Allied Technologies Ltd and Namitech Hol

The judgment

The court's own file

1 COMPETITION TRIBUNAL REPUBLIC OF SOUTH AFRICA Case No: 37/LM/Jul03 In the large merger between: Allied Technologies (Pty) Ltd and NamITech Holdings Limited ______________________________________________________________ Reasons for Decision CONDITIONAL APPROVAL The proposed transaction between Allied Technologies (Pty) Ltd and Namitech Holdings Limited was conditionally approved by the Tribunal on the 5 February 2004. The reasons for this decision follow. BACKGROUND AND SU MMARY 1. This merger involves the acquisition by Altech, a member of the Altron group of companies, of Namitech, a subsidiary of the packaging company, Nampak. Altech will acquire Nampak’s 51.08% shareholding in Namitech. Altech will also acquire the 28% stake in Namitech that is held by Clidet, a special purpose vehicle through which a black economic empowerment entity, Pamodzi, participates in the ownership of Namitech. Therefore, post-merger, Namitech will, prior to the re- introduction of Pamodzi Investment Holdings Limited (“PIH”), become a 79.08% held subsidiary of Altech. 1 2. The products implicated in this transaction are the ubiquitous plastic cards predominantly (although not exclusively) used, in their most 1 Presently, Pamodzi holds its 28% interest in Namitech via Clidet, a special purpose vehicle commonly used for BEE funding structures. Altech has signed agreements for the acquisition of Pamodzi’s indirect shareholding of 28% however it is expressly intended that Pamodzi should, post- merger, retain its stake in Namitech. However, in the post-merger structure, Pamodzi, instead of holding its investment in Namitech via an SPV, will subscribe for new preferred ordinary shares in Namitech directly. Pamodzi will therefore be entitled to hold 28% of Namitech’s total number of shares and to have 28% of the total votes of Namitech shareholders. There are other favourable benefits accruing to Pamodzi via the new structure which need not detain us here. The upshot is that it allows Pamodzi to become direct shareholders in Namitech with a much reduced financial outlay and reduces its debt burden. Pamodzi will therefore have the same voting and other rights as under the existing arrangement. 2 basic form, as membership and loyalty cards (for which the blank card is simply embellished with artwork that identifies the purpose and origin of the card and which are termed ‘basic cards’) and as the basic input to which magnetic strips (‘magstripe cards’) and silicon chips (‘smart cards’) are applied. Magstripe cards are most commonly associated with a variety of bank cards – ATM cards, credit and debit cards and garage cards. The magnetic strip enables the cards to be ‘read’ by ATM and credit card terminals. The silicon chip embedded in a smart card enables the card to perform certain advanced functions and is most commonly associated with telecommunications where these cards are used as SIM cards in the mobile segment of the market and as pre-paid cards for use in fixed line public telephones. It is predicted that the banking sector will soon ‘migrate’ to smart card use. 3. Note that the Commission, in its definition of the relevant market, distinguishes between ‘secure’ and ‘non-secure’ basic and magstripe cards. For the most part, secure cards are those magstripe cards used in the banking sector. There are clearly security considerations that attach to the production and distribution of all bank cards. However, particular security requirements attach to the production of credit and debit cards where the card manufacturer requires formal certification from the card association, for example, Visa or Master Card or American Express 2 . The Commission classified non-secure PVC cards as those cards used in retail, medical, leisure and other types of loyalty type applications. As will be elaborated below, a particular level of security accreditation – so-called EMV certification - is required for the production of (credit and debit) smart cards for the banking sector. 4. For each of these cards to function effectively, they have to be ‘personalised’, that is, information pertinent to their ownership and functions has to be incorporated on the card in some form or another. The technology required for personalization varies significantly – in the case of the basic card this simply involves a laser printing or embossing process, whereas the personalization of magstripe and smart cards requires significantly more sophisticated know-how and equipment. 5. Namitech, the target firm, is a leading South African producer of basic blank PVC cards, magstripe cards, and smart cards. Namitech both manufactures the card and offers personalization in respect of each of these card types. 6. Africard, a subsidiary of Altech, the acquiring firm, is also a significant South African producer of the basic card as well as magstripe cards. It also offers personalization services in respect of these card types. 2 These requirements include strict provisions around all aspects of securing vendor premises. They range from specifications regarding selection and training of employees, to specific requirements regarding securing the vendor’s premises and maintaining high security restricted areas, to stipulations with respect to monitoring production procedures and maintaining rigorous audit trails and product storage and shipping requirements. MasterCard Security Standards for Vendors July 2003. 3 However it does not produce smart cards, although it does import a relatively small number of these products which it supplies to the gaming industry. 7. Currently, then, only non-secure smart cards are being supplied locally, mainly for the telecommunications, loyalty, security and leisure industries. 8. The Altron group – of which Altech is part – is heavily invested in supplying telecommunications equipment but has little exposure to mobile telecommunications. This transaction will, through Namitech’s powerful position in the provision of smart cards to the mobile telecoms industry (commonly referred to as ‘SIM cards’), provide it with instant entry to that segment of the industry. However, it is important to note that this transaction coincides with the imminent migration of banking cards from magstripe to smart cards. 9. The Commission was initially concerned that this merger would result in an unacceptable degree of concentration in the market for the manufacture and supply of basic PVC cards, and, particularly, for the supply of magstripe cards (predominantly used in the financial sector). In addition, the Commission was concerned that the post-merger market position of the merged entity would, upon expected deployment of smart cards in the financial sector, create a platform for the monopolisation of this potentially huge market. 10. Furthermore, vertical concerns arose from the fact that Altech, through its ACS division, is the exclusive agent for Datacard personalization equipment. The concern was that the merged entity would be in a strong position to raise the costs of their rivals many of whom use Datacard equipment. Moreover, the fact that after-sales service is provided exclusively by the agent, will, it was feared, give Altech personnel access to confidential business information when they enter their rivals’ plants to service their equipment. 