Mercantile Bank, A division of Capitec Bank Limited and Others v Surve and Others (206/CAC/Oct22 ; 208/CAC/Oct22 ; 209/CAC/Oct22 ; 210/CAC/Oct22 ; IR153Dec21) [2023] ZACAC 2; [2023] 3 CPLR 33 (CAC) (17 June 2023)

IN THE COMPETITION APPEAL COURT OF SOUTH AFRICA

CAC Case No: 206/CAC/Oct22/

208/CAC/Oct22/

210/CAC/Oct22 CT

Case No: IR153Dec21

In the matter between:

MERCANTILE BANK, A DIVISION OF CAPITEC BANK

LIMITED First Appellant

STANDARD BANK OF SOUTH AFRICA LIMITED Second Appellant

ACCESS BANK LIMITED Third Appellant

And

MOHAMMED IQBAL SURVE` First Respondent

SEKUNJALO INVESTMENT HOLDINGS (PTY) LTD AND

ENTITIES WITHIN THE “SEKUNJALO GROUP OF COMPANIES”

LISTED IN ANNEXURE A 2ND to 36th Respondents

NEDBANK LIMITED 37th Respondent

FIRST RAND BANK LIMITED 39th Respondent

ABSA BANK LIMITED 40th Respondent

SASFIN BANK LIMITED 41st Respondent

INVESTEC BANK LIMITED 42nd Respondent

BIDVEST BANK LIMITED 43rd Respondent

THE COMPETITION COMMISSION 45th Respondent

ORDER:

Court: Competition Appeal Court

Coram: Manoim JP, Mlambo AJA, Spilg AJA

Heard: 30 -31 March 2023

Delivered: 17 July 2023

Summary: Application for interim relief for refusal to deal. Whether such conduct constitutes a concerted practice in terms of sections

4(1)(a), or section 8(1)(c) and section 8(1)(d)(ii) of the Competition Act, 89 of 1998 – need to establish at least a prima facie case of anticompetitive effect – need for theory of harm– Whether a respondent in an appeal against an order of the Tribunal has established that relief was final – Court finding it was in interests of justice to allow appeal – lis alibi pendens –same relief in other courts does not constitute lis alibi pendens as Competition Act considerations are exclusive

On appeal from the Competition Tribunal

1. The appeal of the three appellants is upheld.

2. The Tribunal’s orders in respect of paragraphs 1.2 (re Standard Bank), 1.5 (re Mercantile) and 1.8 (re Access Bank) are set aside.

JUDGMENT:

Manoim JP (Mlambo AJA concurring)

Introduction

[1] In this matter three banks seek to set aside an interim relief order made by the Competition Tribunal that obliges them, variously, to reinstate or not to close bank accounts, with firms comprising part of what is known as the Sekunjalo Group. The practice complained of is referred to in competition law parlance as a ‘refusal to deal’.

[2] When it brought its application for interim relief the Sekunjalo Group sought relief against nine banks. The Tribunal granted relief against eight of the nine.[1] Although eight banks were implicated in the Tribunal’s order, only three have sought to set it aside in this matter. I have referred to the Sekunjalo Group as a single entity because it is a useful shorthand for discussing some of the issues. However, there were 36 entities that brought the application to the Tribunal comprising of individuals, trusts, and firms.[2] Nevertheless all made common cause and were represented by one legal team both in the Tribunal and before this court. It is also accepted that they form part of the same control group and are all identified as black owned firms.

[3] This is not a case where all the entities banked with all the banks. The two banks which provided services to most of the entities are Nedbank and ABSA, but they are not appellants. The three appellants only provided some services to some of the entities. The first appellant, Standard Bank Ltd served three entities, whilst the second appellant, Mercantile Bank and the third appellant, Access Bank Ltd, served only one each.[3]

[4] The Tribunal found that in refusing to deal with the Sekunjalo entities all eight banks had contravened section 4(1)(a) and sections 8(1)(c) and 8(1)(d)(ii) of the Competition Act 89 of 1998 (“the Act”). Put more simply in refusing to deal, the banks were found to have acted in co- ordination with one another, and to have acted unilaterally as dominant firms, to abuse a dominant position. The Sekunjalo Group had also sought relief in terms of section 4(1)(b) and 5(1) of the Act based on the same facts but this relief was not granted by the Tribunal.

[5] Each appellant bank has both appealed and reviewed the Tribunal’s order. (For simplicity I refer to the banks simply as ‘appellants’ and where necessary by name). This leads to the first issue we have to decide in this case. The appeals and reviews largely traverse the same issues. The Sekunjalo Group’s first challenge in opposing the relief sought by the appellants is procedural. It asserts that the banks should not succeed on either route.

[6] The Act gives both applicants and respondents in interim relief proceedings before the Tribunal, the right to appeal. However, while that right is unrestricted for the applicant before the Tribunal it is restricted for a respondent, which may only appeal an order that has “a final or irreversible effect.” The appellant banks were the respondents in the Tribunal proceedings. Hence, they needed to show that the Tribunal order had a final or irreversible effect. The Sekunjalo Group argues that they have not and hence the appeals should fail.

[7] But the appellants have also brought a review. Although the Act provides for reviews of interim relief orders without, as in an appeal, distinguishing between an applicant and a respondent, Sekunjalo argues that nevertheless the review right should be construed narrowly for respondents, otherwise there is a danger that it becomes a form of disguised appeal. Put differently, the right to review should not permit an unsuccessful respondent before the Tribunal to recast a case as a review, which would otherwise be impermissible for it as an appeal.

[8] The second range of issues concerns whether a prima facie case has been made out both in fact and law against the banks. Each appellant makes out a similar argument here that no prima facie case of an infringement has been made out against them and hence they should succeed in setting aside the order.

[9] Finally, each bank makes out a narrower case for relief based on facts specific to the particular bank. Here the argument is that the Tribunal erred in throwing all the banks into the same pot for relief and thus failing to consider the individual differences.

[10] For reasons I explain later I will first deal with the prima facie case and then the issue of appeal and review although ordinarily one would deal with these issues first.

