South Africa: Competition Appeal CourtYou are here: SAFLII >> Databases >> South Africa: Competition Appeal Court >> 2011 >>  ZACAC 1 | Noteup | LawCite
Netstar (Pty) Ltd and Others v Competition Commission South Africa and Another (99/CAC/MAY10, 98/CAC/MAY10, 97/CAC/MAY10)  ZACAC 1; 2011 (3) SA 171 (CAC) (15 February 2011)
Download original files
Bookmark/share this page
IN THE COMPETITION APPEAL COURT
REPUBLIC OF SOUTH AFRICA
CAC CASE NO.99/CAC/MAY10
CAC CASE NO.98/CAC/MAY10
CAC CASE NO.97/CAC/MAY10
CT CASE NO.17/CR/MAR05
In the matter between
NETSTAR (PTY) LIMITED ..........................................................First Appellant
MATRIX VEHICLE TRACKING
(PTY) LIMITED …......................................................................Second Appellant
TRACKER (PTY) LIMITED …....................................................Third Appellant
SOUTH AFRICA …......................................................................First Respondent
TRACETEC (PTY) LIMITED …....................................................Second Respondent
J U D G M E N T
WALLIS AJA (DAVIS JP and DAMBUZA JA concurring)
 Prudent motorists insure their motor vehicles, principally against the risk of loss caused by accidents and theft. Finance houses that fund the purchase or lease of motor vehicles usually require the purchaser or lessee to take out such insurance. Short-term insurers provide this type of cover and bear the losses suffered in consequence of claims under the policies they issue to motorists. It is therefore in the interests of short-term insurers to take steps to limit or minimise such claims.
 Once a vehicle has been stolen minimising the loss depends upon how quickly it can be traced and recovered. The ability to do that has been considerably enhanced by the introduction into South Africa about twenty years ago of stolen vehicle recovery systems that enable stolen vehicles to be traced rapidly and recovered. Starting with high value luxury cars but increasingly across a broader range of vehicles insurers have taken steps to ensure that motor vehicles are fitted with these systems. The one route they adopt is to require the installation of such a system and to offer a reduced premium in return. The other is to offer insured motorists a discount on their insurance premiums if they have such a system fitted to their motor vehicles.
 The appellants, Netstar, Matrix and Tracker, were amongst those who established the industry for the provision of stolen vehicle recovery systems. Netstar and Tracker are, and always have been, by a significant margin the largest participants in the industry. Throughout the period relevant to this appeal Matrix was the third largest. During that period they, together with another entity, Bandit, and, from 2002, Global Telematics, formed the stolen vehicle recovery (SVR) committee of the Vehicle Security Association of South Africa (VESA). In February 2005 their conduct as members of that committee was the subject of a complaint by the second respondent, Tracetec, to the Competition Commission. That led in turn to a referral to the Competition Tribunal and, after a lengthy hearing and the consideration of a mass of documents, the issue of a declaratory order that they had contravened s 4(1)(a) of the Competition Act 89 of 1998 (“the Act”). These appeals lie against that order.
 A similar declaratory order was made against VESA at the instance of Tracetec, which obtained leave to intervene in the Commission’s referral primarily for that purpose. VESA has not appealed against the Tribunal’s order so that issue is not before us. Tracetec also obtained leave to pursue the appellants in relation to a claim under s 8 of the Act that the Commission decided not to refer to the Tribunal and on a lesser issue of the period of the alleged contravention of s 4(1)(a). It failed on these issues before the Tribunal and they form the subject of a cross-appeal. However, during the argument on behalf of Matrix we were informed that Tracetec was not persisting in the cross-appeal in relation to s 8. As regards the timing issue it can conveniently be dealt with in conjunction with the general consideration of the merits of the appeals.
 The dispute between the parties arises from the terms upon which providers of SVR systems could become members of the SVR Committee of VESA and have their systems accredited by VESA. During the period from 1996 to February 1999 certain criteria for membership and accreditation were developed and thereafter implemented by VESA. The effect of this, according to the Commission and Tracetec, was to impede newcomers to the industry from entering the market for the provision of SVR systems. The reason is that these criteria or ‘standards’, as they were described, were used by most of the members of the South African Insurance Association (SAIA) in deciding which SVR systems would, when installed in the motor vehicles they insured, qualify the insured to receive a reduced or discounted premium. The complaint by the Commission and Tracetec is that newcomers to the industry could not satisfy the performance component of the standards and hence could not obtain membership of the Committee and VESA accreditation of their systems. It was contended that this blocked their entry into the industry because of the fundamental importance that insurers attached to VESA accreditation. In brief therefore the establishment and implementation of these standards was said to arise from an agreement prohibited under s 4(1)(a) of the Act.
 There is no serious factual dispute on the record and the facts underpinning the complaint can be set out quite briefly. In the mid-1990’s, when the SVR industry was in its infancy in this country, a large number of businesses were offering to provide SVR systems. Short-term insurers were concerned about the quality and reliability of their offerings. Their concern was two-fold. Firstly, if the system did not work or the supplier proved unreliable or went out of business, the insured, who had been compelled or induced to fit the system, would complain to and possibly seek compensation from the insurer. Secondly, if the system did not work or the supplier proved unreliable, the anticipated benefits by way of a reduction in the amount that would have to be paid in respect of theft claims would not materialise.
 During 1996 and 1997 there was some discussion, in meetings arranged by VESA involving various parties interested in the SVR sector, about the establishment of a standard for the supply of SVR systems. This was a natural extension of the existing activities of VESA, which had already established such standards in respect of various vehicle anti-theft devices. Netstar attended an initial meeting and Tracker was represented at the next meeting. In September 1997 VESA published some draft standards, principally of a technical nature. In the preamble the following was said:
‘Stolen vehicle tracking systems are proving to be useful means of stolen vehicle recovery and a number of these systems are appearing in the marketplace. Insurance companies would like to endorse the systems when they are effective, and so it is appropriate that a means to measure the performance of recovery systems be developed. Not only will this allow insurers to introduce incentives for fitting approved systems, it will also provide a measure of protection to the general public who will have some assurance that the systems they select will provide adequate long-term protection to their vehicles.
It is therefore the purpose of this document to specify minimum performance standards for electronic tracking and recovery systems … The emphasis will be on the recovery infrastructure and the ability of the recovery service to [service]1 its customers. This emphasis is important because vehicle recovery systems cannot be sold on a “fit and forget” basis. The customer who buys the system will rightly expect that it will continue to protect his vehicle for the foreseeable future.’
That this accurately reflected the views of a majority of short-term insurers is clear from the evidence of Ms Caroline de Silva, at the time the deputy to the executive director of SAIA, who testified that:
‘The tracking industry was relatively new at the time … but it was proving to be very, very successful in assisting us with vehicle crime. And we had, I suppose, any number of people saying that they could track vehicles, they could recover vehicles and they could do whatever it is that they do. And insurance companies were starting to give discounts to policyholders upon a fitment of a tracking system, because they had proven so capable in recovering vehicles. And like any other after-market fitment of security, the industry wanted some sort of standard or comfort that vehicle tracking was being managed and that we were giving the discounts on the right basis, that we would actually recover that risk, and ask the tracking companies to form under a standard body and of course, the preference for the insurance industry was VESA.’
 SAIA’s desire to establish a uniform standard for SVR systems that it could recommend to motorists and that its members could use as a basis for offering reduced or discounted premiums, was given considerable impetus when, near the end of 1997, Datatrak, a leading supplier of SVR systems, withdrew from the market because its holding company, a public listed company, saw no reasonable prospect of rendering its operations sufficiently profitable in the near-term. Not only was this a source of embarrassment to the members of SAIA who had recommended Datatrak as an SVR supplier but it also led them to realise that technical specifications alone would not suffice to provide a standard for the supply of SVR systems. As Ms da Silva said:
‘There had to be a performance standard and a capability standard and a sustainability standard that we wanted the tracking companies to adhere to’
 This led SAIA to convene a meeting of the five leading SVR companies including the three appellants. The purpose was to persuade them to become members of VESA and participate in the formulation of standards for the supply of SVR systems that would satisfy the requirements of the insurers. The SVR companies were reluctant to do this and suggested that the South African Bureau of Standards be approached instead to formulate standards. However, an approach to the SABS revealed that standards could only be prepared over a three or four year period at very substantial expense and this would not meet the immediate needs of the insurance industry. What SAIA wanted therefore was for the leading tracking companies to participate in VESA in order to enable VESA to produce the standards that SAIA members wanted.
 Although the tracking companies were reluctant to join VESA for this purpose, SAIA was in a strong bargaining position in its efforts to persuade them to do so. As Ms da Silva pointed out the members of SAIA had already written into their policy documents that ‘any aftermarket fitment of vehicle security would be VESA approved’. This applied both to anti-theft devices and to SVR systems. It would, as she said, be ‘far more in the interest of the insurance industry if the tracking companies jointed VESA’. The perspective of the tracking companies in regard to SAIA’s persuasive efforts is reflected in the witness statement of Mr de Clerk, the managing director at the time of Tracker. He pointed out that Tracker’s business model was very dependent on obtaining support from insurers. After Datatrak’s exit from the market SAIA placed the tracking companies under pressure to join VESA and, according to Mr de Clerk:
‘It was stated unequivocally that, if a SVR service provider wanted endorsement by SAIA, they had to be VESA approved. Tracker, having envisaged a business strategy that depended on its co-operation with the insurers, was left with little choice and accordingly joined VESA in or about 1997.’2
 In consequence of Ms da Silva’s efforts the appellants and three or four other tracking companies agreed to participate in a process under the auspices of VESA to try and determine some standards for the SVR industry. An initial meeting was convened by VESA on 16 July 1998. The minutes record that ‘the Insurance industry are pro-active and leading the direction in the development of vehicle security products’ and that VESA was acting on a mandate from SAIA. The chairman, a VESA representative, indicated that its objectives were to:
‘1. Identify the key elements that will demonstrate sustainable loss reduction.