11. However, in November 2003, the parties negotiated a settlement agreement with the Commission, which, they both maintained, would ameliorate the anti-competitive consequences of the merger that had been initially identified by the Commission. THE AGREEMENT BETWEEN THE COMMISSION AND THE MERGING PARTIES 12. The agreement between the Commission and the merging parties provides for the merged entity to dispose of the Africard division. In other words, the acquiring company, Altech, will dispose of the card manufacturer, Africard, that it brought into the merger, while retaining Namitech, the card manufacturer that it has acquired from Nampak. 4 13. The agreement stipulates that Africard be sold to an independent purchaser approved by the Competition Commission. In essence, the Commission must be satisfied that the prospective buyer possesses the financial resources, proven expertise and the incentive to maintain the divested business as a viable and active competitive force in competition with the merging parties. Note that Africard’s personalization bureau will not be included in this sale but will remain part of the newly merged entity under the control of Altech. 14. The parties and Commission contend that this remedy will ensure that, post-divestiture, there will still be two effective competitors producing magnetic stripe cards both of which will possess the financial resources, technical capability and incentive to supply smart cards to the financial sector when the predicted migration from magstripe cards occurred. 15. Furthermore, the merged entity has agreed that it will terminate Altech’s exclusive right to distribute Datacard personalization equipment in the South African market. The Commission and the parties argued that this remedy would provide the opportunity for Altech’s existing customers and the merged entity’s future competitors to obtain these products and the requisite after-sales services without Altech gaining knowledge of their competitors’ market information, as was originally feared. 16. At the hearing it was expressly agreed that only the transaction in its revised form was under consideration. We are thus evaluating the combined impact on competition of the purchase by Altech of Namitech conditional upon the sale of Africard minus its personalization bureau which will remain with Altech. We are also considering the impact on competition of the transaction after the elimination of Altech’s exclusive right to distribute Datacard personalization equipment. THE PARTIES AND THEIR RATIONALE FOR THE MERGER The Primary Acquiring Firm 17. The primary acquiring firm is Allied Technologies Limited (“Altech”). Altech is a subsidiary of Allied Electronics Corporation Limited (“Altron”). Altron is controlled by the Venter family. Altech in turn wholly owns its IT subsidiary, Altech Data. Altech is a high technology business active in providing products and services to the telecommunications, multimedia, IT and electronic sectors. Its subsidiaries are set out in the organogram below, the only one relevant to this transaction being Altech Card Solutions (“ACS”). 18. ACS, in turn, has a number of subsidiaries. These, too, are set out in the following organogram: 5 * Mastercard, Visa and Diners approved facility. 19. ACS, through its various divisions, is active in the manufacturing and personalization of non-chip PVC-based cards. These include basic cards such as retail and loyalty cards as well as magstripe cards for the banking sector. They also supply imported chip cards to Sun International, a gaming and leisure firm. The Primary Target Firm 20. The primary target firm is Namitech Holdings Limited (“Namitech”). Namitech is a subsidiary of Nampak Products Limited (“Nampak”), itself a 100% subsidiary of Nampak Ltd. Namitech manufactures and supplies a range of secure technology services and products for use in ISIS INFORMATION SYSTEMS ACS Smart Card Technologie s Cardtronic Africard * Card Personalisatio n Manufacture and personalisation of non-secure PVC cards to loyalty and SME market. Provides hardware (terminals) for EFTPOS imported from Schlumberger. ** Used in banking and telecoms. Manufacture and personalisation of secure and non-secure PVC cards; supply chip cards to Sun International. Supplies and maintains DataCard personalisation equipment to banking and retail markets. ALTECH ALTECH DATA 6 telecommunications, financial services and gaming and leisure. The shareholding in Namitech is as follows: 21. Nampak, Namitech’s controlling shareholder, is South Africa’s leading packaging company and is increasingly focused on expansion of its international interests in this large sector. Accordingly Namitech is not part of Nampak’s core focus. Namitech derives 95% of its revenue from the supply of smart cards to fixed line (pre-paid vouchers) and mobile (SIM cards) telecommunications operators in South Africa and elsewhere on the African continent. Smart card technologies are procured by Namitech from Gieseke & Devrient (G&D), a large German-based supplier of smart card technologies, which owns, and will, post-merger, continue to own, 15% of Namitech’s equity. Gaming Telecommunications (95% of its revenue derives from this) Financial Leisure NAMITECH Within these industries, it operates in the IT security, payment and transaction sectors. Clidet 426 NamITech Nampak Mphatllalatsane Investment Co. Vehold Nominees 28% 51.08 % Giesecke & Devrient (“G&D”) 2% 3.92% 15% 100% Pamodzi Investments 7 22. In summary then, Altech is one of South Africa’s leading IT companies, and a leading supplier of inputs to the telecommunications industry. Insofar as the lion’s share of Namitech’s revenue derives from the manufacture and supply of smart cards to the mobile telecommunications sector, the transaction will enable Altech to enter a rapidly growing segment of the telecommunications market, namely mobile telecommunications, from which it has, hitherto, largely been absent. Consolidation of its position as supplier to the telecommunications sector is Altech’s stated goal. Altech avers that its inability to supply smart cards to the telecommunications sector is fatally impaired by its inability to find a suitable international technology partner. 23. Both Namitech and Altech are active in the manufacturing and personalization of basic and magstripe cards for banking and other applications. ACS 3 has approximately 40% of the basic and magstripe card market while Namitech accounts for approximately 50% of local basic and magstripe card sales 4 . They are the only two firms in South Africa certified to manufacture magstripe cards for the banking industry. Namitech is the largest supplier of smart cards - SIM cards – used in mobile telephones. THE PRODUCTS 24. Prior to examining the impact of the transaction on competition, it is useful, for ease of exposition, to describe certain key features of the products implicated in this transaction. A degree of repetition is unavoidable. 25. The basic card is merely a PVC card that carries certain generic information. The AA membership card is a good example of the most basic of these cards. 26. Once a magnetic strip is affixed to the basic PVC card it is transformed into a magstripe card. The magnetic strip enables the card to be read by terminals designed for that function and so are widely used by banking institutions that issue them as ATM, garage and debit and credit cards. Magstripe cards are also used in other, non-banking applications. 27. While the basic card and the magstripe card are largely commodity products, the cost of the card is influenced by the variety of colours or designs specified by the customer. Note that, while basic cards and magstripe cards are relatively easily available in the international market, customer preference for proximity to those responsible for the 3 ACS includes Africard and Cardtronic. 