Current status of the orders

[11] The interim relief orders against the three banks operated for a period of six months.[4] The orders were granted on 16 September 2022 and would have expired on 15 March 2023. However, the orders were extended for a further period of six months until mid-September 2023. The extension of the orders for this further period was not opposed by the three banks. Whether an interim relief order can be extended more than once is a matter of contention but not one this court is called upon to decide now. The fact that the orders are still extant till at least September 2023 means the proceedings before us are not moot.

Background

[12] The Sekunjalo Group is a group of companies operating in a variety of different markets but all ultimately part of the same control structure. The first applicant is Dr Iqbal Survé, who deposed to the founding affidavit, and in his personal capacity was the first applicant in the Tribunal proceedings. Dr Survé is the executive Chairperson of the Group and sits on some of its

boards. He explains in his founding affidavit before the Tribunal that the label Sekunjalo Group is a descriptive term not a legal

entity. What they have in common is that Sekunjalo Investment Holdings (“SIH”) has either a direct or indirect associate interest in the companies. The Group identifies itself as a black controlled company and many of its entities are new entrants into the markets they operate in. The Group extends beyond the 34 applicants and comprises 200 companies. These companies have interests in a wide ranging number of markets. Dr Survé identifies some as media, information technology, health services, fishing, and aquaculture.

[13] The Group’s troubles in obtaining banking services started in March 2020 when a government Commission of Enquiry into allegations of corruption at the Public Investment Corporation (PIC) published its report. The Commission was led by the former President of the Supreme Court of Appeal, Justice Mpati, and hence its name, the Mpati Commission. The Commission inter alia concerned itself with the relationship between the Sekunjalo Group and the state owned PIC. Amongst the findings were that:

“In light of the above, the Commission recommends that PIC must conduct a forensic review of all the processes involved and all transactions

entered into with the Sekunjalo Group. Ensure that the PIC obtains company registration numbers of every entity to be able to conduct a forensic investigation as the flow of monies out of and into the group.”[5]

[14] These findings had an immediate impact on the banking sector. On 27 August 2020, ABSA which was the Sekunjalo Group’s primary banker, gave notice that it would close its accounts. ABSA then closed several of the Group’s accounts on 27 February 2021. Over the next fifteen months subsequent to the decision made by ABSA, all eight banks either closed accounts of some of the Sekunjalo entities, or refused to onboard new accounts, or indicated that the accounts were under review.

[15] The banks allege that their concerns emanate originally from the findings of the Mpati Commission, and the consequent publicity, including one on an industry website known as the World Compliance Report. [6]

[16] The banks explained that as registered financial entities they are subject to strict regulatory requirements under a number of statutes.[7] Most placed emphasis on the provisions of the Financial Intelligence Centre Act 38 of 2001 (“FICA”) which obliges banks in certain circumstances to terminate a banking relationship with a client. [8] They also relied on their common law right to terminate a contract with a client unilaterally for reasons that include reputational risk which was recognised by the Supreme Court of Appeal (SCA) in Bredenkamp v Standard Bank of South Africa.[9]

[17] The various banks maintained that the findings by the Mpati Commission and the considerable media attention they attracted, placed an obligation on them to reconsider, or in the case of new clients to consider, their reputational and regulatory risk if they continued or commenced a relationship with entities in the Sekunjalo Group. Much publicity attached to these decisions which some of the late deciding banks considered in their decisions i.e., an admitted awareness of what the other banks were doing.

[18] As some banks closed the doors to them, so the entities in the Sekunjalo group sought banking services from banks they had not previously had any dealings with. Thus, in the case of two of the appellants, Mercantile and Access, although only banking for one of the Sekunjalo entities at the time, they were later each approached to provide services for other entities in the Sekunjalo Group.

[19] Faced with the prospect of being “unbanked” with the doors of financial institutions apparently closing on it sequentially, the Sekunjalo Group embarked on a series of legal actions against various banks to interdict them from closing its entities accounts or to restore them. The court was advised at the time of the hearing before us that these applications were still pending. Thus, the application to the Competition Tribunal in terms of the Competition Act was one of several legal actions. I turn to this point later because one of the points Sekunjalo has raised is the fact that these applications make the present appeal cum review lis alibi pendens.

Case before the Tribunal

[20] The Sekunjalo Group filed a complaint with the Commission on 15 December 2021. The application for interim relief was then filed on 22nd December 2021. In the complaint Sekunjalo did not rely on section 4(1)(a). But this changed in the interim relief application. Its significance is that it was one of the sections the Tribunal relied on to grant the relief that it did against the banks.

[21] The Tribunal heard the case over two days from 7th to 8th March 2022. It issued its reasons and order on 16th September 2022.

[22] The Tribunal rejected Sekunjalo’ s complaints in terms of section 4(1)(b) and section 5(1). Since there was no cross appeal, these findings need not concern us. What the Tribunal did find was that all eight banks, including the three appellants, had contravened sections 4(1)(a), 8(c) and 8(1)(d)(iii) of the Act, in respect of the same cause of action, namely, a refusal to deal with the Sekunjalo Group or individual companies within that group.

Did the Tribunal findings establish a prima facie case.

[23] In deciding whether or not to grant interim relief the threshold set out in section 49C (2) of the Act states:

The Competition Tribunal—

(a) must give the respondent a reasonable opportunity to be heard, having regard to the urgency of the proceedings; and

(b) may grant an interim order if it is reasonable and just to do so, having regard to the following factors:

(i) The evidence relating to the alleged prohibited practice;

(ii) the need to prevent serious or irreparable damage to the applicant; and

(iii) the balance of convenience.

[24] The Act goes on to state in section 49(C)(3) that:

“In any proceedings in terms of this section, the standard of proof is the same as the standard of proof in a High Court on a common law application for an interim interdict.”

[25] The issue on appeal focussed on sub-paragraph (b)(i) viz. the evidence relating to the alleged prohibited practice. Read with subsection 49C (3) this requirement, all were agreed, meant prima facie evidence.

[26] The Tribunal, as noted, found that the conduct contravened section 4(1)(a) not section 4(1)(b).

