects of the merger would not be outweighed by any efficiencies arising from the merger.
[23]As regards the public interest effects of the merger, the Commission received employment concerns from CEPPWAWU on behalf of the employees of AkzoNobel, from GIWUSA on behalf of the employees of KPAL, and from the employee representatives of the non-unionised employees of KPAL. The Commission found that the employment commitments made by the merger parties did not adequately address these employment concerns, and also found that the other public interest commitments made by the merger parties (in relation to new capital investment, localisation, and participation and ownership by historically disadvantaged persons (“HDPs”)) did not outweigh the anti-competitive effects of the proposed merger.
[24]Whilst the merger parties tendered a supply condition to address the input foreclosure concern raised by the Commission, it did not tender any condition to address the Commission’s unilateral effects concern. The Commission concluded that the conditions proposed by the merger parties did not adequately remedy the anti- competitive effects of the merger, and therefore prohibited it. 7 CENTRAL ISSUES
[25]Having regard to the merger assessment required by section 12A of the Act, 5 the central issues raised by the proposed merger are the following: 25.1. Whether the proposed merger is likely to give rise to unilateral effects in the relevant decorative paint market(s); 25.2. Whether the proposed merger is likely to give rise to input foreclosure in the supply of colourants to decorative paint manufacturers in South Africa; 25.3. Whether any such anti-competitive effects are outweighed by pro-competitive benefits of the proposed merger; 25.4. Whether the proposed merger can or cannot be justified on substantial public interest grounds; and 25.5. Whether any negative competition or public interest effects of the proposed merger are adequately addressed by the conditions proposed by the merger parties.
[26]We consider these issues in turn below. First, however, we make some preliminary observations regarding merger analysis in differentiated product markets. MERGER ANALYSIS IN DIFFERENTIATED PRODUCT MARKETS
[27]It was common cause between the Commission and the merger parties that the relevant decorative paints market(s) in this case (however defined) contain differentiated products, in which some products are closer substitutes to each other than others. As discussed further below, this differentiation is based on various factors, including brand image, (actual and perceived) quality, length of guarantees, service levels and price.
[28]In differentiated product markets, competition analysis is generally less concerned with reliance on formal market definition and market shares, and more concerned with evidence of closeness of competition and other constraints on the merger parties (which can be interpreted without having defined formal relevant markets). As explained in the CMA Merger Assessment Guidelines: 5 See Imerys South Africa (Pty) Ltd and Another v Competition Commission (147/CAC/Oct16, IM013May15) [2017] ZACAC 1 (2 March 2017), paras 36-42. 8 “While market definition can sometimes be a useful tool, it is not an end in itself. The outcome of any market definition exercise does not determine the outcome of the CMA’s analysis of the competitive effects of the merger in any mechanistic way. In assessing whether a merger may give rise to a [significant lessening of competition], the CMA may take into account constraints outside the relevant market, segmentation within the relevant market, or other ways in which some constraints are more important than others. 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. There may be no need for the CMA’s assessment of competitive effects to be based on a highly specific description of any particular market definition (including, for example, descriptions of the precise boundaries of the relevant markets and bright-line determinations of whether particular products or services fall within the relevant market). The CMA may take a simple approach to defining the market – for example, by describing the market as comprising the most important constraints on the merger firms that have been identified in the CMA’s assessment of competitive effects.” 6
[29]The Tribunal followed a similar approach in WeBuyCars: “[I]t is useful to identify those firms and products that place a competitive constraint on the merging products and services and then rank them based on the magnitude of that constraint. In this regard, identifying close competitors allows the Tribunal to assess the likely effects of a merger without having to definitively conclude on the relevant market.” 7
[30]Following this approach, we do not believe it is necessary in this case to make any definitive findings on whether the parts of the broader decorative coatings market in 6 CMA Merger Assessment Guidelines, supra, at paras 9.4 and 9.5. 7 MIH eCommerce Holdings Pty Ltd t/a OLX South Africa v WeBuyCars Pty Ltd (LM183Sep18) [2020] ZACT 93 (7 December 2020) at para 173. See also Primedia Ltd and Others v Competition Commission and Another (39/AM/MAY06) [2008] ZACT 30 (9 May 2008) at paras 65-66. 9 which the Commission contends there are most likely to be anti-competitive effects – namely, the premium and specified trade segments – constitute self-standing markets. The relevant question is rather whether customers and consumers purchasing products in those segments regard the merger parties’ brands as the closest competitors and, if so, what the relative strength is of any other constraints on those brands. 8
[31]It is well-established 9 that, in differentiated product markets, unilateral effects are more likely to arise where the merger parties are close competitors or where their products are close substitutes. The more closely the merger parties’ products compete, the greater the likelihood of unilateral effects because the merged entity will recapture a more significant share of the sales lost in response to a price increase (or another kind of worsening in their offerings), making the price rise more profitable. By the same token, the incentive of the merged entity to raise prices is more likely to be constrained when rival firms produce close substitutes to the products of the merger parties.
[32]Closeness of competition is thus a relative concept. The merger parties need not be each other’s closest competitors in order for unilateral effects to arise. It is sufficient that the merger parties compete closely and that the remaining competitive constraints are not sufficient to offset the loss of competition between them resulting from the merger. It is therefore necessary to consider the overall closeness of competition between the merger parties in the context of the other constraints that would remain post-merger. As the Tribunal noted in Future Life, “what matters is the strength of competition between the two firms' products in relation to that with other products of non-merging firms in the candidate relevant market”. 10
[33]The types of evidence to assess closeness of competition are diverse and may vary from case to case. Evidence considered by the Tribunal in previous mergers of this kind includes the views of customers and competitors, functionality, pricing, internal documents, and other relevant indicators. 11 A similar approach is followed by other competition authorities. 12 8 See, eg, Wieland-Werke AG v European Commission, T-251/19, paras. 64, 71. 9 See generally, Guidelines on the assessment of horizontal mergers under the Council Regulation on the control of concentrations between undertakings (OJ 2004 C 31, p. 5) (“EU Horizontal Merger Guidelines”), at para 28; Commission v CK Telecoms UK Investments (Case C-376/20 P) Court of Justice, 13 July 2023, at paras 187-189; CMA Merger Assessment Guidelines (CMA129) 18 March 2021, at paras 4.6 - 4.10. 10 Pioneer Foods (Pty) Ltd v Future Life Health Products (Pty) Ltd (LM017May15) [2015] ZACT 130 (22 December 2015) at para 46. 11 Future Life, supra, at paras 48-81. 12 See, e.g., CMA Merger Assessment Guidelines, supra, at para 4.13. 10
[34]The Tribunal has previously considered that some of the most persuasive evidence on the issue of closeness of competition can be derived from the internal documents of the firms in the market, and in particular, those documents that were not prepared in contemplation of the merger hearing. 13 As the Tribunal explained in WeBuyCars, internal documents have the benefit that they “reveal the true opinions of the executives concerned, before an awareness about how they would be interpreted by the competition authorities or any other regulatory bodies had caused them to self-edit, or to be much more circumspect about how they expressed their opinions”. 14
[35]This is not to say, of course, that internal documents are necessarily conclusive; it is always necessary to understand the context in which such documents were produced, the information on which they were based, and their consistency with the other available evidence. However, they are nevertheless an important source of evidence to be considered in the assessment of closeness of competition. 15
[36]Whilst a central element of the competitive analysis, closeness of competition is not the only consideration. Merger assessment is generally forward-looking and must therefore also consider potential new entry and expansion by rivals post-merger. 16
[37]It is also necessary to consider countervailing buyer power, although it is generally considered that forms of buyer power that do not result in new entry – e.g., buyer power based on a customer’s size, sophistication, or ability to switch easily – are unlikely to prevent any anti-competitive merger effect. This is because a customer’s buyer power depends on the availability of good alternatives, which, in the context of an anti- competitive effect, will have been reduced. In this sense, market power and buyer power are two sides of the same coin. 17
[38]We will consider the evidence on all of these issues below. First, however, we outline the parties’ opposing contentions on market definition and market shares in the relevant decorative coatings market(s). 13 Primedia, supra, at para 59; WeBuyCars, supra, at paras 174-175. 14 WeBuyCars, supra, at para 175. 15 See also in this regard, Wieland-Werke AG v European Commission (Case T-251/19), General Court, 18 May 2022, at paras 116-118. 16 See CMA Merger Assessment Guidelines, supra, at paras 4.16-4.19; EU Horizontal Merger Guidelines, supra, at paras 68-73. See also Expert Report of Mr Smith, Trial Bundle A at p 464, para 94. 17 See CMA Merger Assessment Guidelines, supra, at at para 4.20; EU Horizontal Merger Guidelines, supra, at paras 64-67. See also Expert Report of Mr Smith, Trial Bundle A at p 464, para 94. 11 MARKET DEFINITION AND MARKET SHARES IN THE DECORATIVE COATINGS MARKET(S) Market definition
[39]The Commission’s unilateral effects case before the Tribunal was, in essence, that: 39.1. There are distinct markets for the manufacture and supply of decorative coatings products to (i) the retail channel and (ii) the specified trade channel, and within each of these markets there is further segmentation by product quality tiers (premium, mid-tier and economy). 18 39.2. The retail channel is one in which retail outlets sell decorative paint to individual end-consumers and small contractors, whilst the specified trade channel is one in which manufacturers sell decorative paint directly to managers of large-scale residential, commercial and public infrastructure projects for which that paint has been specified. 19 The Commission contended that, within each of these channels, the suite of decorative coatings supplied by manufacturers is stratified according to product quality tiers to cater to the full range of customer preferences and LSM categories. 20 39.3. The merger parties, with their respective Dulux and Plascon brands, are effective competitors in all segments of the decorative paints market, and are each other’s closest competitors, particularly in the premium and specified trade segments. 21
[40]There was agreement between the Commission and the merger parties that the relevant downstream product market is limited to the manufacture and supply of decorative coatings products, and that the geographic ambit of the relevant market is national. 22
[41]However, all other aspects of the Commission’s case were disputed by the merger parties, who contended that: 18 Expert Presentation of Mr Njisane, Exhibit 15, at slide 6; Njisane, Hearing Transcript at pp 1256 - 1257. 19 Expert Presentation of Mr Njisane, Exhibit 15, at slide 7. 20 Expert Presentation of Mr Njisane, Exhibit 15, at slide 8. 21 Expert Presentation of Mr Njisane, Exhibit 15, at slides 3 and 20-37. 22 Expert Presentation of Mr Njisane, Exhibit 15, at slide 10; Expert Presentation of Mr Smith, Exhibit 14, at slide 12. Smith, Hearing Transcript at p 1228. See also Njisane, Hearing Transcript at p 1256-7. 12 41.1. There is a single relevant product market for all decorative coatings. Mr Smith stated that there is no robust and consistent evidence to justify defining separate product markets, whether by price/quality tier or by sales channel. 23 Rather, the decorative paints market is comprised of a continuous chain of substitution, in which customers and consumers constantly trade off differences in price against quality factors such as spread rate, durability and guarantees, which impact on the ultimate cost of using a particular product. 24 41.2. Furthermore, different customer types, such as trade customers, constantly switch between products sold in different channels, and all producers can readily switch between supplying different kinds of paint to different channels and consumer categories. Mr Smith added that supply-side substitution by manufacturers between different paint products is quick, easy and continuous. 25 41.3. Whilst the merger parties are close and effective competitors of each other in the relevant market, they are not “uniquely close” competitors, and the merged entity will continue to face strong and effective competitive constraints post-merger from a multitude of competing suppliers (and also as a result of the countervailing power of retailers, and the potential for expansion and new entry).
[42]During the course of the hearing, the merger parties clarified that, for a consumer whose first choice is a high quality/ high price Dulux or Plascon product, they did not contend that all other brands in the market would be equally close competitors on the basis of the price/quality continuum. Mr Smith acknowledged that the most likely substitutes for an individual purchase decision are probably from the closest adjacent products: “So, the chain is not direct substitution right from the top right to the bottom. It is most concentrated on adjacent products.” 26
[43]Similarly, the merger parties’ counsel acknowledged in debate with Mr Mahabeer: “[N]o-one in the marketplace disagrees with the obvious. And if you are talking in the motor vehicle space, common sense tells you a Rolls Royce is in a different segment to a 100 cc hatchback. No need to have an economics degree to figure that out. In the 23 Expert Presentation of Mr Smith, Exhibit 14, at slide 12; Smith, Hearing Transcript at pp 1227-1228. 24 Expert Presentation of Mr Smith, Exhibit 14, slide 17; Smith, Hearing Transcript at pp 1287-1289. See also Smidt witness statement, Trial Bundle, Part A, p 275, para 78; Lalla witness statement, Trial Bundle, Part A, p 366, para 31. 25 Expert Presentation of Mr Smith, Exhibit 14, at slide 13; Smith, Hearing Transcript at pp 1227-1229. 26 Smith, Hearing Transcript at pp 1603-1604. 13 same way one would say well, a Dulux Luxurious Silk is in a different space in the market compared to any one of the economy brands, any economy products of a number of the manufacturers. You would say, they are at other ends of the spectrum. It is obvious. One is low cost, low quality .. one is high quality, high cost. About that there can be no debate but that is only to talk about the parameters. Where the debates will begin is when you start narrowing across the range.” 27
[44]This acknowledgement closely mirrors the Tribunal’s observation in Daimler Chrysler that “[a] customer who wishes to buy a Mercedes-Benz would not likely substitute to a Toyota Corolla model”, 28 and in Distillers that: “just as ‘common sense’ would reject an argument for placing a multi-million Rand Rolls Royce in the same market as the humble Opel Corsa so too does it rebel at the notion that Remy Martin competes with Wellington (or Cape to Rio) or Chivas Regal with First Watch (or Russian Bear)”. 29
[45]Thus clarified, we understood the merger parties’ case to be that the premium Dulux and Plascon brands at the top end of the price/quality continuum are not necessarily constrained by economy brands at the bottom end, but that they are constrained by products that are more proximately located to them in the continuum.
[46]We will consider the relevant evidence in this regard further below in these reasons. However, we note at the outset that, in circumstances where the merger parties’ brands are found to be each other’s closest competitors in terms of the relevant price and quality metrics, persuasive evidence would be required that more distant competitors will continue to constrain the merged entity sufficiently strongly and effectively that the merger is unlikely to give rise to unilateral effects. 30 Market shares
[47]There was also disagreement between the Commission and the merger parties on the respective market shares of the merger parties and other manufacturers in the different candidate product markets. This was a function not only of definitional disputes as to 27 Mahabeer, Hearing Transcript at pp 1041-1042 28 DaimlerChrysler South Africa (Pty) Ltd/Sandown Motor Holdings (Pty) Ltd 44/LM/Jul01 p12. 29 Distillers Corporation (SA) Ltd v Stellenbosch Farmers Winery Group Ltd 08/LM/Feb02 para 150. 30 See CMA Merger Assessment Guidelines, supra, at para 4.10. 14 which players and products should be included in each candidate market, but also of inconsistent and disputed documentary evidence and data.
[48]As regards calculation metrics, Mr Njisane argued that market shares should be calculated in value as opposed to volume terms, because value is a better indicator of market power. 31 Mr Smith appeared to agree that, in principle, value shares were more probative than volume shares, but stated that in this case volume data was more robust (“everyone can see what a litre is”) and for that reason should be used instead. 32
[49]Mr Njisane and Mr Smith both acknowledged that market shares were difficult to estimate given the limitations in the available data. 33 However, Mr Njisane testified that, irrespective of whether value or volume data were used, and irrespective of the source of the data, they all reflected a consistent trend over a number of years, namely that the merger parties’ products were the first and second largest competitors, and that the merged entity would be significantly larger than any other player, in the broad decorative coatings market, and particularly in the premium and specified trade segments. 34
[50]Mr Njisane calculated revenue market shares for AkzoNobel and KPAL in the broad decorative coatings market for each of the years 2019 to 2021 based on three sets of figures, namely (i) the merger parties’ revenues as a percentage of KPAL’s estimate of the total market size (retail only), (ii) the merger parties’ revenues as a percentage of AkzoNobel’s estimate of the total market size (retail only) and (iii) the merger parties’ revenues as a percentage of a total market size calculated by a third party consulting firm called Orr & Boss (retail and trade).