4 Commission’s Recommendations Annexure A.3 8 design and, particularly, personalization of the card, gives local suppliers a competitive edge. 28. The generic PVC card is also utilized as the basic input in the production of smart cards. 29. Smart cards contain an integrated chip that is embedded in the basic card. The parties aver that the cost of the basic card input comprises only 4 -5% of the total cost of the smart card. Smart cards are significantly more technologically sophisticated than magstripe cards, and are defined as intelligent cards with memory, multifunction and greater security features. Mr Vedder of G&D, who testified for the merging parties, described the chip in the smart card as a small PC, replete with an operating system and memory. This permits off-line authentication of the card. Smart cards embody chips of varying sophistication and capability ranging from relatively simple memory cards used principally for memory storage to ‘intelligent’ smart cards that contain a microprocessor capable of storing and securing information. The cost of smart cards predictably varies with the sophistication of the product itself and the level of certification required. 30. In South Africa smart cards are most commonly used in telecommunications applications, notably in the form of SIM cards in the mobile segment of the industry and also as pre-paid cards for use in public phones in the fixed line segment of the industry. Namitech supplies 30% of smart cards to the local mobile phone industry, with global smart card p layers and Prism accounting for 65% and 5% respectively 5 . The logo of the particular network is printed on the card that is then forwarded to the manufacturer responsible for embedding the chip. Note that smart cards for non-banking applications do not carry magnetic stripes. 31. Smart card use is, as we elaborate below, expected to spread into other applications in South Africa. The entry of smart cards into the banking sector is both certain and imminent. Smart cards are also predicted to play a significant role in applications related to welfare payments, national identity cards and e-commerce. Note that the smart card that will be used in banking is, for the foreseeable future, expected also to carry a magnetic stripe. The production process 32. The production process for each of the card types is, up to a point, identical since the same or a similar plastic body is used as the basic input. All raw materials used in the production of the basic card – either PVC or ABS, a slightly less malleable material used for SIM card production - are imported. The aesthetic design features, including, of course, the customer’s logo and hologram, are then affixed to the card. 5 Namitech Witness Statement 9 This is then followed by the addition of the magnetic strip and, in the case of smart cards, by the embedding of the silicon chip. Specific software is loaded onto the smart card for each type of application. 6 33. Finally personalisation of the card follows. This entails loading customer information onto the cards as well as installing the specific applications that will allow scanners to read the cards. In the case of basic loyalty-type cards this is simply a printing or embossing function, however the personalisation of the magnetic strip and the silicon chip naturally involve significantly more complex processes. Personalisation of the magstripe and the silicon chip are distinct process involving specific technologies. The personalisation function need not necessarily be undertaken by the card manufacturer. Indeed, as elaborated below, some of the large banks undertake personalisation of their magstripe cards in-house. Certification 34. While production of all banking cards presupposes security considerations, these are particularly important with respect to credit and debit cards. Hence, in order to produce a debit or credit card formal accreditation /certification is required by a MasterCard or Visa card association. Formal certification is not required to produce other magstripe banking cards, such as garage and ATM cards. Certification effectively constitutes a formal acknowledgement by the certifying authority that the facility conforms to specified security standards. Note that the card fabrication facility and the personalization facility require separate personalization. 35. In order to issue smart cards for banking applications, EMV certification is required. EMV refers to a specification for smart cards that are used as payment cards for Europay, Mastercard and Visa. Therefore, if a firm is in the business of producing EMV chip cards for a bank client, they would be required to have a certification licence for manufacturing the card and a certification licence for personalising the chip. Although the manufacturing and the personalisation activities may be housed under one roof, they are distinctly separate processes with stringent rules governing the movement of cards and personnel between one environment and another. 36. Africard and Namitech are certified to produce and to personalise magstripe cards for Visa and Mastercard. Namitech has EMV certification for the production and personalization of smart cards. Africard does not possess EMV certification. Prism has EMV certification for the manufacture – but not personalization – of smart cards. 6 There are no South African manufacturers of the integrated circuit chips which are embedded onto the smart cards. These are imported from a wide range of overseas smart card technology companies. 10 THE HEARING 37. A pre-hearing was held on 12 November 2003 and the hearing was held on the 13 th to the 16 th January 2004. Closing arguments were delivered on 22 January 2004. 38. The Tribunal called the following witnesses 7 : 1. Mr Andrew Richardson from Roberts Printing Works 8 ; 2. Ms. Sonya Fourie, Head of Card Production and Distribution and Output Management, Nedcor Bank 9 ; 3. Mr. Alwyn Burger Director of Technology and Operations at Standard Bank 10 ; 4. Mr Eric Meniere, Managing Director of Gemplus South Africa 11 ; 5. Mr Mike Serrao, Group Operations Director for Prism Holdings Limited 12 ; 6. Mr Robert Gumede, Chief Executive, Gijima Card Technologies 13 . 7 A representative of Face Technologies was also initially subpoenaed but it was decided not to call him at the hearing due to time constraints and the fact that his evidence was adequately covered by other witnesses. 8 Durban based printer who runs a family business that manufactures non-bank cards, primarily loyalty cards. They have a low volume business and do not compete with the merging parties in the banking card market at all. The evidence of Andrew Richardson of Roberts did not take the matter further since he indicated that being active in the non-secure card market only, Roberts is largely unaffected by the instant transaction. 9 One of South Africa’s big four banks with limited in-house personalization capacity only in respect of its BoE division bank cards. 10 Another one of the big four banks. It is supplied with plastic cards from Altech and Namitech however in respect of magstripe cards, Africard does the personalization. Namitech has been granted the tender to supply them with EMV smart cards within the next few months. 11 Gemplus is a French company, a multinational worldwide leader in smart card manufacture with 16 plants worldwide. Gemplus competes in the smart card manufacturing/supply and personalization markets. It has a close technological association with Gijima in South Africa. 