[27] Section 4(1)(a) states:

“An agreement between, or concerted practice by, firms, or a decision by an association of firms, is prohibited if it is between parties in a horizontal relationship and if—

(a) it has the effect of substantially preventing, or lessening, competition in a market, unless a party to the agreement, concerted practice, or decision can prove that any technological, efficiency or other procompetitive gain resulting from it outweighs that effect; …

[28] Under section 4(1)(a) the conduct complained of could constitute an agreement, decision by an association of firms, or a concerted practice. The Tribunal found that it constituted a concerted practice. There is therefore no need to consider the two other possibilities

[29] A concerted practice is defined in the Act as “[…] cooperative or coordinated conduct between firms, achieved through direct or indirect contact, that replaces their independent action, but which does not amount to an agreement;”

[30] Before coming to a conclusion that a case for a concerted practice had been made out the Tribunal observed the difficulty with coming to such a conclusion in an oligopoly market which it identified the market for banking services to be;

“The concept that a concerted practice may be inferred from behaviour might seem to suggest, at first glance, that interdependent oligopoly behaviour amounts to a concerted practice. However, oligopoly behaviour does not establish a concerted practice unless, given the nature of the market, the behaviour of the firms concerned cannot be explained other than by concerted behaviour.”[10]

[31] This was a salutary note of caution. Unfortunately, as I will demonstrate, this was not followed through by the Tribunal in its analysis. The basis for the Tribunal’s decision that there was a prima facie case of a concerted practice hinged on the following facts: The banks had adopted a common position – to close or threaten to close the Sekunjalo entities’ bank accounts, that they did so within a short time frame (fifteen months) and that despite relying on the fact that they responded to the same regulatory environment this justification was unconvincing.

[32] The conclusion of its reasoning is contained in the following paragraph:

“While parallel conduct is not by itself proof of a concerted practice (such proof can only be uncovered through an investigation by the Commission), given the nature of the banking services market – the undisputed evidence of high barriers to entry and concentration – and the conduct of the banks, we find that the conduct complained of does at a prima facie level establish a case of a concerted practice between firms. This is because the Respondents, acting in concert, closed the bank accounts one after another without an adequate justification. For this finding it is not necessary to show that the Respondents agreed to this.”[11] (My emphasis)

[33] The Tribunal is here dealing with the evidence of the conduct of all eight respondent banks and not just the three appellants. The evidence was that in a period of fifteen months, starting with ABSA bank, all the banks had either terminated their banking relationships with entities in the Sekunjalo Group or when approach to provide banking services had refused to do so.

[34] There is no evidence in the record that any of the banks directly co-ordinated with each other in refusing to deal. Indeed, Standard

Bank had not yet made a decision to refuse banking services at the time the application was heard. In the case of Mercantile, it had not closed down the account of the only entity to whom it was providing services although it had decided not to take on any new ones. This suggests that the banks were more probably behaving independently than acting in co-ordination with one another.[12].

[35] The strongest evidence of the regard that any of the banks had to the actions of the others came from Access bank’s internal emails. These suggest that the bank’s officials had regard to the actions of other banks when considering their position in providing services to one of the Sekunjalo entities. But even if one were to extrapolate from this that such a ‘follow my leader’ approach informed the decision-making of all the appellants, it does not amount to a concerted practice. Firms may frequently have regard to the approach taken by rivals but that at best is conscious parallelism not parallel behaviour and the two should not be confused.

[36] The Tribunal acknowledged that the concertation did not relate to price, market division or collusive tendering (all of which are identified as anticompetitive practices under section 4(1)(b)).[13] But having correctly made this observation it did not identify, as it was obliged to as an essential step in the process of legal

reasoning, what the object of the concertation was. Rather the Tribunal moved on to what the effect of the concertation was; concluding that it was exclusionary because it denied the Sekunjalo Group banking services, making it impossible for them to operate as commercial entities competing in their respective markets.

[37] The Tribunal then went on to consider whether the banks had raised an efficiency justification for their conduct. Whilst acknowledging the importance to the banks of reputational risk, this was rejected because the Tribunal accepted the accusation made by Sekunjalo that the banks had been selective in adopting this policy. It was suggested that firms whose conduct was highly questionable did not have their banking services terminated. Whilst some names were mentioned it was not made clear which banks were providing them with services[14]. The Tribunal then concluded that the applicants had established, on a prima facie basis that there had been a concerted refusal to deal.

[38] The problem with the Tribunals’ approach is that the case falls at the first hurdle. Even under section 4(1)(a) which unlike 4(1)(b) does not itemise specific anticompetitive practices, there needs to be some theory of harm. The subsection refers to the concerted practice having an anticompetitive effect. Making this conclusion based on parallel conduct in a concentrated market does not amount to an explanation of why the conduct is anticompetitive. What the Tribunal did was to conflate an outcome – exclusion from the market – with an anticompetitive effect. While exclusion may be the result of an anticompetitive practice it does not suffice to use it as a substitute for analysing whether, as a fact, there has been an anticompetitive practice; particularly where the Sekunjalo Group did not allege that any of the banks had a direct or indirect interest in any relevant market that was in issue.

[39] There are many reasons why a practice may be anticompetitive – typically it may lead to higher prices, reduced supply, inferior service or quality, or lack of innovation. The refusal to offer services might be anticompetitive, if linked to some other theory of harm such as raising a rival’s costs or attempting to exclude it from a market in which the refusing firm competes with the refused firm. But none is advanced here. The Tribunal does not contend that the exclusion occurs in the banking markets. It occurs in the markets where each entity competes. What this means is that the banks are alleged to have concerted in an upstream market (banking) to exclude customers in downstream markets. This would make the effect of the alleged co-ordination vertical. Granted, it is a legitimate theory of harm to allege that horizontal co-ordination among rivals have vertical effects.

[40] But it is not suggested that any of the appellants competes in the markets in respect of any of the entities they are alleged to have excluded. Moreover, for the most part it appears that the different banks dealt with different entities in the Sekunjalo Group. It is thus not clear why the banks should want to co-ordinate to achieve some anticompetitive outcome against the Group’s various entities in unrelated markets. Such an explanation or theory of harm is completely absent in this case. But given the language of section 4(1)(a) which refers to the practice “… substantially, lessening or preventing competition in a market…” such a case needed to be made out. Since none was alleged on the papers, none could be contended to exist even on a prima facie basis. Section 4(1)(b) itemises three practices which are presumptively anticompetitive, but section 4(1)(a) does not, hence it requires them to be alleged before the respondent firm is required to plead its pro- competitive justification.