[51]Mr Njisane testified that, on all of these measures, AkzoNobel and KPAL have the highest market shares, and the merged entity would have a market share that is far higher than any other manufacturer. On the first two measures, the merged entity would have a market share in the range of 40-50%, and on the third measure a market share in the range of 35-40%. 35 Mr Njisane testified further that these estimates were consistent with the estimates contained in the merger parties’ internal documents and in submissions made by third parties. 36 31 Njisane, Hearing Transcript at p 1372. 32 Smith, Hearing Transcript at pp 1355, 1376. 33 Expert Presentation of Mr Smith, Exhibit 14, at slide 38; Smith, Hearing Transcript at pp 1239, 1354; Njisane, Hearing Transcript at p 1374. 34 Expert Presentation of Mr Njisane, Exhibit 15, at slides 12-17; Njisane, Hearing Transcript at pp 1257 - 1260, 1374-1383, 1663-1664. 35 Expert Presentation of Mr Njisane, Exhibit 15, at slide 12; Njisane, Hearing Transcript at pp 1257-1258, 1373- 1377. 36 Expert Presentation of Mr Njisane, Exhibit 15, at slide 13; Njisane, Hearing Transcript at p 1378. 15
[52]Mr Njisane also estimated market sizes in the premium and mid-tier segments of the decorative coatings market based on (i) KPAL’s segmentation and market size estimate, (ii) AkzoNobel’s segmentation and market size estimate, and (iii) each of KPAL’s and AkzoNobel’s segmentation and Orr & Boss’s estimated total market size. Mr Njisane testified that: 52.1. In the premium segment, AkzoNobel and KPAL have by far the highest market shares, and the merged entity would have a market share that is very considerably higher than any other manufacturer. On the first two measures, the merged entity would have a market share in the range of 50-85% (with the next largest competitor at 5% or less), and on the third measure a market share in the range of 45-70% (with the next largest competitor at under 5%). 37 52.2. In the mid-tier segment, the combined market share of AkzoNobel and KPAL is in the range of 15-40%, with Duram at an equivalent market share and the next largest competitor under 20%. 38
[53]As regards the specified trade segment, Mr Njisane calculated that the merger parties have a combined market share of between 75-80% based on the merger parties internal documents. 39
[54]Mr Smith, for his part, stated that the best basis available was Orr & Boss data which indicated that the merger parties’ combined market share in the total market (retail and trade) was approximately between 20-25% by volume and between 30-35% by value (and that these figures under-represented the sales figures of suppliers such as Duram, Promac and Medal that were provided at the Tribunal hearing) 40 . Mr Smith stated further that there were no robust estimates of shares within any narrower product tiers or categories, and that the documents relied upon by Mr Njisane were based on incomplete data. 41
[55]As discussed above, precise market definitions and market share calculations are inherently challenging, and of limited probative value, in differentiated product markets 37 Expert Presentation of Mr Njisane, Exhibit 15, at slide 14; Njisane, Hearing Transcript at pp 1378-1379. 38 Expert Presentation of Mr Njisane, Exhibit 15, at slide 15; Njisane, Hearing Transcript at p 1380. 39 Expert Presentation of Mr Njisane, Exhibit 15, at slide 16; Njisane, Hearing Transcript at pp 1380 - 1381. 40 It appears, however, that the figures for Promac and Medal include the DOBs they manufacture and not only their own brands (Smith, T1649). 41 Expert Presentation of Mr Smith, Exhibit 14, at slide 38; Smith, Hearing Transcript at pp 1238-1241, 1355, 1358, 1376, 1385-1386. 16 such as those at issue here. We therefore do not reach any firm conclusion in these reasons on precise delineation of the relevant product market(s) or on the market shares of the merger parties and other competitors therein.
[56]However, we agree with Mr Njisane that there appears to be a consistent trend in all of the data and documents referred to above, namely that AkzoNobel and KPAL are significant competitors in the broad decorative coatings market in South Africa, and this was confirmed by the merger parties’ witnesses. Mr Lalla stated that KPAL and AkzoNobel are generally the two largest competitors in the broad decorative coatings market (in both volume and value terms), and would have more than double the market share of the next largest player (Duram) post-merger. 42 Mr Smith acknowledged in his evidence that the merger parties are “substantial competitors” in the broader decorative coatings market in South Africa and that, as a first filter, the concentration that is indicated warrants further consideration. 43
[57]The evidence referred to above also suggests that the merger parties are, by a considerable measure, the largest players in the premium and specified trade segments, and would have a combined market share very significantly greater than Duram.
[58]Against this background, we proceed to consider the evidence of closeness of competition (and other constraints) in the different areas of the broad decorative coatings market.
[59]We address first the retail channel and then the trade channel of the market. CLOSENESS OF COMPETITION – RETAIL CHANNEL Closeness of competition between Dulux and Plascon
[60]In their evidence, the merger parties did not dispute that they are close competitors of each other the decorative coatings market in South Africa.
[61]For instance, Mr Smidt acknowledged that KPAL is an “important competitor” to AkzoNobel in the South African decorative coatings market. He explained that KPAL manufactures and sells a wide range of decorative coating products in South Africa; and that the parties’ Plascon and Dulux brands have competed against each other for many 42 Lalla, Hearing Transcript at pp 282-283. 43 Smith, Hearing Transcript at p 1240. 17 years in South Africa. He also stated that “Plascon remains a strong competitor to Dulux with a comparable range of products and quality of offering”. 44
[62]Mr Smidt contended however that AkzoNobel also competes with a multitude of other paint manufacturers that produce multiple ranges of paints (from economy paints to so- called “premium” quality coatings) for a variety of different applications, and for supply to both retailers and trade customers. 45
[63]Mr Lalla likewise acknowledged that, from KPAL’s perspective, it regards AkzoNobel as a “close competitor” but similarly stated that KPAL competes with a range of other manufacturers. 46 He concurred with Mr Smidt that most paint manufacturers manufacture several different paint ranges – in particular, a “premium” range, a “mid- level” range and an “economy” range – that are targeted at different price points and are intended to appeal to consumers in different LSM groups. 47 Mr Lalla explained further that KPAL has historically focused on the high-end of the market, seeking to produce best-in-class “premium” paints aimed at LSM 8 to 10 consumers; and has priced its products accordingly. 48 Segmentation of the market
[64]We pause at this juncture to consider the significance of the segmentation of the decorative coatings market into “premium”, “mid-level” and “economy” tiers. As indicated above, Messrs Smidt and Lalla both acknowledge that the decorative coatings market is segmented in this manner from a manufacturer perspective. This is also clear from the internal documents of both merger parties, which consistently reflect this segmentation in their respective market assessments and analyses, together with the positioning of their (and other manufacturers’) brands within those segments. 49 44 Smidt witness statement, Trial Bundle, Part A, p 256, paras 22-26; T502-503. 45 Smidt witness statement, Trial Bundle, Part A, p 256, para 22. 46 Lalla, Hearing Transcript at p 144. 47 Lalla witness statement, Trial Bundle, Part A, pp 364-365, para 24. See various examples at Smidt witness statement, Trial Bundle, Part A, pp 333-349. 48 Lalla witness statement, Trial Bundle, Part A, p 367, para 34. 49 See, eg, Trial Bundle, Part B at p 742, Part B at p 786, Part B at p808, Part B at p 813, Part B at p 1891, Part C at p 2297, Part C at p 5715, Part C at p 8602-8607. Mr Smith identified some examples of particular brands being categorised inconsistently in certain internal documents (Expert Presentation of Mr Smith, Exhibit 14, slide 21), but these appear to us to be isolated examples and do not detract from the general position reflected in the merger parties’ documents. 18
[65]This segmentation was also confirmed by other manufacturers who testified at the hearing (Mr Stekhoven of Duram, 50 Mr Madhu of Medal, 51 and Mr Prost of Promac 52 ), all of which produce their own separate “premium”, “mid-level” and “economy” ranges of paints.
[66]The evidence indicates that each manufacturer’s “premium” range products are typically positioned as higher price, higher quality products with longer manufacturer guarantees than those products in that manufacturer’s “mid-tier” range; and that the products in such manufacturer’s “mid-tier” range are distinguished from those in its “economy” range on the same basis. 53
[67]As outlined above, the Commission contended that this segmentation is reflective of different product markets. The merger parties, on the other hand, disputed that this market segmentation had any competitive salience. Mr Lalla stated that the segmentation of the market by manufacturers was only for “convenience” 54 whilst Mr Smidt went so far as to contend that it was “ill-conceived” because there are no distinct boundaries between the different quality segments from a product perspective 55 . They argued that there is a continuous chain of substitution between the different segments, in which customers and consumers constantly trade off differences in price against quality and other characteristics.
[68]In support of this contention, Mr Smidt stressed in his evidence that there is no “formal” or “objective” standard for what constitutes a “premium”, “mid-tier” or “economy” paint: “Each supplier will have its own subjective measurements and definitions of brand and quality, and so . . . will each final customer that purchases decorative coating products. Therefore, what is considered by one supplier or customer to be a “premium” product might be considered by another as a “mid-tier” or “economy” product – and so on.” 56 50 Stekhoven witness statement, Trial Bundle, Part A, pp 407-410, paras 12-23; Stekhoven, Hearing Transcript at p 785. 51 Madhu witness statement, Trial Bundle, Part A, p 426, paras 8-19. 52 Prost witness statement, Exhibit 12, para 2. See also Hearing Transcript, Prost, pp 1145-1146. 53 Smidt witness statement, Trial Bundle, Part A, p 254, para 15; Lalla witness statement, Trial Bundle, Part A, pp 364-366, paras 24-28; Stekhoven witness statement, Trial Bundle, Part A, pp 407-410, paras 12-23; Madhu witness statement, Trial Bundle, Part A, p 426, para 8-19. 54 Lalla witness statement, Trial Bundle, Part A, p 366, para 31. 55 Smidt witness statement, Trial Bundle, Part A, p 254, para 14. See also Smidt, Hearing Transcript at pp 492- 493. 56 Smidt witness statement, Trial Bundle, Part A, pp 254-255, para 17. See also Lalla witness statement, Trial Bundle, Part A, p 365, para 24; Lalla, Hearing Transcript at p 143. 19
[69]To illustrate this, the merger parties referred to various examples of products that are positioned by different manufacturers as “premium” or “mid-tier” but which have widely varying (and, in some instances, overlapping) price points, manufacturer guarantee periods and the like. 57
[70]However, it does not appear to us that the manner in which manufacturers – and the merger parties themselves – segment the decorative coatings market should be disregarded as readily as the merger parties suggest. In our view, the difficulty of drawing precise boundaries between different market segments does not necessarily mean that the segmentation lacks competitive significance. This is particularly so in markets where (as in this case) purchasing behaviour is largely driven by customer perception, which is by its nature based on a holistic assessment of many different factors. 58 In these circumstances, it seems unsurprising that, as Mr Stekhoven put it, an “accepted understanding” of different quality tiers might exist notwithstanding the absence of any precise, or uniform, definition of each of the different quality tiers. 59
[71]As the Tribunal observed in a previous merger involving cores and tubes: “[I]t is indeed not easy to specify a precise point of delineation between these market segments. Counsel for the parties insists that because a specific delineation proves elusive, we are then left with a single market for cores and tubes with the various categories simply falling along a single, seamless spectrum. It is somewhat akin to defining an elephant – while this may be a difficult task, it is nevertheless easy to recognize an elephant when one happens upon one. A failure to accurately define an elephant does not simply place it along a continuum of four-legged beasts, the one substantively indistinguishable from the other. So with cores and tubes – every witness who testified before the Tribunal (including those representing the merger parties) constantly referred to two distinct market segments.” 60
[72]Likewise, in this case, the fact that all manufacturers segment the market into different quality tiers suggests, at least on the face of it, that there is competitive significance to this differentiation. Why else would all manufacturers produce separate ranges of 57 See, eg, Expert Presentation of Mr Smith, Exhibit 14, slides 14-22. 58 DaimlerChrysler South Africa (Pty) Ltd/Sandown Motor Holdings (Pty) Ltd 44/LM/Jul01 p11. 59 Hearing Transcript, Stekhoven, p 785. 60 Mondi Limited and Kohler Cores and Tubes a division of Kohler Packaging Limited (06/LM/Jan02) [2002] ZACT 40 (20 June 2002) para 37. 20 products targeted at different tiers, and the merger parties’ internal documents analyse competition within each of these different tiers?
[73]This is particularly so given that the different ranges manufactured by the various manufacturers are intended to enable them to meet different customer preferences and LSM categories, and are also marketed by retailers on that basis. 61 For example, Mr Mahabeer of Build-It testified that: “There are essentially three product tiers that decorative coating products can be segmented [into]: the premium, the mid and the economy tier. These product tier segments cater to different end-consumers with respect to price and quality. In terms of price, Build-It categorises the product tier segments to cater to specific LSM groups. The premium tier segment can be categorised to serve end-consumers classified between LSM 8 to 10, but mostly between LSM 9 and 10 group. The mid-tier segment can be categorised to serve end-consumers classified between LSM 6 to 8, and the economy tier segment caters to end-consumers falling under LSM 5 and below.” 62
[74]Mr Smidt acknowledged that, whilst the categorisation process is a subjective one, a manufacturer’s decision as to what product to produce, and where to position it, will be informed by its perception of customer and consumer demand. 63 The evidence of Mr Stekhoven and Mr Madhu was to the same effect. 64 In the words of Mr Stekhoven: “We as paint companies decided on the tiers because we want to differentiate and have a bite at each segment of the market.” 65
[75]A good example of this was provided by Mr Carter, who explained that BBS Mica decided to expand its decorative paints offering to include Plascon products, driven in large part by demand for perceived premium products by wealthy consumers (LSM category 8 - 10) in the Ballito area. Mr Carter contrasted this with the position in stores he operated in rural areas such as Nongoma and Eshowe, where there is virtually no 61 Lalla witness statement, Trial Bundle, Part A, pp 364-365, para 24. See also Njisane, Hearing Transcript at p 1345. 62 Mahabeer witness statement, Trial Bundle, Part A, pp 393-394, paras 9-10. See also Stekhoven witness statement, Trial Bundle, Part A, p 410, para 23. 63 Smidt witness statement, Trial Bundle, Part A, pp 254-255, para 17. See also Hearing Transcript, Smidt, pp 598-602; Hearing Transcript, Greeff, p 1119. 64 Hearing Transcript, Stekhoven, pp 716, 850; Hearing Transcript, Madhu, p 882. 65 Hearing Transcript, Stekhoven, p 850. 21 demand for premium range products because they are too expensive for the consumers in that area, who fall into LSM category 3 and below. 66 Mr Smidt confirmed that the ranges of products that retailers elect to sell in their stores are largely driven by the demands of consumers in the catchment areas of those different stores. 67
[76]All this evidence appears to us to be inconsistent with the notion of a single broad market in which all products compete equally on the basis of price/ quality trade-off. It is indicative rather of a market in which there are distinct quality segments (albeit with imprecisely defined boundaries) in each of which segments manufacturers compete with a distinct range of products for a distinct target market of customers, defined primarily by reference to LSM categories.