2. Find a simple but very effective method to monitor claim performance.’
Thereafter a series of workshops was held attended by representatives of a number of different SVR suppliers. According to Mr Maskrey of VESA, who attended all of the meetings on behalf of VESA, it attempted to involve as many SVR companies as possible. Representatives of VESA and Ms da Silva, on behalf of SAIA, were kept informed of progress. The workshops broadly covered two areas. The one was the question of standards where it seems that the suggestions in a memorandum prepared by Netstar, the largest operator in the industry, and provided to Mr Chris Bezuidenhout of SANTAM, the largest short-term insurer in the country were used as a working base.3 The other related to the establishment of a vehicle tracking committee within VESA and the constitution and rules of such a committee. It is apparent from minutes of these workshops that representatives of VESA were present at all the workshops and that the board of VESA was kept abreast of developments in the workshops.
 It is not clear from the documents when finality was reached on the contents of the standards. It may well have been, as the Tribunal inferred, in about September 1998. Some things are however clear. First they were the product of a process in which not only the three appellants but also other SVR and tracking companies were involved as well as representatives of VESA and SAIA. Second the standards were created before the establishment of a vehicle tracking committee, much less an SVR committee, within VESA and were accordingly not the product of the deliberations of that committee. Third the workshop process culminated in parties being invited in November 1998 to become members of VESA and the vehicle tracking committee. Fourth the applications that were submitted were submitted to VESA, which dealt with the approvals. Accordingly the Board of VESA must have had and approved the standards because they could not otherwise have taken the decisions that they did in regard to membership and approval. That is apparent from the provisional approval of Bandit. Fifth the decision that there should be two committees, namely, a fleet management committee and an SVR committee was a decision by the Board of VESA taken at some time shortly before the launch of the committees.
 In February 1999 it was decided by VESA in conjunction with SAIA to hold a launch function on 16 March 1999. On that day the chief executive of SAIA issued a press release in which he said the following:
‘Motor vehicle tracking and recovery and fleet management companies were now being regulated by an independent body, launched today by the Vehicle Security Association of South Africa and the SA Insurance Association.
If the owner of a motor vehicle or fleet uses the system of an approved tracking and recovery company or has one installed, the short-term insurance industry may offer a discount on premiums, according to SAIA chief executive, Barry Scott. …
He said VESA already developed a certification system for immobilisers and gearlocks used by SAIA members.
“We approached them to plan and implement regulations for the tracking industry. VESA then approached companies active in the field and the result was the formation of this independent regulatory body,” Scott said.’
The press release then identified Tracker, Netstar, Matrix, Capital Air (Pty) Limited and Bandit as the approved companies for tracking and recovery. Other companies were identified as approved in the fleet management area.
 The issue of approval mentioned in the press release was a tricky one because Bandit was merely granted provisional approval. The reason was that the performance standards required for full approval that the SVR supplier should have a minimum client base of 3000; should have been in operation for at least one year; should have made at least 100 recoveries and should have achieved a specific recovery rate based on industry performance. A new company starting out in the industry would not be able to satisfy these requirements. To that end the standards included a category of provisional approval based on the applicant’s infrastructure rather than its performance record. Bandit fell in this latter category and had been granted approval on a provisional basis.
 Provisional approval was problematic as far as the members of SAIA were concerned. Ms da Silva expressed the difficulty in the following terms;
‘There was always a real doubt that provisional approval will be workable. The insurance industry was giving upfront discounts that were quite significant in order for SVR fitment. Their feeling was that they would be very unwilling to give discounts where somebody only had provisional status, because they didn’t have the same comfort that those companies could do the recoveries they wanted. So, they had, I suppose, all along indicated to VESA that they had difficulties with the provisional and full approval approach.’
These concerns were conveyed both to the SVR committee and to the board of directors of VESA. In August 1999 Mr Oosthuizen, the chief executive of VESA, wrote to the members of the SVR committee concerning a provision of the terms of provisional approval that would require demonstration of an established performance capacity. At a meeting of the VESA board on 7 September 1999 Mr Oosthuizen reported that provisional approval had not been well received by the insurance industry. On 13 September 1999 the SVR committee revised the provisional membership requirements over the objections of Netstar and Tracker, which appealed to the VESA board. On 8 November 1999, at a joint meeting of the VESA board and the SVR committee, Ms da Silva conveyed SAIA’s view that an SVR system should either be approved or rejected and that the insurers did not agree with the principle of ‘semi-approval’. At a further combined meeting on 22 November 1999 it was recorded that the SAIA Motor Advisory Committee requested that provisional approval status be discontinued. Thereafter the matter was further considered by the VESA board, which took legal advice on the question. On 25 April 2000 Mr Oosthuizen wrote to the SVR committee informing them that;
‘In the category of Stolen Vehicle Recovery there will be only one approval status (provisional approval having fallen away). All new applicants will be required to comply with the criteria for what was previously termed “full approval”.’
From that time onwards there was no longer any provisional approval category.
 From April 2000 until August 2003 the basic requirements for membership of the SVR committee and accreditation of an SVR system by VESA remained the same. They were that the system had been fitted in 3000 vehicles; that the operator had been in operation for over a year; that the supplier had effected over 100 recoveries with that system and that the recovery rate fell within a defined industry average. Within the committee itself this caused some difficulties when members sought to introduce systems based on new technology, where the fitment and recovery requirements were not yet satisfied. Disputes over these issues highlighted the internal stresses within the committee, the members of which were competitors. In my view, however, it is unnecessary to explore these disputes. The reason is that under s 4(1)(a) of the Act the Court is concerned with whether an agreement ‘has the effect of substantially preventing or lessening competition in a market’ and the complaint in this case is that the alleged agreement had an exclusionary effect on new entrants to the industry seeking membership of the SVR committee and VESA accreditation. Internal wrangling among the existing SVR committee members has no bearing on this. The point in issue is whether potential entrants to the industry were excluded or hampered in their efforts to enter the market by the existence of these standards, not whether those who were already part of the SVR committee and enjoying VESA accreditation were self-serving in their dealings with one another.
 The situation of aspirant new members remained unchanged throughout the period from April 2000 to August 2003. Unless they achieved the prescribed number of fitments and recoveries they were not admitted as members and did not obtain VESA accreditation. That situation only changed in July 2003 when the requirements for admission were altered by permitting SVR suppliers that were unable to satisfy the performance criteria to provide a guarantee of R2 million to VESA in lieu of satisfying those performance criteria. After this several more SVR suppliers became members of the SVR committee and obtained VESA accreditation. This in turn caused considerable dissatisfaction to the appellants resulting in their joint withdrawal from VESA and the SVR committee in May 2004.
 In summary therefore the factual position is that the SVR committee was established and membership of the committee and accreditation of SVR systems by VESA depended upon the ability of a SVR supplier to satisfy the performance criteria. From the outset (although there was a minor difficulty, soon resolved, involving Matrix in April 2000 when the SVR committee was separated from the Fleet Management Committee of VESA) the appellants were members of the SVR committee. Their systems were accredited by VESA. Once the category of provisional membership fell away in April 2000 only a few entities made enquiries about or sought membership of the SVR committee. None of them could satisfy the performance criteria and none of them was admitted or given VESA accreditation for their SVR systems. That situation only changed in August 2003 with the introduction of the financial guarantee and shortly after that the appellants withdrew from both the committee and VESA. In effect that rendered VESA accreditation of SVR systems a dead letter because the short-term insurers indicated to the appellants that they would continue to recognise their systems as justifying reduced or discounted insurance premiums.
THE COMPLAINT AND THE TRIBUNAL’S DECISION
 It is convenient against that background to consider the terms of the Commission’s complaint and the determination by the Tribunal. In doing so it must be borne in mind that the Act only came into operation on 1 September 1999 so that activities by the appellants in the SVR committee prior to that date do not fall within the ambit of the Act.
 The Commission’s complaint is set out in the affidavit of its inspector, Mr Hendrik Senekal. He recorded that Tracetec had lodged a complaint with the Commission that the appellants (and two other parties who subsequently fell away) were engaging in prohibited practices in contravention of the provisions of s 4(1)(a) of the Act by:
‘Collectively agreeing on accreditation criteria for VESA membership of the Stolen Vehicle Recovery Committee (“SVR Committee”) which had the effect of obstructing new entrants from entering into and/or effectively participating in the market for the tracking and recovery of motor vehicles to the detriment of consumers.’
Mr Senekal said that the Commission had found that the appellants had engaged in a restrictive horizontal practice as referred to in s 4(1)(a) of the Act:
‘… in that the agreement between or concerted practice by the respondents impeded or prevented Tracetec and other competitors such as Karoocell (now Mobile Traker Manufacturing (Pty) Limited “Mobile Traker” (sic)), Cartrack (Pty) Limited (“Cartrack”) and Cellstop (Pty) Limited (“Cellstop”) from entering or expanding within the relevant market.’