[41] The Tribunal however jumped this necessary hurdle and proceeded to ‘claw back’ an inference of anticompetitive practice by rejecting the banks regulatory compliance justification. This approach is impermissible. The onus in section 4(1)(a) requires the party alleging it to make out why the practice is anticompetitive

before the respondent is required to rely on the pro-competitive proviso. A mere observation of parallel behaviour coupled with a conclusion that it is exclusionary does not suffice. Otherwise, every firm that cut off a debtor at the same time as its competitors would be vulnerable to accusations of perpetrating a concerted practice. The party alleging must make out a case of why the parallelism

creates some form of competitive harm, even if it be only on a prima facie basis for the purposes of an interim interdict.

[42] In the well-known decision of the European Court of Justice in the so-called Woodpulp case the approach to parallel conduct evidence was expressed in this way:

“In determining the probative value of those different factors, it must be noted that parallel conduct cannot be regarded as furnishing proof of concertation unless concertation constitutes the only plausible explanation for such conduct”.[15]

[43] On the evidence before the Tribunal, it was only the banks that have given an explanation for their conduct. Sekunjalo gives an observation of the conduct not an explanation of why it is anticompetitive. Simply to allege that it is anticompetitive because the outcome was exclusionary is not sufficient; not even at a prima facie interim interdict stage since the issue is whether a legally competent basis has been alleged. There is no alternative explanation of why the banks would want this outcome. Thus, faced with that scenario the Tribunal could not infer there was, a fortiori, an anticompetitive explanation for the conduct simply because it rejected the plausibility of the banks justification.

[44] We find that there was no legal let alone factual basis to find that prima facie the appellant banks had contravened section 4(1)(a) by engaging in a refusal to deal with Sekunjalo entities.

Abuse of dominance.

[45] It is trite law that in order for a firm to transgress the dominance provisions of the Act it must be a dominant firm. Section 7 provides for three possibilities for a firm to be regarded as dominant. The first two, although reversing the onus, depend on market shares, [16] and the third, section 7(c), says a firm may be considered dominant if “it holds less than 35% of that market, but has market power".

[46] Market power is defined in the Act as: “[…] the power of a firm to control prices, to exclude competition or to behave to an appreciable extent independently of its competitors, customers or suppliers;”

[47] This means that even if there is no evidence that the firms’ market share exceeds one of the threshold requirements in sub-sections 7(a) or (b), it still might be considered dominant under 7(c) if it can be shown to have market power. The onus to prove market power rests on the party alleging this. In the present case Sekunjalo did not rely on market share for its allegation of dominance and instead relied on section 7(c).

[48] The Tribunal accepted that dominance depended on the application of section 7(c) and thus it had to determine whether each of the banks had market power. However, in making the assessment the Tribunal considered the banking sector as a whole and not the situation of the individual banks. Thus, it reached the following conclusion:

“The banking services market is highly concentrated. The Applicants’ unchallenged evidence is that Nedbank, Standard Bank, FirstRand, ABSA, and Investec collectively have a market share of about 90%[17] This suggests that the market is oligopolistic, with a fringe of smaller players. A combination of factors such as high concentration,

high barriers to entry, and a weak position of customers, suggests that the Respondents have an appreciable degree of market power.[18]

[49] The Tribunal went on to conclude:

“We do not have to decide whether the Respondent Banks are collectively dominant since on the prima facie evidence before us they

individually possess market power.”[19]

[50] This is a thin case on which to base a finding of the existence of market power, even by the less demanding standards of interim relief. Firstly, by adopting an overall approach and not dealing with each individual bank the Tribunal did not deal with the fact that two of the appellants, Access Bank and Mercantile, have market shares of around 1%. It would be hard to conceive of market power being exercised at such low levels. It was suggested in argument that the theory of a ‘lucky monopolist’ could apply. This is based on this court’s finding in the Babelegi case. [20] That case concerned a firm’s excessive pricing of face masks during the Covid 19 pandemic. The Court found that the respondent firm which was otherwise not dominant prior to the pandemic, had become a dominant firm because of the demand spike and supply constraints created by the pandemic and restrictive regulations. Hence it was a ‘lucky monopolist’ and able to extract an excessive price. This theory might hold in this case if the other seventy registered banks were refusing or unable to supply, leaving the remaining firms – the appellants - with market power.[21] However this theory has not been made out in the papers nor does the Tribunal advance this theory. None of the appellants can be regarded as a lucky monopolist.

[51] Second, switching costs is not a relevant issue on the facts of this case. It is a refusal to deal case, not a case where the dominant firm leverages market power over customers because of their reluctance to switch. The evidence in this case is that Sekunjalo was willing to switch or forced to switch when other banks began to refuse to deal with them. Thus, switching costs in this case was not an appropriate factor for inferring market power.

[52] But even if the dominance finding is sufficient for the purpose of interim relief, which I do not accept, there are still problems with the next step which is to determine whether the alleged dominance had been abused. The Tribunal found that Sekunjalo had made out a prima facie case for a contravention of section 8(1)(c) and 8(1)(d)(ii).

[53] Section 8(1)(c) states:

“It is prohibited for a dominant firm to— (a)…

(b) …;

(c) engage in an exclusionary act, other than an act listed in paragraph (d), if the anticompetitive effect of that act outweighs its technological, efficiency or other procompetitive gain;”

[54] Section 8(1)(d)(ii) states:

“It is prohibited for a dominant firm to —engage in any of the following exclusionary acts, unless the firm concerned can show

technological, efficiency or other procompetitive gains which outweigh the anticompetitive effect of its act—

(i) …

(ii) refusing to supply scarce goods or services to a competitor or customer when supplying those goods or services is economically feasible;”

[55] The Tribunal adopted the same analysis of the facts to come to its conclusions under both 8(1)(c) and 8(1)(d)(ii). The reasoning went as follows: Banking services are an essential service for the entities to compete in their respective markets. Without this input they are likely to be forced to exit their markets. If they exit there will be less competition in the markets the entities serve. The entities are black owned firms and hence, adopting a transformative constitutional approach, exclusion should be looked at with greater sensitivity to its effect on the excluded firm. The Tribunal then relies for this on the decision of this court in the eMedia case where interim relief was granted in a case concerning sub-section 8(1)(d)(ii) where a black owned firm alleged it was subject to exclusionary conduct by the denial of service from a dominant supplier. [22]

[56] But there is an important distinction between this case and that of eMedia. There, the dominant firm, and the firm allegedly being excluded, competed in the same market where the excluded firm was a significant rival.