[77]However, we agree with the merger parties that the mere fact that a particular manufacturer might label certain of its products, or ranges, as “premium”, or “mid-tier”, does not necessarily mean that they will be perceived and accepted by customers and consumers as such. The evidence suggests rather that – perhaps precisely because there are no objective or formal criteria to qualify for one segment or another – the segment in which a particular product or range in fact competes is ultimately determined not by the manufacturer’s attempted positioning of that product or range, but by its actual quality and, perhaps even more importantly, its perceived quality in the eyes of consumers.
[78]Put differently, the mere fact that a particular manufacturer may market a range of products as “premium” does not mean that it will be perceived as such by consumers, and accordingly does not mean that it will compete closely with all other products that are marketed as “premium” products by their respective manufacturers. Mr Stekhoven agreed that “one manufacturer’s premium might not be another manufacturer’s premium”. 68
[79]Conversely, products positioned by manufacturers as mid-tier products may potentially be regarded as consumers as premium products. An example of this is Plascon Polvin, which was historically positioned as a mid-tier brand, but which KPAL repositioned as a premium brand when it come under pressure from Duram in the mid-tier segment and research indicated that the market in fact regarded it as a premium product. 69 66 Carter witness statement, Trial Bundle, Part A, p 376, paras 6-8. See also Hearing Transcript, Carter, pp 412- 415, 459-460. 67 Hearing Transcript, Smidt, pp 679-681. 68 Hearing Transcript, Stekhoven, p 786. 69 Hearing Transcript, Lalla, pp 126-127, pp 314-315. 22
[80]What then determines perceptions of quality in the decorative coatings market? The evidence suggests that it is a multi-dimensional determination, in which various related factors play a role. 70 These include product intrinsics (ingredients, spread rate, opacity, durability, finish and the like) and also external indicators of quality such as the length of guarantee provided by the manufacturer, packaging, marketing, and price point. 71 However, perhaps the most important indicator of quality, in the premium segment in particular, is brand equity, and specifically the association of the brand with high quality in the eyes of consumers. 72
[81]Mr Stekhoven of Duram explained in this regard that “[b]rand equity plays a crucial role in the perception of a particular brand”. He explained that this is generally driven by the marketing efforts and investments a manufacturer is prepared to make to develop its brand equity and to position its products in the different tiers of the market. 73
[82]Mr Madhu of Medal agreed that “brand equity is an important aspect with regard to the marketing and sales of decorative coating products” and influences the ability of manufacturers to compete in different segments of the market. 74
[83]Mr Mahabeer of Build-It also confirmed this: “End-consumers tend to be aligned to the product that they already know. They want to purchase right the first time as failure to do so could have major financial implications for them. Thus, brand equity plays a significant role in the ability to enter the market for paint products.” 75 70 Smith, Hearing Transcript at p 1287. 71 Smidt witness statement, Trial Bundle, Part A, p 254, para 15; Lalla witness statement, Trial Bundle, Part A, pp 364-366, paras 24-28; Lalla, Hearing Transcript at p 140; Stekhoven witness statement, Trial Bundle, Part A, pp 407-410, paras 12-23; Madhu witness statement, Trial Bundle, Part A, p 426, para 8-19; Mahabeer, T979-980; Expert Presentation of Mr Smith, Exhibit 14, slide 13. 72 Smidt witness statement, Trial Bundle, Part A, p p 255, para 19; Lalla witness statement, Trial Bundle, Part A, p 365, para 27; Stekhoven witness statement, Trial Bundle, Part A, pp 407-411, paras 13-24; Madhu witness statement, Trial Bundle, Part A, pp 427-432, paras 9-28; Mahabeer witness statement, Trial Bundle, Part A, p 397, para 20. 73 Stekhoven witness statement, Trial Bundle, Part A, p 411, para 24. 74 Madhu witness statement, Trial Bundle, Part A, p 430, para 20. 75 Mahabeer witness statement, Trial Bundle, Part A, p 399, para 27. 23 Brand equity of Plascon and Dulux, and barriers to entry faced by Duram and others
[84]The evidence furthermore shows that the Plascon and Dulux brands enjoy high (and unparalleled) brand equity as top-quality paint brands in South Africa, which has been developed by significant investment over many years. 76 As Mr Stekhoven explained: “Plascon and Dulux brands have always been associated with high-end, quality products which is not the same for other coating brands who may generally be associated with lesser quality product offerings. As a result, the Plascon and Dulux brands are generally accepted as premium tier and this is reflected in their ability to command premium pricing relative to other players in the industry.” 77
[85]The relative power of the Plascon and Dulux brands is reflected in a study commissioned by Dulux in 2022 which reflects Plascon and Dulux as having exponentially higher top of mind awareness and spontaneous awareness than Duram or any of the other brands measured. 78
[86]It is also evident from the difficulties that newer brands, which lack the long-established equity of Plascon and Dulux, have faced in trying to compete with Plascon and Dulux in the premium segment of the market, as set out below. Duram
[87]Mr Stekhoven testifed that the brand equity of Plascon and Dulux constitutes a significant barrier to Duram successfully entering and growing in the premium segment of the market: “The biggest challenge to Duram's growth in the premium tier space is the legacy brand awareness of the merger parties. Plascon and Dulux have for many years invested significantly in brand and colour marketing in the premium product tier. This serves to ensure that they remain strong brands in both the premium and the medium tiers.” 79 76 Stekhoven, Hearing Transcript at pp 748-749. 77 Stekhoven witness statement, Trial Bundle, Part A, p 411, para 24. 78 C2647; Expert Presentation of Mr Njisane, Exhibit 15, slide 22; Njisane, Hearing Transcript at pp 1265-1266, 1411. 79 Stekhoven witness statement, Trial Bundle, Part A, p 413, para 34. 24
[88]Mr Stekhoven elaborated that, in the premium segment of the market, Plascon and Dulux are the market leaders, followed by Duram “with a significantly smaller market share”. 80 He explained the relative positioning of Duram to Dulux and Plascon as being comparable to the positioning of Toyota to BMW and Mercedes-Benz in the motor vehicle market – whilst all three are functionally very similar, the latter brands are regarded as more premium than the former. 81
[89]As a result, notwithstanding Duram’s efforts to enter the premium segment in competition with Plascon and Dulux since 2014 – including by directly targeting certain of their products (for example, it appears that Duram’s Armaguard directly targeted Dulux’s Weatherguard and Plascon’s Micatex products), 82 and offering equivalent guarantees and paint quality – Duram has been forced to price its (small range of) premium paint products at a lower price point in order to attract sales (as discussed further below), and even then it still has a “very limited presence” in the premium segment. 83
[90]In support of their contention that Duram is a close competitor of Plascon and Dulux, the merger parties referred to an incident in 2015 when KPAL experienced difficulties in implementing a SAP inventory and accounting system. Mr Lalla explained that this impacted on KPAL’s ability to deliver its products to a substantial number of retailers, causing significant retailer losses and dissatisfaction. Mr Lalla stated that Duram was the primary beneficiary of KPAL’s difficulties – it quickly increased production to meet retailer demand, and used the opportunity to increase in major retailers, which it has since maintained and grown. 84
[91]Mr Carter confirmed that, when the SAP incident took place in 2015, it was mostly Duram that filled the gap left by Plascon's inability to supply. 85 However, Mr Carter clarified in his oral evidence that his experience of the SAP event was limited to his Mica store in Nongoma which, because of its low LSM target market, only stocked mid-tier and economy products (predominantly Duram, Promac and Excelsior) and no premium 80 Stekhoven witness statement, Trial Bundle, Part A, p 413, para 32; Hearing Transcript at p 714. 81 Stekhoven witness statement, Trial Bundle, Part A, p 421, para 64. 82 See Trial Bundle, Part C, pp 3931, 6302, 6328. 83 Stekhoven witness statement, Trial Bundle, Part A, p 410, para 23; p 413, paras 32-34; p 420, paras 58-59; p 421, paras 62-63. See also Stekhoven, Hearing Transcript at pp 796-797, 845, 853. 84 Lalla witness statement, Trial Bundle, Part A, pp 368-369, para 39. See also Bornman witness statement, Trial Bundle, Part A, pp 383-384, para 9. 85 Carter witness statement, Trial Bundle, Part A, p 379, para 15. 25 segment products. 86 The only Plascon products that were stocked in the Nongoma store at the time were a “handful of lines” for government buildings. 87 Mr Carter accordingly acknowledged that the only Duram response to the SAP incident of which he was personally aware concerned mid-tier products. 88
[92]In his witness statement, Mr Bornman of Pits Paints also stated that Duram benefited more than other brands in his stores as a result of the SAP incident. 89 However, he then clarified that Duram’s product range in 2015 was a “very much mid-tier range”. 90 Mr Bornman also explained that, because Pits Paints was a Dulux-preferred supplier, “Dulux already had our trade to a certain extent and Plascon filled the gaps. We had to have another supplier to fill in on the Plascon gap”. 91
[93]Mr Mahabeer of Build-It testified that Duram was “in the right place at the right time” when the SAP incident occurred, but also stated that Duram’s range is mid-tier rather than premium in nature, and that Build-It primarily stocks Plascon as its premium range brand. 92 Asked why Build-it did not pivot to Dulux when KPAL’s SAP incident occurred, Mr Mahabeer responded that AkzoNobel was not in a position to service all its stores, and that there were historical strategic differences between the two companies. He also explained that Build-It reverted to Plascon when it was able to resume supply of products in the premium segment. 93
[94]Having regard to this evidence, and Mr Stekhoven’s evidence regarding the difficulty Duram has experienced in penetrating the premium segment of the market, we are not persuaded that the 2015 SAP incident demonstrates that Duram competes as closely with Dulux and Plascon in the premium segment as they do with each other. It appears to have been a once-off situation where retailers wanted an alternative to Dulux while Plascon was temporarily out of the market, and Duram’s mid-tier brands, which had just entered the market, were regarded as the best replacement at the time. When supply of Plascon’s brands resumed, there is no evidence that Duram grew its presence in the 86 Carter, Hearing Transcript at pp 410-411. See also Carter witness statement, Trial Bundle, Part A, p 376, para 8. 87 Carter, Hearing Transcript at pp 442-443. 88 Carter, Hearing Transcript at pp 444-445. 89 Bornman witness statement, Trial Bundle, Part A, pp 383-384, para 9. 90 Bornman, Hearing Transcript at pp 340. 91 Bornman, Hearing Transcript at pp 341. 92 Mahabeer, Hearing Transcript at pp 977-983, 990-991. 93 Mahabeer, Hearing Transcript at pp 977-983, 990-991, 1018-1019, 1089-1092. 26 premium segment to any extent. As set out above, Mr Stekhoven’s evidence was to the contrary.
[95]The merger parties’ internal documents also suggest that Duram is not a close competitor of Dulux and Plascon in the premium segment of the market.
[96]Another document that received attention in evidence was an AkzoNobel presentation 96
[97]Mr Smidt explained in his evidence that this document was prepared for senior management in Holland and was not intended to be an exhaustive analysis of all competitors in the market, but simply to portray the 97
[98]However, this explanation does not detract from the tiered manner in which AkzoNobel views the market; the fact that Plascon and Dulux are reflected as the only brands that are generally positioned in the premium segment; the fact that Duram is generally positioned as a mid-tier brand; and the fact that DOBs are generally positioned as mid- tier and economy brands. If the premium ranges of Duram or any DOBs had a significant presence in the premium segment, one would have expected, even on the basis of Mr Smidt’s explanation, to see those brands also reflected in the premium segment of the market. Medal
[99]Medal’s Mr Madhu also spoke to the barriers created by the strength of the Plascon and Dulux brands in the premium segment. He explained that Medal manufactures DOBs 94 Trial Bundle, Part C, p 3328. 95 Trial Bundle, Part C, p 1330. 96 Deep Dive Paint Sub-Sahara Africa” dated 7 February 2022, Trial Bundle, Part C, p 2269 ff at p 2296, slide 29. 97 Hearing Transcript, Smidt, pp 588-589, 603-606, 673-674. 27 for Cashbuild and Build-It, but also decided to launch its own Medal brand in about 2006. However, despite intensive marketing over a period of 15 years, Medal has made very limited inroads into the premium segment (Mr Madhu estimated that Medal has achieved a premium segment share of below 5%, which represents% of its sales) and has therefore decided to focus on middle to lower-income consumers (LSM category 7 and below) in the mid-tier and economy segments of the market instead. 98
[100]Like Mr Stekhoven, Mr Madhu attributes Medal’s lack of success in the premium segment to the inability of Medal to establish brand equity that is comparable to the brand heritage of Dulux and Plascon, in which the merger parties continue to invest heavily. He testified that, without similar brand equity, entrant brands are not attractive to consumers in the premium segment, even at lower prices, and face difficulties in obtaining shelf space from retailers. He added that it is also for this reason that DOBs are targeted at the mid-tier and economy segments of the market. 99
[101]Mr Madhu testified further that, because Plascon and Dulux are “household brands” with a long history in South Africa, they are “incredibly strong” in the premium segment, with Duram a distant third and Medal even further behind. Mr Madhu explained that Medal’s “premium” range of products, despite being positioned and marketed by Medal as such, is regarded by retailers as a mid-tier brand because it does not have the same brand equity as Plascon and Dulux. Medal has therefore had to price its paint products at significantly lower price points in an attempt to drive sales. Even then, however, and despite extensive advertising, it has “failed hopelessly” to make an effective entry into the premium segment. 100
[102]Mr Madhu therefore expressed the view that Plascon and Dulux “own” the premium segment of the market, and would be able to increase the prices of their products in that segment post-merger by removing each of the brands’ closest competitor. 101 Promac
[103]The other manufacturer who testified at the hearing, Mr Prost of Promac, was called by the merger parties in relation to the proposed divestiture of the Micatex sub-brand (discussed further below). However, the motivation that Mr Prost gave for Promac’s 98 Madhu witness statement, Trial Bundle, Part A, p 426, para 7; pp 429-430, paras 18-22; Madhu, Hearing Transcript at p 870. 99 Madhu witness statement, Trial Bundle, Part A, pp 429-430, paras 19-20; p 433, paras 32-34. 100 Madhu witness statement, Trial Bundle, Part A, pp 431-432, paras 25-28; Madhu, Hearing Transcript at pp 878- 879, 887-889, 933-934, 945, 959. 101 Madhu, Hearing Transcript at pp 962-963, T970. 28 interest in acquiring the Micatex sub-brand is of significance in considering the brand strength of the Dulux and Plascon brands.
[104]Mr Prost claimed that Promac has its own premium, mid-tier and economy ranges which compete directly with the merger parties’ equivalent products in each quality segment, albeit at lower prices. He gave, as an example of this, Promac’s exterior paint, Micaguard, which was introduced in 2001 as a competitor to Dulux’s Weatherguard and Plascon’s Micatex products. 102 Mr Prost acknowledged, however, that despite having been in the market for over 20 years, Micaguard is “not as successful and well-known to consumers as the Micatex brand” which has a “very long history and brand recognition with consumers”. Mr Prost explained that, for this reason, the acquisition of Micatex would result in an immediate boost to Promac's overall volumes, effectively doubling its production of textured paints. 103 .
