BNP Paribas v Competition Commission

The Constitutional Court’s decision in the BNP Paribas / Competition Commission matter arose from allegations that several banks – initially 18 were identified – colluded to manipulate the USD/ZAR exchange rate from 2007 until at least September 2013. The case traversed the Tribunal, the Competition Appeal Court (CAC) and ultimately the Constitutional Court over a period of nearly a decade, largely on preliminary jurisdictional and pleading issues. For deal teams, the case is a useful reminder that competition-law risk in South African transactions may arise from conduct outside South Africa where that conduct is alleged to affect a South African market.

The jurisdictional anchor is section 3(1) of the Competition Act, which applies to all economic activity within South Africa or having an effect within South Africa. In this litigation, however, “effect” was not treated as a loose concept. The operative framework required “qualified effects” – i.e. that it was foreseeable that the prohibited conduct would have a direct or immediate, and substantial, effect in South Africa (known as the QE test). In addition, for foreign entities without local domicile (so-called “peregrini”), personal jurisdiction required proof of “adequate connecting factors” between the Commission’s complaint and the Tribunal as a forum (the ACF test). In practice, the key connecting factor was whether the foreign bank participated in a conspiracy that included South African banks. In transactional terms, the diligence question is not only where the conduct occurred, but whether it may have affected South African customers, pricing, trading conditions or market outcomes in a way that could satisfy those jurisdictional requirements.

The Court drew an important distinction between different categories of foreign respondent. “Pure peregrini” – foreign firms without any branch or representative office in South Africa – faced the highest jurisdictional bar: the Commission had to show both that the impugned conduct had qualified effects in South Africa and that the firm had adequate connecting factors to South African jurisdiction. “Local peregrini” – foreign firms with a local branch – were treated more favourably for personal jurisdiction but still needed to show qualified effects for subject matter jurisdiction. For deal teams assessing a target with offshore operations, understanding where the target’s entities sit on this spectrum can be important in sizing competition-law exposure.

If section 3(1) explains why offshore conduct can still be relevant, section 4(1)(b) frames the substantive risk. It prohibits restrictive horizontal practices such as price-fixing, market division and collusive tendering. In a deal context, historic exposure to those issues may affect warranties, indemnities, disclosures, and valuation.

The Court also confirmed that a holding company cannot be included as a respondent merely because it is the parent of an implicated bank. The CAC held that where the Commission’s allegations were directed at the subsidiary’s conduct, the case could not be sustained against the holding company on that basis alone. In a deal context, this is relevant to how competition risk is allocated among entities within a group: the mere existence of a parent-subsidiary relationship does not automatically extend competition liability upward.

The judgment is especially useful for transactional diligence because the alleged conduct was not said to arise only from formal agreements. The Commission relied heavily on trader communications, including Bloomberg chatrooms such as the “Old Gits” and “ZAR” chatrooms. It alleged that these chatrooms were a primary mode of communication between traders and that messages were visible to participants whether they were actively posting or passively present. The Court catalogued some 162 specific incidents of alleged communication over the period 2007 to 2013, organised under headings such as sharing of information on bid-offer spreads, arrangements to coordinate trading, manipulation of the USD/ZAR rate at the FIX, and sharing of competitively sensitive information.

This changes the due diligence focus from formal documentation to the way employees actually communicate. Competition risk may sit in informal communications, shared platforms and exchanges of market-sensitive information. However, the judgment also highlights a critical evidential boundary: the Commission must be able to distinguish between information that is publicly available – for example, indicative exchange rates posted on the Reuters information platform, visible to all market participants – and information that could only be the product of collusive activity. In higher-risk sectors, deal teams should look beyond policies and standard warranties and consider how employees communicate with competitors, how pricing information is shared, and whether communication channels create section 4(1)(b) exposure. As a practical matter, diligence should also consider whether the target can credibly distinguish between legitimate market data and communications that evidence coordination.

The significance of the chatroom allegations was amplified by their link to a broader alleged “single overarching conspiracy” (SOC) to manipulate and distort normal competitive conditions in trading the USD/ZAR currency pair. Drawing on EU competition law, the Court endorsed three requirements for an SOC: (1) a common anti-competitive objective, being the existence of an overall plan pursuing a common economic objective; (2) each firm’s intentional contribution by its own conduct to the common objective; and (3) that the firm was either aware of the actual conduct planned or put into effect by other firms in pursuit of the same objective, or could reasonably have foreseen it and was prepared to take the risk. The Court further confirmed that a firm may be held liable for a single and continuous infringement even if it did not participate in all of the offending conduct, provided it had actual or constructive knowledge of the wider scheme.

