Competition Commission of South Africa v NPC-Cimpor (Pty) Limited and Others (178/CAC/DEC19) [2020] ZACAC 8 (7 August 2020)

THE COMPETITION APPEAL COURT OF SOUTH AFRICA

HELD IN CAPE TOWN

CAC CASE NO: 178/CAC/DEC19

In the matter between

THE COMPETITION COMMISSION OF SOUTH AFRICA Appellant

and

NPC-CIMPOR (PROPRIETARY) LIMITED First

Respondent

AFRISAM (SOUTH AFRICA) Second

Respondent

PROPRIETARY) LIMITED

LAFARGE SOUTH AFRICA

(PROPRIETARY) LIMITED Third

Respondent

PPC LIMITED Fourth

Respondent

JUDGMENT

Date: 07 August 2020

DAVIS JP

[1] This appeal turns on the question of whether first respondent was party to an agreement or alternatively a concerted practice as prohibited in terms of s 4(1) (b) of the Competition Act 89 of 1998 (the Act ) and which continued until 2009. The appellant (the Commission) contends that this agreement or concerted practice commenced prior to the introduction of the Act on 1 September and continued at least until 2009. It involved the four respondents in indirectly fixing cement prices in contravention of s 4 (1) (b) (i) of the Act and dividing the cement market through, inter alia, the allocation of market shares in contravention of s 4(1) (b) (ii) of the Act.

[2] The appeal concerns a number of important legal issues relating to the scope and meaning of s 4(1) (b) of the Act, including whether the first respondent’s conduct constituted a prohibited practice and even if it did whether it continued after 2 June 2005, being the initiation date of the complaint. The complaint upon which the litigation has been predicated was initiated in June 2008 by the Commission for the purposes of investigating allege cartel behaviour in the cement and cement extender industry. Upon the commencement of the investigation second, third and fourth respondents have all confessed to having participated in cartel conduct in breach of s 4(1) (b) of the Act. Leniency was granted to fourth respondent in late 2009 and then consent agreements were concluded between the Commission and second and third respondents in 2011 and 2012 respectively. Seven years after the initiation of the complaint and three years after the conclusion of the final consent agreement, that is with third respondent in March 2012 the Commission referred a complaint to the Tribunal in February 2015 against the first respondent. The Competition Tribunal (the Tribunal) dismissed this referral in December 2019, finding that the Commission had failed to prove an ongoing agreement or concerted practice in the light of its failure to show that the initial agreement upon which its entire case

had been predicated had been concluded in breach of s 4(1) (b) of the Act. It is against this finding that the Commission has approached

this court on appeal.

The factual matrix

[3] According to the founding affidavit deposed to by Katlego Monareng on behalf of the Commission. For many years dating back to the 1940s, cement producers in South Africa had been granted an exemption in terms of the then legislation to conduct the manufacturing and distribution of cement by way of a lawful cartel. The main features of this cartel were an agreement on market shares largely based on each cement producer’s original production capacity, the division of South Africa into two major regions, the northern region and the southern region, a centralised sales and distribution system and a unitary price model known as a Twycross pricing model in terms of which the Lafarge factory in Lichtenburg was used to determine a base price.

[4] The Competition Board which is the predecessor to the Commission withdrew this exemption in 1995. This required the main producers of cement to find a fresh solution.

[5] On 15 December 1995 at a meeting of the South African Cement Producers Association (SACPA) it was agreed amongst the cement producers to set market share limits as follows:

Alpha 35 – 36%

Lafarge 22-23%

PPC 42-43 %

According to the note of the meeting, the following was also agreed:

‘The above market share included one-third of NPC’s sales to each producer. Lafarge received a larger market share envelope than its 21, 64% average for the period 1985 to 1995 due to its acquisition of African Portland Cement (Otjiwarongo, Namibia) and Petermix (Pietersburg and Nelspruit). Alpha had acquired third-party blenders Inter-Africa (Swaziland) and Potgietersrust Cement Distributors. The above resulted in PPC agreeing to a lower market share envelope than its 43,00% average during the period 1985 to 1995. Accurate market shares as agreed to by the companies at that point in time were therefore: -

Alpha 35,33%

Lafarge 22,33%

PPC 42,33%’

[6] Mr. Collin Jones the executive of fourth respondent was responsible for sales and marketing testified that ‘the information exchange was a critical part of the arrangement. A key thing that they had to put in place was a replacement for the data collection process, which prior to that had been done under the auspices of SACPA and the Cement Distributors (South Africa)(Pty) Ltd (CDSA) and which allowed the participants to monitor market shares and it had to be a process that would withstand scrutiny. Deloitte was approached ultimately and they were asked to become custodians of the data, so to speak, and so instead of each of the parties submitting the data to SACPA, a lot of discussion went into the format of the data that would be submitted to Deloitte and what consolidated data Deloitte would circulate to the members.’ (my emphasis)

[7] By 1996 a price war had broken out, particularly as a result of fourth respondent aggressively competing with the other producers of cement. A series of preliminary meetings then took place in 1998 in order to restore stability to the market so that the global ‘cement rules of the game’ could be concluded in order to ensure that the market returned to stability. In August of 1998 representatives of first second, third and fourth respondent attended a two day meeting near Port Shepstone in KwaZulu-Natal (the Port Shepstone meeting). The aim of this meeting was to discuss the return to stability in the market. Present at the meeting were Mr Strauss, representing first respondent, Mr Fourie the third respondent, Mr Pienaar the second respondent and Mr Jones the fourth respondent. According to the Commission, the meeting culminated in what was referred to in the proceedings as the Port Shepstone agreement and which included the following:

‘1. market share allocation in line with the market share allocation under the lawful cartel for the SACU market, being all of South Africa, Lesotho, Botswana, Namibia and Swaziland, although also referred to by the cement producers as the “national” market.

2. the target market shares of the cement producers under the 1995 agreement were refined.

3. market share allocation for each producer per province, with provincial market shares added up to the previous “national” cartel market shares;

4. the pricing parameters for different types of cement;

5. the scaling back of marketing and distribution activities, coupled with agreed closure of certain depots in certain regions; and

6. cement producers not offering special discounts on higher quality cement.’