[105]This evidence, although given for other purposes, also indicates that, notwithstanding that other manufacturers might have introduced brands to compete in the premium segment with Plascon and Micatex, those efforts have been largely unsuccessful, despite being sold at lower price points, because of the brand strength of Dulux and Plascon in that segment. Retailers
[106]The evidence of retailers who testified at the hearing corroborates this. Mr Mahabeer explained that Build-It chooses different brands in each of the quality segments of the market, and regards Plascon and Dulux as the only competitors in the premium segment because of their “dominant presence in the premium segment with their strong brand heritage and brand equity”. Therefore, Build-It has head office contracts with both brands, but has selected Plascon over Dulux to be the single premium segment brand it carries in its stores. Mr Mahabeer stated that both the quality and the pricing of the Plascon and Dulux products in the premium tier is the same, but Build-It regards Plascon as providing slightly better service than Dulux. 104 Mr Mahabeer stated that customers seeking Dulux products are generally willing to buy the Plascon equivalent that is available in store. 105 102 Prost witness statement, Exhibit 12, para 2. See also Hearing Transcript, Prost, pp 1145-1146. 103 Prost witness statement, Exhibit 12, paras 9 and 11. 104 Mahabeer witness statement, Trial Bundle, Part A, pp 396-397, para 20; Mahabeer, Hearing Transcript at pp 980-981. 105 Mahabeer witness statement, Trial Bundle, Part A, pp 396-398, paras 20-22; Mahabeer, Hearing Transcript at pp 977-981. 29
[107]Mr Mahabeer testified further that Build-It attempted to introduce a premium range of DOB products into the premium segment, but failed to make an effective entry into that segment despite extensive marketing efforts (including offering lengthy guarantees and reduced pricing) over a period of 10 years. He likewise attributed this to the brand equity enjoyed by Plascon and Dulux in the premium segment, where a customer “buys the brand that he knows”. 106
[108]Mr Mahabeer acknowledged that Duram has a few products that cater to the premium segment, but stated that it is an “insignificant” player in the premium segment relative to Plascon and Dulux. 107 Mr Mahabeer explained that the development of brand equity requires, inter alia, significant commitments in terms of time, energy and money in training and educating stores, sales staff and customers about the manufacturer's products. He testified further that Dulux and Plascon were the first entrants in the premium segment of the market, and are the market leaders with the most brand equity with end-consumers. He noted that Duram is now beginning to generate brand equity, but stated that this is mostly in the mid-tier segment where the focus of its efforts lies. 108
[109]The evidence of Mr Carter is also illustrative of the relative brand strength of Plascon and Dulux in the premium segment. He explained that BBS Mica was opened in Ballito, KwaZulu-Natal in September 2021 in close proximity to another store that is part of the Mica buying group, namely the Mica store in the Ballito Lifestyle Centre (“Mica LC”). 109 When BBS Mica first opened, it did not stock any Dulux or Plascon paints, and instead carried the mid-tier and premium ranges of Duram, Promac and Excelsior because BBS Mica wanted to differentiate its offering from that of Mica LC (which stocked Dulux and Plascon in its premium offering). 110
[110]However, BBS Mica subsequently decided to expand its offering to include Plascon products in February 2023, driven in large part by demand for perceived premium products by wealthy consumers (LSM category 8 - 10) in the Ballito area. Between the introduction of Plascon in February 2023 and the Tribunal hearing in September 2023, Plascon grew to account for 8-10% of BBS Mica’s store’s total paint sales. 111 106 Mahabeer, Hearing Transcript at pp 987-988, 1093. 107 Mahabeer witness statement, Trial Bundle, Part A, p 398, para 23. 108 Mahabeer witness statement, Trial Bundle, Part A, p 399, paras 27-29; p 401, para 35; Mahabeer, Hearing Transcript at pp 982-983. 109 Carter witness statement, Trial Bundle, Part A, pp 375-376, para 4. 110 Carter witness statement, Trial Bundle, Part A, p 376, para 6; Hearing Transcript, Carter, pp 453-455. 111 Hearing Transcript, Carter, pp 439, 448, 459. 30
[111]Mr Carter acknowledged in his oral evidence that “Plascon really is a leader in the paint industry and it is hard to compete with them”. 112 As a result, a significant number of consumers in the premium segment are not open to buy other brands that might be available in-store (“their eyes glaze over”). 113
[112]Mr Bornman explained that Pits Paints is a paint retailer predominantly servicing the Limpopo, Mpumalanga and Western Cape regions. Pits Paints is also a Dulux-preferred supplier, and is therefore contractually incentivised to stock and sell Dulux products to its customers. Dulux products accordingly represent over% of Pits Paints’ total paint sales, but it also carries a range of Excelsior products for consumers who cannot afford Dulux paints, as well as small quantities of Plascon products. 114 Mr Bornman acknowledged that customers generally prefer to stick with brands they know, and that Plascon and Dulux are well-known household brands due to the significant investment that has been made in them over many years. 115
[113]Mr Greeff explained that EST is a buying group whose members are typically small stores located in rural areas. 116 He testified that the Dulux and Plascon brands are “pivotal” to EST’s members, and that he could not imagine many stores being able to survive without one or both of them. Mr Greeff explained that this was because of the (actual and perceived) quality of those brands, and their brand equity with consumers. 117 Mr Greeff stated further that it would take years for Duram to develop its brand sufficiently to become a close competitor to the merged entity in the premium segment, and that the other brands were so far behind Dulux and Plascon as to be immaterial. 118
[114]Other customers and competitors interviewed by the Commission during its investigation also generally confirmed that the decorative coatings market is segmented into quality tiers, and that Plascon and Dulux are the two largest and closest competitors in the broad market, and in the premium segment in particular. 119 112 Hearing Transcript, Carter, p 424. 113 Hearing Transcript, Carter, pp 415-416, pp 431-433. 114 Bornman witness statement, Trial Bundle, Part A, p 382, paras 2-4; Hearing Transcript, Bornman, pp 339-351, 386. See also Exhibits 3 and 4. 115 Hearing Transcript, Bornman, pp 358-359 and 361. 116 Hearing Transcript, Greeff, pp 1101-1102. 117 Hearing Transcript, Greeff, pp 1103-1104 and 1108. 118 Hearing Transcript, Greeff, pp 1107-1110, 1128-1130, and 1133. 119 Expert Presentation of Mr Njisane, Exhibit 15, slides 28-29. 31
[115]The above evidence thus suggests that Dulux and Plascon are the pre-eminent quality- associated brands in the decorative coatings market in South Africa with unrivalled brand equity in the eyes of consumers, particularly in the premium segment of the market. This position has been established and entrenched by significant ongoing investment in these brands over many years. The only other brand that appears to have made any significant entry into the premium segment is Duram, 120 but the evidence suggests that it is currently a very distant third player in the premium segment, in terms of both brand equity and market share. Supply-side substitution and new entry
[116]The above evidence also suggests that brand equity is a major inhibitor not only of demand-side substitution, but also of supply-side substitution in the premium segment of the market. Both Mr Smidt and Mr Lalla testified that there are low (if any) technical barriers to the production of premium-quality paint. They explained that, although production processes may vary, the production process for paint is well-established, and raw materials suppliers can supply the inputs necessary to enable any manufacturer to make any high quality paint product. Mr Smidt also referred to various recent entrants into the decorative coatings market. 121 Mr Carter of BBS Mica agreed that “anyone can make good paint”. 122 It was on this basis that Mr Smith submitted that that there is ready supply-side substitution into the premium segment of the market. 123
[117]However, it does not appear that the intrinsic quality of the premium Dulux and Plascon products is as easy for other brands to match as the merger parties suggest. As noted above, the merger parties maintain that their products are in fact of a higher quality than any other products in the market, which raises the question why, if the quality of the merger parties’ products was easy to replicate, this has not yet been done.
[118]Second, there is no suggestion that any of the new entrants referred to by Mr Smidt have, apart from Duram, made any entry into the premium segment, and we have discussed above the amount of time it has taken Duram to make even the limited entry it has achieved in the premium segment. 120 See, eg, Trial Bundle, Part B at pp 808, 2331. 121 Smidt witness statement, Trial Bundle, Part A, p 255, para 20; pp 289-291, paras107-111; Smidt, Hearing Transcript at pp 565-567. See also Lalla witness statement, Trial Bundle, Part A, p 367, para 33. 122 Carter witness statement, Trial Bundle, Part A, p 378, para 11. 123 Expert Presentation of Mr Njisane, Exhibit 15, at slide 13; Smith, Hearing Transcript at pp 1229-1230, 1326- 1328. 32
[119]Third, and related to the above, Mr Smith’s supply-side substitution analysis ignores the critical role that brand equity plays in the premium segment of the market. The difficulties experienced by Duram, Medal and Build-It all suggest that, even with a high-quality paint range, equivalent guarantees and competitive prices, entry into the premium segment is unlikely to be successful without significant brand equity, which generally takes many years, if not decades, to develop. 124 The pricing power of Dulux and Plascon
[120]A clear indicator of the unrivalled position enjoyed by the Dulux and Plascon brands in the premium segment is the fact that they have been able to command, and maintain, a significant price premium to all other brands, including the so-called “premium” brands of Duram and other manufacturers. Whether this price premium is a function of actual quality advantages (higher quality ingredients, better spread rate etc.) or perceived quality advantages (brand equity), or both, it indicates that Dulux and Plascon are regarded as offering the highest quality products in the premium segment, for which consumers are willing to pay a premium. By the same token, other manufacturers’ “premium” products are not regarded as having the same quality as equivalent Dulux and Plascon products, and they are accordingly forced to price their products at a considerable discount in order to make them attractive to customers. As a result, Dulux and Plascon are priced closest to each other in the premium segment, with equivalent “premium” Duram products being positioned at considerably lower price points. 125
[121]Mr Smidt confirmed in his evidence that, from a brand equity perspective, Dulux’s closest competitor is Plascon (it is “very, very close . . . because of a 100 years of heritage”). He explained further that this both requires the two brands to compete closely on price “to be competitive”; and also permits Dulux and Plascon to charge higher prices for equivalent products than any of their competitors. 126
[122]As noted above, a 2023 AkzoNobel competitor analysis states that 127 Mr Smidt confirmed in relation to this document that Duram products are generally cheaper than equivalent Dulux products in the premium segment, and that this is a function of actual and perceived quality, and service 124 Njisane, Hearing Transcript at pp 1329-1330. 125 Smidt witness statement, Trial Bundle, Part A, p 255, paras 18-19, and p 259, para 35; Smidt, Hearing Transcript at p 516; Lalla witness statement, Trial Bundle, Part A, pp 365- 366, paras 26 and 29-31; Lalla, Hearing Transcript at pp 206, 215-227. See also Trial Bundle, Part C, pp 8604, 6292. 126 Smidt, Hearing Transcript at pp 619-620, 627-628, 634-635. See also Trial Bundle, Part C, p 3327. 127 Trial Bundle, Part C, p 3328. 33 factors, including that “they do not have the brand equity we have” and accordingly “they cannot price at the price point we can price at, in our premium space”. 128
[123]Mr Lalla similarly stated in his evidence that one of the reasons why Plascon is able to command premium prices for its products is that it has a brand heritage that is over 130 years old. 129 He also confirmed that the competitor which prices closest to Plascon for equivalent products is Dulux, and that Duram prices its “premium” products at a relative discount to equivalent Dulux and Plascon products. This is also reflected in KPAL’s internal documents. 130 These documents also reflect that KPAL generally compares Plascon’s prices against those of Dulux and Duram in the premium and mid-tier segments. 131
[124]Mr Lalla explained that KPAL’s price comparison documents are based on Builders Warehouse data and that KPAL does not have the capability to track pricing throughout the market. 132 We accept that these price comparisons do not necessarily reflect all the players in the relevant markets, but they remain probative, in our view, of which brands KPAL regards as its main competitors.
[125]Mr Stekhoven confirmed that Duram prices its products at a discount of approximately 15-25% to equivalent Dulux and Plascon products in the light of the strength of the latter two brands. 133 He stated further that, if Duram attempted to increase the price point of its products, it would “hit up” against the brand presence of Dulux and Plascon, and lose sales volumes. 134
[126]The price premium that Plascon and Dulux are able to command over other “premium” products because of their relative brand strength was also confirmed by Mr Carter of BBS Mica 135 and Mr Prost of Promac 136 .
[127]The merger parties argued that the significant price gap between Dulux’s products and those of Duram does not take into account the better “spread rate” of Dulux products, which results in less Dulux paint being required to paint the same surface area and lower 128 Smidt, Hearing Transcript at pp 608-617. See also Hearing Transcript at p 516. 129 Lalla, Hearing Transcript at p 206. 130 Lalla, Hearing Transcript at pp 206, 215-227. See also Trial Bundle, Part C, pp 6292-6293, 8602-8607. 131 See also Lalla, Hearing Transcript at pp 229-233, Trial Bundle, Part C, pp 6282-6287. 132 Lalla, Hearing Transcript at p 307. 133 Stekhoven, Hearing Transcript at pp 777, 784. 134 Stekhoven, Hearing Transcript at pp 846-847. 135 Carter, Hearing Transcript at p 452. 136 Prost, Hearing Transcript at pp 1200-1201. 34 labour costs. They contended that, if this is taken into account, the price gap between Dulux and Duram products would reduce considerably, or even disappear. 137
[128]While we accept that spread rate is relevant to the ultimate cost of a product, it was not clear to us to what extent this is in fact factored into a consumer’s assessment of price, and there was no evidence that this accounts for the considerable price gap between the premium products of Plascon and Dulux on the one hand, and those of Duram (and other brands) on the other. On the contrary, the evidence above indicates that Duram and other products are deliberately positioned at materially lower price points than the premium products of Dulux and Plascon because of their much weaker brand equity, and are perceived by consumers as mid-tier in nature.
[129]Therefore, insofar as the spread rate factor exists, it simply serves to illustrate the quality and price differences between the merger parties’ premium brands and other brands in the market; it does not demonstrate any closeness of competition between them.
[130]In his evidence, Mr Smith testified that there is “no magic in price levels” because “economy”, “mid-tier” and “premium” products are sold across a continuous range of prices (calculated on a price per litre basis) without any clean breaks between quality tiers. 138 137 Smidt witness statement, Trial Bundle, Part A, p 259, paras 35-36; Smidt, Hearing Transcript at pp 493-497; Lalla witness statement, Trial Bundle, Part A, p 365, para 26; Lalla, Hearing Transcript at pp 143-144. 138 Expert Presentation of Mr Smith, Exhibit 14, slide 14; Smith, Hearing Transcript at p 1230-1231. 35
[131]However, as Mr Smith acknowledged, what one manufacturer positions as “premium” or “mid-tier” might not be regarded as such in the market. In addition, an average price per litre calculation masks differences between prices of different-sized containers (20 litre, 5 litre etc.), and there may be composition problems with the identification of the product types themselves. 139 These difficulties notwithstanding, the above graph confirms, in our view, that there is a distinct segmentation of price ranges between economy, mid- tier and premium brands. The fact that there might be overlaps between the segments is not surprising given the subjective nature of segmentation in this market, and does not detract from the fact that mid-tier brands are overwhelmingly priced below premium brands, and economy brands below mid-tier brands.
[132]Mr Smith also testified that there are significant price differences between different brands, and that even individual brands (which may include multiple products) 140 are sold at a wide range of different prices. 141 However, both of these propositions are consistent with the price segmentation between quality tiers referred to above. What is of relevance is not the existence of a price range itself, but how the price ranges of the different brands compare with one another.
