Brief Summary

Competition — Merger Approval — Holdco acquiring Lanseria International Airport and Execujet Airline Investments — Competition Tribunal approving merger with conditions to mitigate potential anti-competitive effects due to cross-shareholding by Public Investment Corporation — Tribunal finding no significant lessening of competition likely as a result of merger, despite concerns over information sharing and potential unilateral control.

Comprehensive Summary

Summary of Judgment

1. Introduction

These proceedings concerned a merger approval before the Competition Tribunal of South Africa in terms of the Competition Act 89 of 1998. The Tribunal was required to determine whether the proposed transaction raised competition concerns and, if so, whether those concerns could be addressed through conditions.

The acquiring firm was Holdco, a company to be incorporated and not yet trading at the time of the proceedings. The target firms were Lanseria International Airport (Pty) Ltd and Execujet Airline Investments (Pty) Ltd (together referred to in the reasons as the entities through which the Lanseria airport business would be acquired). The composition of Holdco’s shareholders was central to the analysis, particularly the role of the Public Investment Corporation SOC Limited (PIC), which held interests both in Holdco (and thus the target) and separately in Airports Company South Africa Limited (ACSA).

The matter came before the Tribunal following the Competition Commission’s recommendation in terms of section 14A(1)(b) of the Act. The merger was heard and decided on 29 May 2013, with reasons issued on 26 July 2013 and non-confidential reasons issued on 1 October 2013. The Tribunal ultimately approved the merger subject to conditions, directed at addressing risks linked to cross-shareholding and information flows.

The dispute’s general subject-matter was the competitive interaction between Lanseria International Airport (LIA) and OR Tambo International Airport (ORTIA) (owned through ACSA), and the potential for the PIC’s cross-interest to generate either unilateral incentives (through joint profit maximisation) or co-ordinated effects (through information exchange and governance links).

2. Material Facts

Holdco, the acquiring vehicle, had three shareholder groupings: the PIC, the Pan African Infrastructure Development Fund (PAIDF), and a BEE consortium. The PIC’s stake in Holdco was 37.5%, with PAIDF holding 37.5% and the BEE consortium 25%. Through the transaction, the PIC would thus hold an effective 37.5% interest in the Lanseria business post-implementation (as described in the reasons, including the contemplated restructuring).

Separately from the transaction, the PIC also held an interest in ACSA, amounting to 20%, with rights to appoint 3 out of 12 directors. ACSA’s majority shareholder was the Department of Transport (approximately 74.6%). On the facts accepted by the Tribunal, the PIC did not have sole control of ACSA, but it had joint control together with the Minister of Transport, coupled with a significant financial interest.

The transaction structure was that Holdco would acquire 50% of the shares in LIA and 100% of the shares in Execujet. After implementation, Holdco could restructure the assets so that LIA’s property and business (other than shares held in Lanseria Airport) would be transferred to a new wholly-owned subsidiary of Holdco (referred to as “Opco” in the reasons).

A contextual fact relied upon by the Tribunal was that LIA and ORTIA were not identical offerings. Both provided general and scheduled aviation services, but LIA’s scheduled services were limited to domestic and regional flights, whereas ORTIA also offered long-haul international flights and interlining capabilities.

On the question of control over LIA post-transaction, the merging parties contended (and the Commission agreed) that the PIC’s shareholding would not confer de jure control and that the funding arrangements were said to be arms-length and not to confer special control rights if other shareholders complied with funding obligations. The Tribunal stated that it would assume (without having reviewed the funding agreements) that these contentions were correct for purposes of its analysis. However, the Tribunal also considered that, even absent de jure control, factual circumstances (including the PIC’s role as funder, the relative size and dependence of minority shareholders, and the PIC’s expertise) could support a probability of de facto unilateral influence over LIA if the PIC wished to exercise it.

The Tribunal relied on several facts to assess incentive effects. It accepted that ORTIA’s tariffs for regulated services were constrained because such tariffs are regulated by the Airports Regulator, meaning a strategy to shift demand via ORTIA price increases would not be feasible through ACSA-controlled regulated tariffs. It further relied on the distinction between ORTIA’s outsourcing model and LIA’s integrated service provision: ORTIA’s unregulated services were largely provided by third parties or airlines, and these unregulated services were said to comprise a very large portion (approximately 87% to 93%) of an airline’s operating costs at ORTIA, whereas LIA itself provided many services that ORTIA outsourced.