 The Commission found that the appellants were all members of VESA’s SVR committee from the time of its establishment until their resignation. In order to obtain VESA accreditation firms needed to become members of VESA within specified categories relating to the nature of the product concerned. Mr Senekal went on:
’10.5 In order to enter the market and compete effectively in the relevant market, Tracetec and other competitors and/or potential competitors such as Mobile Traker, Cartrack and Cellstop required the accreditation of most major insurance companies. This was so because tracking devices are mainly installed as a result of insurers’ requirements and insurance companies only offered discounts on insurance premiums to insured persons if they used approved and/or accredited products and services. In practice, firms that competed in the relevant market had to obtain VESA membership in the SVR category in order to obtain approval from the insurance companies since the insurance companies do not recognise and/or accredit the products of non-VESA members. Without VESA accreditation in the SVR category it was therefore virtually impossible for firms to enter or expand within the relevant market.’
Mr Senekal then referred to the performance criteria of a minimum of 3000 subscribers and 100 stolen vehicle recoveries and said that before July 2003 this:
‘… raised an almost insurmountable barrier to entry to potential new entrants into the relevant market, since without VESA accreditation in the SVR category, firms seeking accreditation were unable to build up a sufficiently large client base to meet the … 3000 subscriber requirement or conduct the required 100 recoveries. The respondents, during the duration of their membership of VESA’s SVR committee, up until at least July 2003, therefore excluded new entrants from the relevant market by setting and/or attaining criteria, which prevented potential new entrants from obtaining VESA accreditation in the SVR category.’
 The Commission’s case was therefore that the performance criteria and particularly the requirement that an SVR supplier have 3000 customers and over 100 recoveries were exclusionary. It relied upon both an agreement and a concerted practice in terms of s 4(1)(a) of the Act. In response to requests for particulars by each of the appellants as to the terms of the agreement said to constitute a contravention of s 4(1)(a) of the Act they said that:
‘The terms of the agreement between the respondents were that membership of VESA (and accordingly, accreditation by VESA) in the SVR category … should only be granted to new entrants who could satisfy [the specified criteria].’
In regard to the terms of the alleged concerted practice it was alleged that the members of the SVR committee co-operated in establishing the criteria for membership of the SVR committee and accordingly for accreditation by VESA and thereafter co-operated ‘in resisting and/or frustrating and/or ignoring attempts by the Board of VESA to ameliorate the established criteria’.
 In substance the Tribunal upheld this complaint, although the terms of its declaratory order in some respects differ from it and in others are confusing. It held that the complaint was established for the entire period from 1 September 1999, when the Act came into force, until August 2003 when the giving of a guarantee was introduced as an alternative to compliance with the performance criteria. It did not regard the period of provisional membership as being relevant to ameliorate the anti-competitive effect of the performance criteria because ‘provisional membership was not something meaningful and when Bandit tried to rely on it for marketing it was chastised’. The correctness of that decision lies at the heart of these appeals.
 A contravention of s 4(1)(a) of the Act arises in three circumstances. These are where there is an agreement, as defined in s 1(1)(ii) of the Act, that has the effect of substantially preventing or lessening competition in a market; where there is a concerted practice, as defined in s 1(1)(vi) of the Act, having the same effect and where a decision by an association of firms has that effect. Both Tracetec in its original complaint and the Commission in its reference to the Tribunal alleged that there was either an agreement between or a concerted practice by the appellants that constituted the alleged contravention of s 4(1)(a). No reliance was placed on a decision of firms insofar as the three appellants were concerned and the endeavour by Tracetec to rely on it in its heads of argument was not pursued in oral argument.
 The distinction between an agreement and a concerted practice is important as appears from the definition of the two expressions.4 These read:
‘”Agreement”, when used in relation to a prohibited practice, includes a contract, arrangement or understanding, whether or not legally enforceable’
‘“Concerted practice” means co-operative, or co-ordinated conduct between firms, achieved through direct or indirect contact, that replaces their independent action, but which does not amount to an agreement.’
A concerted practice arises from the conduct of the parties and does not amount to an agreement. A possible example might be the type of cartel arrangement where a market leader signals a price increase by way of public announcement and, in accordance with long-standing practice in the industry, the other participants follow its lead. However care must be taken not to confuse independent conduct with interdependent conduct.5 It suffices for present purposes to say that the emphasis is on the conduct of the parties. By contrast an agreement arises from the actions of and discussions among the parties directed at arriving at an arrangement that will bind them either contractually or by virtue of moral suasion or commercial interest. It may be a contract, which is legally binding, or an arrangement or understanding that is not, but which the parties regard as binding upon them. Its essence is that the parties have reached some kind of consensus.6 No doubt in many cases the same evidence may be relied upon as pointing towards either an agreement or a concerted practice. However, sight should not be lost of the fact that they are different. The definition of an agreement extends the concept beyond a contractual arrangement. However, what it requires is still a form of arrangement that the parties regard as binding upon both themselves and the other parties to the agreement. Absent such an arrangement there is no agreement even in the more extended sense embodied in the definition. By contrast a concerted practice examines the conduct of the parties to determine whether it is co-ordinated conduct or they are acting in concert. The absence of any arrangement between them or any belief that they are obliged to act in that fashion is immaterial.
 Observing the distinction between an agreement and a concerted practice is essential to a proper analysis of allegedly anti-competitive conduct. If the argument or analysis runs the two together that leads to confusion and error. In particular, and that is evident in this case, it leads to an incorrect approach to and evaluation of the evidence. The case for a concerted practice is based on evidence that assesses the nature of the conduct of the firms said to be party to the practice. By contrast the case for an agreement examines whether an agreement as defined was concluded and that focuses on the existence of consensus between the parties. Even where reliance is placed on the same evidence in support of these distinct cases it requires separate evaluation. There are also jurisdictional and procedural reasons for a careful observance of the distinction. The jurisdictional reason lies in the path by which a complaint reaches the Competition Tribunal. The process starts with the Commissioner initiating a complaint in terms of s 49B(1) of the Act or some other person submitting a complaint to the Commission under s 49B(2)(b) of the Act. In either case the complaint must be investigated and, if the Commission concludes that a prohibited practice has been established, must be referred to the Tribunal under s 50 of the Act. The Tribunal’s jurisdiction is confined to a consideration of the complaint so referred and the terms of that complaint are likewise constrained by the terms of the complaint initiated by the Commissioner or made by some other person. Accordingly if the original ground for the complaint is that there was a prohibited agreement the Tribunal cannot determine it on the basis that there was a concerted practice or vice versa.
 As regards the procedural aspect this court held in Sappi7 that the Commission has no general power to consider anti-competitive behaviour but can only deal with a complaint that alleges a contravention by the target of the complaint of a specific applicable provision of the Act. In Woodlands Dairy8 the initiation of a complaint was likened to a summons in that it must contain sufficient particularity and clarity to survive the test of legality and intelligibility. This is not to say that what is required of a complaint is the level of precision that is demanded in pleadings. Where the complainant is a lay person that would be to demand more than can reasonably be expected. What is required is that the conduct said to contravene the Act must be expressed with sufficient clarity for the party against whom that allegation is made to know what the charge is and be able to prepare to meet and rebut it.9 It is true that the competition issues upon which the Tribunal is called to adjudicate may be broader, more general and less clear-cut than those that arise in a conventional civil case in the High Court. That does not mean, however, that broad and unspecific generalities should take the place of a properly articulated complaint before the Tribunal to which the target of the complaint can respond. It was previously argued in regard to hearings before the Industrial Court under the Labour Relations Act 28 of 1956 that a broader and more tolerant approach should be taken in those tribunals because of the lack of sophistication and experience of employees in advancing their claims. That approach was roundly rejected by the then Appellant Division.10 There is even less excuse for it to be adopted in competition matters, where the parties are usually sophisticated and have, as here, the services of experienced attorneys and counsel. There is no excuse for resorting to vague generalities instead of formulating complaints accurately and in detail in accordance with the provisions of the Act. It is the Tribunal’s responsibility to ensure that complaints brought before it are pursued on this basis and also to ensure that it determines only the complaint laid before it, not something else that arises or occurs to it in the course of the proceedings.
 Proof of the existence of an agreement or concerted practice, as defined, is the first issue to be determined in an enquiry into an alleged contravention of s 4(1)(a) of the Act. The second is whether the proven agreement or concerted practice has the effect of substantially preventing or lessening competition in a market. There are three elements to this. The first requires an identification of the relevant market. The second requires proof that there has been some prevention or lessening of competition in that market and the third that this is an effect of the agreement or concerted practice. The last is a question of causation. Whilst these three elements are notionally distinct, in practice it may prove difficult to separate them.
 In the present case the identification of the market has not posed any particular difficulties. The market is the market in respect of the supply and operation of stolen vehicle recovery systems. Its geographic extent is the whole of the Republic. Whether there has in some respect been a substantial prevention or lessening of competition in that market is a factual enquiry.11 Unless that enquiry generates an affirmative answer the existence of an agreement or concerted practice is irrelevant, at least insofar as s 4(1)(a) is concerned. There may be some overlap between prevention and lessening of competition. For example, an agreement between existing competitors may both lessen the level of existing competition and prevent competition in the future. It could be argued that this should be taken as a composite term and not as two distinct concepts. An attempt to resolve the issues raised by this language would require us to consider the jurisprudence in the European Union on the nature of prevention of competition. It is unnecessary for present purposes for us to enter upon that enquiry and preferable to leave it for consideration on another more appropriate occasion. In this case there were very specific allegations of the nature of the prevention or lessening of competition relied upon by the Commission and Tracetec and I confine myself to a consideration of whether that case was established by the evidence.