[133]In addition, KPAL’s own pricing analyses referred to above show that, at both an individual product level and a product category level, the prices of equivalent Plascon and Dulux products are positioned the closest to each other, with Duram priced at a significant discount. This is also evident from the pricing analysis conducted by the Commission. 142
[134]Having regard to all the above evidence, we agree with Mr Njisane’s conclusion that: “Duram is there as a competitor, but it is a far distance from [where] Micatex or Dulux sit and you can see that consumers themselves, even where Duram has positioned a product head on, consumers are willing to pay an additional R100 for a Plascon brand as opposed to a Duram, . . . what that tells me when I read that is that they may either not see it as a good, an efficient substitute, so they are willing to pay a higher price rather than turning to a Duram product as an alternative.” 143 139 Smith, Hearing Transcript at pp 1230-1231, 1286; Njisane, Hearing Transcript at pp 1347-1348. 140 Expert Presentation of Mr Smith, Exhibit 14, slides 18-20; Smith, Hearing Transcript at pp 1234-1235. 141 Expert Presentation of Mr Smith, Exhibit 14, slides 15-16; Smith, Hearing Transcript at pp 1231-1235, 1286- 1287. 142 Expert Presentation of Mr Njisane, Exhibit 15, slides 23-25; Njisane, Hearing Transcript at pp 1347-1348. 143 Njisane, Hearing Transcript at pp 1644-1645. 36 Dealer-owned brands (“DOBs”) and countervailing power
[135]In addition to Duram, the merger parties argued that the various retailer DOBs that have entered the market in recent years also compete closely with Plascon and Dulux in the premium segment. 144
[136]Mr Smidt explained that in South Africa, decorative coatings are primarily supplied to retailers, including (i) large sales outlets (“LSOs”) such as Builders Warehouse, Cashbuild, Checkers Hyper, Leroy Merlin, Makro and Game; (ii) small independent retail outlets (“SSOs”); and (iii) franchise operations such as Mica, Build-It and Buco that use a buying group model to negotiate prices on their behalf. 145
[137]Mr Smidt testified that retailers account for the vast majority of decorative paint sales in South Africa, and that there are several factors that give DOBs a significant competitive advantage over third party paint brands in the outlets of the retailers to which they belong. Retailers control the shelf space allocation in their stores, and they are strongly incentivized to give their own DOBs the best, most visible shelf space. Retailers are also well-positioned to market their DOBs over third-party brands when engaging with customers in-store. Mr Smidt stated that retailers are able to use their power as “gate keepers” to switch customers to their own brands, or other brands on which they earn the most margin, if they so wish, or to delist brands if they are unhappy with their pricing. 146
[138]Mr Smidt testified that Massmart has been particularly successful in this regard, with its Fired Earth DOB. 147 Mr Smidt explained that Massmart accounts for a very large number of the sales outlets for decorative paint in South Africa, with over 230 Builders Warehouse, Builders Express, Makro and Game stores around the country, and that it primarily stocks Dulux, Plascon, Duram and Fired Earth in its stores. 148 144 Smidt witness statement, Trial Bundle, Part A, pp 265-266, para 61; T527-528. 145 Smidt witness statement, Trial Bundle, Part A, pp 276-277, para 86. 146 Smidt witness statement, Trial Bundle, Part A, p 266, para 62; pp 276-281, paras 86-91; Smidt, Hearing Transcript at p 536-538. See also Lalla, Hearing Transcript at pp 149-150. See also Expert Presentation of Mr Smith, Exhibit 14, slides 65-69; Smith, Hearing Transcript at pp 1248-1249, 1423-1427. 147 Smidt witness statement, Trial Bundle, Part A, pp 266-267, paras 63-65. 148 Smidt witness statement, Trial Bundle, Part A, p 260, para 42; pp 281-282, paras 92-95; Smidt, Hearing Transcript at pp 551-552; Expert Presentation of Mr Smith, Exhibit 14, slide 55. 37
[139]However, whilst it appears that Fired Earth has, as a result of the retailer power of Massmart, made a successful entry into the broad decorative coatings market, there is no evidence that Fired Earth (or any of the retailer DOBs) have made any significant entry into the premium segment of the market.
[140]For instance, Mr Stekhoven testified that Fired Earth competes mostly in the mid-tier segment, and he is not even aware whether it has a premium range of products. 149 Consistent with this, Mr Lalla explained that KPAL produces the tint bases for Fired Earth according to a Duram specification rather than a Plascon specification because Massmart is targeting a lower quality, lower price consumer market than Plascon. 150 KPAL’s internal documents also refer to Fired Earth as being a competitor in the mid- tier segment. 151
[141]The evidence of other witnesses corroborated this. Mr Madhu testified that DOBs are not generally perceived as premium segment products even if they are marketed by their suppliers as such. 152 In addition, we have referred above to Build-It’s inability to enter the premium segment with its “premium” range of DOB products. 153 Mr Mahabeer explained that Build-It’s DOBs are, as a result, currently positioned only in the economy and mid-tier segments of the market and are targeted at customers who cannot afford the premium segment products of Plascon and Dulux. As such, Build-It regards its DOBs as a complement to, rather than a substitute for, Placon and Dulux. 154
[142]Again, this is not to say that DOBs are not successful and growing brands in their respective retail outlets. However, the above evidence suggests that any such growth is taking place primarily in the mid-tier and economy segments of the market rather than in the premium segment. No other brands appear to have made any material entry into the premium segment, despite their best efforts to do so, often over considerable periods of time.
[143]As regards countervailing power, it seems clear from the evidence that Massmart and other large retail groups exercise considerable bargaining power in their dealings with decorative paint manufacturers. In his witness statement, Mr Smidt cited various 149 Stekhoven, Hearing Transcript at p 766. 150 Lalla, Hearing Transcript at pp 267-270. 151 Lalla, Hearing Transcript at pp 272-274, Trial Bundle, Part C, p 6353. 152 Madhu, Hearing Transcript at pp 875-877, 938-940. 153 Mahabeer, Hearing Transcript at pp 987-988, 1093. 154 Mahabeer, Hearing Transcript at pp 973-974, 985, 999. 38 instances where retailers replaced AkzoNobel’s products with products of manufacturers other than KPAL. 155
[144]However, we are unable to determine from this anecdotal evidence how widespread it is, or, more importantly, what the reasons were for this conduct. As discussed further below, it may be a function of income effects arising from the declining South African economy or other factors unrelated to relative rises in the prices of the merger parties’ products. In circumstances where there are no (or very few) alternatives to Plascon and Dulux in the premium segment, it does not appear to us that retailers would be able to materially constrain the merger parties’ conduct in that segment post-merger. 156 As noted earlier in these reasons, forms of buyer power that do not result in new entry are unlikely to prevent any anti-competitive merger effect. This is because a retailer’s buyer power depends on the availability of good alternatives to the merger parties’ products, which, in the context of an anti-competitive effect, will necessarily have been reduced. 157
[145]It might be said that DOBs are precisely the kinds of new entry that retailers could promote post-merger, as they have done already. However, as discussed above, the evidence suggests that DOBs, including Massmart’s Fired Earth, have not to date been able to materially penetrate the premium segment of the market given the importance of brand equity in that market, and there is no reason to believe that their ability to do so would be any greater after the merger.
[146]Therefore, it does not appear to us that, on the facts of this case, the existence of countervailing power advances the debate on anti-competitive effects. 158
[147]The stances of the retailers who testified at the hearing were mixed. Whilst Mr Mahabeer and Mr Greeff expressed concerns about the unilateral effects of the merger in the premium segment, Messrs Carter and Bornman appeared sanguine about the ability of retailers to discipline the merged entity if it attempted to raise prices post-merger, by shifting customers to other brands. The views of Messrs Carter and Bornman might be influenced by the fact that they each have strong preferences for only one of the merger parties’ brands in their stores (Plascon and Dulux respectively), but the stance of Massmart as a major customer of both merger parties is less clear. Massmart itself did not testify at the hearing, and its submission to the Commission does not appear to 155 Smidt witness statement, Trial Bundle, Part A, pp 279-281, paras 90-91. See also Expert Presentation of Mr Smith, Exhibit 14, slides 65-69; Smith, Hearing Transcript at p 1248-1249, 1423-1427. 156 Njisane, Hearing Transcript at p 1261. 157 See CMA Merger Assessment Guidelines, supra, at at para 4.20; EU Horizontal Merger Guidelines, supra, at paras 64-67. See also Expert Report of Mr Smith, Trial Bundle A at p 464, para 94. 158 Njisane, Hearing Transcript at p 1441. 39 reflect any material concerns with the unilateral aspects of the merger despite its position that the merger parties are close competitors. 159
[148]Given that Massmart was not called to explain the submission it made to the Commission, we were not able to interrogate the reasons for its stance on the merger. It appears to us, however, that the position of retail customers such as Massmart may be different from that of end-consumers in that, while a post-merger price rise in the premium segment would clearly harm end-consumers in that segment, it might enable retailers to switch those end-consumers to other (mid-tier) brands, including their own DOBs, that provide the retailers with equivalent, or greater, margins.
[149]It therefore does not appear to us that the stance of Massmart in this matter is necessarily inconsistent with the other evidence we have heard regarding the anti- competitive effects of the merger in the premium segment of the retail channel. Other brands
[150]The merger parties also referred in their evidence to other numerous decorative coatings manufacturers in South Africa, most of which were said to have their own respective “premium”, “mid-tier” and “economy” product ranges. Apart from Duram, Medal, Promac and the DOBs referred to above, these included a long list of national and regional manufacturers such Excelsior, Dekade, Versus, Universal, Warrior, Stevensons, Dekro, Shaves and many others. 160 Mr Smidt stated that there are over 70 decorative coatings manufacturers registered on the website of the South African Paint Manufacturers’ Association (SAPMA). 161
[151]The merger parties contended that all these brands compete effectively with Plascon and Dulux in the premium segment, and referred to numerous internal documents of the merger parties in which various of these brands are referred to, and analysed, as “competitors”, “threats” or “benchmarks”. 162 159 Trial Bundle, Part B, pp 4945-4947. 160 Smidt witness statement, Trial Bundle, Part A, pp 264-265, paras 56-60 and pp 268-272, para 67; Hearing Transcript at pp 525-532. See also Bornman witness statement, Trial Bundle, Part A, pp 384-386, paras 10-14; Expert Presentation of Mr Smith, Exhibit 14, slides 38-41; Smith, Hearing Transcript at pp 1240, 1360-1361. 161 Smidt witness statement, Trial Bundle, Part A, p 256, para 22. 162 See, eg, Lalla, Hearing Transcript at pp 186-199, Trial Bundle, Part B, pp 2335-2337, Trial Bundle, Part B, pp 3106-3405; Hearing Transcript at pp 196-197, Trial Bundle, Part C, pp 6108-6113, Trial Bundle, Part C, pp 6582- 6587; Smidt, Hearing Transcript at pp 590-594, Trial Bundle, Part C, pp 1024-1030, 1183-1191, 1967-1972, 3925- 3942, 5855-5861. 40
[152]However, none of these documents reflect any of these other brands (many of which appear to be regional or local in nature) 163 as being close competitors of Dulux and Plascon in the premium segment. Recall that the Commission’s case is not that there is no competitive interaction at all between the merger parties’ brands and other brands in the market; it is that Dulux and Plascon are the closest competitors to each other, particularly in the premium segment, and that the strength of this competitive constraint would not be maintained by other brands post-merger. As we read them, none of the “competitor” documents referred to by the merger parties speak to the relative closeness of competition between the merger parties’ brands and other brands in the premium segment of the market. Where there is reference in the merger parties’ documents to the premium segment, they overwhelmingly reflect Plascon and Dulux (and occasionally Duram) as the only participants in that segment of the market. 164 The Commission’s diversion ratio analysis
[153]Evidence of cross-price elasticity, or diversion ratios, is often regarded as the most probative quantitative evidence of whether two products are close competitors.
[154]In this case, the Commission estimated diversion ratios between the merger parties and other manufacturers in the broad decorative coatings market based on its market share estimates referred to above. This analysis indicated that the merger parties were more likely to lose sales to each other than to other competitors following a price rise, which suggested that the merger would likely result in anti-competitive unilateral effects. The Commission also conducted UPP (Upward Pricing Pressure) and GUPPI (Gross Upward Pricing Pressure) tests to quantify the likelihood and quantum of such a price increase on the basis of the same market share data, which indicated that the merger was likely to give rise to significant price increases. 165
[155]However, as Mr Smith observed, these analyses were based simply on market share data, and therefore do not advance the debate on closeness of competition any further than the market share data itself, the limitations of which have been addressed above. Mr Smith also disputed the margin assumptions in the Commission’s model. 166 We 163 See Exhibit 18. 164 See, eg, Lalla, Hearing Transcript at pp 186-187, Trial Bundle, Part B, pp 2330-2331. See also Trial Bundle, Part B, p 808, Trial Bundle, Part C, p 2296, Trial Bundle, Part C, 8602-8607. 165 Expert Presentation of Mr Njisane, Exhibit 15, slides 38-41; Njisane, Hearing Transcript at pp 1270, 1415-1421. 166 Smith, Hearing Transcript at pp 1240, 1422-1424; 1428-1431. 41 therefore do not place any reliance on the results of these analyses for our findings in this matter. Stocking of brands
[156]In support of their contention that Dulux and Plascon are not uniquely close competitors, the merger parties testified that there are various retail stores in which the Dulux and Plascon brands are not the largest selling paints, and also stores where one or even both brands are not sold at all. 167 Mr Smith contended on this basis that the merger parties predominantly serve different customers. 168
[157]We do not believe this conclusion can be inferred from the evidence. Not only is it inconsistent with the evidence referred to above (including that of Mr Smidt and Mr Lalla themselves) that Dulux and Plascon are close competitors of each other across the decorative coatings market, but it ignores the fact that stores sell products in all quality tiers relevant to their target markets. Therefore, stores with more of a mid-tier or economy target market would be expected to sell proportionately less Dulux and Plascon products than stores targeted at high LSM consumers. As explained above, stores in low LSM areas may not sell any premium products at all. 169
[158]Furthermore, the fact that certain stores may stock only Plascon or only Dulux products, but not both, may be attributable to strategic reasons, or space constraints, or the fact that certain stores are Plascon or Dulux-preferred suppliers. For example, as explained above, Build-It has a one brand strategy per quality tier, and prefers Plascon to Dulux in its premium segment for various reasons. This clearly does not mean that Plascon and Dulux are not close competitors. On the contrary, Mr Mahabeer’s evidence was that Build-It regards Plascon and Dulux as the only competitors in the premium segment and for that reason it has head office contracts with both brands; it has simply selected Plascon in the competition to be the single premium tier brand that Build-It carries in its stores. 170
[159]Mr Carter explained BBS Mica’s decision to stock only Plascon products in its store on a similar basis. While BBS Mica wanted to differentiate its offering from that of Mica LC 167 Smidt witness statement, Trial Bundle, Part A, pp 282-283, para 97; Smidt, Hearing Transcript at p 561; Lalla Hearing Transcript at p 149. 168 Expert Presentation of Mr Smith, Exhibit 14, slide 55; Smith, Hearing Transcript at pp 1241-1244, 1394-1397. 169 See also Lalla, Hearing Transcript at p 266 (referring to Cashbuild). 170 Mahabeer witness statement, Trial Bundle, Part A, pp 396-398, paras 20-22; Mahabeer, Hearing Transcript at pp 977-981. 42 (which stocked both Dulux and Plascon in its premium offering), 171 it decided that it needed to have a premium segment brand in its store in order to attract wealthy consumers, and chose Plascon rather than Dulux for this purpose. 172
[160]These examples, therefore, do not, in our view, reflect any absence of close competition between Dulux and Plascon but rather an outcome of that competitive interaction. The merger parties’ switching evidence and the “buying down” phenomenon
[161]Mr Smith also argued that, even where there are overlaps between the merger parties’ brands in retail outlets, there is no evidence of switching between them. Mr Smith relied in this regard on a number of graphs showing that, when one of the merger parties’ premium products lost sales in a particular retailer, there was no evidence that the equivalent products of the other merger party gained sales; on the contrary, the sales of the other merger party’s equivalent products declined as well. 173
[162]Mr Njisane raised various data objections to Mr Smith’s analyses, including that they were selective, and unjustifiably grouped different products together with the result that the comparisons were not “like for like”. 174
[163]More fundamentally, however, Mr Njisane argued that that there are income effects that are likely driving the trends in Mr Smith’s analyses rather than an absence of substitution effects between the merger parties’ brands.