[8] Until October 2002 first respondent was jointly owned by second, third and fourth respondents equal shares. The chairperson of first respondent was appointed by the second respondent. In October 2002, Cimpor acquired the shares in NPC from each of second, third and fourth respondents. Upon the acquisition of the shares in first respondent, Cimpor appointed its own directors to the board of first respondent. All of the directors who had been appointed by second, third and fourth respondents resigned. Strauss however remained as the managing director of first respondent.

[9] The Commission contends that, notwithstanding the takeover and the change in the composition of the board, that first respondent continued to abide by the cartel arrangements which had been agreed to at the Port Shepstone meeting, submitted sales to Deloitte and continued to target its allocated market shares.

[10] A meeting on 1 April 2003 was held in Paris between representatives of Cimpor which included Mr Strauss of first respondent and representatives of Lafarge including Mr Fourie. It appeared that Lafarge International had a shareholding in Cimpor and for this reason, representatives of third respondent attended the meeting. The minutes of the meeting reflect the following:

‘Lafarge indicated that they have surplus clinker in South Africa. Their estimates indicate that NPC will need more clinker in 2004 and they could be in a position to supply. This will be a relatively short term arrangement, but could postpone the need for NPC to invest in more clinker capacity. Cimpor indicated that more clinker milling capacity would also be required and asked if the Richards Bay grinding plant would be for sale. Lafarge indicated that they are not considering the disposal of Richards Bay depot.’

[11] In August 2005 an agreement was reached between first and third respondents in terms of which the former would import clinker from third respondent to grind on behalf of first respondent. Cement was also sold to first respondent using third respondent’s capacity at Richards Bay. A minute of the meeting on 24 August 2005 reflects that the arrangement was to have no impact on selling prices or market share.

[12] On the basis of these facts, the Commission contends that the Port Shepstone discussions had taken place in the shadow of a series of collusive discussions which preceded it and the need, which the various cement producers considered to be vital, for the establishment of a new cartel arrangement following the actions of the Competition Board which had brought the lawful cartel to an end.

[13] The Commission contends an agreement was reached at Port Shepstone, which consensus constituted an agreement to establish a cartel and at which meeting Mr Strauss of first respondent was an active participant; that is Mr Strauss as managing director of first respondent actively brought first respondent into the cartel arrangement in which the other three respondents were also

participants. Central to the agreement were the allocated market targets to second, third and fourth respondents and further that the first respondent would target a market share between 10% and 12% of the total market in the SACU region.

[14] The Commission also contended that the Port Shepstone agreement included the allocation of territories between the four respondents. In particular, the Western Cape was allocated to the fourth respondent together with the territory known as the Eastern Cape, being half of the province of the Eastern Cape whereas the other half was allocated to second respondent together with Swaziland and Lesotho. Botswana was allocated to the fourth respondent in exchange for it withdrawing from the market in Northern KZN to the benefit of third respondent. In particular, the Commission emphasised that there was an agreement that second, third and fourth respondents were all to withdraw from the territory that the cement producers had defined as South KZN which was then allocated to first respondent which generated the majority of its production in sales of cement in this area.

[15] The Commission then turned to the exchange of information between the parties. It contended that to ensure that this agreement could be monitored, the four parties had adopted an information exchange framework in terms of which they ‘inter alia, provided detailed sales volume information to an industry association known as the Concrete and Cement Institute of South Africa (C&CI), in terms thereof which set of received a highly desegregated total figures which allowed each of them to monitor changes to their market shares on a monthly (and later weekly) basis.

[16] In this connection, the Commission placed considerable emphasis on the evidence of Mr Jones who, as I have said, referred to the information exchange as ‘a critical part of the arrangement’

[17] In his witness statement Mr Jones amplified on this point as follows:

‘It is important to note that the mechanism in place by the early 2000s were sufficient to maintain the agreement without the need for representatives from cement producers to meet in person. The mechanisms that I am referring to here are the information exchange framework as well as the mechanism of the C&CI to maintain the framework. Without the information exchanges through the C&CI, the monitoring of market shares would have largely been guesswork.’

[18] The Commission contended that all of the cement producers, including first respondent, continued to submit their respective data to the C&CI until 2009 when the Commission prevented the continuation of this information exchange.

[19] In summary: it was the Commission’s case that, following the Port Shepstone agreement, the first respondent operated almost entirely in Southern KZN until at least 2009. Together with its competitors, it continued to submit its sales data with C&CI until 2009 whereby it retained the means to monitor its target market shares. Its market share remained constantly within the target bracket of 10 – 12 %, even after it was taken over by Cimpor. Furthermore, first respondent delayed expansion of its production capacity by some two years relying rather on cooperation from third respondent which, in turn, had sacrificed its own capacity which it might have been able to employ to take market share away from first respondent.

[20] There was also some suggestion, based upon a memorandum from second respondent (referred to as exhibit 4 in the proceedings), that Cimpor would not have purchased the total shareholding in first respondent for the price ultimately paid, unless there had been some assurance given that the other cement producers would not encroach upon first respondent’s territory after the sale. Thus, it was for this reason that Cimpor had ultimately paid a premium for its acquisition of the shares in first respondent. This argument was not foreshadowed in the case presented by the Commission as set out in the founding affidavit nor in the replying affidavit where the following does appear.

‘The Commission is not aware of the reasons behind the sale of NPC to Cimpor and puts NPC to the proof thereof.’

The Tribunal’s decision

[21] The essence of the Tribunal’s finding that first respondent was not a party to any agreement reached at Port Shepstone is to be located in the considerable emphasis it placed on the fact that, until 2002,first respondent was a wholly owned subsidiary of the three respondents Accordingly it found:

‘There was no meaningful competition between NPC and its shareholders, and they could not be considered to be in a horizontal relationship because NPC was selling and making profits for its shareholder. NPC was not independent of the shareholders who decided what NPC would buy from them and the price of those purchases. Management was unhappy about this fact but could not do anything about it.’