 The references to preventing and lessening competition are qualified by the adverb ‘substantially’. Whether any particular prevention or lessening of competition is substantial will be a question of fact in every case. However, the existence of this qualification demonstrates that what is required is something that is neither speculative nor trivial. It may be notional or hypothetical in the sense that there is no actual instance of a person being prevented by the agreement or concerted practice from entering the industry concerned. Nonetheless, if the evidence is strong enough to show that, but for that agreement or concerted practice, persons would probably have entered the industry and engaged in competition, the requirements of a substantial prevention of competition will almost certainly be satisfied. However, that may be a difficult case to prove. It will require consideration of not only the agreement or concerted practice, but also other obstacles that would have confronted aspirant entrants to the market. Issues such as the availability of capital, the need to establish an infrastructure, the availability of suitably qualified staff and access to appropriate technology may all be relevant to the question whether it is realistic to say that new entrants would have come forward when as a matter of fact they did not do so.
 Once an agreement or concerted practice has been established it must be linked to the prevention or lessening of competition. This raises the question of causation embodied in the words ‘if it has the effect of’. The agreement or concerted practice must be linked to the prevention or lessening of competition. Whilst in practice this is not dealt with as a discrete issue it is necessary from an analytical perspective to recognise it as separate from the other two questions as it may raise distinct issues. This court has not previously considered the question of causation in relation to this section of the Act. What is required in order to justify a finding that the prevention or lessening of competition in a market is an effect of an agreement or concerted practice? Clearly the agreement or concerted practice must be a necessary cause (causa sine qua non) of the lessening or prevention of competition, without which it would not have occurred. This is the same enquiry as to factual causation that is conducted in other areas of the law, such as the law of delict.12 One asks whether ‘but for’ the agreement or concerted practice the prevention or lessening of competition would have occurred. If the answer is in the affirmative then that prevention or lessening of competition was not an effect of the agreement or concerted practice. However, a negative answer does not finally dispose of the question of causation.
 A market is a complex concept and many factors operate to influence what happens in a market. Often it will be clear that the only reason that there has been some prevention or lessening of competition in a market will be the particular agreement or concerted practice identified in the complaint. The link of cause and effect will be clear. On other occasions, of which the present case is an example, there may be other factors that are also a necessary cause of the prevention or lessening of competition. That is the argument for the appellants. They contend that if there was any prevention or lessening of competition in the market for SVR systems it was the decisions by SAIA and its members that had that effect, not any agreement or concerted practice by the appellants. How is this resolved if both causes satisfy the ‘but for’ test?
 The section does not contemplate concurrent causes but requires a determination of whether the agreement or concerted practice had the effect of preventing or lessening competition in the market. The language used is not consistent with the agreement or concerted practice being one of several concurrent causes of the prevention or lessening of competition in the market. It requires the agreement or concerted practice to have the required detrimental effect. In my view what is required by this provision is that the primary or substantial cause of the prevention or lessening of competition must be identified. Where an agreement and some other separate action by a different party are identified as both having been necessary conditions for the prevention or lessening of competition it is only where the agreement or concerted practice plays the dominant role that it falls within the section. In other words it is necessary to consider whether the lessening or prevention of competition is so closely connected to the agreement or concerted practice that it can properly be said that the former was the effect of the latter and any other action was merely ancillary. In this investigation it is necessary to have regard to the intervention or involvement of third parties and any other factors that may have had a causative effect in bringing about the lessening or prevention of competition. This is similar to, but not the same, as the question of legal causation that arises in the delictual context. In the context of competition law and the requirements of the Act it is only where the substantial prevention or lessening of competition is the direct and predominant consequence of the agreement or concerted practice that s 4(1(a) is contravened.
 Against the background of that analysis the following questions fall to be considered in the present case:
(a) Were the appellants party to an agreement or concerted practice as alleged by the Commission?
(b) Was there a substantial prevention or lessening of competition in the South African market for the supply and operation of stolen vehicle recovery systems?
(c) If so, was that substantial prevention or lessening of competition an effect of the agreement or concerted practice, within the meaning of that expression in s 4(1)(a)?
Each of these will be considered in turn.
AGREEMENT OR CONCERTED PRACTICE
 Regrettably the Tribunal’s decision fails to recognise the need to distinguish clearly between an agreement and a concerted practice. The order it granted declared that the actions of the appellants ‘in concluding an agreement and/or engaging in a concerted practice’ amounted to a contravention of s 4(1)(a). The use of ‘and/or’ betrays its lack of clarity on this point.13 The Tribunal also overlooked the fact that the basis upon which the Commission alleged that there had been an agreement and the basis for its allegation of a concerted practice were different. In its discussion of the issue under the heading ‘Was there an agreement or concerted practice?’ the topic of a concerted practice is not discussed and the conclusion appears to have been that there was an agreement. The alleged grounds of a concerted practice were not considered. If the Tribunal intended simply to find that there was an agreement there is the further difficulty that it is not apparent from its determination when this agreement was concluded or what its terms were.
 The relevant passage in the Tribunal’s determination reads as follows:
It is common cause that the three SVR respondents are competitors in the SVR market. The agreement alleged to be a restrictive practice is the performance standard for membership of the VESA SVR committee. The SVR respondents do not concede that they are a party to an agreement in respect of these standards. It is hard to see how they can credibly make this assertion. The SVR respondents all joined VESA with a view to it setting performance standards. The early history of the Tracking and Recovery sub-committee showed that when the SVR respondents did not reach agreement no standard was determined. It was only when all three returned after being prodded by SAIA that a standard was agreed upon and that required consensus being reached by these three firms and in particular Nestar and Tracker. Although other firms had brief roles in the committee they played as extras. The real roles and hence the agreement on the standards came about when the three SVR respondents reached an agreement all three found acceptable. The agreement was finally determined sometime in September 1998 and then implemented in February 1999, with the press release. It was, as the chronology shows, the sub-committee which set the standard and then proceeded to make it public – prior to its endorsement by the executive of VESA – something it is unclear ever happened. At various times this standard was amended by the Committee to make it easier for the three firms to get approval for their new products whilst at the same time denying this latitude to new entrants. The SVR respondents then voted in favour of raising the admission threshold by abolishing provisional membership and then repeatedly resisting the suggestion of a financial guarantee until conceding to this in August 2003 at the eleventh hour.
As the history has shown, when major decisions needed to be taken they often consulted outside of the committee and only amongst the three of them. VESA permitted the SVR committee to set the standard and the SVR respondents were the key protagonists in that consensus – hardly surprising as they represented at that time more than 90% of the industry.
When the consensus in the SVR committee broke down at a time when the SVR respondents could be outvoted by the other members of the committee, they resigned, collectively endorsing the same reasons for their resignations and adopting the same press statement. As they had operated in setting the performance standard so they operated to prevent the standard that had been for once imposed upon them by outside pressure, from becoming a means of easing entry into the market. As we have seen, once they left VESA it ceased effectively to become the approval body for the industry. There is therefore little doubt that the standard set for approval based on a performance standard was reached as a result of inter alia the agreement of the three SVR respondents. Since they did so as a VESA sub-committee, which VESA permitted to be held out to the world as its standard, it must be regarded as a decision of an association of firms as well. We deal more fully with VESA below.
 There are a number of problems with this passage. Certain factual statements are simply incorrect. There was no formal Tracking and Recovery sub-committee within VESA prior to the workshops being convened and the appellants were not members of VESA prior to or while the workshops were conducted. The Fleet Management and SVR committees were created at the end of this process not at the beginning and only then did the appellants apply to join VESA. The standards were not publicly launched without the endorsement of VESA. In fact VESA, in conjunction with SAIA, organised the launch. VESA clearly endorsed and enforced the standards. Equally problematic is the statement that the agreement is the performance standard. That cannot be correct. The standard alone cannot be an agreement nor is it the agreement alleged by the Commission to exist. That agreement was that the appellants had agreed that membership of VESA in the SVR category and accordingly accreditation by VESA should only be granted to new entrants who could satisfy the standard. The question before the Tribunal was whether that agreement had been established.
 The key finding by the Tribunal appears to be encapsulated in the following extract from the quoted passage:
‘It was only when all three returned after being prodded by SAIA that a standard was agreed upon and that required consensus being reached by these three firms and in particular Netstar and Tracker. Although other firms had brief roles in the committee they played as extras. The real roles and hence the agreement on the standards came about when the three SVR respondents reached an agreement all three found acceptable.’
With respect to the Tribunal this involves a considerable confusion of thought and a misreading of the evidence. The minutes show that at least seven vehicle tracking companies, including the appellants, participated in the workshops. There is nothing to indicate that the other four merely stood by and ‘played as extras’. The role of both SAIA and VESA in these meetings is ignored. It is a non sequitur to say that because the three appellants found the ultimate set of standards agreeable that meant that the standards were constituted by way of an agreement among the three of them. It is also not factually correct. All those who participated in developing the standards agreed to them. The minutes show relatively little disagreement even by companies that could not satisfy the standards.14 Nor is there any basis for the finding that the agreement of the three appellants was essential to the establishment of the requisite standards. If Matrix, for example, had refused to accept the standards there is no reason to think that this would have derailed the entire enterprise. It is mere speculation to say what the impact would have been if one of the two largest companies, Netstar or Tracker, had declined to participate, because that eventuality did not arise and there were obvious commercial reasons why the one would not withdraw without the other.