12 Prism is a South African IP company in the sense that it has its own locally grown smart card technology systems for both the GSM cellular and banking industries. One of its major clients is MTN. Prism is both a customer and competitor of the merging parties. It competes in the mobile telephony business with Namitech insofar as both are involved in the supply of GSM SIM cards to mobile operators. It is a customer in the sense that it does not produce plastic card bodies itself, but sources 90% of these card requirements from Altech and NamITech. Prism has Visa certification for the manufacture – but not personalization – of smart cards. 13 Gijima Africa Smart Card Technologies (“Gijima”) is a wholly-owned subsidiary of the Gijima Technologies Group. It is a Black Economic Empowerment smart card manufacturing company. Gijima is focused on telecommunications and does not produce bank cards at all. 11 39. The merging parties called the following witnesses: 1. Mr Craig Venter, CEO Altech; 2. Mr Rex Tomlinson, CEO of Namitech and an Executive Director of Nampak Limited; 3. Ms Diane Jackson, Consultant at Lexecon Limited, UK; 4. Mr Klaus Vedder, Executive Vice President, the Head of the Telecommunications Division and the Deputy to the Executive Director on the Board of the Group Giesecke & Devrient 14 ; 5. Mr Brian Van Rooyen, Group CEO and Acting Chairman of Labat Africa Limited 15 . 6. Alexander Georgiev, Marketing Director of Labat Card Technologies 16 40. The Commission did not call any witnesses. THE IMPACT OF THE MERGER ON COMPETITION Introduction 41. In its original form – that is prior to the conditions agreed between the parties and the Commission – the transaction portended the merging of South Africa’s largest producers of basic PVC cards and magstripe cards. Between them the merging parties also account for a significant share of the market for the provision of card personalization services, although Africard’s share of this latter market is quite significantly larger than that of Namitech. Africard has the largest personalization bureau in South Africa and therefore the greater share of personalization of magstripe cards than does Namitech, Namitech accounts for 1.2% of secure card personalization whilst Africard (ACS) accounts 41%. It has Visa, Diners, MasterCard and American Express certification. 14 G&D is a world leader in the development and supply of smart card technology. Namitech has, through its technology and shareholder relationship with G&D, exclusive use of most of this technology. 15 A black economic empowerment firm involved in IT and high-tech manufacturing. Labat houses SAMES, the only chip manufacturing plant in South Africa, a leader in the design, manufacture and distribution of electronic chips for a range of products from water metering devices to security devices. Labat has for some time been seeking to extend its activities into the card market in order to gain entry into the smart card market with its existing chip manufacturing capacity. 14 Previously head of NAV Consulting, a consultancy that has been incorporated into Labat Card Technologies. 12 42. The Commission appeared to find that, although the actual level of import competition in the secure and non-secure market is limited, there is some import penetration in respect of non-secure cards (non- bank). Prism’s evidence appears to confirm the view that, depending on the exchange rate position, internationally manufactured basic cards may be highly competitive with the locally produced variety. 43. The Commission was, however, more concerned about the merged entity’s position in the supply of secure magstripe cards, that is, magstripe cards supplied predominantly to the banking industry. Imports, in particular, did not seem to offer a realistic counter to an exercise of market power in respect of these products. It is principally this concern that the divestiture of Africard is intended to address. It will, on the face of it, restore the market to its pre-merger situation with the two pre-eminent pre-merger producers of magstripe cards, namely Africard and Namitech continuing their existence as independent, competing entities. However, there are two crucial distinctions between the pre- and post-merger positions: ??Firstly, Nampak will exit the market, leaving its erstwhile subsidiary, Namitech, under the control of Altech, while Africard, Altech’s erstwhile subsidiary, will be in the hands of a third party. ??Secondly, the divested Africard will not retain its personalization capacity and contracts. In terms of the agreement between the parties and the Commission these will remain with Altech – in other words Africard and Namitech’s personalization capacities would be merged. 44. We have to ask ourselves whether Africard, in new hands and without its personalization capacity, will provide at least the same level of competition to Namitech as it did in its pre-merger incarnation. Or expressed more accurately and in the language of the Act, we must determine whether the level of competition likely to be provided by post-merger Africard will not have substantially lessened relative to that prevailing pre-merger. 45. In the smart card market, on the other hand, there are no significant horizontal overlaps. Africard is a very limited participant in the smart card market, with its sole activity restricted to the on-sale of imported smart cards to the leisure industry. Namitech, of course, has a powerful position in the supply of smart cards to the telecommunications market. Hence, on the face of it, no horizontal concerns arise in respect of the smart card market. It should also be noted that, while undoubtedly powerful, Namitech is faced with robust competition in the telecommunications smart card market from, inter alia, Gijima, Prism and Gemplus. 46. Finally, the merger generates certain vertical concerns related to Altech’s role as an exclusive distribution agent for Datacard smart card personalization equipment. On the face of it, these concerns are eliminated by the condition agreed between the Commission and the 13 merging parties which requires Altech to relinquish its exclusive agency agreement with Datacard. 47. However, even after the imposition of the aforementioned conditions, there are, indeed, concerns that remain in each of these areas. These are ??In relation to magstripe cards, will the purchaser of the Africard division be able to sustain the level of competition currently offered by Africard, particularly given Altech’s retention of Africard’s personalization capabilities? ??In relation to smart cards, we must ask whether Altech’s acquisition of Namitech does not remove the most likely entrant, namely Africard, in to the still nascent, but imminently significant, market for the supply of smart cards to the banking sector? In other words, is Africard, in new hands, substantially less likely to enter the banking smart card market than Africard in Altech’s hands? As we shall see, the location of Altech’s personalization bureau also has bearing on this question. ??Finally, we must ask whether by formally relinquishing its exclusive agency arrangement with Datacard, Altech relinquishes its actual position in relation to that equipment. That is, we must ask whether Altech’s competitors will not remain dependent on Altech to service their Datacard equipment even if Altech no longer has the exclusive right to perform these services. 48. We will proceed to answer each of these questions. In relation to magstripe cards, will the purchaser of the Africard division be able to sustain the level of competition currently offered by Africard, particularly given Altech’s retention of Africard’s personalization capabilities? 