The Tribunal also relied on the fact that LIA’s scheduled domestic customer base was very limited, at the time comprising only two commercial customers (airlines) for scheduled domestic airline flights, and that a significant tariff increase at LIA risked the loss of those customers rather than partial diversion.

Finally, a further material fact was that the PIC had committed, under transaction-related agreements (identified in the reasons as the Sale of Shares and Claims Agreement and the LIA Group Relationships Agreement), to contribute to substantial investment in LIA, described as having a collective value of approximately R1.1 billion, including infrastructure expansion (apron expansion and construction of a new terminal, control tower, fire station, and runway). The Tribunal treated this investment posture as relevant to assessing likely incentives post-merger.

3. Legal Issues

The central legal questions were whether the merger was likely to result in a substantial lessening or prevention of competition within the meaning of section 12A(1) of the Competition Act, and if so, whether the competition concerns could be adequately addressed through merger conditions under section 16(2)(b).

The dispute primarily concerned the application of competition-law principles to a specific ownership and governance structure. It turned on mixed questions of fact and law, including an assessment of (i) the nature and implications of minority shareholdings and board appointment rights, (ii) incentive effects arising from cross-shareholdings in competing assets, and (iii) the risk of co-ordination through information exchange mechanisms.

Two theory-of-harm questions framed the Tribunal’s analysis. The first was whether the PIC’s cross-interest could create a unilateral incentive for LIA to raise tariffs in a manner consistent with joint profit maximisation, on the basis that lost customers might divert to ORTIA (in which the PIC also had a financial interest). The second was whether the PIC’s presence across both investments could create co-ordinated effects through the exchange of competitively sensitive information, particularly through board representation and internal reporting structures.

To the extent the Tribunal considered whether the PIC might have de facto control over LIA, that question was relevant as background to the incentive analysis, but the Tribunal expressly indicated it was not necessary to decide conclusively whether de facto control would exist in order to resolve the competition assessment.

4. Court’s Reasoning

The Tribunal approached the analysis by identifying that the merger’s competition concern was not a conventional horizontal overlap between the acquiring and target firms as operating competitors, but rather the PIC’s cross-shareholding linking interests in ACSA/ORTIA and LIA. The Tribunal therefore examined whether that link could change competitive incentives (unilateral effects) or enable co-ordination (co-ordinated effects).

On the issue of control, the Tribunal accepted (for purposes of analysis) the merging parties’ and Commission’s position that the PIC would not acquire de jure control through its 37.5% stake, and that the funding arrangements were said not to confer formal control rights. At the same time, the Tribunal reasoned that de jure control was not the sole lens: it considered that the PIC could potentially exercise de facto control given its funding role, the relative size and sophistication of minority investors, possible dependence on the PIC beyond this transaction, and the PIC’s sector expertise. The Tribunal nonetheless treated the decisive question as whether competition harm would likely arise even if such influence existed.

In assessing the joint profit maximisation concern, the Tribunal reasoned that the strategy hypothesised would require the PIC to benefit from demand diverted from LIA to ORTIA. However, a key constraint identified was that ORTIA’s regulated tariffs could not be increased at will because they are regulated by the Airports Regulator, limiting one dimension of profit recapture through ACSA. The Tribunal also reasoned that the PIC’s shareholding was smaller in ACSA (20%) than in LIA (37.5%), reducing the strength of any incentive to sacrifice profits at LIA for recaptured profits via ACSA.

A further major consideration was the difference in revenue capture between the airports due to outsourcing versus insourcing of services. The Tribunal accepted that ORTIA’s unregulated services were largely provided by third parties or the airlines themselves, whereas LIA provided many services directly. It treated this structural difference as undermining the profitability of any diversion strategy, because a substantial portion of aviation-related costs and associated revenues at ORTIA would not accrue to ACSA in the same way that comparable revenues could accrue to LIA where services are provided internally.

The Tribunal also evaluated demand-side fragility at LIA. With only two scheduled domestic airline customers, the Tribunal reasoned that meaningful tariff increases at LIA were more likely to cause the airport to lose those customers altogether rather than retain some business at higher margins. In that scenario, the mechanism underlying joint profit maximisation—earning more from remaining customers while recapturing some diverted demand through a stake in the rival—would be unlikely to operate effectively because there might be no retained customer base at LIA from which to extract higher margins.