 The Tribunal also relied on events after the standards had been produced and launched to the public and after the SVR committee had been established. It is not clear why it thought these relevant to the existence of an agreement that it held had been finalised in September 1998 and implemented in February 1999. Either the agreement existed at that time or it did not. Subsequent events cannot affect that. The Tribunal’s factual findings in this regard are not supported by the evidence and particularly not the documentary evidence. It is correct that there was debate in the committee about the introduction of new systems by existing members of the committee. However, this was never an issue in relation to new entrants. It is not correct to say that the appellants voted in favour of raising the admission threshold by abolishing provisional membership. This was done by the VESA board at the instigation of SAIA and the SVR committee was simply informed of that decision. It is not correct to say that the appellants resisted the suggestion of a financial guarantee. The minutes reflect that Mr Oosthuizen, the chief executive of VESA, raised this issue at a meeting of the SVR committee on 22 November 2001. The members were asked to provide comment and input regarding financial guarantees in order to address requests from new pending applicants. At the next meeting of the committee on 14 February 2002 Mr Oosthuizen reported that VESA had consulted with its lawyers and ‘it was concluded that it was not viable to proceed with the financial guarantee in lieu of achieving the current specification’. The minutes do not reveal that the issue of a financial guarantee was again raised within the committee until its meeting on 1 July 2003 when it agreed that a financial guarantee would be permitted from applicants who could not satisfy the requirement of 3000 active clients and 100 recoveries. Finally, it is difficult to see how the collective resignation of appellants from the committee provides any evidence of the existence of an agreement.
 It follows that the factual finding of the Tribunal that the standard was set as a result of the agreement of the three appellants cannot stand. Nor do the facts support the allegation by the Commission that there was an agreement between the appellants that membership of VESA (and accordingly accreditation by VESA) in the SVR category should only be granted to new entrants who could satisfy the criteria. They go no further than showing that the appellants, together with other participants in the vehicle tracking industry and with the approval of the VESA board of directors and SAIA, agreed to the adoption of the standards for membership of the SVR committee and accreditation of SVR systems by VESA.
 That conclusion gives rise to a difficult question regarding the application of s 4(1)(a). The section refers to an agreement between parties in a horizontal relationship and s 1(1)(xiii) of the Act defines a horizontal relationship as meaning a relationship between competitors. The only possible agreement that is supported by the facts in this case is an agreement involving a number of competitors in the SVR industry, VESA and SAIA. VESA was seeking to establish itself as a private regulatory body in respect of the SVR industry and SAIA represented the insurance companies, who were the principal drivers in securing that motorists fitted SVR systems to their vehicles. Clearly neither VESA nor SAIA were in a horizontal relationship with the appellants and the other SVR companies. That raises the issue of whether such an agreement can fall within s 4(1)(a) merely because some of the parties thereto are in a horizontal relationship. The point is one of considerable difficulty on which we have not had the benefit of full argument.15 As these appeals can be resolved without reaching a final conclusion on that issue it is preferable not to do so. The judgment proceeds on the assumption that the agreement that has been identified is an agreement capable of falling within s 4(1)(a) but that assumption should not be taken as any indication of a view by the court that this is correct.
SUBSTANTIALLY PREVENTING OR LESSENING COMPETITION.
 The Commission’s complaint was based upon the proposition that in order to enter the market for SVR systems and compete effectively Tracetec and other competitors, or potential competitors, such as Mobile Tracker, Cartrack and Cellstop, required the accreditation of major insurance companies and in order to obtain this VESA membership in the SVR category was essential. It alleged that without VESA accreditation in the SVR category it was virtually impossible for firms to enter or expand within the relevant market. In particulars sought by Netstar it identified the relevant insurance companies as being:
‘Absa, Mutual & Federal, Auto & General, Hollard, Regent, Budget, Outsurance, Dial Direct and BMW.’
It said that apart from BMW all insurance companies required VESA accreditation. Insofar as prospective applicants for accreditation were concerned it identified these as including Orbtech, Cellstop, Tracetec, Cartrack, Karoocell/Mobile Tracker, Global Telematics, Global Asset Protection, Star Track and LST.
 Although the Tribunal discussed the issue of the prevention or lessening of competition extensively in paragraphs  to  and concluded that;
‘… it was not possible for a firm to expand in the SVR market at the time without having its product approved by VESA in the SVR category.’
there is no mention in the discussion of any of the insurance companies identified in the reply to the request for particulars nor of any of the aspirant entrants into the industry, that had similarly been identified. The reason is apparent namely that the evidence did not support the case advanced by the Commission.
 That is hardly surprising in regard to the insurance companies. The major insurance companies that the Commission said would not accept anything other than VESA accreditation turned out to be largely a list of insurance companies that did not require VESA accreditation. The major company that required VESA accreditation was Santam, the largest short-term insurer in the country.16 The Commission called Mr Pienaar from Santam, which required VESA accreditation. He testified that Auto & General, Budget, First For Women and Outsurance, did not require such accreditation. It is not clear whether they were members of SAIA. It seems probable that the bulk of the members of SAIA required VESA accreditation of SVR systems if they were to attract a reduced or discounted premium, but Mr Pienaar agreed with the proposition that it overstated the position to say that ‘only VESA approved devices are commercially viable’. Tracetec identified a number of insurers and brokers as parties from whom it obtained support and that included virtually all those in the list provided by the Commission as requiring VESA accreditation. The evidence that there were a number of insurance companies, insurance brokers and some vehicle suppliers who did not require VESA accreditation was destructive of the Commission’s contention that no such insurance companies existed. There was no evidence before the Tribunal as to the proportion of the insured market held by insurers who required VESA accreditation. It was accordingly impossible to assess whether an aspirant entrant into the market could target these insurance companies and brokers as a way of building up their client base.
 The Tribunal dealt with this difficulty in the following manner. It recorded that SAIA represented the bulk of the short-term insurance industry. It then said:17
‘It was thus not viable for a firm to enter into this market backed only by insurers who were outside of SAIA. No evidence was led that any insurer broke ranks to do so except for the efforts of De Meillon of Auto & General. Even Auto & General had, whilst interested at one stage in Tracetec, relied on Matrix as a partner. Thus Tracetec was not looked at in isolation but through the lens of an incumbent. No more than uninsured motorists would non-SAIA insurers want to consider non-VESA firms unless they were considering a joint venture of the sort contemplated by Auto & General, which in any event did not work out.’
There was no evidence whatsoever justifying the first sentence in this passage. Mr Pienaar had said that Auto & General, Budget, First For Women and Outsurance did not require VESA accreditation. Tracetec had identified a number of other insurers and insurance brokers who did not require VESA accreditation and were allegedly willing to deal with it. In the absence of information showing that the share of the insurance market held by these firms was so tiny that access to them would not enable a new entrant to secure a customer base of 3000 clients there is no justification for the finding that it was not viable for a firm to enter into this market via insurance companies who did not adhere to the SAIA standard.
 The evidence of Mr De Meillon was that, from the point of view of the insurance companies that he represented, VESA accreditation was never an issue or requirement. What he was concerned with, before being willing to recommend a vehicle tracking and recovery product to its insurance customers, was that the provider could provide an effective, viable and sustainable service. The investigations he made, in conjunction with Matrix into the possibility of investing in Tracetec, were as a result of approaches by Tracetec itself. His conclusion was that:
‘In 2002, Tracetec did not have operational and functional technology capable of being marketed and sold nor did it have the financial resources required to implement a viable and sustainable vehicle tracking and recovery service.’
In other words Mr De Meillon’s evidence was that the insurers that he represented, which included Auto & General, Budget, First for Women and Outsurance, were perfectly happy to use an SVR provider that was not accredited by VESA provided their system and support structure was adequate. This provided no support for the conclusion by the Tribunal that it was not viable for a firm to enter into this market without VESA accreditation.
 A consideration of Mr De Meillon’s evidence takes one conveniently to the other aspect of the Commission’s case, namely that there were a number of aspirant entrants to the SVR market whose path was blocked by the existence of the standards. Of course, the primary case was that Tracetec was such an entity. Indeed that was the entire basis for Tracetec’s original complaint to the Commission. However, that case collapsed not so much with a bang as with a whimper when at the outset of the proceedings before the Tribunal leading counsel for Tracetec said the following;
‘… it’s not possible to say what the position might have been in the absence of VESA’s track record criteria in 2001 and 2002. So, we are not going to seek to persuade you that Tracetec would have entered the market, would have been successful during those two years. As a fact we were not able to roll out the physical infrastructure to any significant degree prior to July 2003 when Legatt joined and provided the financing.
So, any statement by Tracetec in its documents to the contrary, in other words that Tracetec was in a position to go out and into (enter?) the market and compete is incorrect and we don’t intend to rely on any of those statements. Spending on infrastructure started in July 2003 when Legatt joined, but it was limited by the fact that there was nobody’s approval until October 2003 when the BMW contract was landed and Tracetec was able then to spend more money and start building up its physical infrastructure.
We are able to say that but for the SVR restrictions, which were in place at that stage, we would have been able to market the product fully by November 2003, October/November 2003.’