[57] No such facts are advanced in this case. Thus, as with the section 4(1)(a) count, there is no theory of harm in respect of the section 8 counts. As Whish and Bailey note in their book:

“Most refusal to supply cases concern a vertically integrated undertaking that is dominant in an upstream market, and which refuses to supply to an existing or new customer in a downstream market.”[23]

[58] Again, it must be repeated that the appellant banks are not competitors of their customers or attempting to leverage their alleged dominance into any market in which their customers compete. Thus, the actions on the present record cannot be explained under any competition theory of harm. None has been offered and no facts were alleged to sustain such a case in law. Instead, the only explanation for the conduct is the one offered by the banks - the refusal to supply was a response to the regulatory climate they faced in respect of serving these customers, post the Mpati Commission report.

[59] Nor is there any analysis of why the removal of services will lead to the exclusion of the respective entities and why their removal will lead to an anticompetitive effect in their respective markets. Indeed, conducting the analysis at such a high level of generality has meant the Tribunal has not considered the specific impacts of each of the appellant’s decisions.

[60] Thus, in the case of Mercantile Bank the Tribunal ordered it to keep open an account for an entity known as Health System Technology (HST). However, it is now common cause that Mercantile had never closed this account. Although HST is mentioned in the founding affidavit before the Tribunal, Sekunjalo complains that it was Bidvest Bank and ABSA which were alleged to have denied HST facilities, not Mercantile. The complaint made against Mercantile was that between May and August 2022 it had declined to provide services to other Sekunjalo entities namely, AEEI, Afrinat and Loot Online. The Tribunal reflects this earlier in its reasons.[24] But its relief was restricted to HST not the refusal in respect of the other entities. Thus, there is no rational connection between the relief sought against Mercantile and that which was granted. Moreover, the service provided to HST was confined to the operation

of a forex trading account.

[61] Access bank only provided services for one entity in the Group, a company called Afrinat. It is not disputed that Access bank provided these services to Afrinat. However, the complaint in the founding affidavit was that the exclusionary effect in respect of Afrinat was that it was unable to secure forex services and thus procure raw materials from overseas suppliers. But Access never provided such a facility, only a transactional banking facility. The Tribunal’s order compels Access bank to restore the services it previously offered. But since it never offered forex services to Afrinat, the order does not remedy the alleged harm in respect of Afrinat. Nor does the order require Access bank to offer banking services to the other entities which it refused to deal with – the relief was confined to Afrinat.[25]

[62] In relation to Standard Bank the situation is more complex. At the time of the hearing, and the Tribunal does acknowledge this, Standard Bank had not yet refused services to any of the clients from the Group that it then served.[26] However it had given an indication that it was reviewing these clients. What Standard Bank had done at the time the application for interim relief was launched was to request information from Sekunjalo as a result of the adverse media reports about the Group. There is some dispute of fact about when the information was then furnished to Standard Bank and indeed whether it was ever adequately provided. Certainly, Standard Bank considered it had not.

[63] Nevertheless, it was only subsequent to the hearing, after conducting a due diligence exercise, that Standard Bank terminated these services. The Tribunal was aware that Standard Bank had not yet terminated these services, but it seems to have regarded the information request as a sufficient basis to grant the relief in the form that it did, which was to interdict Standard Bank from closing the accounts held with it. Standard Bank contends that the Tribunal could not have given such an order at the time since at the time of the hearing these accounts were still open and there was no reason, prima facie or otherwise, to suggest that it would not bring an independent mind to bear on whether to continue operating the account.

[64] Standard Bank through the review wanted its due diligence efforts to be placed on record to achieve two objectives: To suggest it did not act in concert with the other banks but followed its own process. Second, to make a factual case under FICA that it was not given a discretion to refuse to provide the services but was, having conducted the due diligence and found the information wanting, obliged to do so. Sekunjalo strongly opposed the admission of this new evidence either by way of appeal or review. However, for reasons given earlier this case can be decided on the assumption that Standard Bank would have closed the accounts as it is now common cause it did.[27] However Standard was not the exclusive banker of any of these three entities all of whom also had accounts with other banks. These facts were not considered and needed to be as part of an exclusionary analysis of Standard Bank’s potential abuse of dominance.

[65] Finally, the real question was whether a theory of harm was made out in respect of any of the appellants and the answer to that is in the negative. Abuse of dominance requires an analysis of the unilateral actions of the respondent firm. By relying on generalised facts and applying it to all the respondent banks the Tribunal failed to do so. In conclusion then, there was no prima facie case made out sustainable in law of an abuse of dominance established in respect of any of the appellants. The case failed as the appellants were neither shown prima facie to be dominant, let alone if they had been, to have abused that dominance in relation to the entities for which relief was granted.

Conclusion on prima facie case

[66] The Tribunal appears to have been alive to these weaknesses in the competition theory of harm but sought to compensate for it in two ways. The first was to place reliance on the fact that an applicant only needs to make out a prima facie case. The second was that the factors which require consideration for interim relief must be viewed cumulatively and the presence of one strong factor can make up for the weakness of any other. It certainly considered that the evidence of irreparable harm to the entities was strong as was the balance of convenience favouring the granting of the relief. As a legal approach to considering interim relief applications both propositions are correct. But a distinction needs to be made between a prima facie case though open to some doubt (as per the Webster v Mitchell formulation) and a non-existent one. Even in a case requiring only prima facie evidence the complainant has to place some facts before the decision maker to pass muster. Here, for the reasons I have explained, Sekunjalo has not made out a case that the refusals to deal were anticompetitive.

[67] There is therefore no prima facie case made out on the papers in terms of a prohibited practice by the appellants either in terms of section 4(1)(a) or sections 8(1)(c) or 8(1)(d)(ii). For that reason, no order for interim relief should have been granted against the appellants.