[22] It then held that, once Cimpor had acquired the entire shareholding of first respondent, it ‘was independent of its founding shareholders and stopped providing its former shareholders with its company, financial and sales information. It submitted the latter only to the C&CI, confidentially through Deloitte. It saw only the published C&CI statistics which it used independently to plan its production, sales and marketing. It also prepare a flash report each month which provided the directors and management with an overview of the expected monthly results. Cimpor focussed on national and not regional markets and staff were incentivised to grow national market shares. Post the Cimpor takeover, NPC was susceptible to a possible contravention of the Act in its own right. However, there is no evidence that the newly constituted NPC implemented the prior market sharing arrangements. There is also no evidence of collusive meetings involving NPC of the kind that characterised the pre-2002 period where the founding shareholders met to reach an agreement or to confirm the so-called global rules.’

[23] To the argument that Mr Strauss had been an active participant in the Port Shepstone meeting, the Tribunal found that as the representative of a company wholly owned by the three cartelists, being second, third and fourth respondent, ‘he was obliged to carry out the instructions given to him by the other three and had to implement the decisions taken by them in relation to NPC. He appeared to resent that and the fact that his recommendations to the shareholders were not acted upon.’

[24] Turning to the information exchange and the submission of the data to the C&CI, the Tribunal referred to the Commission’s own witness Ms Corrie, an employee of fourth respondent who testified that the C&CI data was aggregated information which could not be employed to determine, for example, first respondent’s actual share of the market and whether it was enjoying a growth or decline in sales. In addition, she confirmed that, between 1999 and 2001/2, she had attended regularly scheduled meetings with second and third respondent at which targeted market shares, the level or pricing and the related matters were discussed but at which meetings the first respondent was invited. Thus, if evidence of the information exchange had sufficed to further the collusive agreement, the Tribunal found the meetings which had been described by Ms Corrie would have been unnecessary. As the Tribunal put it: In other words, the information exchange fell short of the information required by the three firms to further the market sharing arrangements and, hence they needed further, clandestine meetings to gain additional information and to facilitate communication by direct contact.

[25] For these reasons, the Tribunal found that the 1998 Port Shepstone agreement could not be held to be an agreement within the meaning of that term as employed in s 4 (1) (b) of the Act. Further, in the absence of additional evidence, it was not possible to conclude that there had been an agreement for the purposes of s 4 (1) (b) during the period of the complaint; that is between 2005 and 2009.

The appeal

[26] On appeal, Mr Gotz who appeared together with Mr Marolen submitted on the strength of the European Court’ of First Instance’s decision in Heinz van Landewyck SARL v Commission 2009/78 at para 8 that an agreement will be concluded for the purposes of a cartel arrangement where firms have ‘expressed their joint intention to conduct themselves in the market in a specific way. Mr Gotz submitted this is precisely what unfolded at the Port Shepstone meeting where representatives of the four firms present jointly expressed the intension to divide the market by allocating market shares in territories between themselves. In the view of Mr Gotz, the evidence, particularly of the representatives of second to fourth respondents, confirmed that it was their joint intention to conduct themselves in a market specific way which included the participation of first respondent which continued to operate at its full productive capacity at that time which enabled it to maintain the historic market share that was understood to be between 10% to 12 % of the SACU market.

[27] Mr Gotz submitted further that the participants agreed that second to fourth respondents would withdraw from Southern KZN, effectively allocating that territory exclusively to the first respondent. While this agreement was entered into before the Act took effect on 1 September 1999 the consensus reached at that meeting continued to hold throughout the relevant period of the complaint. Mr Gotz placed emphasis on the evidence of Mr Strauss that an agreement had been reached at Port Shepstone and that first respondent would maintain a particular market share, that he had participated in discussions at the meeting and that he had not objected in any way to the agreement that was reached at the meeting.

[28] The agreement which had been reached at that meeting at which Mr Strauss of first respondent had participated sufficed to bring the four respondents conduct within the scope of s4 (1)(b) of the Act. In the view of Mr Gotz, whatever the consequences of the takeover by Cimpor, the fact remained that first respondent submitted sales figures to the C&CI and carefully monitored its national share throughout the relevant period. It continued to target a market share of 10 to 12 % which remained remarkably constant throughout the relevant period. This conduct had to be viewed against the backdrop of the evidence of the nature, purpose and effect of the Port Shepstone agreement and the exchange of information which constituted the key mechanism by which the parties were able to maintain the agreement or to put it in the words of Mr Jones ‘without the information exchanges through the C&CI the monitoring of market shares would have largely been guess work.’ In short, first respondent continued throughout the complaint period to comport its conduct in terms of the agreement which had been reached at Port Shepstone.

Evaluation of Port Shepstone Agreement

[29] The fundamental premise upon which the Commission’s case rests is that a continuous thread must be drawn from the conclusion of an agreement reached in Port Shepstone in 1998, and that the conduct of first respondent, , albeit under the control of Cimpor from 2005 to 2009 manifested clear compliance with the terms of this agreement. In so submitting the Commission relies on the approach adopted by this Court in Netstar (Pty) Ltd v Competition Commission (Case number: 97/CAC/May 10 at para 25) where an agreement was defined as ‘actions of and discussions among the parties directed at arriving at an arrangement that will bind them either contractually or by virtue of moral suasion or commercial interest. It may be a contract which is legally binding or an arrangement or understanding that is not for which the parties regard as binding upon them. Its essence is that the parties reach some kind of consensus.’

[30] On the strength of this dictum, Mr Gotz submitted that the Port Shepstone agreement clearly involved discussions among as well as actions which were directed at achieving a consensus between the parties. In his view. the evidence of representatives of second, third and fourth respondents namely Jones, Pienaar and Fourie, clearly confirmed the joint intention of the parties to conduct themselves in the market in a specific agreed way .This included the fact that first respondent would operate to its full productive capacity at that time, would maintain a historic market share which was known to the parties as between 10 to 12 % of the total SACU market and that the other participants, being second third and fourth respondents, would all withdraw from the Southern KZN effectively thereby allocating that territory to first respondent.