 Notwithstanding that concession, the accuracy of Tracetec’s claims in documents it issued to investors and in its endeavours to obtain membership of VESA, whether in the SVR committee or otherwise, were extensively canvassed in the evidence and particularly in the cross-examination of Mr Pickering. It was demonstrated beyond doubt that the claims Tracetec had persistently made in regard to its product and its ability to enter the market at the relevant time were simply, and on some occasions deliberately, untruthful. The concession by leading counsel at the outset of the case was not only fully warranted but was thereafter borne out by the demonstrated facts. Insofar as the case before the Tribunal rested upon the standards for VESA accreditation obstructing Tracetec in its endeavours to enter the market that case was abandoned by Tracetec and destroyed by the evidence. It is unfathomable in those circumstances on what basis the Tribunal could conclude in paragraph  that Tracetec was disputing the allegation that it was in no position to enter the market. It was also wrong for it to say:
‘It is not necessary for us to determine whether Tracetec could have successfully entered the market at the relevant time, but for the VESA approval criteria. Rather the question is posed in more general terms; could a reasonably efficient firm or a firm at least as efficient as one of the SVR respondents have entered the market without VESA approval or, if not, could it have obtained VESA approval within a reasonable time period?’
The complaint by the Commission and Tracetec had not been brought before the Tribunal on that footing and it was not open to the Tribunal, once Tracetec’s grounds of complaint proved unfounded, to create a different case on a purely hypothetical basis that was not the case that the appellants were required to meet. In addition, the manner in which the Tribunal formulated this question was founded on a premise about which there was no evidence whatsoever and which squarely confronted it with the question of causation. The underlying premise is the presumed existence of a reasonably efficient firm, or one at least as efficient as one of the appellants, wishing to enter the industry. In the absence of any evidence that such a firm existed a discussion of the question whether the absence of VESA accreditation constituted an obstacle to its entering the market was entirely academic.
 Reverting to the case that the Commission actually sought to advance, although it listed an impressive number of firms in its further particulars, it produced witness statements from only two, Karoocell and Cartrack, and was not in a position to call the witness from Karoocell. It followed that once Tracetec made its concession the only tangible evidence placed before the Commission in support of the proposition that the standards substantially prevented competition in the market was that of Mr de Wet from Cartrack. An examination of his evidence is therefore necessary, although the Tribunal did not rely upon it in this part of the enquiry.
 Mr de Wet testified that Cartrack was established in July 2001. It was a start-up enterprise with a product it had itself developed. It made no formal application for admission to the SVR committee and VESA before 2003. According to him, he spoke to Mr Oosthuizen, the chief executive of VESA, in the latter stages of 2001 about obtaining VESA accreditation for Cartrack’s product and was told that Cartrack did not comply with the standards. No written application was ever made. He says that approaches made to insurers by Cartrack were always rebuffed on the basis that their product was not VESA accredited. However, he gives no indication of how extensive these approaches were. Instead Cartrack moved into other neighbouring countries during 2002 and 2003 and applied for membership of the SVR committee when the requirements were relaxed to permit the furnishing of a guarantee. Even then it took ten months after its application was approved to launch its product in South Africa.
 It is difficult to assess what weight to attach to the evidence of Mr de Wet. The problem is that it suffers from a complete lack of verifiable detail. In his statement he mentions a number of insurers that were approached in January 2004 after Cartrack had obtained VESA accreditation. However, he gives no detail in either his statement or his evidence of which insurers he approached between 2001 and 2003. What he said in his witness statement was that it was common knowledge at the time that the South African insurance industry would only accredit products that were VESA approved. Mr Pienaar, who was far better qualified than he to speak to this issue, regarded this as an overstatement. Mr de Wet said that the SVR market was driven by the insurance industry and that once one started a marketing drive it was the insurance companies that had to be approached for approval. He went on:
‘… if you didn’t have VESA approval at the time, you basically wouldn’t even be able to talk to them about it.
That was always the first question that they asked you, anybody would ask you, are you VESA approved. If you weren’t VESA approved, that would basically be the end of the conversation. It wasn’t just the insurance industry, but that was by far the major issue.’
However, this was only one of the hurdles facing Cartrack. A second was in persuading manufacturers that its product could be fitted in their vehicles and a third was the need to obtain information from the cellular networks to operate the system. That in turn required Cartrack to persuade the cellular networks that it had a viable business.
 The route adopted by Cartrack was to launch its operations in countries other than South Africa. Mr de Wet gave no evidence regarding the basis upon which this was done or whether it required the co-operation of insurance companies operating in those countries. One significant aspect of his evidence is that Cartrack only launched its operations in South Africa two months after the appellants withdrew from VESA and the SVR committee so that VESA accreditation ceased to be a significant factor in developing a business in the SVR market. That flows from the fact that the members of SAIA were happy to support the appellants even though they were no longer members of VESA. They were aware of the circumstances in which the appellants had withdrawn from VESA and their concerns that access to the SVR committee by way of a financial guarantee was no guarantor of an SVR firm’s performance, stability or reliability. It is difficult then to attribute any part of the breakthrough achieved by Cartrack to its acquisition of VESA accreditation. Mr de Wet says that by January 2005 Cartrack had over 4500 South African members and was installing 800 units per month. In 2007 when he signed his statement the number of members had risen to 45 000 and they were installing 3000 units per month. When he gave evidence in November 2008 Cartrack had just under 110 000 subscribers and was installing just over 5000 new systems every month. Its major growth and development has therefore been entirely during the period when, according to the Tribunal’s findings, VESA had ceased effectively to be the approval body for the industry. It is difficult to see on what basis therefore Cartrack’s development in South Africa after its launch in July 2004 can be attributed to its having secured VESA membership and accreditation.
 That difficulty is compounded by an examination of the available information regarding the growth of the market for SVR systems over this period. The figures contained in the report by Mr Hodge, an economist who gave evidence on behalf of Netstar and Tracker, was that the market for SVR systems was expanding rapidly. That expansion continued, if anything even more dramatically, after July 2003, when membership of the SVR committee was made easier in consequence of the introduction of the alternative of a guarantee. It continued after the three appellants withdrew from VESA and the latter’s regulatory role was effectively ended. The picture is therefore of a rapidly expanding market in which new entrants, such as Cartrack, were able to expand their customer base rapidly even though VESA membership had become largely irrelevant. Those facts weigh against a conclusion that during the period when the standards were being applied by VESA from March 2000 to July 2003 the conditions operated to prevent competition.
 On this issue it is also appropriate to consider the firms identified by the Commission as having complained about the 3000 subscribers requirement. Global Telematics obtained VESA approval in the fleet management category in February 2001. In May 2001 it requested membership of the SVR category but at that stage did not satisfy the requirement of 3000 clients. By May 2002, however, it had achieved the requisite number of clients and was admitted to the SVR category. Its experience does not suggest that it was impossible for new entrants to achieve the standards as it did so during their heyday. Karoocell (which later changed its name to Mobile Tracker) only approached VESA for membership of the SVR committee at some stage near the middle of 2003 and it was an early beneficiary of the introduction of the financial guarantee. In the witness statement of its Mr Tafur, other than broad generalities, there is no indication that it took any steps prior to 2003 to enter this particular market. Mr Tafur was not available to be called as a witness and the Tribunal ruled that his statement should be disregarded.
 The other firms identified by the Commission can be dealt with reasonably quickly. Cellstop applied for VESA approval in February 2003 and their application was referred to the VESA board by the SVR committee. They were offered membership in the fleet management category but rejected this. They offered a financial guarantee and when the standards were changed to permit this they became a member of the SVR committee in June 2004. Orbtech was a member of the fleet management committee but resigned in February 2002. It only applied for membership of the SVR committee in September 2003. Mobile Data was a joint venture that was only launched in May 2004. LST only applied for membership in January 2004 and was only able to tender a guarantee payable over 6 months. The minutes reflect that there were concerns about its product from a technical perspective. Global Asset Protection expressed interest in about July 2003 but do not appear to have pursued it. No information was placed before the Tribunal in regard to Star Track.
 Again it is not surprising that the Tribunal did not place any reliance on the experience of these firms. There was not the slightest indication that they were in a position to enter the SVR market during the period from 2000 to July 2003. To the extent that any of them were trying to break into that market there was no evidence of their capacity to do so, what steps they took to achieve entry and, to the extent that this was the position, the reason for their failure to achieve entry. Insofar as there was any evidence it lay in the experience of Global Telematics, which was that, notwithstanding the fact that it was not a member of VESA in the SVR category, it was able to achieve the admission criteria during the period between May 2001 and May 2002. It seems likely that it did this by developing its existing fleet management business. That this was a possible way of obtaining entry into the insured SVR market was in accordance with the evidence of Mr Hodge on behalf of Netstar and Tracker. However, the Tribunal attached no weight to this evidence.
 In those circumstances the evidence before the Tribunal did not sustain the factual basis for the Commission’s complaint. The evidence in regard to the actual situation in the market for SVR systems during the period from March 2000 to July 2003 did not disclose that the existence of the standards substantially prevented or lessened competition in that market. The major contender for such a situation was Tracetec and its concession at the commencement of proceedings fatally damaged the Commission’s case in that regard. Beyond that there was only the limited evidence from Cartrack and some evidence about Global Telematics, both of which developed their business notwithstanding the existence of the standards.