[164]The witnesses of the merger parties and the Commission all agreed that there has been a significant shift in demand from premium to mid-tier paints in South Africa in recent years, which was referred to as the “buying down” phenomenon. 175 Mr Lalla explained that, as macroeconomic conditions have deteriorated in South Africa, consumers have become more price-sensitive and less willing to pay for quality differentials, and therefore have been switching away from expensive, premium paints to more affordable options. 176 171 Carter witness statement, Trial Bundle, Part A, p 376, para 6; Hearing Transcript, Carter, pp 453-455. 172 Hearing Transcript, Carter, pp 439, 448, 459. 173 Expert Presentation of Mr Smith, Exhibit 14, slides 56-60; Smith, Hearing Transcript at pp 1246-1247, 1397- 1406. 174 Njisane, Hearing Transcript at pp 1407-1408. 175 Smidt witness statement, Trial Bundle, Part A, p 256, para 25; Hearing Transcript at p 503. 176 Lalla witness statement, Trial Bundle, Part A, pp 367-368, para 35. 43
[165]Numerous of the merger parties’ documents also referred to this buying down trend. 177 Mr Smith’s analysis, which covered the period 2018 to 2022, also showed a substantial shift in demand away from higher-priced to lower-priced brands in the market. 178
[166]This “buying down” trend, which appears to have commenced in about 2015, played into the hands of Duram, which (as discussed above) was entering the market with a strong mid-tier offering at that time. AkzoNobel, for its part, responded to the trend by adding the Dulux mother brand to its existing mid-tier RockGrip product, and introducing a new MaxiCover sub-brand in the mid-tier segment, in order to leverage the equity of its Dulux mother brand into the mid-tier segment. 179 KPAL did not have a mid-tier offering at the time and therefore launched the Plascon TradePro sub-brand into the mid-tier segment in order to respond to the growing demand for mid-tier products. 180 181
[167]The merger parties relied on this “buying down” phenomenon to argue that the increasing price-sensitivity of consumers has rendered mid-tier products an effective competitive constraint on Dulux and Plascon in the premium segment. However, it is important to distinguish in this context between income effects and substitution effects. As Mr Njisane explained, an income effect reflects a change in purchasing behaviour of consumers in response to a change in their income. A substitution effect, on the other hand, reflects a change in purchasing behaviour in response to a rise in prices or other exercise of market power. 182 Mr Njisane explained further that competition analysis is concerned with substitution effects and not income effects – not only do income effects change in line with general economic conditions, but the fact that certain consumers may “buy down” from premium to mid-tier (or economy) products in a declining economy 177 Lalla, Hearing Transcript at p 175, B814. See also Lalla Hearing Transcript at p 178-180, Trial Bundle, Part B, p 1891; Hearing Transcript at p 182, Trial Bundle, Part B, p 2330; Hearing Transcript at pp 198-199, Trial Bundle, Part C, p 7498. 178 Expert Presentation of Mr Smith, Exhibit 14, slides 28-33; Smith, Hearing Transcript at pp 1238-1239, 1351- 1352. 179 Smidt T646-653, T662-668, C38, C3670; Stekhoven witness statement, Trial Bundle, Part A, p 412, para 31. 180 Lalla witness statement, Trial Bundle, Part A, pp 367-368, paras 35-36; Lalla, Hearing Transcript at pp 135-136, 166, 180, 195; Trial Bundle, Part B, pp 761, 767-769; Trial Bundle, Part C, pp 5697, 5705. 181 See also Mahabeer witness statement, Trial Bundle, Part A, p 400, paras 30-31; Mahabeer, Hearing Transcript at p 987; Madhu witness statement, Trial Bundle, Part A, p 431, paras 25-26; Madhu, Hearing Transcript at p 904. 182 Njiane, Hearing Transcript at p 1269. 44 does not detract from the fact that there remains a premium segment of the market in which consumers do not regard a mid-tier product as a satisfactory substitute for a premium one. 183
[168]Indeed, the fact that the merger parties decided to reposition different products in the mid-tier segment rather than simply to reduce the prices of their premium products is testament to the enduring salience of the different quality tiers in the market (as is the decision by KPAL to reposition Polvin as a premium segment product in response to the competitive pressure it was experiencing from Duram in the mid-tier segment).
[169]We understood Mr Smith to agree with the distinction between income and substitution effects, and with the fact that the move from premium to mid-tier products largely reflects an income effect. He argued though that, in a depressed economy with increasingly price-sensitive consumers, there are also more likely to be substitution effects in response to relative price changes in premium and mid-tier products.
[170]However, as was put to Mr Smith, his analyses do not demonstrate substitution as opposed to income effects. 184 In particular, they do not show that the shift from premium to mid-price products is a function of relative price changes in premium and mid-tier products rather than the general buying down phenomenon that the merger parties acknowledged has been a fundamental feature of the decorative coatings market since 2015. Indeed, Mr Lalla stated in his evidence that the declining performance of Plascon at Massmart is due to the fact that KPAL “do not have the portfolio that is required for this market shift”. 185 Similarly, Mr Smidt explained that Game (one of the retailers covered by Mr Smith’s analyses) has, as a function of the declining economy, decided to focus on its own DOBs, with the result that Dulux has lost sales in that retailer. 186
[171]There are similar difficulties with the switching evidence referred to by Mr Smidt. In his witness statement, Mr Smidt cited various instances where retailers replaced AkzoNobel’s products with products of manufacturers other than KPAL. 187 However, there is again no evidence that this switching was a function of substitution rather than income effects arising from the declining South African economy. 183 Njisane, Hearing Transcript at pp 1269, 1328-1329. See also Smidt, Hearing Transcript at pp 647-657. 184 Smith, Hearing Transcript at pp 1401-1403. 185 Lalla, Hearing Transcript at p 150. 186 Smidt, Hearing Transcript at p 520. 187 Smidt witness statement, Trial Bundle, Part A, pp 279-281, paras 90-91. 45
[172]It is significant in this regard that no evidence was provided by the merger parties of switching in the other direction, i.e., retailers switching mid-tier products for the merger parties’ premium tier products – which might have been expected if the observed retailer behaviour reflected substitution rather than income effects.
[173]Therefore, while the Commission’s case was lacking in positive evidence of switching, it does not appear to us that Mr Smith’s analyses disprove the factual evidence above regarding the closeness of competition between the merger parties’ brands. Conclusion on closeness of competition in the retail channel
[174]Having regard to all the above evidence, it appears to us that: 174.1. Dulux and Plascon are the closest competitors in the premium segment of the decorative coatings market from both a price and perceived quality perspective. This is largely a function of the unparalleled equity of the Dulux and Plascon brands, which has been developed by very significant investment over many years. 174.2. The next closest competitor to Dulux and Plascon in the premium segment of the market is Duram, but it is a very distant third in terms of its perceived quality, price and market share in that segment. Other brands, including Massmart’s Fired Earth DOB, are even further behind, and arguably do not participate in the premium segment at all. Critically, none of these other brands has equity that is even close to that of Dulux and Plascon, and they have accordingly not been able to compete closely with those brands in the premium segment, even at significantly lower prices. 174.3. The available evidence does not suggest that the competitive constraints that Duram and other brands may impose on the merged entity post-merger would match the strength and effectiveness of the constraints that Dulux and Plascon currently impose on each other in the premium segment of the market. 174.4. While large retail outlets appear to have significant countervailing power, this would not constrain the merged entity in the premium segment of the market in circumstances where (as here) there are no close substitutes to Dulux and Plascon in the premium segment to which those retailers could switch end-consumers. 174.5. In the circumstances, the proposed merger would likely give rise to significant unilateral effects in the premium segment of the market. 46 CLOSENESS OF COMPETITION – TRADE CHANNEL Retail vs trade channel
[175]Mr Smidt explained that the trade segment refers to the sales channel in which paint products are supplied to tradespeople, ranging from small contractors (colloquially referred to as the “bakkie brigade”) 188 to large painting contractors and construction companies which typically purchase a range of products for a “project” such as a residential house or complex, or a corporate or government building. 189
[176]Mr Smidt testified further that the overall split between trade and retail sales in South Africa has been relatively consistent over the years, with retail accounting for around 40- 50% of sales, and the balance being accounted for by the trade channel. 190
[177]It was common cause that the trade channel, like the retail channel, is segmented into premium, mid-tier and economy tiers, and that manufacturers, including AkzoNobel and KPAL, manufacture specific brands for the trade channel (e.g., Plascon Professional and Dulux Trade) which have slightly different formulations to their equivalent retail brands. 191 Mr Smidt explained that a high proportion of paints used in the specified trade segment is comprised of premium paints because of the high potential cost associated with something going wrong with a large painting project. 192
[178]Mr Smidt emphasised, however, that trade customers also have the option to purchase retail brand paints if they wish. 193 As he explained: “Although most manufacturers produce a “trade” or “professional” range targeted at tradespeople, the quality of the paint in the bucket is no different to that which is available in the retail trade; and the products that are available in retail channels are generally also made available in the direct trade channel, and vice versa. For example, your typical household consumer can also purchase Dulux Trade at a retail store. There are therefore no differences in terms of the variety of brands available, product prices, and product quality available through each channel.” 194 188 As noted above, the Commission categorises the informal trade as part of the retail channel, but nothing turns on this categorization for purposes of our analysis. 189 Smidt witness statement, Trial Bundle, Part A, pp 283-284, para 98. 190 Smidt witness statement, Trial Bundle, Part A, p 284, para 99. 191 Lalla, Hearing Transcript at pp 176, 191-192; Smidt, Hearing Transcript at pp 562, 683-685. 192 Smidt, Hearing Transcript at p 497. 193 Smidt witness statement, Trial Bundle, Part A, p 285, para 103. 194 Smidt witness statement, Trial Bundle, Part A, p 285, para 103. 47
[179]Mr Smidt also explained that trade customers are able to collect their paint products directly from the manufacturer's factory, from a “point of sale” (“POS”) retail outlet that is in partnership with (or appointed by) the manufacturer, or directly from a retail store if they so wish. Mr Smith testified that a lot of trade purchases takes place through retail outlets, 195 and Mr Smidt indicated that most of AkzoNobel’s sales are made through retail outlets. 196
[180]The merger parties testified that, given the overlaps between retail and trade brands and sales outlets, sales in the trade channel cannot be clearly separated from the retail channel, and vice versa. For the same reasons, prices of products in the retail and trade channels track each other very closely. 197 Specified trade segment
[181]It was common cause that there is a sub-segment of the wider trade sales channel in which manufacturers market their products to architects, painting contractors, and developers with a view to getting their paint products “specified” (i.e., named on an architect’s plan) for large construction projects. 198 This is referred to as the “specified trade” segment.
[182]It was also common cause that the approach to marketing and selling products in the specified trade segment differs significantly from the manner in which products are sold in the retail channel. 199 It typically involves manufacturers establishing dedicated specification and support teams to identify the most appropriate paints from their portfolios for the project in question; to approach architects, quantity surveyors and other specifiers to have their respective products specified for the project; to seek to ensure that the contractor does not switch their specified paint products for other brands (the merger parties stressed that contractors are able to switch paints after specification, particularly where the specification permits the use of an “equivalent” paint); to attend on-site meetings; and thereafter to assist in the application and maintenance process. Manufacturers thus offer a comprehensive turnkey solution in the specified trade 195 Expert Presentation of Mr Smith, Exhibit 14, at slides 25-26; Smith, Hearing Transcript at p 1236-1238, 1312. 196 Witness statement of Mr Smidt, Trial Bundle, Part A, p 277, para 87, and p 284, para 100.2. 197 Witness statement of Mr Smidt, Trial Bundle, Part A, pp 284-289, paras 100-106; Smidt, T562-564; Lalla, Hearing Transcript at pp 155-157. See also Expert Presentation of Mr Smith, Exhibit 14, at slides 23-24; Smith, Hearing Transcript at pp 1298-1306. 198 Witness statement of Mr Smidt, Trial Bundle, Part A, p 285, para 102. 199 Witness statement of Mr Smidt, Trial Bundle, Part A, p 285, para 103. 48 segment, and also provide “full spectrum” or “360 degree” guarantees 200 for all aspects of their work, including the paint itself and its application. 201
[183]As noted above, it is the specified trade component of the trade channel that is the focus of the Commission’s case in this matter. The Commission contends that Plascon and Dulux are the largest and closest competitors in the specified trade channel, with a combined market share of almost 80%. 202 The merger parties, on the other hand, stated that there are numerous national and regional paint manufacturers that compete actively in the specified trade segment. 203
[184]Market shares in the specified trade channel were, again, a matter of considerable contestation between the Commission and the merger parties. In support of his estimate, Mr Njisane relied on various pieces of evidence.
[185]First, Mr Njisane referred to various internal documents of KPAL.
[186]Mr Njisane also referred to monthly surveys by a third-party consultancy called Industry Insight, which reflects Plascon’s “Market Exposure Rate” (“MER”) to specified trade projects. 200 See Exhibit 5. 201 Smidt, Hearing Transcript at pp 639-643; Lalla, Hearing Transcript at pp 152-154, 193-194; Stekhoven, Hearing Transcript at pp 741-742; Prost, Hearing Transcript at pp 1149-1151; 1178-1181. 202 Expert Presentation of Mr Njisane, Exhibit 15, at slides 12-17. 203 Witness statement of Mr Smidt, Trial Bundle, Part A, pp 285-289, paras 104-105; Expert Presentation of Mr Smith, Exhibit 14, at slides 42-53. 204 Trial Bundle, Part C, pp 7096 at 7135; Lalla, Hearing Transcript at pp 317-318. 205 Expert Presentation of Mr Njisane, Exhibit 15, at slide 16. 206 Expert Presentation of Mr Njisane, Exhibit 15, at slide 16. 49
[187]The merger parties argued that little weight should be placed on KPAL’s market share figures because they are based on (i) KPAL’s sales directly to contractors and to POS stores (i.e. to contractors with which KPAL has direct accounts), and on (ii) MER figures for specified trade projects. 209 As regards KPAL’s trade sales figures, Mr Lalla explained that those include both specified and informal (unspecified) trade sales at POS stores (which KPAL is unable to distinguish), and also exclude specified trade sales that might be made through the retail channel. 210 As regards the MER figures, Mr Lalla stated that they do not reflect market shares, but simply reflect the extent of involvement of suppliers and brands in construction projects across South Africa based on specification. It appears that the MER calculation is based on the overall value of the projects in question rather than the volume or value of the products supplied by each supplier for such projects. 211
[188]We do not agree that these figures should be disregarded as the merger parties suggest. Notwithstanding their limitations, they are, on KPAL’s own version, the best figures it has access to, and KPAL itself regards them as sufficiently probative that it relies on them in its internal documents. Furthermore, Mr Lalla explained that KPAL does not rely on these figures blindly; rather, it has regard to both these data sources and to input from its sales teams in order to calculate its specified trade market shares. 212
[189]Mr Lalla’s oral evidence was also consistent with these internal documents. He confirmed that architects prefer to specify Plascon and Dulux, and that these are the two largest manufacturers, in the specified trade segment. 213 He added that Dulux, Plascon and Prominent are the only manufacturers that offer equivalent extensive guarantees for specified trade projects. 214 207 Lalla, Hearing Transcript at pp 261-263; Trial Bundle, Part C, p 7443. 208 Lalla, Hearing Transcript at pp 249-263; Trial Bundle, Part C, pp 6611, 7270. 209 Lalla, Hearing Transcript at pp 155-157, 173-174, 244. 210 Lalla, Hearing Transcript at pp 155-157, 173-174, 244, 254-255. 211 Lalla, Hearing Transcript at pp 173-174, 249-255. See also Trial Bundle, Part C, pp 7268-7270. 212 Lalla, Hearing Transcript at pp 173-174. See also Trial Bundle, Part C, pp 7268-7270. 213 Lalla, Hearing Transcript at pp 255, 257, 263. 214 Lalla, Hearing Transcript at pp 2487-249, 309-310. 50
[190]Mr Smidt likewise confirmed that Dulux’s most significant competitor in the specified trade segment is Plascon. 215 He was also referred in this regard to an AkzoNobel internal document which reflects two tables under the heading “Specifications conversion rate”, one for 2020 and the other for 2021. 216
[191]These tables reflect that: 191.1.