[31] By contrast, Mr Maenetje who appeared together with Mr Turner on behalf of the first respondent relied on the judgment of the Court of Justice of the European Union in Soliver NB v European Commission (10 October 2014). The case involved alleged agreements and concerted practices between various undertakings which were active in the production and the distribution of car glass. The applicant, a small glass manufacturer which was active in the automobile industry was found by the European Commission to have participated in cartel activity between 19 November 2001 to 11 March 2003. The Commission contended that a group of car glass suppliers had throughout the relevant period monitored their respective market shares and further the three major manufacturers (which excluded the applicant) had participated in trilateral meetings sometimes called “club meetings”. These meetings between competitors comprised of discussions that took place with regard to the supply of car glass for current or future car models, the evaluation and monitoring of market shares, the allocation of supplies of car glass to manufacturers and the exchange of information on prices as well as the exchange of other commercially sensitive information and the coordination of the pricing and supply strategies of those various competitors.

[32] At some point the applicant began to participate in the cartel, having been contacted by one of the large manufacturers. According to the court, the large manufacturers exploited the fact that the applicant was dependent on the producers of the raw material as it lacked in house production of flat glass.

[33] Central to the applicant’s argument before the Court was a denial that it intended to contribute by its own conduct with the common objectives pursued by all the participants in the cartel, that it was aware of the planned or actual conduct of the other undertakings in pursuit of these objectives or that it could reasonably have foreseen the nature of the cartel activity and that it was prepared to take the risk. Of particular relevance was a submission that since the applicant did not participate in any of the so-called club meetings it was not aware of the existence of systematic and detailed agreements between the other three covering the entire market for car glass in the European Union.

[34] The Court set out the relevant law as follows:

‘The undertaking must therefore be aware of the general scope and the essential characteristics of the cartel as a whole… where that is the case, the fact that an undertaking did not take part in all the constituent elements of a cartel or that it played only a minor role in the elements in which it did participate must be taken into consideration only when the gravity of the infringement is assessed and, as the case may be, in determining the amount of the fine… It is undisputed that the applicant did not take part in all of the constituent elements of the infringement. Moreover the Commission acknowledged that the applicant had not participated in any of the actual meetings of the representatives of the club either prior to the period in which according to the Commission the applicant participated in the cartel or during that period itself. It follows, in accordance with the principles set out …, that in order to establish the applicants participation in the single and continuous infringement to which the contested decision relates, the Commission must show not only the anti-competitive nature of the applicants contacts… but also that the applicant was aware or could reasonably be expected to be aware of first the fact that those contacts were intended to contribute to achieving the cartels overall plan and secondly the general and the essential characteristics of the cartel.’ (paras 64-67)

[35] Based on this approach which adds a necessary gloss to the Netstar dictum of this Court, Mr Maenetje submitted that the cartel was constituted of second, third and fourth respondents. Furthermore the Port Shepstone meeting was not the only meeting between these three parties. Between 1997 until 2006/7, multiple meetings were held between these parties at different levels of management. During 1998 and 1999 a group of middle managers met, including Messrs Pienaar, Fourie and Jones. In late 1998 and during 1999 engagements took place at a senior management level with Mr Doyle on behalf of the second respondent, Mr Leo on behalf of the third respondent and Mr Gomersall on behalf of the fourth respondent.

[36] There was some effort to contend on the basis of a written note of the development of cartel discussions by Mr Blackbeard that the Port Shepstone meeting which had not been mentioned in the note was therefore not critical to the constitution of a cartel. But, even if it was the vital meeting, Mr Maenetje contends that first respondent was only included in two meetings and only one in respect of which evidence was led and that was the Port Shepstone meeting.

[37] First respondent’s case based upon the Soliver dictum was to the effect that the attendance by Mr Strauss, on behalf of a wholly owned subsidiary of the three admitted cartelists at the Port Shepstone meeting could not of itself make first respondent a member of the cartel. This submission is buttressed by the further observation that the content of the discussions that which was ultimately concluded concerned the implementation by the three shareholders of the first respondent of a division of the national

market ,an agreement on an exchange of confidential information and communication between operational personnel to ensure the implementation and policing of these allocations. By contrast, as a wholly owned subsidiary of the three cartelists, first respondent was obliged to provide its shareholders with all relevant information and it was therefore any agreement regarding these matters as they concerned the first respondent was never necessary for the purpose of the implementation of an allocation or the monitoring of the agreement. In short, the respondent’s case was that at that meeting in Port Shepstone, first respondent was no more than a witness to an agreement between the other three respondents. It was not asked to agree to maintain any national market share in that it did not have one. The 10 %to 12 % share simply comprised its share of the KwaZulu-Natal market translated into the factor that the other firms employed to measure market share, namely a percentage of a national market share.

[38] Until the share of first respondent were acquired by Cimpor, second respondent was the appointed manager of first respondent and the three shareholders constituted a board of directors which had full strategic control over first respondent in addition to which it possessed the right of complete access to all first respondent’s confidential commercial information. Having decided that first respondent would be restricted to an operation in the southern part of KwaZulu-Natal, there was no need to do more in that it was operating at full capacity, in a geographical location which was distant from the contested inland market where all the key producers, being the three cartelist operated.

[39] According to Mr Strauss, upon his return to first respondent after the 1998 meeting, it was not necessary to provide any instruction to anyone employed by first respondent to carry out any cartel related activity. All that happened was that first respondent continued with its business activities in exactly the manner as it had conducted them prior to the meeting,

[40] In this connection Mr Strauss’ evidence is of particular significance.

Notwithstanding the attempt by Mr Gotz to seek to exploit it for the advantage of the Commission’s case. The following passage is instructive:

‘At the 1998 meeting the representative of PPC, Alfa and Lafarge decided that each of their companies and NPC should maintain certain market shares in designated geographic areas. Where a company’s market share and a particular geographic area exceeded the agreed market share a company would increase prices in that area with a view to reducing sales and market share in that area.’

[41] In other words, the agreement was struck between the three shareholders which then imposed their bargain on their wholly owned subsidiary, first respondent.