Is the decision of the Tribunal appealable or reviewable?

[68] I now turn to the issue of whether the appellants have made out a case for appeal or review given that the decision of the Tribunal was for interim not final relief.

[69] Sekunjalo argues that none of the appellants has met the threshold requirements for an appeal set out in subsection 49C (8) of the Act. That subsection states:

“The respondent may appeal to the CAC in terms of this section against any order that has a final or irreversible effect.” (My emphasis).

[70] On the face of it this language appears tautologous. After all what could be more final than an order which is not irreversible. Yet because of the ‘or’ these words must be read disjunctively. Unterhalter AJA has offered a sensible interpretive solution to this problem when in Business Connexion, an interim relief case in this court, he explained that:

“(a) The interim order is rendered final in effect because the prohibited practice and the relief to which it gives rise will not be considered by the Tribunal because no referral is likely to be made or the Tribunal purports to decide an issue with finality by way of interim relief that it would be required to decide on a referral to it. This second variety of finality will likely constitute an ultra vires decision that is reviewable, but it may be also be (sic) appealed. The interim order has an irreversible effect where it materially disadvantages the competitive position of the respondent in the market and the disadvantage is not likely to be undone should the respondent prevail before the Tribunal upon the hearing of the referral or should the referral never occur.[28]

[71] The appellants in this matter contend that the appeal is competent because the effect of the Tribunal order is irreversible. If they have to comply with the terms of the order, they will be in contravention of their regulatory and governance obligations. The consequence of that is reputational damage. That damage they argue is irreversible even if the interim relief does not become final relief.

[72] Sekunjalo argued that the reputational damage argument is exaggerated. The banks would be seen to be complying with an order of a regulator which they had attempted to oppose. Any sense of their being complicit with providing a facility in breach of the law is fully met by the argument that they had to comply with another regulator. There is no suggestion at this stage that banking regulators are acting against any bank for providing services to the Sekunjalo Group.

[73] This slender reliance on irreversibility may well explain why ex abundanti cautela the appellants have also brought a review for essentially the same relief as would be brought in an appeal. The review focuses on both a legality and PAJA review of the Tribunal.

[74] However, this does not mean that an appeal is not competent in the present case. A more solid basis for considering the case appealable is also found in Business Connexion. Unterhalter AJA despite having formulated a narrow test in the passage I cited earlier, goes on to take a more holistic approach when he observed that:

“a proper understanding of what constitutes a final or irreversible effect must reflect the need to permit this Court to correct error when particular failures of justice would otherwise result.”

[75] This test is consistent with what the Constitutional Court laid down in ITAC v SCAW, where it held that the appealability of an interim order made by a High Court will depend on what the interests of justice require.[29]

[76] In the present case we have found the actions of the banks did not only not contravene the Act but were carried out in what the banks perceived was necessary compliance with their regulatory obligation. The banks obtained no financial benefit for their actions. It would amount to a failure of justice if the appellants, in seeking to comply with one regulatory regime (banking legislation) remained subject to an order, under another regulatory regime, albeit only interim, that without a sufficient legal basis, required them not to comply with what they believe are their obligations under the former. Hence, I find the order for interim relief is appealable.

[77] Finally on this score, I consider the issue of deference to the Tribunal which was also raised. It is correct that in the case of Mediclinic the Constitutional Court held in overturning a decision of this court that in turn had overturned a decision of the Tribunal that:

“Interference with factual findings by appellate courts would thus be justified only in the event of a misdirection or a clearly wrong decision. And this is to be done for the sole purpose of achieving justice.”[30]

[78] But here we have found that the decision was clearly wrong and would lead to an injustice. Nor have we sought to second guess the Tribunal’s economic expertise. To the extent that the Tribunal made an economic analysis about parallel conduct in oligopoly markets, in the passage from the paragraph in its reasons that we quoted earlier, we have accepted that analysis.[31] The problem is that the Tribunal did not follow it through in its own analysis.

[79] It is understandable that the Tribunal was concerned about the effect on the Sekunjalo Group of being ‘unbanked’ when access to banking is essential to being able to compete in a market. But that concern may well be addressed by the other challenges we understand the Group has made in terms of the Promotion of Equality and Prevention of Unfair Discrimination Act 4 of 2000 and the common law. This case is confined to the question of whether that refusal can be remedied under the Competition Act, and I find that based on the present record, it cannot. The conclusion therefore is that the matter should properly be determined on appeal. However, if I am wrong, then on the ground of irrationality and for the same reasons given earlier regarding the irrationality of the Tribunal’s decision, and it exceeding the lawful parameters of the relevant provisions of the Act, the appellants’ review of the Tribunal would fall to be upheld and it would be an act of supererogation to remit it back.

Lis alibi pendens.

[80] Sekunjalo also raised a defence of lis alibi pendens. Aside from bringing an application before the Tribunal it has also brought proceedings before the Equality Court and the ordinary High Court - in each instance apparently seeking interim interdictory or mandatory relief. What is surprising is that Sekunjalo raised this issue given that it was responsible for bringing all these applications.

[81] It therefore should not be open to Sekunjalo to raise lis alibi pendens, save possibly to the limited extent that the High Court proceedings also involve a review. However, that is a review of the bank’s

decision to terminate a contractual arrangement, while the review before us is that of the Tribunal’s decision. Moreover, even if the facts before each court now seized with the respondents’ applications are the same, the outcomes are based on different considerations.

[82] Before this court the issues are confined to considerations of whether the appellants conduct was anti-competitive. Those considerations fall within the exclusive jurisdiction of the Competition Tribunal and this court in terms of section 62(1) of the Act. Conversely, the Tribunal and this court do not have jurisdiction over the subject matter in the cases brought by Sekunjalo before the Equality Court and the High Court.

[83] Accordingly, the principles and law (whether circumscribed by statute or common law) to be applied before each forum will be different to this forum even if the result may ultimately be the same. [32]

Costs

[84] Costs must follow the result. The question is whether having decided that this matter can be decided as an appeal the respondents should also pay the costs of the review. The Act allows a party to bring both an appeal and a review. The fact that the review points were in a large part identical to those raised in the appeal should not mean that the appellants should be penalised.

