Monday, September 01, 2008

Striking the right balance between regulation and the creation of a merger- friendly environment can be – and, indeed, frequently is – tricky.
More often than not, the South African competition authorities tend to succeed. Yet the Competition Tribunal’s approach to the notion of control, which triggers a merger notification to the Competition Commission, is reluctant to stray from the letter of the law and casts the net extremely widely.
The result, evident from several cases, is that firms are often required to notify mergers, going through the rigmarole of a merger assessment, which fail to result in an acquisition of control in the true sense and consequently do not alter the market structure.
Merger control is designed to maintain a competitive market structure. In healthy commercial markets, characterised by the fluid transfer of capital and assets, merger control avoids market structures that could create or strengthen market power.1
In Distillers Corporation (SA) and Another/Bulmer (SA) and Another, the Competition Appeal Court noted: “…merger control…allow[s] the relevant competition authorities to examine a wide range of transactions which could result in an alteration of market structure and in particular reduce the level of competition in the relevant market.”
Consequently, control is only triggered when the transaction presents the essential features of a merger – that is, according to section 12(1)(a) of the Competition Act No. 89 of 1998, when “one or more firms directly or indirectly acquire or establish direct or indirect control over the whole or part of the business of another firm”.
The acquisition of control is, therefore, the trigger which could potentially alter the competitive landscape, requiring the merger to be assessed prior to implementation. According to section 12(2), control is deemed to exist when:
“(2) A person controls a firm if that person:

(a) beneficially owns more than one half of the issued share capital of the firm;
(b) is entitled to vote a majority of the votes cast at a general meeting of the firm, or has the ability to control the voting of a majority of those votes, either directly or through a controlled entity of that person;
(c) is able to appoint or to veto the appointment of a majority of the directors of the firm;
(d) is a holding company and the firm is a subsidiary of that company as contemplated in section 1(3)(a) of the Companies Act, 1973 (Act No. 61 of 1973);
(e) in the case of a firm that is a trust, has the ability to control the majority of the votes of the trustees or to appoint or change the majority of the beneficiaries of the trust;
(f) in the case of a close corporation owns the majority of members’ interest or controls directly or has the right to control the majority of member’s votes in the close corporation; or
(g) has the ability to materially influence the policy of the firm in a manner comparable to a person who in ordinary commercial practice, can exercise an element of control referred to in paragraphs (a) to (f).  