[42] When these competing arguments are evaluated, at best for the Commission is the conclusion that Mr Strauss on behalf of NPC said nothing to gainsay the drawing in of his firm into the agreement struck by the other three respondents. This is not entirely surprising, bearing in mind that at that point the Act had not yet been implemented. Nonetheless, given that Mr Srauss had said nothing in protest and therefore both knew about the agreement and never distanced himself therefrom, the Commission’s case rests on the basis of a duty to speak against the participation in the cartel if a party wants to show that it was not a party thereto. By its silence, first respondent, notwithstanding its inability to resist the will of its shareholders, constituted a party to the agreement. It was precisely this conclusion which the Tribunal refused to accept. It took account of the lack of independence of first respondent from its shareholders, which were clearly participants in the cartel. It noted that any exchange of information between first respondent and second, third and fourth respondents was not dependent upon the conclusion of a cartel agreement but as of right as shareholders and because its representatives constituted the board of first respondent.

[43] The Commission sought to counter the argument that the first respondent was a wholly owned subsidiary of three firms that were participants in a cartel by emphasising both Mr Strauss’ knowledge of the need to end the price war, his participation in the deliberations the Port Shepstone meeting and the organisational autonomy possessed by first respondent to compete with its

shareholders. But what is never answered in the Commission’s case is the implication of the many meetings that continued to take place concerning cartel activity in the absence of first respondent nor the fact that there was no need to obtain agreement from first respondent to share confidential information, given its position as a wholly owned subsidiary nor the autonomy it appeared not to possess in the allocation of market share.

[44] Even if this Court assumes in favour of the Commission that an agreement was reached at Port Shepstone which included the first respondent, the question arises as to whether the conduct of first respondent constituted participation in a cartel for the purposes of s 4 (1) (b) of the Act subsequent to the purchase of its shares by Cimpor from the three cartelists being second, third and fourth respondents.

[45] It is to that issue that I must now turn.

The takeover and its consequences

[46] The Commission contended that there were no negative implications for its case as a result of the 2002 takeover of first respondent. Mr Gotz submitted that a mere change in the shareholding did not justify prior unlawful conduct in that to hold to the contrary would permit a cartel member to escape liability by ensuring that its shares were transferred to another entity prior to the referral of a complaint. What was required of NPC under its new shareholding to justify the argument that it was not part of the cartel was ‘a firm repudiation of its continued adherence to the Port Shepstone agreement.’ This would be in keeping with this Court’s dictum in MacNeil Agencies (Pty) Ltd v the Competition Commission of South Africa [2013] ZACAC 3 at para 64: ‘under certain circumstances our law imposes on a person the duty to speak and a failure to do so where the duty exists may amount to an objective manifestation of consent regardless of the subjective intention of the silent party.’

[47] It was suggested by the Commission in argument that the price paid by Cimpor for first respondent’s shares included a premium paid to keep the second, third and fourth respondent ,being the sellers away from competing in the market of first respondent. The only evidence upon which the Commission could point to justify this submission was an internal memorandum which appeared within the organisation of second respondent and in terms of which an unknown author speculated as to what a potential third party buyer might pay for the shares in first respondent. All that is contained in the record concerning this document, referred to as exhibit 4 in the record of evidence, was that it was authored by certain Alpha 1 and was dated 17 July 2000. It contemplates the possible sale of first respondent and then: ‘Sell to a Fourth Player The problem will be to find a fourth player willing to run the risk of low real returns unless there is an implicit non-competition agreement which defeats the objectives of the new Act.

A new player

o does not have the cement capacity to retaliate in the North to any significant extent;

o its only resort is a price war

…

Sale to one of existing players

This could be more viable than selling to an outsider as:-

o an existing player would have more potential to retaliate than an outsider by damaging competitors’ Northern markets

o but will only be considered viable if there is implicit non-competition.

Status Quo

This will however require:-

o removal of shareholder directors from Board and appointing outside directors;

o loss of control unless shareholders have the right to fire directors if the Company does not perform;

o implicit non-interference in NPC natural market.’

[48] Mr Gotz referred to his cross examination of Mr Strauss and in particular the questions put to Mr Strauss about the approximate

purchase price that Cimpor paid for the shares being R 984 million which was considerably more than the R 500 m that had been referred

to in exhibit 4. In particular he emphasised the question put to Mr Strauss by the Tribunal in which the former conceded that it appeared that Cimpor had paid a premium for the shares that it acquired.

[49] A reading of Mr Strauss’ evidence shows that he was hardly a model of clarity with regard to the payment of a premium. In response to a question from Mr Daniels of the Tribunal, Mr Strauss said:

‘What I learned later from Cimpor is that they had certain, call it rules of thumb. So, they would say if you buy a Ready Mix company, you would pay so many EBITDA’s. If you buy a cement company, you would pay so many EBITDA’s. So, they would related the price to the profitability of the company. That was their view. Whether they … I don’t know if they did the IRR under these scenarios.’

[50] Pressed by Mr Daniels as to whether Cimpor would have wanted to ensure that no one attacked the market of first respondent in Southern Kwa- Zulu Natal he said:

‘Well I don’t know what assumptions they made regarding the competition but they certainly knew what the total market was and what the competition was doing.’

[51] There were two further difficulties with regard to this aspect of the Commission’s case. In the first place the question of the premium was not pleaded. It was not the case brought by the Commission before the Tribunal. Secondly, the note authored by an anonymous writer could not be examined by the Tribunal and as the first respondent expert economist Mr Smith pointed out, when the note was prepared in 2000, it could not be said with certainty that the market conditions upon which the note was predicated were the same as those which applied when the Cimpor sale took place. In short, this component of the Commission’s case appeared to be an afterthought. At the very least, it should have been pleaded and evidence with regard to a fair market value and/or premium for the shares in the year of the sale, 2002, should have been produced by the party on whom the onus rested to show the existence of a cartel.

[52] Mr Gotz referred to a meeting in Lisbon which took place immediately after the Cimpor takeover in December 2002 at which Mr Fourie of the third respondent was also in attendance. It appears that Mr Fourie attended as a representative of third respondent which held shares in Cimpor. Notwithstanding first respondent’s limited productive capacity, a decision to expand its production facility was not taken at the Lisbon meeting although Mr Strauss had motivated for it.