49. As already indicated, Namitech and Africard are the overwhelmingly dominant suppliers of magstripe cards to the banking sector. Africard’s major customers are ABSA, FNB, Standard Bank and Nedbank. Africard accounts for some 40% of the banking magstripe market. Namitech’s major bank customers are ABSA, Standard Bank and FNB. It accounts for some 60% of the market. Card manufacture and personalization are not necessarily performed by the same supplier. In fact Africard has a significantly larger share of bank card personalization than does Namitech and certain of the large banks undertake personalisation in-house. 50. There is no apparent reason to suppose that the banks would facilitate the monopolization of one of their supplier markets through turning their backs on the post-merger Africard. Commonsense dictates – and this view was, indeed, offered in evidence - that the banks would not wish to rely on a single supplier of an important input, particularly for a service like personalization, where speed and proximity are crucial. 14 51. There is little reason to suppose that Africard, in hands other than Altech’s, would not be able to offer an efficient source of supply of magstripe cards. As already pointed out, this is a commodity business with mature technologies. In fact it appears that the contracts for the supply of the banking magstripe card bodies, for the short-term at least, will remain with Africard. Moreover Labat, the likely purchaser of Africard, will b estow empowerment credentials on the card manufacturer, credentials lacking in its erstwhile owner, the wholly white owned and controlled, Altech. Given the added significance assigned by the Financial Services Charter to empowerment criteria in procurement, there are grounds then for believing that Africard will be more competitive in Labat’s livery than in that of Altech. 52. However there is one area of concern, one element of the transaction that may threaten Labat/ Africard’s position as a supplier of magstripe cards to the banking sector, and that is the retention by Altech of its erstwhile subsidiary’s personalization bureau. Bear in mind that Africard was, pre-merger, far and away the market leader in the provision of personalization services to the banks. 53. The merging parties sought to place the personalization activity in what they viewed as the proper perspective by pointing out that it represented revenues of some R7 million, which, in the context of the entire transaction, is insignificant. Altech’s desire to retain the personalization bureau was, averred the Altech CEO, simply dictated by the fact that the Africard personalization machinery was of a more recent vintage than that of Namitech. He also conceded that Africard’s personalization bureau was a good business, earning solid returns in a sector characterized by low returns, even if, in the total scheme of things, it was small beer. In short his desire to retain Africard’s personalization bureau was, he insisted, dictated by narrow but perfectly legitimate commercial considerations. There were no strategic, anti-competitive designs at stake. 54. Indeed, Labat’s CEO insisted that Altech’s retention of Africard’s personalization bureau actually presented an opportunity to his company. He averred that he had already entered into a technology partnership with an internationally recognized player in this field, a partnership which included the transfer of personalization equipment. 17 He averred that this meant that it would be Africard, in the Labat stable, that would now possess the most contemporary personalization 17 The evidence of Labat CEO, van Rooyen, asserted that the technology agreement was all but formally concluded and would be imminently made public. Our decision is heavily influenced by this evidence provided under oath. The parties, too, relied on this in support of their argument regarding Labat’s viability as a purchaser. We are concerned that, at the time of writing, the promised announcement has yet to be made. The Commission, when assessing the viability of the purchaser of Africard, is advised to monitor this closely. 15 equipment and would immediately be in a position to compete against Altech’s personalization bureau with its relatively outdated equipment. 18 55. Several witnesses, notably Meniere of Gemplus and Gumede of Gijima, took issue with these contentions. These witnesses argued eloquently that the strategic significance of personalization was belied by its relatively small contribution to revenue. Both witnesses identified particular value in the place occupied by the personalization function in the card’s value chain. Personalisation is the direct interface between the card producer and the bank. It is the personalization provider who is entrusted with highly sensitive customer information belonging to the banks. This introduced an element of trust into the relationship between the personalization provider and the bank, in contrast with that between the provider of a mere plastic card, a commodity product, and the customer. Personalisation allows a company to extend and secure its customer base as a platform for launching other services to them. 56. As a result, contended these witnesses, Labat/Africard would be left with the low value end of the banking magstripe business. They would supply the plastic card, with the magnetic strip affixed. However, even if Labat/Africard retained the formal contracts with its banking customers for the supply of the non-personalised magstripe card bodies, it, that is, Labat/Africard, would then have to deliver this card to the supplier contracted to supply the personalization service. Labat/Africard would effectively become a supplier of low value inputs to the supplier of the personalization service. In this circumstance it would make perfect sense for the bank to simply contract for the personalized card with the entity capable of producing the completed card, which would bear responsibility for contracting with those who, like Labat/Africard, were only capable of producing the intermediate input, namely the card body with the non-personalised magnetic stripe. 57. Meniere, supported by the witness from Prism, also questioned the contention – advanced by both Labat and Altech - that the establishment of a successful personalization plant simply required the purchase of the requisite equipment. He argued persuasively that it required considerable learning and internal re-organisation and, particularly, the recruitment and training of personnel, as well as significant technological upgrading. T hese alternate views are ultimately manifest in disagreement over the length of time required to set up a viable, functioning personalization plant with Meniere insisting that it would take at least one year and more likely two years. 18 This seems to be a case of Labat making the best of a bad thing. Although we attempted to ascertain whether the decision to retain the Africard personalization bureau in Altech’s hand arose in consequence of Altech’s unwillingness to sell or Labat’s willingness to buy, the witnesses’ answers failed to clarify the origins of this decision.. On a reading of the record, however, it is reasonably clear that Labat wished to buy the bureau but Altech refused to sell. Noteworthy is the fact that in the documentation of the merchant bank charged with finding buyers for Africard it is specified that the sale would not include the personalization bureau. 