 Some reliance was placed on the fact that, once the standards were altered by permitting a guarantee, several firms applied for and obtained VESA membership. However that is no indication that the standards had prevented them from competing in the market. It only indicates that they perceived a possible advantage in VESA membership. On its own it gives no indication that in the absence of the standards there would have been a greater level of competition in the market. Had there been no standards the appellants could quite freely have marketed their products as being the only ones with an established track record in terms of the number of vehicles to which their systems were fitted and their rates of recovery. That would have confronted the new entrants with precisely the same difficulty in dealing with the insurance companies, namely that their products were unknown and untested, with no track record to back them up. They were entirely dependent on persuading insurers, brokers, fleet managers and the market generally that they should to adopt a colloquialism ‘be given a chance’. Both Ms de Silva and Mr Pienaar said that, in the absence of the formulation of standards through VESA, insurers would have formulated their own standards before recommending the installation of a tracking system and giving a reduced or discounted premium for doing so. In view of the expense involved in doing that the probability is that they would have preferred the safe offerings of the large companies, namely the appellants, to those of new start-ups who could give them no comfort in regard to their long-term viability and effectiveness. In other words in the absence of some evidence that tended to show that insurers would, but for the standards, have been willing to use new entrants, the case is taken no further by these companies joining VESA by way of the mechanism of providing a guarantee.
 In my view where the Tribunal went wrong is that it approached this question on a theoretical basis concerning the potential effect of standards in a market instead of examining the factual basis of the complaint that had been referred to it. It commenced its discussion of the issue by saying that:
‘The theory of harm advanced in this case is that the standards operated to exclude rival firms from effectively entering in the market to compete with incumbents who had been approved in terms of the standard.’
From there it engaged in a largely academic discussion of the effect of standards on competition and sought to draw a distinction between benign and malign standard setting. In doing so it appears to have been significantly influenced by the approach in other jurisdictions and in particular the various tests adopted by courts in the United States of America in implementing its rule of reason approach to standard setting.
 The flaw in this approach is twofold. Firstly, it drew the Tribunal away from a consideration of the facts of the complaint that had been referred to it with the result that it did not consider whether the Commission’s evidence supported the complaint. Secondly, it substituted a theoretical investigation of the effects of standard setting for a factual enquiry into the question whether there had in fact been any substantial prevention or lessening of competition in the market for the supply of SVR systems as a result of the application of the VESA standards. This is not to say that reference to foreign authorities and the approach taken in other courts may not be helpful in standard setting cases. However, where the basis of the complaint is that particular firms identified in the complaint were constrained in their efforts to enter an identifiable market the enquiry for the Tribunal must revolve around those factual allegations not hypothetical tests for determining the general impact of standards in a market. Those tests may be helpful in other cases but they found little purchase in this case where the Commission’s complaint was that specific entities had been prevented from entering the market.
 In the light of these problems it is necessary once again to emphasise that the Tribunal is not at large to decide whether conduct is anti-competitive and then to formulate reasons for that finding. It is, as this Court said in Mittal Steel,18 bound to apply the Act and engage with the issues as they arise from a proper construction of the Act’s provisions. It does so in the light of a specific complaint that has been referred to it for determination and its only function is to determine whether, in the light of the Act’s provisions and the evidence placed before it or obtained by it pursuant to the exercise of its inquisitorial powers, that complaint is made out. Not only did it not do that in this case, its approach to the evidence was flawed. A reading of its decision in the light of the record conveys that it highlighted every piece of evidence that could be construed as casting the conduct of the appellants in an unfavourable light and disregarded or discounted all the evidence to the contrary. There is considerable force in the criticism by the appellants that in places it substituted speculation for legitimate inference. In conjunction with its theoretical approach to the task before it that leaves the unfortunate impression that it was concerned to mould the evidence to fit its theory of harm and the principles it was seeking to apply rather than adjudicating on the case before it. That is something that should be strictly guarded against in any adjudicative body.
 In Tracetec’s argument it was submitted that the standards were irrational and anti-competitive. Reliance was also placed on the repeatedly stated concerns of the appellants, as well as members of the VESA board, that the standards might be seen as anti-competitive. However this does not assist. The fact that the appellants were concerned not to be seen as engaging in anti-competitive conduct and were anxious about the consequences if they did so cannot create an agreement or concerted practice, nor can it demonstrate, absent clear independent evidence, that competition was in fact prevented or lessened. And unless there was an agreement or concerted practice and competition was prevented or lessened thereby the irrationality of the standards was immaterial. It is entirely proper for firms engaged in a market to have an ongoing concern not to fall foul of the provisions of the Act and the fact that they have such concerns and adopt a conservative approach to that possibility does not of itself demonstrate that the concerns are well-founded. Whilst a business that sets out to behave in an anti-competitive way is unlikely to receive sympathy from the Tribunal when it contends that it failed in its endeavours a firm that is anxious not to do so cannot be condemned by reason of that anxiety.
 In my view there was no sufficient basis in the evidence for the Tribunal to conclude that there was a substantial prevention or lessening of competition in the market for the supply of SVR systems during the period from April 2000 to July 2003. The evidence does not show that there existed in that market during that period any actual or potential entrant that was hampered in its efforts to enter the market by the existence of the standards. As discussed in the previous section of the judgment the complaint was made and referred to the Tribunal principally on the footing that Tracetec was prevented from entering the market by the existence of the standards and its consequent inability to obtain membership of the SVR committee and VESA accreditation in that category. That complaint collapsed and the other evidence tendered by Tracetec and the Commission was insufficient to show that any of the parties identified by the Commission had been hampered in securing entry to or competing in the market for SVR systems as a result of their lack of VESA accreditation. It was accordingly insufficient to resuscitate the complaint. The resort to a hypothetical ‘reasonably efficient firm or a firm at least as efficient as the SVR respondents’ was not only not open on the terms of the reference but the evidence failed to disclose that such a firm existed or could have been brought into being had the standards not been in force.
 On that ground alone the necessary causal link between the agreement that I have assumed for purposes of this judgment existed among the appellants and any prevention or lessening of competition in the market does not exist because no such interference with competition was established. However even if it could be said that there was some interference with competition in my view it is plain that the necessary causal connection between that and any conduct on the part of the appellants has not been established.
 I reach that conclusion because of the key role that SAIA and the insurers that required VESA accreditation for SVR systems played in the process. Whilst the insurance companies were not themselves the customers for those systems they drove the decisions by their clients both to install such systems and as to the suppliers with whom they would contract. That is because the reduced or discounted premiums would only be made available to persons seeking insurance if the system they installed was one that their insurer approved. Accordingly if an insurer required a VESA accredited system that is what would be installed.
 It necessarily follows that the effect of the standards does not arise from their mere existence but from the stance that insurance companies took to them. It was the insurance companies in the first instance and not the tracking companies that required standards. On the face of it the appellants were perfectly happy to continue to develop and market their businesses without such standards. No doubt in the course of doing so they would have boasted of their track records and stressed to the insurers the advantages of recommending their systems ahead of those of competitors. There are documents in the record that demonstrate that this was the marketing strategy of both Netstar and Tracker. When confronted with newcomers to the industry they would no doubt have emphasised the latter’s lack of a proven track record and the limitation of their coverage and any other weaknesses they perceived in their offerings. A warning against making use of potential ‘fly by nights’ would have been sounded. That is what competitive marketing involves. Had the insurance companies through SAIA not made it clear that they wanted standards it seems unlikely that the appellants would have become involved in developing standards at all. If standards had been developed in that situation it would have been by the insurance companies themselves or by a separate agency.
 Not only did SAIA drive the process that lead to the creation of the standards but it was involved in determining their content. When it regarded the proposals as unduly stringent for its purposes it required them to be watered down and Netstar accepted this. Once the standards had been developed and the SVR committee established it was left to the individual insurance companies to decide whether or not they would require VESA accreditation for SVR systems. Mr Pienaar made it clear that Santam’s decision to do so was Santam’s own decision not that of the appellants. Some insurance companies were clearly prepared to make use of SVR systems that were not accredited, as were some insurance brokerages. The latter were important because they directed clients to the insurance companies. Of central importance is that the mere existence of the standards had no effect at all. What was crucially important was that the insurance companies accepted the standards as providing the measure of comfort that they were seeking so that they would require their clients to install VESA approved SVR systems.
 Mr Loxton SC, on behalf of Tracetec, focussed in his submissions on this issue of causation. His approach, like that of this judgment, was first to apply the ‘but for’ test and then to examine the impact of other factors. He accepted that the insurers’ decision to accept the standards and require that VESA accredited SVR systems be installed in their clients’ vehicles was itself a causa sine qua non of the effects of the standards in the market. As he put it in response to a question from the bench: ‘If SAIA was not on board there would be no effect in the market.’ It is the acceptance by SAIA of the standards and the decision of their members to make use of VESA accreditation that caused the effect in the market.
 It is probably incorrect to describe the role of SAIA and its members as an intervening cause of any market effect as they participated throughout and their involvement cannot be separated from those of VESA and the persons who participated in preparing the standards. However s 4(1)(a) does not appear to contemplate concurrent necessary causes one of which contravenes the section and the other of which does not. In any event that is an artificial approach to the section. As dealt with earlier the proper construction of the section requires that the dominant or real and substantial cause of any prevention or lessening of competition in the market be identified. In the present case that involves asking whether it was an agreement among the appellants that did this or the conduct of SAIA and its members.