[53] On 1 April 2003 a further meeting was held in Paris which was attended inter alia by Mr Strauss and Mr Fourie. The question of cooperation between the two companies certainly raised that the minute of the meeting reflects the following:

‘Lafarge indicated that they have surplus clinkers capacity in South Africa. Their estimates indicate that NPC will need more clinker in 2004 and they could be in a position to supply. This will be a relatively short term arrangement but could postpone the need for NPC to invest in more clinker capacity. Cimpor that more clinker milling capacity would also be required and asked if the Richards Bay grinding plant would be for sale. Lafarge indicated that they are not considering the disposal of the Richards Bay depot.’

[54] Following upon this meeting in April 2003, first and third respondent concluded a supply agreement in terms of which the former undertook to mill clinker at its Richards Bay plant and to supply first respondent with cement. Mr Gotz referred in addition to the fact that first respondent supplied both second and third respondents with cement for their Ready Mix operations in Southern and Northern KZN. In other words, these competitors constituted customers of first respondent. Further, the cement was sold to them at prices lower than the prices it offered to third party customers and in certain years did not impose any price increases with regard to these sales.

[55] In summary, the Commission contended that the competitors of first respondent appeared to sacrifice capacity that could otherwise have been used to take market share away from first respondent, and further that first respondent used a significant portion of its capacity to supply its competitors when it could have redirected that capacity to supply third party customers. These facts, read with the continued exchange of information and the maintenance of first respondent’s market share, confirmed the continued existence of first respondent’s participation in a cartel which had its origins in the Port Shepstone agreement.

[56] To this argument, first respondent responds that there was a clear proposal to expand capacity by way of expansion of a fully-fledged clinker and cement factory at Simuma. According to Mr Strauss, Cimpor required studies to be done before investing an approximate amount of R 1 b in the expansion project. The forecast and assumptions relevant to first respondents investigation in 2003 were different from those which had applied in 1996 when the issues of expansion was raised previously.

[57] There was evidence that throughout the period commencing in the 1990’s and continuing until to 2009, first respondent supplied cement to Ready Mix Concrete Operations owned by second and third respondents in Durban. These operations were major downstream

customers located close to its factory and they bought large volumes and indeed did receive preferential prices.

[58] The difficulty confronting this Court in an evaluation of the competing arguments relating to the supply of cement by first respondent to second and third respondents was that there was no evidence produced by the Commission to justify the argument that the provision of this concrete formed part of any agreement either concluded at Port Shepstone or otherwise. Furthermore, no witnesses from either the second or the third respondent were called to explain how these agreements form part of the cartel. Significantly, it would have been possible for the Commission to have produced the kind of evidence which might well have shone better light, viewed from the vantage points of the Commission’s case, on this issue. For example, Ms Saizedes who was employed by third respondent throughout the relevant period made no mention of this arrangement or implications thereof in her witness statement, notwithstanding that she referred in her witness statement to the cement producers laboured under capacity constraints due to the growth in demand of cement.’ She was not called to give evidence on behalf of the Commission which was clearly an option available to the Commission as were the employees of second and third respondents.

[59] The regrettable consequence is that insufficient evidence was produced by the party on whom the onus rested to draw out what possible implications this supply of product by first respondent could have held for the overall case dealing with first respondent’s

participation in a cartel. While a court is entitled to draw an inference from the proven facts, this inference must be the more

natural or plausible conclusion among several other possibilities; that is it must be the more plausible acceptable or credible

inference that can be drawn by a court. See Accident Guarantee Corporation Ltd v Koch 1963 (4) SA 147 (A) at 159.

[60] Mr Maenetje submitted that an equally plausible inference to be drawn was that these agreements had absolutely nothing to do with the cartel and had they held such importance, the Commission, having access to the relevant witnesses, would have called one or more of them to testify before the Tribunal. Indeed, it would be a dangerous precedent for this court to engage in speculation in favour of the Commission simply because, understandably all competition authorities including this Court, must adopt as strict as possible approach to cartel behaviour being the most pernicious of all anti-competitive practices.

The balance of the Commission’s case

[61] In a Directive issued by this Court to facilitate the oral hearing, the parties were asked as to whether from 2005 to 2009, there was evidence that the first respondent remained a member of the cartel. To this critical question the Commission answered as follows: First respondent submitted sales to C&CI and monitored its national market share during this period. It continued to target a market share between 10 to 12% throughout the period. These practices had to be viewed against the backdrop of the evidence, nature, purpose and effect of the Port Shepstone agreement and the consequent exchange of information. In support of this argument, the evidence of Mr Jones was again cited; in particular his testimony that ‘without the information exchanges through the C&CI the monitoring of market share would have largely been guess work.’ Thus, by continuing to submit information to the C&CI, the first respondent through Mr Strauss continued to act in accordance with this understanding/ concerted practice.

[62] Mr Gotz submitted that the information exchange was of the utmost importance. To this end he referred to the evidence of Mr Strauss; in particular the following passages:

‘Mr Strauss, we need not beat about the bush, what I want to put to you is that NPC used the information in order to track its market shares.

MR STRAUSS: Mr Chairman, I can answer that, but the answer is not one word. I can say no, and if you want me to qualify it I can.

ADV GOTZ: Sorry, I’m not sure I understand why you’re being ambivalent Mr Strauss. Either the question is yes or no, or the answer is yes or no?

MR STRAUSS: We did not …no, the answer has to be qualified. We can see our weekly market shares. The fact of the matter is there’s not a lot that you can do about it. PPC wanted weekly stats because they were under pressure from their shareholders, why they couldn’t supply the market, how big is the market, what is happening? We had no way and it wasn’t in NPC’s habit to just change prices on a weekly basis. We didn’t do that, but we had to know what our … our objective was, as we said before, to run at full capacity.

If we saw that we’re falling behind on that, then it’s useful for us to understand where did we lose? Did we lost it in the South or the North, so we can go and take action on that? That was our overall objective. The market share is an outcome of that strategy, it’s not the target.’

And further:

‘You see Mr Strauss you’re monitoring your market share and you tell the Commission that you know it moves in the range 9% to 12%?

MR STRAUSS: Yes.