[85] A court which enjoys both appeal and full review jurisdiction, as this court does, should not adopt a highly technical approach as it might result in failing to provide access to justice. [33]

[86] A court should also recognise the dilemma that faces a party who genuinely wishes to challenge both the regularity of the proceedings and the merits of the case. If it is driven by ulterior motives, then that will be readily discernible to a court and suitable orders of costs can be made to discourage similar abuse. This is not such a case.

N. Manoim

Judge President

I concur

D. Mlambo

Acting Judge of Appeal

ORDER

1. The appeal of the three appellants is upheld

2. The Tribunal’s orders in respect of paragraphs 1.2 (re Standard Bank), 1.5 (re Mercantile) and

1.8 (re Access Bank) are set aside.

3. The appellants are entitled to the costs of the appeal and review including the costs of two counsel where employed.

Spilg AJA (separate concurring)

[87] I have had the privilege of reading the decision of the Judge President and concur with the grounds set out for upholding the appeal, failing which in finding on the alternative basis relied on that the reviews fall to be upheld.

[88] There is a matter I wish to touch on briefly. It concerns the position of Standard Bank, including my willingness to have received the further evidence sought to be introduced regarding events after the parties had argued the case before the Tribunal.

[89] I am of the view that Sekunjalo acted precipitously in joining Standard Bank. At the time it brought the application, Standard Bank was still awaiting a response from Sekunjalo to a series of incisive questions arising from the Mpati Report and Sekunjalo’s earlier but clearly inadequate reply to previous enquiries from the bank.

[90] Of particular concern, bearing in mind the contents of the Mpati Report[34], were issues relating to verifying the identity of the ultimate beneficial owners, whether future transactions would be consistent with the bank’s knowledge of a prospective client having regard to the nature of the intergroup transactions and where the bank was unable to conduct ongoing due diligence in respect of the banking relationship with the client due to the Sekunjalo Group’s failure to adequately respond to the bank’s enquiries.

[91] A number of concerns also arise from Sekunjalo’s failure to have replied to Standard Bank before bringing its application. The more obvious one is that had it properly responded to the enquiries timeously and had Standard Bank decided not to continue providing banking services to Sekunjalo then the matter would have been argued on principled and not speculative grounds. Sekunjalo would then have demonstrated that either it could on a prima facie basis deal with the bank’s position, and in particular that there was no ground to raise FICA compliance red flags. Of course, had it done so there may have been a knock on effect because if Sekunjalo was unable to deal adequately with Standard Bank’s request for further information then its case against the other banks may also have been affected.

[92] Moreover, it was a simple matter for Sekunjalo to have proceeded against the other banks and then join Standard Bank once it had replied to the bank’s further, and quite legitimate enquiries, and received their response.

[93] I would therefore have found that the case was bought prematurely against Standard Bank. In bringing its case prematurely, Sekunjalo precluded a proper ventilation of the issues in relation to that bank. For this reason, it does not lie in Sekunjalo’s mouth to preclude the bank from introducing evidence of the former’s response to Standard Bank’s enquiries, the bank’s evaluation of the response and the processes it applied leading to its decision to terminate Sekunjalo’s banking services.

[94] These events occurred after the parties had presented argument before the Tribunal. Sekunjalo only had itself to blame for the inordinate delay (in marked contrast to the urgency with which Sekunjalo professed its need to approach the Tribunal on urgency) in responding to legitimate and highly relevant information it was requested to produce to enable Standard Bank to consider its position within the context of legislation affecting the conducting of its banking business within the four corners of such laws.

[95] The further evidence sought to be introduced was highly relevant and Sekunjalo was not prejudiced because it could have responded had it so wished. Had it been necessary to do so, I would therefore have allowed the introduction of this further evidence.

B Splig

Acting Judge of Appeal

Counsel for the First Appellant: Adv

G. Engelbrecht SC

Instructed by: Werksmans

Attorneys

Counsel for the Second Appellant: Adv S. Budlender SC, Assisted by Adv P.

Ngcongo and Adv I. Cloete

Instructed by: Herbert

Smith Freehills South Africa

Counsel for the Third Appellant: Adv

A. Subel SC, Assisted by

Adv A. Botha SC and Adv T. Marolen

Instructed by: Lawtons

Africa

Counsel for the First Respondent: Adv H. Maenetje SC Assisted by

Adv V. Ngalwana SC, Adv K. Monareng

And Adv E.N. Sithole

Date of hearing: 30 – 31 March 2023

Date of judgment: 17 July 2023

ANNEXURE “A”

SEKUNJALO INVESTMENT HOLDINGS (PTY) LTD 2nd Respondent AFRICAN EQUITY EMPOWERMENT INVESTMENT LTD 3rd Respondent AFRINAT (PTY) LTD 4th Respondent BIOCLONES (PTY) LTD 5th Respondent SEKPHARMA (PTY) LTD 6th Respondent ORLEANS COSMETICS (PTY) LTD 7th Respondent ESP AFRICA (PTY) LTD 8th Respondent PREMIER FISHING AND BRANDS LTD 9th Respondent PREMIER FISHING SA (PTY) LTD 10th Respondent MARINE GROWERS (PTY) LTD 11th Respondent TALHADO FISHING ENTERPRISES (PTY) LTD 12th Respondent AYO THECHNOLOGY SOLUTIONS LTD 13th Respondent HEALTH SYSTEMS TECHNOLOGIES (PTY) LTD 14th Respondent GLOBAL COMMAND & CONTROL TECHNOLOGIES (PTY) LTD 15th Respondent KALULA COMMUNICATIONS (PTY) LTD 16th Respondent KATHEA COMMUNICATIONS (PTY) LTD 17th Respondent SEKUNJALO PROPERTIES (PTY) LTD 18th Respondent 3LAWS CAPITAL (PTY) LTD 19th Respondent CAPE SUNSET VILLAS (PTY) LTD 20th Respondent SILO CAPE WATERFRONT (PTY) LTD 21st Respondent AFRICAN NEWS AGENCY (PTY) LTD 22nd Respondent SOUTH AFRICAN PRESS ASSOCIATION (PTY) LTD 23rd Respondent MAGIC 828 (PTY) LTD 24th Respondent INDEPENDENT NEWSPAPERS (PTY) LTD 25th Respondent INDEPENDENT MEDIA CONSORTIUM (PTY) LTD 26th Respondent SAGARMATHA TECHNOLOGIES LTD 27th Respondent LOOT ONLINE (PTY) LTD 28th Respondent SURVE` PHILANTHROPIES 29th Respondent SEKUNJALO DEVELOPMENT FOUNDATION TRUST 30th Respondent DR IQBAL SERVE` BURSARY TRUST 31st Respondent SOCIAL ENTREPRENEURSHIP FOUNDATION TRUST 32nd Respondent HARAAS TRUST 33rd Respondent LINACRE INVESTMENTS (PTY) LTD 34th Respondent KILOMAX INVESTMENTS (PTY) LTD 35th Respondent BUSINESS VENTURE INVESTMENTS NO 1126 (PTY) LTD 36th Respondent

