A question that arises regularly in commercial practice is when ordinary contractual restraints – such as non-competes, exclusivity agreements, supply commitments, settlement terms and joint venture obligations – risk characterisation as horizontal market division under the South African Competition Act 89 of 1998 (as amended) (Act).
The issue is not always obvious, as restraints of this nature are often included in ordinary commercial agreements for legitimate reasons, including the protection of confidential information, transaction-specific investment, goodwill, know-how or the integrity of a broader commercial arrangement.
However, where a restraint operates in substance to allocate customers, suppliers, territories or specific goods or services between actual or potential competitors, it may attract scrutiny under section 4(1)(b)(ii).
The regulatory framework
Section 4(1)(b) of the Act prohibits a defined set of horizontal practices between firms in a horizontal relationship, including the allocation of customers, suppliers, territories or specific goods and services.
The prohibition is per se, in that there is no rule-of-reason inquiry, and unlike section 4(1)(a) - no opportunity to demonstrate offsetting efficiencies or pro-competitive effects. Once the conduct is properly characterised as falling within section 4(1)(b)(ii), liability follows without further balancing.
The contested issues, in practice, are accordingly threefold:
- whether there is an agreement, decision or concerted practice between the parties;
- whether the parties are in a horizontal relationship; and
- whether the restraint amounts in substance to market allocation rather than legitimate commercial cooperation.
Each of these is a matter of substance over form – a principle the Competition Tribunal and Competition Appeal Court have consistently reinforced.
A clause does not escape section 4(1)(b)(ii) merely because it is housed in a supply contract, a distribution agreement, a settlement, a licence or a joint venture document, and an arrangement need not be reduced to writing to qualify as an agreement or concerted practice for purposes of the Act.
Of these three elements, the characterisation of the parties’ relationship is often most contended – and the most consequential – in practice.
Nicholas De Decker 16 Feb 2026 When are parties “horizontal”?
This is the analytical heart of most cases of this nature, as a relationship can be vertical in one respect and horizontal, or potentially horizontal, in another.
A supplier, distributor, technology partner, licensee or joint venture participant may also be an actual competitor or potential competitor of its counterparty if it has the capability, incentive and reasonable strategic option to enter the adjacent market.
The “potential competitor” maxim matters particularly here, as a firm need not currently sell into a market to be a potential competitor; it is enough that, but for the restraint, the firm could realistically have entered within a reasonable timeframe, with the assets, capabilities and intent to do so.
The factors that typically inform this assessment include the firm’s existing technical and operational capabilities, the barriers to entry in the relevant market, the firm’s financial resources and strategic priorities, and any contemporaneous evidence of an intention to enter, such as internal strategy documents, board approvals, product development activity or commercial overtures to prospective customers.
A clause does not lose its character merely because of the document in which it appears, and the question is not how the parties describe their relationship in the recitals but how it looks when one asks whether each party could realistically have competed with the other absent the restraint.
Understanding whether parties are horizontal is only part of the inquiry. The next question is where, in the architecture of a commercial transaction, market-division risk is most likely to arise.
Where the risk hides
Market-division risk in commercial contracts is rarely labelled as such, and it tends to surface in clauses serving other apparent purposes, such as:
- A manufacturing agreement that prevents a supplier from entering its customer’s downstream market;
- A distribution arrangement that allocates customers between firms that could otherwise compete;
- A settlement agreement that resolves a dispute by one party agreeing to exit a market;
- A technology licence that prevents the licensee from developing a competing platform; or
- An outsourcing arrangement that constrains the service provider from serving certain competitors or from building its own offering.
None of these structures are inherently unlawful, and each requires a competition law overlay that asks whether, in substance, the restraint allocates markets, customers, suppliers or services between actual or potential competitors.
The evidentiary significance of internal correspondence and contemporaneous deal documentation should not be underestimated, as the way in which a restraint is described in negotiation correspondence, board memoranda and internal communications often informs how it is later characterised.
For example, language framing a restraint in terms of parties “staying in their lanes” or “not stepping on each other’s toes” can be prove deeply problematic at the evidentiary stage, revealing an allocative intent that more carefully calibrated contractual drafting cannot retrospectively cure.
The ancillary restraints question
The proper test is thus not whether a restraint exists, but whether it is genuinely ancillary to a legitimate purpose and proportionate to the protection that purpose requires.
For instance, a restraint protecting confidential information, the value of a business sold as a going concern, transaction-specific investment or the integrity of a bona fide collaboration is more readily defensible than one that prevents entry into a market unrelated to the core transaction, lasts longer than necessary, sweeps across overly broad territories or customer groups, or shields a party from competition generally.
The questions to ask are therefore:
- what legitimate transaction does this restraint support;
- would the transaction proceed without it;
- is it limited in duration, territory and scope to what is reasonably necessary; and
- is there a less restrictive way to achieve the same objective?
The proportionality of each limb of the restraint matters, and a restraint that is broader than necessary in any one respect (whether duration, geographic reach, product scope or customer category) will be more difficult to defend, even where its underlying purpose is legitimate.
A restraint that cannot survive these questions is unlikely to be defensible as ancillary.
Nicholas De Decker 18 Oct 2024 Practical implications
For in-house counsel and transaction advisors, four key points carry through:
- Firstly, the form of the agreement in question is not determinative – as section 4(1)(b)(ii) of the Act looks to substance, and a clause embedded in a supply, distribution, settlement or licensing agreement will be analysed on the same basis as a standalone restraint;
- Secondly, contemporaneous documentation matters, and articulating the legitimate purpose of a restraint in the agreement and in the deal structure is far more credible than reconstruction after the fact;
- Thirdly, legacy agreements warrant attention, as conduct under agreements that have since terminated can become the subject of regulatory scrutiny long after the fact, and a periodic review of long-term supply, distribution, technology, settlement and joint venture arrangements is a sound governance practice; and
- Fourthly, where a problematic restraint is identified through such a review, early engagement with experienced competition counsel is advisable, and the Competition Commission’s Corporate Leniency Policy provides a structured route for parties wishing to disclose participation in prohibited horizontal conduct, and may, depending on the circumstances, mitigate or eliminate firm-level exposure to administrative penalties.
Conclusion
The central point is that competition law risk in South Africa is not confined to obvious cartel conduct. It can also arise from restraints embedded in ordinary commercial agreements, particularly where a supplier, distributor, customer, licensee or strategic partner is also an actual or potential competitor.
Drawing the line correctly – at the drafting stage, in contemporaneous documentation and across the full portfolio of commercial agreements – is the most reliable way to keep legitimate commercial restraints on the right side of the Act, and to avoid the firm-level and individual consequences that may follow from a contrary finding.