 In my view it is clear that the conduct that had the closest connection with any prevention or lessening of competition was that of SAIA and its members. Whilst the preparation of the standards was a necessary element, that flowed from SAIA’s needs not those of the appellants. SAIA participated in the preparation of the standards, assisted in their launch and its members, or the bulk of them, decided to use VESA accreditation in giving reduced or discounted premiums. The moment SAIA changed that stance the standards could no longer play a role in determining competitive conditions in the market, as occurred when it agreed to continue to endorse the appellants’ systems after they withdrew from VESA. As counsel put it, without SAIA there was no effect in the market. It follows that any prevention or lessening of competition in the market as a result of these standards being implemented was brought about by SAIA and its members and was not an effect of any agreement among the appellants.
 That conclusion is consistent with the evidence in regard to the competitive effect of standards that was placed before the Tribunal by Mr Hodge. He made the point, referring to United States sources, that standards cannot generally be viewed as anti-competitive when the party or parties that drive their implementation in practice are the customers or users of those standards. In this case it is SAIA and its members that played that role, acting on behalf of their clients and in their own interests. Mr Hodge also pointed out that there was never an obligation on potential entrants to comply with these standards. They could always seek to advance their business through other channels, as Global Telematics and Cartrack did, and it was always open to them to seek to persuade the insurance companies that their systems, albeit new and relatively untried, had advantages that made them worthy of endorsement.
 For those reasons no contravention of s 4(1)(a) was established during the period from April 2000 to August 2003. That conclusion means that there was no contravention during the earlier period from September 1999 to April 2000, when provisional approval was possible and no contravention after August 2003 when the financial guarantee was introduced, as it is accepted that if anything any barriers to entry were less during those periods. The appeals must therefore succeed and the determination by the Tribunal be set aside. The cross-appeal must fail. As Tracetec has been unsuccessful in its attempts to sustain the Tribunal’s determination it must pay the costs of the successful appellants. Those costs will include in the case of Netstar and Tracker the costs of two counsel.
 Before completing this judgment it is necessary to make some comment for the future guidance of practitioners in this Court in regard to the production of records and heads of argument. In regard to the record we were confronted with 52 volumes running to 5873 pages. In case that was thought insufficient three further volumes totalling 465 pages were added when the heads of argument were delivered. No significant reference was made to these volumes in either the heads or the oral argument. Rule 20(1)(h) of the rules of this Court provides that only documents referred to in the Tribunal proceedings should be included in the record. This was ignored and we were presented with vast swathes of transcripts of meetings when at most a page or two had been referred to in evidence. Rules 20(2)(b)(vii) and (viii) say that neither opening addresses nor closing argument should be included in the record. Both were, adding 418 pages to the record. Rule 20(4) says that the documents referred to in the Tribunal must be arranged in chronological order. They were not. That meant that they could not be perused quickly in order to get an overall impression from the documents of the course of events.
Rule 20(5) provides that:
This was done imperfectly by way of the index (although the members of the Court were not told this and accordingly had to discover it for themselves). However the index provided no means of cross-referencing a document to the passage or passages in the record where a witness referred to it. By way of example Mr Pickering referred to a VESA board meeting on 20 August 2003. The transcript of the meeting runs to 130 pages over two volumes. Apart from the occasional first name the speakers are not identified and one has to try and identify them from the cross-examination. Finding the passages on which he was being examined was almost impossible. The reason references are required to be in the record of the evidence is to make finding documents while reading relatively easy. Instead it was rendered extremely difficult as one had to go back to the index in order to find the document being referred to and then search the document for the relevant passage. The difficulty was compounded by the fact that the documents were not in the same form as the bundles presented to the Tribunal, nor did they contain the original references that they had borne in those proceedings. In the witness statement prepared by Price Metrics portions constituting confidential material were omitted. All of this made it exceptionally difficult to read the record and resulted in a far greater expenditure of time than would have been necessary had the rules been followed. And this in a case where all counsel agreed that there was little or no dispute of fact on the papers. Notwithstanding that fact there was no helpful response to the Judge-President’s directive that we should be told what needed to be read and what was irrelevant and could be disregarded.
 Turning to the heads of argument they were in some respects excessive and missed the point of heads.19 In total they ran to over 700 pages with the shortest being 69 pages and the longest a touch under 200 pages. This mountain of paper arrived in the three weeks immediately prior to the hearing together, in the one case, with a further three volumes of the record. It is simply impossible for judges of this Court, who have no separate time for reading and preparation from their duties in the provincial divisions, to prepare adequately on the basis of these volumes of paper. If this case were to go further to the SCA the parties would be confined to 40 pages for their heads of argument and in our own revised practice directive 70 pages is set as a maximum, thereby recognising that a greater engagement with the evidence may be necessary in competition cases. In addition our practice note requires the provision of a chronology and will now be further revised to mandate the provision of a core bundle of documents and oblige counsel to advise the court of the portions of the record that do not need to be read for the purposes of the appeal. In future, practitioners in this Court will be required to ensure that their heads of argument comply fully with the requirements for such heads laid down in the revised practice note.
 I contemplated whether the successful parties should be deprived of some of their costs arising out of the situation described in paragraphs  to . Not without some hesitation, as counsel who appeared in these appeals are senior and experienced counsel, their attorneys are from major firms and the Commission itself and all concerned should have known better, I have decided not to do so on this occasion as the parties had not been warned that the Court might take this approach. In future, however, if there is non-compliance with the rules in regard to the preparation of the record or non-compliance with the requirements now laid down for the preparation of heads of argument practitioners must be aware that adverse costs orders may follow.
 The following order is made.
The appeals are upheld.
The Tribunal’s orders in paragraphs [316.1] and [316.3] of its determination are set aside and replaced by the following order:
‘(i) The Commission’s complaint referral against Netstar, Matrix and Tracker is dismissed.
(ii) Tracetec’s complaints against Netstar. Matrix and Tracker are dismissed, with costs, such costs to include those consequent upon the employment of two counsel where two counsel were employed. Mr Hodge is declared a necessary witness.’
The cross-appeal by the second respondent is dismissed.
The second respondent is to pay the appellants’ costs of the appeal and the cross appeal, such costs to include in the case of Netstar and Tracker those consequent upon the employment of two counsel.
M J D WALLIS
ACTING JUDGE OF APPEAL
DATES OF HEARING 29 and 30 NOVEMBER 2010
DATE OF JUDGMENT 15 FEBRUARY 2011
FIRST APPELLANT’S COUNSEL John Campbell SC (with him Anthony Gotz)
FIRST APPELLANT’S ATTORNEYS Bowman Gilfillan Inc
SECOND APPELLANT’S COUNSEL Jerome Wilson
SECOND APPELLANT’S ATTORNEYS Werksmans
THIRD APPELLANT’S COUNSEL Mike van der Nest SC (with him Alfred Cockrell SC)
THIRD APPELLANT’S ATTORNEYS Cliffe Dekker Hofmeyer
FIRST RESPONDENT’S COUNSEL D I Berger SC (With him P L Mokoena)
FIRST RESPONDENT’S ATTORNEYS Competition Commission
SECOND RESPONDENT’S COUNSEL C D Loxton SC (with him R M Pearse)
SECOND RESPONDENT’S ATTORNEYS Knowles Husain Lindsay Inc
2Whilst Mr de Clerk did not ultimately give evidence his statement was available to the Tribunal and it was entitled to have regard to it in terms of s 54(f) of the Act.
3The Tribunal found (para ) that this letter, which was copied to Tracker and Matrix, was probably ‘the product of joint discussion between the three firms and was being sent to Bezuidenhout for his buy-in’. There is no evidence in support of this and it is nothing more than speculation.
4It is not correct to say, as do Philip Sutherland and Katherine Kemp, Competition Law of South A frica, (loose-leaf, issue 10) 5-17 that ‘there is no need to draw a clear line between arrangements and concerted practices’ as such a line is drawn by the language of the definitions. Nor are concerted practices merely diluted agreements (5-25).
5Sutherland and Kemp, supra,5-9 to 5-11.
6Sutherland and Kemp, supra, 5-14 to 5-15 where they deal with the approach of the European Commission and the European Court of First Instance to the topic of agreement.
9The essential question is whether the issue was raised with sufficient clarity not whether it was described by a term understood in the area of competition law. See Senwes Ltd v Competition Commission of South Africa  ZACAC 4, paras  to .
11This may be, but is not necessarily, the same as anti-competitive consequences to which it is equated by the authors of Sutherland and Kemp, supra, 5-60 to 5-61. It is safer in the circumstances to focus on the expression actually used in the section rather than replace it with one that may not be identical in meaning. For the same reason there are dangers in the uncritical adoption of the American terminology of a ‘rule of reason’ approach to describe the scope of s 4(1)(a). Unlike the Sherman Act, which these authors say (2-11) is ‘widely formulated and it leaves much to the imagination of the courts’ the South African statute is fairly precise in describing what constitutes a horizontal or vertical restrictive practice. If each case is examined in the light of the statutory requirements there is less prospect of falling into error.
14The principal disagreement was between Netstar and SAIA over the suggestion that 5000 fitments were necessary. SAIA did not agree and this was watered down to 3000.
15Sutherland and Kemp are emphatic that only horizontal relationships are covered by the section (5-6 to 5-8) but do not deal with this type of situation.
16According to Mr Pienaar Santam, Mutual and Federal and Zurich together had slightly less than 50% of the market for motor vehicle insurance at the time. There was no evidence of the stance taken by Mutual and Federal and Zurich to VESA accreditation.
17In para  of its Determination.