The Court made it clear that isolated chatroom participation is unlikely, on its own, to establish involvement in a single overarching conspiracy. Repeated communications with competitors, discussions about customer quotes or spreads, coordinated conduct, common communication platforms and links between employees across entities present a different picture and warrant closer scrutiny during due diligence. If offshore conduct may have affected a South African market, potential section 4(1)(b) exposure should form part of the competition diligence exercise.

Prescription is another issue that should be tested during due diligence. Section 67(1) bars complaints relating to prohibited practices that ceased more than three years before the Commission initiated its complaint. The Court confirmed that the relevant date is the Commission's original initiation, not any subsequent joinder or amendment, and that additional respondents may be joined after referral where their conduct falls within the scope of that original complaint. Historic conduct may therefore continue to create competition law exposure, which should be reflected in the transaction's warranty and indemnity framework.

The judgment also reinforces the importance of finality in competition proceedings. The Commission's attempt to revisit the interpretation of section 3(1), which had already been determined in an earlier Competition Appeal Court decision (CAC I), was barred by the principles of res judicata and peremption. The Constitutional Court held that, having elected not to appeal and having complied with the earlier judgment, the Commission was bound by it. For parties involved in competition proceedings, strategic decisions, including whether to appeal an adverse ruling, can have lasting implications for future enforcement action.

Related theme: merger notification, prior implementation and penalty exposure

A separate but related transactional theme is merger-control compliance. The Competition Commission’s March 2019 guidelines on administrative penalties for failure to notify mergers and prior implementation were issued under section 79(1) of the Competition Act and are intended to provide transparency, certainty and objectivity in the Commission’s approach to penalties. They are not binding on the Commission, the Tribunal or the CAC, and do not fetter their discretion, but they give deal teams a practical enforcement yardstick when assessing competition-law risk in transactions.

The statutory starting point is straightforward: parties to an intermediate or large merger must notify the Commission in the prescribed manner and may not implement the merger until it has been approved, with or without conditions, by the relevant competition authority. The Tribunal may impose an administrative penalty where parties fail to give the required merger notice, or proceed to implement a notifiable merger without the required approval.

The penalty methodology is also important for transaction planning. The Commission first considers the nature of the contravention, then uses a base amount equal to double the applicable filing fee, adds a duration-based amount, adjusts for aggravating or mitigating factors, and then applies the statutory cap where necessary. For contraventions of up to one year, each month attracts an additional 50% of the base amount; for contraventions exceeding one year but less than two years, 75% of the base amount is added per month; and for contraventions exceeding two years, 100% of the base amount is added per month.

The guidelines also highlight a sharper enforcement risk for deliberate conduct. If the relevant conduct amounts to a section 4(1)(b) contravention, such as competitors agreeing prices or allocating customers before merger approval, the Commission will assess that conduct under section 4(1)(b) and determine any penalty under its prohibited-practices penalty guidelines. If the Commission considers the conduct wilful or deliberate, the merger-penalty guidelines will not apply and the Commission may seek the maximum allowable penalty as well as divestiture where appropriate.

For deal teams, the practical risk is often not the signing of the sale agreement itself, but conduct that looks like pre-approval control or integration. The guidelines list examples including involvement in day-to-day operations, marketing the merging parties as a single entity before approval, integration or consolidation of operations, influence over strategic decisions, pricing, customer relations, procurement, production, product development, appointment of directors, receipt of profits or other payments connected with the transferred firm’s performance, and non-refundable advance purchase-price arrangements that amount to material influence. They also recognise that planning steps for post-merger integration are different from implementing that planning before approval is obtained.

That has direct drafting and diligence consequences. Conditions precedent, interim covenants, information-sharing protocols, clean-team arrangements, board-observer rights, veto rights, payment mechanics and integration-planning provisions should be tested not only against ordinary deal protection concerns, but also against whether they could confer material influence or amount to implementation before approval. A buyer should be able to show that it sought competition-law advice, approached the Commission proactively where necessary, co-operated with the authorities, acted transparently, and preserved evidence supporting any bona fide failure to notify, because those matters may be mitigating factors in the Commission’s penalty assessment.

Settlement and financial capacity are relevant, but should not be treated as a substitute for compliance. The Commission may, at its sole discretion, offer a settlement discount of up to 50%, with earlier settlements generally more likely to attract a greater discount. It may consider ability to pay only in exceptional circumstances, on objective evidence that the penalty would irretrievably jeopardise the firm’s economic viability and cause it to exit the market; a loss-making position alone may not be enough. In most deals, therefore, the better risk allocation is preventative: identify notifiable control changes early, maintain the standstill obligation, and document the basis for any conclusion that notification is not required.