[1] This was Investec Bank. The Tribunal explains that this was because the wrong entity had been cited; Investec Bank and not Investec

Securities which had terminated one of the Sekunjalo entities accounts.

[2] The Tribunal excluded from its order of relief the personal bank accounts of the first applicant a quo, Dr Iqbal Survé, who had accounts with Nedbank on the basis that these were personal accounts and were thus irrelevant to the ability of the Sekunjalo

Group firms to compete in the markets in which they operated.

[3] Mercantile in the course of the proceedings merged with Capitec Bank. Whilst this was initially made a joinder issue before the Tribunal, Mercantile proceeded to defend itself on the merits and we were not asked to decide this point. I have approached the decision on the basis that Mercantile has been properly cited.

[4] This is by way of operation of section 49(C)(5) which states: “If an interim order has been granted, and a hearing into that matter has not been concluded within six months after the date of that order, the Competition Tribunal, on good cause shown, may extend the interim order for a further period not exceeding six months.”

[5] Mpati report, paragraph 72. In paragraph 58 the Report also states: “The evidence gleaned from various bank statements show that there has been significant movement of the funds between different related parties. This created the impression of funds in bank accounts but, in reality, this was only the case at specific moments in time.”

[6] Access bank stated that it performed a screening exercise on this website for one of the entities AEEI and the result came back “alert threshold met.” Access states that this raised a red flag for it. Access Bank Answering Affidavit paragraphs 126-128.

[7] Since this is not in dispute, it is not necessary to deal with all the applicable legislation . The banks mentioned inter alia that they must comply with the requirements of the Prevention of Organised Crime Act 121 of 1998, the Prevention and Combatting of Corrupt Activities Act 2 of 2004, the Banks Act 84 of 1990, the Code of Banking Practice and the Basel Committees Guidelines on “Sound Management of Risks related to Money Laundering and Financing of Terrorist Activities”.

[8] Section 21E of FICA.

[9] 2010(4) SA 468 (SCA).

[10] Paragraph 131 of the decision .

[11] Paragraph 35.

[12] It is correct that Access Bank took on one of the entities, Afrinat, as a customer in May 2021, which was at a time when some of the other banks had refused to deal with the Sekunjalo Group. However, Afrinat had not disclosed to Access Bank that it was part of the Sekunjalo Group.

[13] Paragraph 139.

[14] The PIC was mentioned as being a client of the major banks. However, unlike the Sekunjalo entities which are all subject to the control of Dr Survé, the PIC is subject to external oversight and as a State-owned enterprise is subject ultimately to State control.

[15] A. Ahlström Osakeyhtiö and others v Commission of the European Communities. EU:C: 1993:120 paragraph 71.

[16] Section 7 states: “A firm is dominant in a market if— (a) it has at least 45% of that market; (b) it has at least 35%, but less than 45%, of that market, unless it can show that it does not have market power; or (c) it holds less than 35% of that market but has market power".

[16] Section 7 states: “A firm is dominant in a market if—

(a) it has at least 45% of that market;

(b) it has at least 35%, but less than 45%, of that market, unless it can show that it does not have market power; or

(c) it holds less than 35% of that market but has market power".

[17] This conclusion is based on the Commission’s Banking Enquiry report that was issued in June 2008.

[18] Tribunal reasons, paragraph 170.3

[19] Ibid, paragraph 171.

[20] Babelegi Industrial Workwear and Industrial Supplies CC v the Competition Commission, Case No. 186/CAC/JUN20

[21] The Tribunal mentions that there are approximately seventy registered banks in South Africa. Tribunal reasons paragraph 176.

[22] eMedia Investments (Pty) Ltd SA v Multichoice (Pty) Ltd and Another (201/CAC/Jun22).

[23] Whish and Bailey, Competition Law , (Oxford) Tenth edition, page 734.

[24] Paragraph 96.1

[25] This is the allegation referred to in paragraph 96.3 of the reasons.

[26] They were Orleans Cosmetics, Independent Newspapers and Loot Online.

[27] For this reason, I have not needed to deal with the issue of whether the additional evidence which Standard Bank sought to introduce should be admitted. Spilg AJA deals with this issue in his separate concurring decision.

[28] Business Connexion (Pty) Ltd v Vexall (Pty) Ltd and another [2020] 2 CPLR 490 (CAC) (“Business Connexion”) para 41

[29] International Trade Administration Commission v SCAW South Africa (Pty) Ltd 2012 (4) SA 618 (CC) (“ITAC v SCAW”) at para 53.

[30] Competition Commission of South Africa v Mediclinic Southern Africa (Pty) Limited 2021 JDR 3149 (CC); 2022(4) SA 323 (CC).

[31] Paragraph 131 of the Tribunal’s reasons.

[32] Compare Visagie v Health Professions Council of South Africa and Others 2023 (2) SA 626 (GP) at para 15 where in an analogous situation Potterill J said that: The approach by the two presiding officers, and the principles applied in the review and the appeal, would thus be wholly different, albeit the result may be the same. I am thus satisfied that lis pendens herein is not applicable

[33] See the observations in Liberty Life Association of Africa v Kachelhoffer NO 2001 (3) SA 1094 (C) at 1111C-E. This case also acknowledged circumstances where a party may appeal on the merits of the decision and still bring a review where grounds for doing so also existed.

[34] See for instance the extracts mentioned in para 13 and footnote to the main judgment.