16 58. However, while we are persuaded that Labat and the merging parties have understated the length of time and the quantum and quality of effort required to establish an efficient personalization plant from scratch, we are equally persuaded that Meniere has exaggerated the difficulties that Labat, in particular, will confront. Certainly the time frame suggested by Gumede, the CEO of Gijima, Gemplus’ partner, accords more closely with the view of van Rooyen and the merging parties than with that of Meniere. 59. Indeed it appears that the more optimistic predictions – those that predict a more rapid re-entry by the post-merger Africard into personalization - are supported by other evidence. Firstly, Labat is a black empowerment firm and, as such, is an attractive supplier to the banks. Provided that Labat evidences its clear intention to re-establish Africard’s personalization capabilities, there seems to be no reason why the banks would drop Labat/Africard thus inviting dependence on a single supplier of a strategic input, namely personalization services. For a time this may necessitate Labat/Africard fulfilling its obligations to the banks through sub-contracting out its personalization obligations to another party, even Altech, but, we repeat, short of some cogent reason for doubting Labat/Africard’s capacity to perform effectively, we cannot understand why the banks would wish to terminate their contracts with Labat/Africard. 60. Secondly, there is evidence of Labat’s intention to re-build its personalization capacity. We were told by the CEO of Labat that his company has entered into a partnership with an international card manufacturer and that its partner has already begun the process of transferring personalization equipment to Labat. Note that the technology partner identified by Labat has an established track record in the supply of personalized magstripe cards to the banking sector. 61. Thirdly, if we accept Meniere and Gumede’s arguments regarding the strategic nature of personalization, then we must equally acknowledge that the incentive for Labat to re-establish Africard’s personalization capacity is substantial. It will, following this argument, provide Labat with entrée to the imminently lucrative banking smart card market. Although there is evidence that s uggests that Labat’s focus is the national identity document segment of the smart card market, the opening up of the banking sector to smart card penetration will be too good to ignore. We point out once again that Labat’s technology partner is firmly entrenched in the banking smart card market elsewhere. Firms like Prism or Gijima who are not in magstripe cards may not be willing to force their way into banking smart cards by ‘retreating’ into magstripe production. But for a firm like Labat/Africard, with a considerable share of the magstripe market and no apparent reason to withdraw from it, the way forward, the route into the banking smart card market, lies through building relations with the banks and that road, in turn, passes through personalization. 17 62. Note also that contrary to the impression that the merging parties initially sought to convey, the magstripe market has considerable life in it yet. Both Vedder of G&D and Prism testified that banking smart cards in South Africa will, for the foreseeable future, include a magnetic strip because until all card acceptance equipment is upgraded to accept EMV chip cards for the banking industry, the terminal infrastructure will still need to read the data that is resident on the magstripe. There is, in other words, plenty of incentive on the part of those already in magstripe production to remain there, the incentives to be found in a still active magstripe market, not to mention the lure of the smart card market. 63. In short, we are persuaded that Labat/Africard will indeed re-enter the personalization market thus ensuring the re-establishment of a relatively competitive market in secure magstripe production. 64. Our assessment of the competitive state of the banking magstripe market is also influenced by the fact that two of the four largest banks – namely Absa and FNB – undertake magstripe personalization in- house, while a third, Nedbank has, through its acquisition of BOE, taken on some personalization capacity. We are happy to accept with Meniere that this is anomalous and that even these banks will gradually move to outsource this function. But this, if anything, reinforces our view of a still robust market for magstripe card production and personalization. 65. Altech’s motive in retaining Africard’s personalization bureau may well have been anti-competitive – we are inclined to believe that it was. It is, however, unlikely to succeed in its objectives. In relation to smart cards, will Altech’s acquisition of Namitech, not remove the most likely entrant, namely Africard, into the still nascent, but imminently significant, market for the supply of smart cards to the banking sector? 66. As already indicated, Namitech has substantial exposure to the smart card telecommunications market. While it is not alone in this – Gijima, Gemplus and Prism also have substantial shares of this market – it alone has made inroads into the embryonic South African market for banking smart cards. It has secured contracts from Standard Bank and ABSA. It will undertake the rollout of manufacture and personalization of the smart cards for Standard Bank as well as the manufacture of smart cards for ABSA’s roll out. 19 It has also secured a pilot for FNB which could, potentially, lead to a rollout. Indeed it has secured all the banking smart card business that has thus far been contracted out. Namitech has EMV certification. 19 While ABSA has magstripe personalization capacity, it is not clear, from a perusal of the record, whether it also possesses the plant or know-how to personalize smart cards. 18 67. Africard, on the other hand, but for the importation and on-sale of a small number of smart cards for the gaming industry, has no presence in this sector. It is not in possession of EMV certification. 68. On the face of it, then, the transaction’s impact on this segment is neutral. In the early days of banking smart card production, Namitech has a monopoly. The transaction will alter the identity – from Nampak to Altech – of the recipient of the rent, but monopoly will prevail after the transaction with no greater or lesser degree of security than had been the case before the transaction. 69. However, potential competition is at issue here. 70. The evidence indicates that a would-be entrant into the banking smart card market would have to possess deep pockets, technological experience and competence and the confidence of the customer base, the banks. Since the adoption of the Financial Services Charter, it is also advantageous, indeed probably essential, to have a black empowerment partner. 71. Altech possesses all of these aforementioned attributes, except for a black empowerment partner. It clearly possesses both financial strength and an established relationship with the banks. Although its technological competence and experience is unquestioned it does not have direct smart card experience. This shortcoming could be addressed by entering into a technology partnership with an international firm experienced in this area. This is the route pursued, with considerable success, by Gijima through a partnership with Gemplus and Namitech in partnership with G&D. However, Altech attributes its inability to enter the telecommunications smart card market precisely to its failure to secure a technology partner. It has apparently been in lengthy, but fruitless, discussions with Schlumberger, a French multinational smart card producer. Gemplus had attempted to interest Altech in a banking smart card partnership but because the former’s established relationship with Gijima meant that this partnership would not be extended into the telecommunications realm, Altech’s avowed rationale for its interest in smart cards, the discussion came to nothing. 