The Tribunal also placed significant weight on the PIC’s committed investment programme in LIA (approximately R1.1 billion), which contemplated expanding and upgrading facilities. The Tribunal treated these commitments as inconsistent with a diversionary or harvest strategy; the investment plans were characterised as aimed at expanding capacity and increasing traffic, which would not align with a theory predicting the deliberate raising of tariffs at LIA to push demand toward ORTIA.

On the second concern—information exchange and cross-directorships—the Tribunal treated the possibility of sharing competitively sensitive information between rival airports (in respect of certain services) as a meaningful risk. However, it accepted that this risk had been addressed through undertakings that were incorporated as conditions to the merger approval. These conditions were designed to ensure separation of the PIC’s oversight structures for the two investments, restrict common directorships, and constrain how competitively sensitive information could be communicated internally within the PIC, including by aggregation and closed sessions.

In sum, the Tribunal concluded that even if the PIC could, as a practical matter, exert significant influence over LIA, the evidence and structural features relied upon made it unlikely that the merger would result in a significant lessening of competition through altered incentives. Any residual co-ordination risk via information flows was treated as manageable through the imposed conditions.

5. Outcome and Relief

The Tribunal approved the merger with conditions in terms of section 16(2)(b) of the Competition Act. It directed that a Merger Clearance Certificate be issued in terms of Competition Tribunal Rule 35(5)(a).

The conditions (set out in Annexure A to the order) focused on preventing common directorships between ACSA and LIA while the PIC retained director-appointment rights at ACSA, implementing internal separation of the PIC’s investments into different divisions with safeguards, limiting the reporting of competitively sensitive information, and placing specific restrictions relating to information about airline negotiations and potential acquisition of BEE consortium shares in the event of default. The conditions also included compliance monitoring obligations, including an affidavit by the PIC’s Chief Executive Officer confirming implementation of specified undertakings within a set timeframe, and ongoing notification obligations to the Commission upon disposal of interests.

The order as provided in the judgment text did not reflect a distinct costs order as between the parties.

Cases Cited

No reported cases were cited in the judgment text provided.

Legislation Cited

Competition Act 89 of 1998 (as amended), including section 12A(1), section 14A(1)(b), and section 16(2)(b).

Rules of Court Cited

Competition Tribunal Rule 35(5)(a).

Rule 39 of the Rules for the Conduct of Proceedings in the Competition Commission (as referenced in the conditions for dealing with an apparent breach).

Held

The Tribunal held that the merger between Holdco and the Lanseria International Airport/Execujet target firms should be approved subject to conditions, because the potential competition risks arising from the PIC’s cross-shareholding in ACSA (associated with ORTIA) and its interest in LIA were not likely to result in a substantial lessening of competition on the facts relied upon, and because the principal residual risk—information exchange and governance overlap—could be addressed through enforceable conditions.

It further held that conditions restricting common directorships, requiring internal separation of PIC investment oversight structures, and limiting the handling and dissemination of competitively sensitive information were appropriate to address the identified concerns, together with monitoring and notification measures.

LEGAL PRINCIPLES

The Tribunal applied the principle that merger assessment under the Competition Act requires an evaluation of whether a transaction is likely to lead to a substantial lessening or prevention of competition, including through unilateral and co-ordinated effects, and that these effects may arise not only from direct operational overlaps but also from ownership links and governance arrangements that can alter incentives or facilitate co-ordination.

It applied the analytical concept that a minority interest in a competitor may, in appropriate circumstances, create incentives aligned with joint profit maximisation, but that such a theory must be evaluated against market structure and revenue mechanisms, including the relative size of shareholdings, the presence of regulation (such as regulated tariffs), and how revenues are captured in practice (including differences between outsourced and in-house service provision).

It also applied the principle that competition risk may arise from information flows and governance connections (including directorships and internal reporting structures) where firms are competitors in respect of some services, and that such risks can be addressed through behavioural conditions designed to prevent the dissemination of competitively sensitive information, ensure separation of oversight structures, and restrict governance overlap, supported by monitoring and compliance mechanisms.

Holco v Lanseria International Airport (Pty) Ltd and Another (016261) [2013] ZACT 99 (1 October 2013)