Section 12(2) only provides guidance on instances of control, which are not exhaustive of the circumstances in which control arises. Indeed, the acquisition of control may extend to any circumstances satisfying the definition of a merger in section 12(1)(a).
The Tribunal, however, takes a literal approach to the assessment of control, tending to assess the acquisition of control on the basis that each instance identified in section 12(2) constitutes a “bright line” of control, which, once crossed, triggers a notification requirement to the Commission.
Legal advisors thus face the difficulty of advising whether a transaction is required to be notified in circumstances where although a “bright line” in section 12(2) has been crossed, a de facto change in control has not taken place and the structure of the market remains unaltered.
This approach, which sacrifices substance for legal form, contradicts long-established South African legal principles, whereby courts examine the substance of transactions and “will not be deceived by the form of the transaction: it will render aside the veil in which the transaction is wrapped and examines its nature and substance”. 2
In the context of competition law, this could spell the premature or unnecessary assessment of transactions, which de facto do not fall within the definition of a merger.
The Tribunal’s policy is that notification is required to be as extensive as possible. According to the Tribunal in Bulmer, if this were not the case, mergers that might adversely affect the market could go undetected because the section 12 jurisdictional barriers had been set too high.
Nevertheless, its substantive analysis found: “…a change in direct control presumptively triggers the obligation to notify. However, we recognise that a limited class of transactions exists where that obligation may be negated if there is irrefutable evidence that indirect control remains unaffected. This is the case of firms who form part of a single economic unit, because the change in the direct form of control is illusory and has not altered the substance of control that both antedated and postdates the transaction”.
Subsequently, however, the Tribunal strayed from this analysis. In Ethos, it found that Ethos’s acquisition of additional shares in Tsebo, which would increase its shareholding to just above 50,1%, constituted an acquisition of sole control, which was required to be notified.
Tsebo was already jointly controlled by Ethos and two other shareholders, Siphumelele and Nozala, through an arrangement that had been notified to the Commission a few years before. Although the acquisition of the additional shares by Ethos amounted to control in terms of section 12(2)(a), Tsebo’s de facto control structure remained unchanged post transaction, because prior to the transaction, the assent of 67% of the shareholders was required for all material decisions. Post transaction, this requirement remained unchanged.
Consequently, with just more than 50,1% of Tsebo, Ethos was unable to influence Tsebo’s day-to-day activities alone. The market structure in which Tsebo operated would have remained unchanged, negating the need for a competitive assessment.
This does not suggest that parties should self assess their transactions. Bear in mind that Ethos was a merger in which the acquisition of joint control had already been notified to the Commission three years prior, when Tsebo was first acquired.
The refusal to consider the factual scenario in Ethos must be rationalised against the Tribunal’s approach in Caxton, when it did undertake a factual enquiry and assessed the de facto control structure of the transaction to ascertain that the transaction did not amount to a merger as defined in the Act..
In Caxton, the Tribunal found that an acquisition of additional shares by Naspers in M-Net, in terms of which Naspers’s direct and indirect interests cumulatively exceeded 50,1% of M-Net’s shares, did not amount to an acquisition of control.
The Tribunal found that as Naspers’s indirect interest of 26% in M-Net was the subject of joint control, Naspers did not, on its own, control the indirect holding in M-Net. Consequently, it would not have been in a position to exercise economic and political control over the whole interest.
Although we agree with the Tribunal’s reasoning, we disagree with the Tribunal’s subsequent finding that an increase in shareholding from 51% to 75% did not give a qualitatively new form of control to that which was previously there.
It found that “[i]f that is the case, then every company that has control at 51% and increases that control to beyond 75% must notify the Commission of that transaction”.
However, in considering Ethos, an acquisition of 75% of the shares would have resulted in Ethos having sole control of Tsebo, as Ethos alone would have met the 67% voting requirement for all material decisions. 3
Ethos would have been in a position to direct Tsebo according to its discretion alone, whereas with 51% of Tsebo, Ethos only had joint control. In that case, an increase in its shareholding from 51% to 75% would have been a qualitatively new form of control, which would have impacted the structure of the market differently to the form of control which Ethos enjoyed jointly with the other shareholders.
Subsequently, in Cape Empowerment Trust v Sanlam & Sancino Projects, the Tribunal maintained that the sum of the par value of a shareholder’s ordinary shares and preference shares would amount to the acquisition of control in terms of section 12(2)(a) if the sum exceeded 50,1% of the issued share capital of the target firm.
This, even though preference shares do not confer on the holder the right to control the company on a day-to-day basis or have any say in its affairs unless it attempts to change the rights attaching to the preference shares or fails to pay the preferred dividend.4
The Tribunal’s judgment suggests that an aggregation of the ordinary and preference shares triggers the Act’s notification requirements, even though such shareholding does not amount to a de facto acquisition of control, and that the holder of the right may have no say whatsoever in the day-to-day activities of the target firm.
The cases illustrate that the Tribunal takes a formalistic and arithmetic approach to the notion of control. Consequently, the Commission is required to investigate “mergers” in which there has not been an actual acquisition of control, but merely an illusory acquisition of control.5
It’s an approach whereby the Commission may render its investigation prematurely in advance of a change in the structure of the control that may only occur at a later stage through a de facto acquisition of control, or which may never occur.
To remedy this absurd situation, the Tribunal should conduct a substantive investigation into each merger case and analyse the de facto control structure pre- and post-merger. 
Daniela Mariotti is an associate in the Competition Law Department at Bowman Gilfillan
1 R Whish Competition Law Fifth Edition 2005 p. 787 – 788; Jones and Sufrin EC Competition Law 2nd Edition 2004 p. 848 – 850 quoting D Turner “Conglomerate Mergers and section 7 of the Clayton Act” (1965) 78 Harvard LR 1313, 1317. 
2 Mariotti, Norton and Roets Ethos Decision: Bright Lines in Control Provisions Seem To Be Anything But Bright Paper delivered to the Competition Commission, Competition Tribunal and Mandela Institute Conference on Competition Law, Economics and Policy in South Africa on 21 May 2007.
3 See Mariotti, Norton and Roets above. 
4 Legh Cape Empowerment Trust: A Case of Undue Preference Competition Law Sibergramme 1/2007 28 May 2007 p. 4.
5 See Mariotti, Norton and Roets above.