72. It is frankly difficult to accept that a firm with Altech’s financial and technological strength and its established position in this market, cannot procure a technology partner. While it sought to portray potential partners with whom it had entered into discussion as unreasonable or unwilling to contemplate a genuine partnership, it is not possible for us to pass judgment on the character of Altech’s demands. Altech in fact acknowledged that in the event that this transaction fell through, they would re-double their efforts to find a partner. Finding an appropriate black empowerment partner would present few problems. 19 73. We must then conclude that, in the absence of this transaction, Altech is a likely new entrant into the smart card market. It has made no secret of its desire to penetrate the telecommunications smart card market and, given its current exposure to the bank card market, it would clearly wish to take advantage of the new opportunities in this sector. 74. And what of Labat/Africard? Labat has clearly signalled its intention to enter the smart card market although its focus has clearly been on the use of smart cards in national identifications systems. Is it likely to participate in the banking smart card market? 75. Applying the same set of considerations outlined above, we see that, although Labat is financially not in the same league as Altech, it is a listed company with access to capital markets. We have no reason to believe that financial considerations would constrain Labat’s entry into the banking smart card market. 76. While Labat does not have Altech’s technological breadth or experience, it is nevertheless a technology company that includes SAMES, South Africa’s only silicon chip manufacturer, in its stable. While this does not mean that SAMES is a candidate for manufacturing the chips embedded in banking or telecommunications smart cards, there are other synergies implied by the presence of SAMES in the Labat group – it appears, for example that there is space to set up a personalization plant in the SAMES plant, a secure facility; SAMES has, we are told, chip embedding capacity. But beyond whatever direct synergies may exist it is reasonably clear that Labat’s activities make it a credible home for a smart card facility. 77. It has also, as already indicated, entered into a technology partnership with an international smart card manufacturer. Again, while this partnership appears to have been constructed around entry into the national identification smart card market there appears to be little reason why it should be confined to this niche particularly in the face of lucrative opportunities in the banking market. Moreover, Van Rooyen avers that Labat’s partner has significant banking experience, indeed that several of its operations elsewhere are EMV accredited. 78. What of Labat’s relationship to the banking customer base? Its black empowerment credentials are firmly established and this, as already observed, should commend it to the banks. It has, in Africard’s magstripe business, established contractual relations with several of the large banks. We have noted the threat posed to this relationship by Altech’s retention of Africard’s personalization bureau, however we are ultimately satisfied that Labat – partly incentivised by the imminent opportunities in banking smart cards – will re-establish its newly acquired subsidiary’s personalization capacity. Note too that Labat has entered into a contract with Itala Bank, a small, Kwazulu Natal-based rural bank, for the supply of smart cards. 20 79. Nor is Labat the only potential new entrant. Giijima is clearly contemplating entry and, in so doing, it would clearly enjoy the support of its telecommunications partner, Gemplus. So too, we believe, is Prism which has, it appears, EMV accreditation for smart card manufacture. Prism’s respresentative in evidence at the hearing remarked that despite the requirement of large-scale investment into personalization infrastructure, given the business opportunity, Prism and probably Gijima could achieve entry. Nor does there seem to be any reason why a number of the other large smart card multinationals should not, in the face of the right incentives, enter this market. Several of them – notably Schlumberger – already have a presence in this market and, Altech’s claimed difficulties notwithstanding, should be willing to find local partners with local knowledge and black empowerment credentials. Certainly, the incentives are clearly on the nearby horizon – migration to smart cards in banking is imminent and the prospect for smart card use for national identification and related purposes is clearly firmly on the agenda here and elsewhere. 80. Gemplus, Gijima and Prism purport to find significant obstacles to new entry. With a single important exception, we do not find these persuasive. While Gemplus would clearly not wish to invest in new productive capacity, it would, equally clearly, support Gijima’s entry into the banking market. We understand that the difficulties of EMV accreditation have been understated by the merging parties and Labat but nor do they seem as onerous as that suggested by these witnesses. The general difficulties of establishing a smart card manufacturing and personalization facility may well constrain Labat’s entry to a greater extent than it is willing to acknowledge, but it should not significantly constrain Gemplus, Gijima or Prism, all of whom have consummate experience of smart card technology and production. 81. The exception – and we acknowledge the significance of this – is, what may be termed, market access. The protagonists of the merger have made much of the consumer power of the banks, their ‘countervailing power’, in the parlance of anti-trust language. They insist that the banks will not countenance an exercise of market power on the part of their suppliers. Their ability to resist any attempt at market power is underwritten by their ability, in the final resort, to use their purchasing power to induce new entry into the market. In fact, it is argued, that the banks will pre-empt this by continuing to ensure that there is always more than a single supplier of any important input. 82. While this argument appears eminently sensible it was not always borne out in the submissions made by the banks at the hearings. 83. The Tribunal has, elsewhere, articulated its general scepticism of the countervailing power argument. A powerful purchaser’s first and easiest defence against an exercise of market power by a monopolistic supplier may well be to pass an input price increase on to its 21 customers, rather than to attempt to organise its supply chain by ensuring the participation of more than one supplier. That is, the response may well be rooted in the purchaser’s own market power vis- à-vis its customers rather than its purchasing power vis-à- vis its suppliers. The structure of the South African banking market lends itself to the co-ordinated passing-on of supply price increases to banking customers. 