ADV GOTZ: You are utilising the C&CI data that you receive every month in order to monitor this market share, correct?

MR STRAUSS: Yes, of course, we’re monitoring the market as well.’

[63] In addition Mr Gotz submitted that first respondent’s market share had remained relatively constant throughout the period, being between 10 to 12% share of the national market. Therefore, it was consistently within the targeted range for the relevant period. In his view, this evidence showed the continuation of the behaviour which was in compliance with the understanding which could trace its existence back to the Port Shepstone agreement. In this connection reference was made by Mr Gotz to the dicta of this Court in Videx Wire Product (Pty) v Competition Commission of South Africa [2014] ZACAC 1 at para 16:

‘The content of the overarching agreement thus inferred might not be an agreement on the specifics of prices to be charged / tendered or of customer allocation (that would happen at meetings held from time to time) but rather than understanding that the firms will benefit from ongoing cooperation on these matters and will thus remain in communication and have an open door for purposes of working out details as occasion demands.

[64] In summary, the argument is that the fact that throughout the period of 2003 to 2009 the first respondent continued to exchange information through the C&CI, that it continued to target a market share and succeeded in achieving that market share between 10 to 12%, notwithstanding changes in the demand for cement relative to its productive capacity, cumulatively manifested a continued adherence to the Port Shepstone agreement. This it was argued was buttressed by the evidence relating to first respondent’s

meetings with third respondent and various other cooperative arrangement which took place subsequent to 2003.

The evaluation of the Commission’s case read as a whole

[65] An evaluation of the Commission’s case depends critically on the weight this Court gives to the two core elements of the Commission’s case as applied to the compliant period; being the exchange of information and the stability of its market share. In other words, even if this Court is prepared to assume in favour of the Commission that the first respondent was a party to the cartel agreement forged at Port Shepstone in 1998, the question arises as to what evidence exists to show that, post the Cipcor takeover, first respondent continued to be part of the cartel.

[66] Absent satisfactory evidence about the reason for any premium paid by Cipcor for the shares bought or the basis of the supply agreements between first, second and third respondents, the Commission’s case must stand or fall on the legal implications of the exchange of information and the market share of first respondent.

[67] It is common cause that first respondent submitted sales data to the C&CI. But it appears that first respondent would send data to Deloitte during the relevant period which would aggregate this data and then publish the aggregate statistics to reflect ‘inter alia’ a national sales totals, regional sales totals ,specific product totals (for example pure cement being separate from extended cement). Deloitte would also provide each firm with its own data organised in a manner which would enable that firm to determine its market share as the published aggregate data. Other than giving back its own data to each firm, Deloitte kept the individual firms’ data confidential and did not even provide this data to the C&CI.

[68] Of equal importance was evidence from both Ms Corrie and Ms Beverin that the C&CI statistics did not provide insight into what other firms were doing and that these statistics could not be used to calculate the share of competitors in the market. Ms Corrie’s testimony in this connection is of particular importance for the evaluation of the significance of the Commission’s argument that the exchange of information between the parties was evidence of cartel activity. For this reason, I set out an extensive passage from her evidence to the Tribunal :

ADV TURNER; ‘And transparency, there are two levels of transparency, the first level of transparency is the C&CI statistics, you have details of the full market and you have details of your production in that market, correct?

MS CORRIE: Correct.

ADV TURNER: That first level doesn’t give you insight into what the others are doing?

MS CORRIE: No.

ADV TURNER: The second level of transparency is when they give you the data and you have full transparency into what they are doing?

ADV TURNER: And you achieved that but you achieved it only with AfriSan and Lafarge

ADV TURNER: You didn’t achieve transparency with NPC? MS CORRIE: I didn’t, no.

ADV TURNER: And predictability, you’ve explained in your witness statement is that you had an open line to your competitors so that if something happened you could call them or you could call a meeting and you could assess the situation in consultation?

ADV TURNER: That gave you predictability? MS CORRIE: Correct.

ADV TURNER: So if something happened and it didn’t seem in line with your arrangements, you could pick up the phone and you did pick up the phone, correct?

ADV TURNER: And you checked whether what they were doing and you could then react, you could react and you could react by maintaining stability?

ADV TURNER: And that was the cartel conduct that you say was in place from 1999 through to when the meetings stopped in around 2006/7/8?

MS CORRIE: The informal get-togethers and informal cartel.

ADV TURNER: Yes. So if I understand just the progression, you needed to have the meetings around 1999 through to 2002 and after NPC was sold, you needed to make sure that everybody knew the rule correct?

ADV TURNER: And as the rules were established and NPC had been sold, you needed fewer meetings, because everyone knew the rules?

ADV TURNER: And when there was a hiccup or there was a problem in the market, you could pick up the phone, speak to you opposite number either at AfriSam or Lafarge or both and you could resolve that quickly, because you had established that relationship of trust?

ADV TURNER: You achieved none of that with NPC? MS CORRIE: I didn’t communicate with NPC at all.’

[69] This evidence highlights the problem of attaching significance to the system of the exchange of information for the purposes of a s4(1)(b) case. As the extensive passage of Ms Corrie’s testimony before the Tribunal clearly reveals, the information submitted to the C&CI and distributed to Deloitte was hardly of the kind that could be employed by cartel members to bring a rogue member into line. As Ms Corrie, who was a key witness for the Commission conceded in her testimony, the nature of the information provided could not contribute to the implementation or the enforcement of the cartel arrangements; in particular as it affected the alleged participation of first respondent

[70] It is clear from the available evidence that various meetings that took place, post the Port Shepstone agreement, which did not include first respondent. These comprised a regular series of secret meetings between representatives of the three major cement producers, being the second to fourth respondents. It was at these meetings that the three cartelists exchanged confidential sales information the content and nature of which manifestly fell outside of the C&CI process.

[71] Returning to the dictum in Videx supra at para 16, this evidence does suggest ‘some sort of ongoing cooperation’ in circumstances where second, third and fourth respondent had frequent discussion and engagements at which direct exchanges of commercially sensitive and confidential information was exchanged through these meetings. This was described in some detail by Ms Corrie but, critically, in circumstances where the first respondent was not involved in any of these processes.