[192]Mr Smith correctly pointed out that the percentage of Dulux conversions in these tenders appears high, given that it suggests that in 2020 Dulux won approximately%, and in 2020 over %, of all tenders that were not still pending, which is materially higher than the figures reflected in the KPAL documents referred to above. As Mr Smith suggested, this might be because the tenders in the AkzoNobel tables were identified by specifiers with links to AkzoNobel, and therefore represent a small and unrepresentative sample of the overall market in each of those years. 217
[193]This notwithstanding, the AkzoNobel tables are in our view nevertheless insightful in that they show that, of the tenders reflected, Plascon was the only competitor identified by 215 Smidt, Hearing Transcript at pp 636-637. 216 Smidt, Hearing Transcript at pp 636-638; Expert Presentation of Mr Njisane, Exhibit 15, slide 30. 217 Smith and Njisane, Hearing Transcript at pp 1314-1316; 1358; 1388-1391. 51 name, and appears to be the competitor to which Dulux lost the most tenders. Indeed, in 2021 Dulux lost more of the identified tenders to Plascon than to all other competitors combined. Mr Smidt confirmed this in his oral evidence. 218
[194]The Commission also referred to a survey of preferred and non-preferred applicators in the specified trade segment that was commissioned by KPAL in 2022. 219
[195]The above evidence indicates not only that Dulux and Plascon are by far the largest (and closest) competitors in the specified trade segment of the market, but that there are significant barriers to entering this market. These include, in particular, the cost and time associated with establishing a dedicated and specialised specification and support team, and the strong preference that specifiers have for the “tried and tested” products of Plascon and Dulux for projects that carry significant risks for specifiers and asset owners if the paint products fail for any reason.
[196]Dulux itself referred to the difficulties associated with entering the specified trade segment of the market. Mr Smidt testified that it was a “long burn” even for AkzoNobel to enter the specified trade segment, because it required the establishment of a dedicated specification team, the building of relationships with specifiers, the driving of Dulux as a brand in that channel, and preparedness to provide “360 degree” guarantees. 220
[197]Mr Prost testified that Promac is active in the specified trade sector, and has a specification team comprised of 5 full-time employees and 15 other personnel who assist on projects. Promac also offers a complete turn-key solution (from specification to assistance with application and maintenance) and “full spectrum” guarantees to developers for all aspects of the project. Mr Prost emphasised that it “takes a lot” to put together a specified trade team because of the specialist skills required. 221 218 Smidt, Hearing Transcript at pp 636-638. 219 Smith, Hearing Transcript at pp 1588-1590; Trial Bundle, Part C, p 6489 at pp 6493-6497. 220 Smidt, Hearing Transcript at p 643. 221 Witness statement of Mr Prost, Exhibit 12, para 4; Prost, Hearing Transcript at pp 1149-1151; 1178-1181. See also Exhibit 5. 52
[198]However, despite all the investment Promac has made in developing its specified trade offering, Mr Prost explained that it has proved “very, very difficult” for Promac to enter the specified trade segment because of the strong preference architects have for Dulux and Plascon based on their established reputations in the market. 222 He elaborated that architects generally specify Plascon and Dulux because those brands have always worked, and they are unwilling to take a chance with new brands, even at cheaper price points, because of the risks associated with paint failures on large projects. Architects therefore “cover themselves” knowing that “the job is going to run smoothly if it is a Plascon [or] a Dulux product that has been applied to the wall”. Promac therefore tries to approach asset owners or managers directly in order to promote its products, but apparently with limited success. 223
[199]Mr Stekhoven testified that Duram has decided not to participate in the specified trade segment at all. He explained that Duram has made this decision because of the expense and time involved in establishing a dedicated specification and support capability; the time required to establish relationships with specifiers; the potential liabilities under the “360 degree” guarantees required in this segment; and the risk of damaging Duram’s relationships with its retailer customers. 224 Mr Stekhoven stressed in this regard the difficulties associated with trying to persuade specifiers to approve brands other than Plascon and Dulux in the specified trade segment, which he referred to as the “trust factor”. 225
[200]Mr Madhu likewise testified that Medal does not participate in the specified trade segment because of the established relationships that Plascon and Dulux have built up with specifiers over many years, and the difficulty of persuading specifiers to specify any other brands. Mr Madhu explained that, in the light of the size of the projects involved, specifiers generally “stick to what they know best”. 226
[201]Mr Mahabeer similarly confirmed that it is “very difficult” to enter the specified trade segment, and that Build-It has failed “dismally” in doing so with its DOBs, because of the preference of specifiers for Plascon and Dulux. 227 222 Prost, Hearing Transcript at pp 1176-1177. 223 Prost, Hearing Transcript at pp 1150-1151, 1176-1177. 224 Stekhoven witness statement, Trial Bundle, Part A, p 407, para 10; Stekhoven, Hearing Transcript at pp 734- 736, 741-742, 812-813. 225 Stekhoven, Hearing Transcript at p 856. 226 Madhu, Hearing Transcript at p 886. 227 Mahabeer, Hearing Transcript at pp 989-990. 53
[202]Other customers and competitors interviewed by the Commission during the course of its investigation also stated that Plascon and Dulux are by far the largest players in the specified trade segment. 228
[203]Mr Smith argued that, because of the significant overlaps between sales of paint products in retail and trade channels, the retail and trade channels directly constrain one another; and therefore, even if Plascon and Dulux are the largest brands in the specified trade segment, their pricing in the specified trade segment is constrained by competition in the retail channel. 229
[204]In our view, there are three main difficulties with this contention. The first is that competition in the specified trade segment is not limited to price. As traversed above, non-price factors such as quality and service levels appear to play a major role in this segment of the market. Therefore, if the merger were to proceed, all these non-price aspects of competition between Plascon and Dulux, as the two largest and closest competitors in the specified trade segment, would be lost.
[205]Second, even insofar as price is concerned, the evidence was that decorative coatings products are generally sold at a slightly cheaper price in the trade channel than in the retail channel because branding plays less of a role. 230 In the circumstances, it appears that there would be scope for Plascon and Dulux to increase prices in the trade channel even within the constraints imposed by the retail channel.
[206]Third, we have found above that competition in at least the premium segment of the retail channel would be lessened as a result of the proposed merger. Therefore, insofar as pricing in the specified trade segment is constrained by pricing in the premium segment of the retail channel (as noted above, Mr Smidt testified that there is a preference for high-quality paints in the specified trade segment), 231 the latter constraint would be softened post-merger.
[207]In all the circumstances, therefore, we agree with the Commission that the proposed merger would also likely have a substantial anti-competitive effect in the specified trade segment of the decorative coatings market. 228 Expert Presentation of Mr Njisane, Exhibit 15, slides 32-33. 229 Expert Presentation of Mr Smith, Exhibit 14, at slides 23-26; Smith, Hearing Transcript at pp 1298-1306. 230 Stekhoven, Hearing Transcript at pp 812-813; Prost, Hearing Transcript at p 1153. 231 Smidt, Hearing Transcript at p 497. 54 COLOURANTS
[208]In addition to its unilateral theory of harm, the Commission contended that the proposed merger would likely give rise to input foreclosure in the supply of colourants to decorative paint manufacturers in South Africa.
[209]This concern arises from the fact that KPAL owns and operates ICC Kansai Colourants (Pty) Ltd (“ICC”), a firm that manufactures colourants that are used in the manufacture of ready-mix paints and for in-store tinting. The Commission contended that the merged entity would have both the ability and incentive to foreclose its downstream competitors from access to colourants, and that this would have an anti-competitive effect in the downstream decorative coatings market. Mr Njisane clarified in his oral testimony that the Commission’s concern related specifically to the supply of colourants for in-store tinting and not for the in-plant manufacture of ready-mixed products. 232
[210]The merger parties disputed that ICC would have any ability or incentive to foreclose downstream competitors, and tendered to continue supplying all ICC’s existing customers with colourants for a five-year period post-merger on the same commercial terms (subject to inflationary increases in the costs of production).
[211]Mr Lalla testified that, in addition to ICC, there are various local colourants suppliers such as Clariant, Chromaflo and Colour Supreme, as well as large, international suppliers, that supply both in-plant and in-store colourants. He stated further that, while it would take approximately 6 months (Mr Smidt estimated up to one year) 233 for a decorative coatings manufacturer to change colourants suppliers, this is not a significant obstacle, and that many paint manufacturers have done so in the past. 234
[212]As regards incentives, Mr Lalla stated that ICC currently has significant spare production capacity (for both in-plant and in-store colourants), and would retain sufficient capacity to continue servicing all its downstream customers even if it also supplied AkzoNobel with colourants post-merger. ICC would therefore have every incentive, given its high fixed cost base, to continue to service downstream customers post-merger and not to foreclose them. 235 232 Expert Presentation of Mr Njisane, Exhibit 15, slides 55-58; Njisane, Hearing Transcript at pp 1277-1278. 233 Smidt, Hearing Transcript at p 659. 234 Witness statement of Mr Lalla, Trial Bundle, Part A, pp 362-363, paras 13-15; Lalla, Hearing Transcript at pp 110, 115-116, 201, 313. 235 Witness statement of Mr Lalla, Trial Bundle, Part A, pp 363-364, paras 17-20; Lalla, Hearing Transcript at pp 116-122. 55
[213]Mr Smith also testified that the merged entity’s incentives to foreclose would be weak, 236
[214]Mr Stekhoven was the only witness who expressed concerns regarding the supply of colourants from ICC. He explained that Duram sources its in-plant colourants (where price is more critical) from a number of suppliers, including ICC, but that ICC is Duram’s only supplier of in-store colourants because it regards the quality of ICC’s in-store colourants, and the service provided by ICC, as superior to that of other suppliers. 237
[215]When asked to comment on AkzoNobel’s proposed 5-year remedy, Mr Stekhoven responded that it would give Duram sufficient time to change suppliers or commence manufacturing its own colourants (he indicated that a period of approximately three years would be required). 238
[216]Having regard to the above, we do not believe there is sufficient evidence before us to substantiate the Commission’s input foreclosure concern. Furthermore, based on the evidence of Mr Stekhoven, it appears that the 5-year supply condition tendered by AkzoNobel would potentially be sufficient to address any concerns in this regard (subject to refinement of its terms).
[217]We therefore do not rest any weight for our decision in this matter on the Commission’s input foreclosure concern. EFFICIENCIES
[218]The next question is whether the unilateral effects that we have identified in relation to the proposed merger are outweighed by any pro-competitive benefits of the merger.
[219]We can be brief on this score. Whilst, as outlined in the “Rationale” section above, Mr Smidt testified that the merger would give rise to various manufacturing and distribution efficiencies, the merger parties did not advance an efficiency defence in this matter, 239 and we believe correctly so. As Mr Njisane pointed out, the efficiencies referred to were not quantified; questions were raised about the extent to which they would be passed 236 Expert Presentation of Mr Smith, Exhibit 14, slides 79-80; Smith, Hearing Transcript at pp 1250-12521, 1491- 1506. 237 Stekhoven, Hearing Transcript at pp 755-757. 238 Stekhoven, Hearing Transcript at pp 780, 859-860, 864-865. 239 Smith, Hearing Transcript at pp 1521. 56 on to consumers; and, more fundamentally, they did not appear sufficiently substantial in magnitude to outweigh the unilateral effects we have identified in respect of the proposed merger. 240
[220]We are therefore of the view that the anti-competitive effects of the proposed merger are not outweighed by any efficiency gains. PUBLIC INTEREST
[221]It is also necessary to consider whether the proposed merger can or cannot be justified on substantial public interest grounds.
[222]The merger parties submitted that the proposed merger would represent a significant investment in South Africa and the other countries in which KPAL and KPEA are active, but they did not contend that this constituted a public interest basis upon which the merger should be approved notwithstanding its anti-competitive effects. The merger parties also tendered various public interest commitments if the merger were approved, which are considered below under the heading of “Conditions”.
[223]The Commission, for its part, contended that the merger would have a negative employment effect because, in terms of the employment commitments made by the merger parties, the merged entity would be entitled to retrench up to 78 employees in middle- and upper-management positions within a period of three years from the closing date, with no commitments after that date. The representatives of CEPPWAWU, GIWUSA and the non-unionised employees of KPAL also expressed concerns about the employment effects of the proposed merger. 241
[224]Given the decision we have made to prohibit the merger on competition grounds, and that the merger parties did not contend that the public interest issues they advanced outweighed the anti-competitive effects of the merger, it is not necessary for us to assess these concerns in any detail. We note, however, that it was not evident from the evidence before us that the merger parties had conducted an adequate assessment of the potential retrenchments that would arise from the merger, or that they had 240 Witness statement of Mr Smidt, Trial Bundle, Part A, pp 252-253, paras 8-10; Smidt, Hearing Transcript at pp 670-671, 690-695; Expert Presentation of Mr Njisane, Exhibit 15, slides 48-49; Njisane, Hearing Transcript at pp 1527-1535. 241 Dlamini, Hearing Transcript at pp 94-104; Varaden, Hearing Transcript at pp 464-473; Appolis, Hearing Transcript at pp 474-487, 1451-1462; Exhibit 23. 57 satisfactorily justified these potential retrenchments. 242 This is a question that would have required further consideration had we not decided to prohibit the merger on competition grounds. PROPOSED CONDITIONS
[225]Finally, it is necessary to consider whether the anti-competitive effects of the proposed merger are adequately addressed by the conditions proposed by the merger parties.
[226]As regards the approach to be followed in circumstances where conditions are tendered to address the negative competition effects of a merger, the CAC stated the following in Imerys: “Where, in the situation just mentioned, 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. I reject the proposition that the Commission bears the burden of proving that the proposed conditions will not adequately address the likely SLC. The Tribunal has the power to prohibit the merger if it is not satisfied that the conditions will adequately remedy the likely SLC. And 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 SLC. 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 SLC. Although the proposed conditions are more likely than not to remedy the likely SLC, there is a reasonable possibility that they will fail to do so. Therefore we prohibit the merger.’ . . . I do not say that the Tribunal would be obliged to reject conditional approval just because there was a reasonable possibility (falling short of a preponderance of probability) that the conditions would fail to remedy the likely SLC. The Tribunal might properly exercise its discretion in such a case to give conditional approval. In exercising its discretion, the Tribunal could be expected to take into account, on the one hand, the precise likelihood and extent of the SLC; and, on the other, the precise extent of the risk that the conditions will fail to remedy the likely SLC. The public interest 242 Metropolitan Holdings Ltd v Momentum Group Ltd (41/LM/Jul10) [2010] ZACT 87 (9 December 2010) at paras 68-77. 58 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 SLC.” 243
[227]In this case, the conditions proposed by the merger parties in their request for consideration included the 5-year colourants supply agreement and the employment condition referred to above, as well as various public interest commitments. These included: 227.1. an employee share ownership plan (“ESOP”) that would hold a 5% shareholding in the South African operations of the merged entity for qualifying employees; 227.2. new capital investments of at least R170 million in productive capacity and upgrading production facilities; 227.3. various initiatives aimed at developing HDP-owned suppliers and increasing levels of local procurement; and 227.4. various other community investments. 244
[228]However, the merger parties did not propose any condition to address the Commission’s unilateral effects concern until 31 August 2023 (days before the commencement of the Tribunal hearing), when they tendered the divestment of KPAL’s Micatex brand, together with enhancements to various of their other commitments. 245 The merger parties tendered further revisions to their proposed conditions on 14 September 2023 (shortly before the completion of the evidential hearing), which addressed the employment concerns raised by Mr Varaden on behalf of the non-unionised employees of KPAL to his satisfaction, but not those raised by GIWUSA and CEPPWAWU.