84. This is borne out by the laconic disregard of the witness from Standard Bank – apparently the largest purchaser of banking cards – at the prospective monopolisation of its card supply chain. Although clear concern at the transaction is evidenced in the written submissions – admittedly made before the agreement to sell Africard – by the time of the hearings this had dissipated, and had been replaced by expressed comfort with the merger. This new confidence in the outcome of the transaction appeared to derive not so much from the agreement to sell Africard, as from assurances, the precise nature of which remain unclear, received from senior Altech representatives. 85. The banks have, as elucidated earlier, in their submissions to the investigation and the hearings, predictably indicated a general preference for multiple suppliers over a single supplier. This bears out our conclusion that Africard will continue to enjoy the custom of the banks. But this stated preference did not translate into concern at the possible monopolisation of the magstripe personalisation services, nor at Namitech’s powerful early position in the market for the supply of banking smart cards. The banks’ concerns at single sources of supply seemed to be rooted more deeply in security of supply considerations – ‘disaster recovery programmes’ were frequently mentioned – than in competition considerations. 86. The end result is that while we are confident that the banks will continue to use multiple suppliers – like Africard and Namitech – in areas where, like Magstripe card supply, they have established relationships, their attitude does not portend well for the prospect of new entry and, thus, for the future shape of the banking smart card market where Namitech have established a clear early advantage. 87. Certainly, all the prospective entrants into this latter market – Prism, Gijima and Gemplus – understandably indicated that they would not undertake the investments necessary to enter the banking smart card market without a reasonable prospect of custom from the banks and they were clearly not confident of their prospects. 88. There was certainly no evidence that the banks were prepared to ensure alternative sources of supply through nurturing new entrants. Gumede’s frustration at the conservatism that underlay the banks’ procurement policies – particularly Standard Bank’s – was palpable. The evidence of the witness from Standard Bank gave a glimpse of the basis for this frustration. Basically, it appeared that, in Namitech, 22 Standard Bank had a supplier with empowerment credentials, and that while the Bank would keep its lines of communication open to other suppliers – in this instance, that Standard Bank’s witness and senior technology manager, Dr. Berger, would continue having an occasional meal with Gumede – it perceived no particular urgency in ensuring competition in the market for smart cards. 89. It was also made clear to us that the ultimate choice of supplier was also influenced by considerations other than price and service quality, for example, the banking business of the supplier. Again, this did not portend for a relatively small prospective new entrant competing with the subsidiary of one of South Africa’s larger companies. 90. In summary then we are somewhat concerned at the future shape of the banking smart card market. Namitech has a clear first mover advantage and it meets the banks’ requirements for an empowered supplier. The banks evince no particular enthusiasm for supporting new suppliers and little concern at the prospective monopolisation of their supply of smart cards. The apparent irrationality in their willingness to accept a single source of supply may be explained by their ability, given the structure of the banking market, to pass on supply price increases to banking customers. 91. In the end, though, in this particular instance we can do little more than signal our concern. As already indicated, we are satisfied that Labat/Africard will be capable of entering the smart card market. We do not doubt the ability of companies like Prism, Gemplus or Gijima – indeed they could only be kept out of this market by the banks’ apparent preference for the ‘quiet life’, for continuing to do business with their established suppliers regardless of the consequences for competition in the smart card market. This would not be changed by proscribing the merger. 92. We do take some comfort from the banks’ concern with security of supply. We are also cognisant that the banking smart card market will represent an attractive prospect for some high quality competitors like Prism, Gijima and Gemplus, not to mention others like Schlumberger, some of whom are familiar with banking technology and the banking market. Aggressive intervention by these companies encouraged by the undoubted attractions of this market may yet trump the conservatism of the banks. Does the termination of the exclusive agency arrangement between Datacard and Altech cure the anti-competitive consequences of this agreement that have been identified by the Commission and other parties? 93. Certain market participants such as Prism and Gemplus, raised the concern that, notwithstanding the condition that obliged Altech to relinquish its exclusive agency arrangement with Datacard, Altech 23 would, in terms of service agreements with Altech, still be able to access their plants and be privy to confidential information such as production runs and gain insight into sensitive information as to customers and volumes. This was unproblematic because Altech was not competing with the likes of Prism and Gemplus for the supply of smart cards. However, as a result of the merger, Altech will, through Namitech, be a major participant in the smart card market. Prism was concerned that insofar as Namitech is in the business of manufacturing SIM cards for operators, the merger would leave it more vulnerable since Altech (a competitor’s) service personnel would have access to their plant and be privy to sensitive information. Gemplus, too, asserted that post the implementation of the condition, it would be threatened by the fact that Datacard would remain in the Altech stable when Altech/Namitech became a smart card player, therefore, a different class of outsider would be gaining access to their facilities to before. 94. On cross-examination both Prism and Gemplus conceded that it was unlikely that Datacard would subject itself to the loss of business clients as a result of its agent’s conflicts of interest. Both Prism and Gemplus and Gijima therefore accepted that there was no reason to believe that they would not be able to negotiate directly with Datacard in the event that it refused to accept Altech’s service. Gemplus also indicated that it had the capacity to undertake the more technologically sophisticated maintenance in-house. Prism, too, appeared ultimately to accept that they would be able to approach Datacard directly and negotiate support services directly with them. It also referred in evidence to an alternate supplier from whom it procured its personalisation equipment, namely Cybernetics in France. 95. Accordingly we are satisfied that the Datacard remedy is sufficient to take care of any anti-competitive concerns arising from this aspect of the merger. Conclusion The Commission has, after consultation with the merging parties, prepared a Remedies Agreement embodying the conditions elaborated above. We have considered the conditions and have accepted them with minor modifications. The agreement was attached as Annexure A and B to the Order handed down by this Tribunal on 5 February 2004. We are accordingly satisfied that the conditions proposed address any anti-competitive harm that may have arisen from this transaction. 17 March 2004 D. Lewis Date 24 Concurring: N. Manoim, T. Orleyn For the merging parties: Adv. D. Unterhalter, instructed by Bowman Gilfillan. For the Commission: Mr H Shozi.