[72] The same problem arises in relation to the second pillar of the Commission‘s case regarding conduct during the complaint period, that is that first respondent continued to target a national market share of between 10 to 12%. There was, however no agreement which allocated a national market share to first respondent neither at the Port Shepstone meeting nor elsewhere. At the time of the Port Shepstone meeting, when first respondent was a wholly owned subsidiary, the three major producers divided the market and simply converted the existing market share of first respondent of Southern KwaZulu Natal into an equivalent national market share. This is how the figure of 10 to 12% was calculated.

[73] Subsequent to the acquisition by Cimpor of the shares in first respondent, a change in the targets for market share to be gained were given to employees of first respondent In terms of the employee KPI scorecards, which formed part of the record, employees of first respondent initially had been incentivised to maintain a target in respect of the regional KZN market share. But from 2003 the same employees were incentivised to increase the national market share of first respondent which was clear evidence that Cimpor had altered the strategic vision for its newly acquired company, being first respondent.

[74] As first respondent’s economic expert, Mr Patrick Smith, pointed out, prior to its acquisition by Cimpor first respondent had shareholders being second, third and fourth respondents failed to initiate the expansion of first respondent’s productive capacity. After its acquisition by Cimpor, the new board of first respondent recognised the problem and set out to alleviate this capacity constraint. Hence first respondent engaged in a number of strategic initiatives to expand its production. It commissioned the first new kiln in South Africa in over 20 years which doubled first respondent’s clinker capacity. It grew its sales volumes, in particular over the period between 2004 to 2008. As Mr Smith noted ‘these initiatives marked a stark contrast in the strategic direction of NPC and demonstrated a clear break from its conduct as a passive controlled subsidiary of the three other respondents. These initiatives also indicated that NPC was trying to compete on its merits to expand outward in order to enable it to win new business.’

Conclusion

[75] To return to the Commission’s case: It alleged that the agreement struck between second, third and fourth respondents in 1998 was also entered into by first respondent, at the relevant time being a wholly owned subsidiary of the other three respondents. At that time first respondent ,as a wholly owned subsidiary, was bound to adhere to the constraints imposed upon it by its shareholders .This meant that it was allocated what was translated into a 10 to 12 % national market share and an obligation was imposed upon it to exchange information, which in any event its shareholders possessed as of right.

[76] Critical to the evaluation of this case is what occurred after the first respondent was uncoupled from its erstwhile shareholders and thus what conduct could be shown to evidence its continued participation in the cartel.Needless to say this examination depends on a conclusion, contrary to that reached by the Tribunal, that first respondent was a party to the 1998 Port Shepstone agreement.

[77] The central plank of the Commission’s case is the information exchange. Apart from the argument that this was used to enforce compliance by all four participants to the cartel including first respondent, the Commission contended that first respondent had used this exchange of information to engage in market share targeting; that is limiting its total production sales as to maintain a national market share as contained in the cartel agreement. But as shown, particularly although not exclusively from the evidence of Ms Corrie, the various respondents only obtained generalised statistics processed through Deloitte. In addition through much of the complaint period, first respondent achieved a relatively constant market share as it was constrained in its production facilities until very late in the period of the complaint.

[78] Neither of these pieces of evidence on their own is sufficient to justify cartel behaviour. Indeed they are in significant part gainsaid by the independent conduct of first respondent as shown through the incentives given to employees and ultimately the production expansion of first respondent subsequent to its acquisition by Cimpor. As I have already noted, to the extent the

Commission’s case could have been buttressed by evidence with regard to a premium paid for the shares of first respondent and the reasons that lay behind the agreements which were entered into between first and third respondents subsequent to the former having been taken over by Cimpor, this was never pleaded nor was any evidence advanced. Where further evidence could have shone more light upon the implications of the supply agreement between competitors after 2003, this was regrettably absent from the case brought by the Commission before the Tribunal.

[79] That cartel activity represents the very worst strain of anti-competitive conduct is surely trite. Courts need to be vigilant in ensuring the prohibition of this conduct. This is again manifestly obvious. Indeed, this Court in cases which were cited in argument in this case, in particular Videx, MacNeil Agencies, and Netstar, supra has developed a responsive jurisprudence for the curbing of cartel activity. But this does not mean that the rule of law does not apply to cartel cases, and can be elided over in favour of a result. That the Commission must discharge the burden that the Act imposes upon it to produce relevant evidence that shows that the nature of the conduct of the impugned party is such that it justifies a finding that the conduct so proved falls within the scope of s 4(1) (b) of the Act.

[80] As the European Court held in Soliver at para 63 with regard to the enforcement of cartel conduct: ‘it is only if the undertaking knew or should have known when it participated in an agreement that in doing so it was joining in the cartel as a whole that its participation in the agreement concerned can constitute the expression of its accession to that cartel …. In other words, the Commission must show that the undertaking intended to contribute by its own conduct to the common objectives pursued by all the participants and that it was aware of the unlawful conduct planned or put into effect by other undertakings in pursuit of the same objectives or that it could reasonably have foreseen it and that it was prepared to take the risk.’

[81] This dictum needs to be read with the approach adopted in Videx, supra at para16 and Netstar, supra at para 25. The overarching

understanding that must be shown to exist between the firms who can be classified as part of a cartel is one which will result in a benefit (or perceived benefit) from the prohibited conduct so that they remain in communication with each other and will be bound by moral suasion or more likely commercial interest .

[82] In the present case, the facts which the Commission proved, for the reasons set out, do not suffice to bring the first respondent within the scope of s 4(1) (b). The evidence presented by the Commission to prove its case does not show that the first respondent by its conduct during the complaint period intended to contribute to the objectives of the three cartelists. This conclusion might not be necessary if the Port Shepstone agreement is held, as did the Tribunal, not to include first respondent as cartel member. But if this finding can be rejected, which rejection requires some measure of generosity to the Commission’s case, the facts as presented do not justify the application to the conduct of first respondent of what are established legal principles regarding the existence of a cartel.

[83] It therefore follows that the appeal is dismissed with costs, including the costs of two counsel.

Davis JP Siwendu and Fisher AJJA concurred