[229]Because the merger parties did not tender the divestiture of the Micatex brand until the eve of the Tribunal hearing, the Commission did not have the opportunity to investigate that remedy in any detail before the hearing commenced, and the evidence on the remedy was accordingly limited to the views of witnesses at the hearing (including Mr 243 Imerys, supra, paras 40-42. 244 Request for Consideration, Trial Bundle, Part A, pp 233-249. 245 Exhibit 21. 59 Prost, who was called by the merger parties to testify regarding Promac’s willingness to acquire the Micatex brand).
[230]The Commission therefore requested, and was afforded an opportunity, to canvass the proposed Micatex remedy more fully with market participants after the completion of the evidential hearing on 15 September 2023, and the merger parties were afforded the opportunity to adduce further evidence on remedies after receipt of the Commission’s submission if they so wished.
[231]The Commission submitted its supplementary report on the proposed Micatex remedy on 16 October 2023, and the merger parties subsequently indicated that they would not adduce any further evidence on the remedy. The matter therefore proceeded to argument on that basis.
[232]This chronology of events illustrates the difficulties that arise when merger parties wait until the last moment to tender substantive remedies. This precludes the other parties in the proceedings from being able to fully assess such remedies in advance of the hearing, and typically results in delays in the completion of the merger hearing, or in the submission of evidence that is not subject to proper interrogation before the Tribunal (as in this case). It is therefore in the interests of merger parties who wish to tender substantive remedies in merger proceedings to submit those to the Tribunal and to the other interested parties in the proceedings with sufficient notice that they can be properly considered in advance of the hearing, and fully interrogated during it.
[233]As regards the merits of the Micatex remedy, the merger parties proposed the divestiture of the “Micatex” brand in South Africa (and certain other countries in Southern Africa) within a after the closing date of the merger, together with a transitional services arrangement for a period of up to 24 months. The proposed condition also stipulated that the purchaser would be required to have the resources, expertise, and ability and incentive, to maintain and develop the “Micatex” brand as a competitor to the merger parties in the decorative coatings market.
[234]It is well-established that divestiture remedies are generally a preferred way in which to address anti-competitive effects arising from mergers, as they seek to address at source the loss of rivalry resulting from the merger. However, this presupposes that the divestment assets, if operated by a suitable purchaser, would restore, on a lasting basis, the competition lost by the proposed merger. For this reason, competition authorities have a clear preference for the divestiture of existing stand-alone businesses, including all the assets and personnel that are necessary to ensure the viability and competitiveness of those businesses post-merger. A divestiture of only certain limited 60 assets (such as a brand) that do not form an existing stand-alone business creates greater risks regarding the viability and competitiveness of the resulting business. Therefore, a limited divestiture package of this sort will generally be accepted only if it can be shown that the resulting business will, in the hands of a suitable purchaser, quickly be able to restore the competition lost by the merger. This may, in certain circumstances, require the approval of a suitable purchaser prior to the approval or implementation of the merger. 246
[235]In this case, therefore, the central question we need to consider is whether there is sufficient evidence before us that the acquisition of the Micatex brand would enable the purchaser to restore, in a short space of time, the competitive constraint that Plascon and Dulux currently impose on each other in the premium and specified trade segments of the decorative coatings market, or at least to do so to a sufficient extent that the merger would no longer result in a significant lessening of competition in those segments.
[236]Mr Lalla testified that Micatex (a premium textured exterior paint) is one of the best- known paint brands in South Africa with a long history, He stated further that ownership of the Micatex brand would give the purchaser immediate distribution in retail and trade outlets across South Africa where Micatex is currently listed, which the purchaser could use to extend the range of products sold under the Micatex brand in due course. 247 Mr Smidt testified to the same effect, and stated that both Promac and Prominent has expressed an interest in purchasing the Micatex brand. 248
[237]Mr Prost confirmed Promac’s interest in the Micatex brand, and explained that, while Promac already has a competing exterior paint called Micaguard, it is not as successful and well-known to consumers as Micatex given the latter’s long brand history. Mr Prost testified further that Promac had sufficient spare manufacturing capacity to take over production of Micatex at its current volumes, and that Micatex would assist Promac to expand its (already extensive) distribution footprint. 249 246 See Commission notice on remedies acceptable under Council Regulation (EC) No 139/2004 and under Commission Regulation (EC) No 802/2004 (2008/C 267/01), paras 22-34, 37-38 and 47-51; CMA Merger Remedies (CMA87) 13 December 2018, paras 5.12-5.14 and 5.28-5.31. 247 Lalla, Hearing Transcript at pp 157-159, 162, 323-324. 248 Smidt, Hearing Transcript at pp 569-570. 249 Witness statement of Mr Prost, Exhibit 12, paras 2-4, 8-13 and 16-17; Prost, Hearing Transcript at pp 1159- 1170. 61
[238]However, the Commission raised two main objections to the proposed Micatex remedy.
[239]The Commission’s first objection was that the Micatex brand would not be as powerful as it currently is without the Plascon “master brand” attached to it. Both Mr Lalla and Mr Smidt acknowledged that, in the premium segment of the market, both KPAL and AkzoNobel follow a “master brand” strategy which highlights the Plascon and Dulux “master brands” on all of their premium products. 250 However, Mr Lalla added that KPAL still supports and leverages its so-called “hero” sub-brands, which include Micatex, and that Micatex retains considerable equity and heritage from the sub-brand advertising that it has historically enjoyed. 251 Mr Prost expressed confidence that Micatex is a strong enough brand to stand on its own, 252 and Mr Smidt agreed that Micatex is an “exceptionally strong subrange store brand on its own”. 253
[240]Various other witnesses were, however, less sanguine. Mr Stekhoven acknowledged that Micatex is the strongest individual paint brand in South Africa, but stated that its strength is inextricably linked with Plascon, in the same way that the “Hilux” sub-brand is inextricably linked with Toyota. As a result, consumers would not necessarily be confident that the quality of the paint was the same if it was owned by another manufacturer. 254 Mr Stekhoven added that retailer margins on Micatex are low, and retailers might therefore take the opportunity to delist Micatex in favour of products that carry higher margins if Micatex did not remain part of the KPAL basket of products. 255
[241]Mr Mahabeer also expressed doubts about the strength of the Micatex brand in the hands of another manufacturer, stating that “It is the Plascon label that carried that brand”. 256 Mr Madhu agreed that retailers and consumers might not continue buying Micatex if it was owned by another manufacturer because: “Micatex is backed by Plascon, without the big red label on the bucket, no one is going to believe that is Micatex. If I go and take Micatex and call it my Medal Micatex, I am 250 Lalla, Hearing Transcript at pp 159-161, 276-277. See Exhibit 2, Trial Bundle, Part B, p 2313, Trial Bundle, Part C, p 7128. 251 Lalla, Hearing Transcript at pp 161, 276-279; Smidt, Hearing Transcript at p 500. See also Trial Bundle, Part B, p 2313. 252 Prost, Hearing Transcript at p 1193. 253 Smidt, Hearing Transcript at pp 664-665. 254 Stekhoven, Hearing Transcript at pp 781-782. 255 Stekhoven, Hearing Transcript at pp 862-863. 256 Mahabeer, Hearing Transcript at pp 1009-1010; Hearing Transcript at pp 1076-1077. 62 not going to have the same leverage as Plascon. It still has to say Plascon Micatex for it to have, for me to get, for it to still remain in the marketplace where it is.” 257
[242]The Commission’s second objection was that it would be difficult for the purchaser to extend the Micatex brand (which is specifically associated with a textured, exterior paint) to replicate the wide existing range of Plascon products in the market.
[243]Mr Lalla and Mr Smidt both stated that it would be possible for the purchaser to extend the Micatex brand to other paints in due course, especially if the purchaser already had other products in its portfolio that could be rebranded. 258
[244]Mr Prost candidly acknowledged that, because the possibility of acquiring the Micatex brand had only recently been raised with Promac, it had not yet been able to give much thought to its future plans for the brand, but he expressed optimism that Promac could potentially expand its current product range under the Micatex brand in the future: “it could go anywhere”. 259
[245]Again, however, Mr Stekhoven was less sure, stating that, while it was “theoretically possible” for the Micatex brand to be extended to a broader range of products, there would be “lots of obstacles” given the association of Micatex with textured exterior paint, and the need to persuade retailers to carry products under the Micatex brand even though they no longer formed part of the KPAL basket. 260 Mr Madhu also expressed doubts in this regard. 261
[246]As noted above, the Commission was afforded an opportunity to test the Micatex remedy with market participants in more detail after the evidential hearing before the Tribunal was completed. In their supplementary report dated 16 October 2023, the Commission recorded the views of a number of market participants regarding the Micatex remedy, which overwhelmingly supported the Commission’s objections referred to above. Concerns were also raised regarding the difficulties that a new purchaser would have in maintaining the Micatex brand in the market with the same level of support, resources and reputation that it currently enjoys from KPAL. The Commission summarised the findings of its investigation as follows: 257 Madhu, Hearing Transcript at p 902. See also Madhu, Hearing Transcript at p 967. 258 Lalla, Hearing Transcript at pp 158, 162, 323-324; Smidt, Hearing Transcript at p 570. 259 Witness statement of Mr Prost, Exhibit 12, para 15; Prost, Hearing Transcript at pp 1185-1187. 260 Stekhoven, Hearing Transcript at p 862. 261 Madhu, Hearing Transcript at pp 902, 967. 63 “The proposed divestiture package is unlikely to be sufficient to address the permanent structural change which will occur through this merger. The proposed divestiture of the Micatex sub-brand will not enable a potential purchaser to be sufficiently competitive against the merged firm and rivals in the market, let alone replicate the competitive force exerted by Plascon pre-merger. In particular, the divestiture is a single sub-brand in a single product category, whereas the merger eliminates competition between the two strongest Masterbrands across the entire portfolio of paint products. Even within the single product line that Micatex competes, market participants indicate that the potential buyer of the Micatex is unlikely to maintain the strength of the Micatex in similar fashion as Plascon, and the expansion of the brand to a Masterbrand is equally unlikely. They indicate that it may be necessary that the merger parties are restrained from competing with the Micatex brand. This highlights the inadequacy of the proposed remedy and its inability to facilitate the restoration of competition that will be lost as a result of the proposed merger. Similarly, the proposed remedy is not capable of practical implementation to resolve the competition concerns given the identified difficulties of maintaining and expanding the Micatex brand to a Masterbrand. Furthermore, market participants expect that retailer customers are more likely to delist the brand given the presence of alternative brands they could turn to or who are already listed for their textured exterior offerings, as well as the low gross margins realised on the product. This is consistent with our understanding that firms compete on the basis of the brand equity derived from their Masterbrands. Disassociation with the Masterbrand thus renders these sub-brands as less strong brands on their own and ultimately implies that the proposed Micatex brand divestiture is not capable of practical implementation in a manner that would restore the lost competition as a result of the proposed merger. Lastly, we understand that developing brand equity is not timely given the period it takes to establish brands in the market. The merger parties and competitors state that brand equity is developed over long periods of time, thus indicating that the proposed remedy, even if it were capable of practical implementation, is unlikely to be timely. In addition, market participants indicated that the proposed transitional services would have to be in place for a relatively substantial period of time in order to enable the new buyer to establish its brand equity. In terms of the clarity of the proposed remedy, 64 market participants indicated that there remain uncertainties regarding the product quality dimensions in respect of the proposed remedy.” 262
[247]In final argument, the merger parties questioned the impartiality and accuracy of the investigation conducted by the Commission, but elected not to adduce further evidence to rebut the contents of the Commission’s supplementary report.
[248]As indicated above, it is unfortunate that, given the late stage at which the merger parties proposed the Micatex remedy, it was not fully interrogated during the Tribunal hearing. However, based on the evidence before us, and having regard to the approach to remedies laid down by the CAC in Imerys, we are not satisfied that the Micatex remedy would adequately remedy the unilateral effects we have identified in relation to the proposed merger. As set out above, the divestiture of brands (as opposed to existing stand-alone businesses) carries heightened risks that the purchaser would not be able to restore the rivalry lost as a result of the merger. Persuasive evidence is therefore required that the divestiture of the Micatex brand to Promac (or any other purchaser) would enable it, in a short time, to restore the competitive constraint that Plascon and Dulux currently impose on each other in the premium and specified trade segments of the market, or at least to do so to a sufficient extent that the merger would no longer result in a significant lessening of competition.
[249]We do not believe that there is persuasive evidence of this sort before us. The evidence referred to above suggests that, whilst Micatex is clearly a very powerful brand, and a very successful product in the market, there is a material risk that, absent the support of the Plascon mother brand, and the financial and technical support it currently receives from KPAL, it would lose at least some of its current brand strength in the eyes of both end-consumers and customers.
[250]Furthermore, and more importantly, there is, in our view, a material risk that the purchaser of the Micatex brand would not, within a short time after the merger, be able to extend the Micatex brand to encompass the wide range of products that KPAL currently sells in the premium and specified trade segments of the market in competition with Dulux. As Mr Prost candidly admitted, Promac has not even given serious thought to this yet, let alone committed to doing so.
[251]Such an initiative would furthermore be a very ambitious, expensive and risky one, dependent as it would be on the development, and acceptance by customers and end- 262 Commission Confidential Supplementary Expert Report, 16 October 2023, paras 98-101. 65 consumers, of a wide range of new (or at least newly branded) products in the premium and specified trade segments, and in the face of the now combined strength of Dulux and Plascon in those segments. As discussed above, the brand equity of Plascon and Dulux in those segments is already a major barrier to entry, and that would be further enhanced if they came together under AkzoNobel. Therefore, even with the benefit of a very well-established sub-brand such as Micatex, the purchaser would face formidable challenges in entering the premium and specified trade segment of the market, let alone in approximating the entrenched position that Plascon currently enjoys in those segments. Even if the purchaser did decide to go down that path, there is, in our view, a significant risk that it would not succeed at all, let alone in a time period that would protect customers and end-consumers from the anti-competitive effects of the merger.
[252]For all these reasons, we conclude that the Micatex remedy does not sufficiently address the anti-competitive unilateral effects of the proposed merger.
[253]Given this conclusion, it is not necessary for us to consider the additional public interest remedies that have been tendered by the merger parties. Positive as those commitments are from a public interest perspective, the merger parties rightly do not suggest that the merger should be approved on the basis of these remedies notwithstanding the anti-competitive effects to which we have found it would give rise. CONCLUSION
[254]For all the reasons set out above, we find that the proposed merger is likely to give rise to substantial unilateral effects in the premium and specified trade segments of the South African decorative coatings market.
[255]We also find that this anti-competitive effect is not outweighed by any pro-competitive or public interest benefits associated with the proposed merger, and that it is not sufficiently addressed by the Micatex remedy that has been proposed by the merger parties.
[256]We therefore prohibit the proposed merger. 18 June 2024 Adv Jerome Wilson SC Date Ms Mondo Mazwai and Prof Liberty Mncube concurring. 66 Tribunal case managers: Juliana Munyembate and Leila Raffee For the merger parties: Adv Michael van der Nest SC and Adv Gavin Marriott instructed by Webber Wentzel. For the Commission: Adv Lebogang Phaladi, Adv Layne Quilliam, Adv Mfundo Ngobese and Adv Mohau Nakeng instructed by Kgaugelo Baloyi Inc.
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.