Daimler Chrysler SA (Pty) Ltd and Sandow

The judgment

The court's own file

COMPETITION TRIBUNAL REPUBLIC OF SOUTH AFRICA Case No: 44/LM/Jul01 In the large merger between: DaimlerChrysler South Africa (Pty) Ltd and Sandown Motor Holdings (Pty) Ltd _______________________________________________________________________ Reasons for Decision _______________________________________________________________________ APPROVAL On 5 November 2001 the Competition Tribunal issued a Merger Clearance Certificate approving the merger between DaimlerChrysler South Africa (Pty) Ltd and Sandown Motor Holdings (Pty) Ltd in terms of section 16(2)(a). The reasons for the approval of the merger appear below. The Parties The primary acquiring firm is DaimlerChrysler South Africa (Pty) Ltd (“DCSA”). DCSA is ultimately controlled by its German parent, DaimlerChrysler AG (“DCAG”), an international motor vehicle manufacturer. Despite DCAG having various other subsidiaries in South Africa, none of these firms are involved in motor vehicle manufacture and supply. DCSA operates in South Africa as a manufacturer/supplier of a range of passenger and commercial vehicles. DCSA manufactures vehicles in South Africa through its 100% subsidiary company, DaimlerChrysler South Africa Manufacturing (“DCSA Manufacturing”). DCSA manufactures the Mitsubishi Colt (L200) in South Africa, as well as the Mercedes-Benz C-Class. 1 All other vehicles supplied by DCSA are imported. DCSA markets a large number of brands within South Africa, constituting a much more extensive range than when it operated just as Mercedes-Benz. These brands are sold into the South African market through a range of authorized dealers, of which SMH is one such dealer. DCSA brands include: 1 This is then sold to DCAG in Germany who in turn sells the vehicles that are destined for the South African market back to DCSA. 2 A. Passenger Cars (“PC”): Table 1: Brands by Segment Category Mercedes-Benz Chrysler Mitsubishi Entry-level cars None None None Small Cars A-Class Chrysler Neon None Lower Middle Cars None None None Upper Middle Cars None None None Large Cars None None None Lower Luxury Cars C-Class None None Upper Luxury Cars E-Class, S - Class None None Lower Speciality Cars None None None Upper Speciality Cars SLK, CLK, CL None None Small Utility None Chrysler Jeep Wrangler Mitsubishi Pajero Lower Middle Utility None Chrysler Jeep Cherokee None Upper Middle Utility M-Class None Mitsubishi Pajero 3500 Small minivans None Chrysler PT Cruiser None Minivans None Chrysler Voyager None B. Light, Medium and Heavy Commercial Vehicles (“CV”) v Mitsubishi Colt LCV’s v Mercedes Benz (Medium and Heavy CV’s) v Mitsubishi Freightliner (Heavy CV’s) C. Buses and Coaches over 10 tons. 2 The primary target firm is Sandown Motor Holdings (Pty) Ltd (“SMH”), a dealer in new and used motor vehicles. 3 SMH is the largest of DCSA’s exclusive dealers. SMH comprises 9 dealerships, located in the Western Province and Gauteng. Their dealer distribution outlets and services offered are arranged as follows: 2 This is a particularly small market, estimated at just over 700 sales in 2000. SMH has very little activity in this market. 3 SMH has 5 subsidiaries, all of which are dormant companies. 3 Table 2: Western Province Dealer- ship Location New PC MB New CV MB New Chrysler & Jeep New Pajero New Colt Used PC Used CV Parts & Service Fore- court Orbit Motors N1 City, Cape Town X X X X X Eikestad Motors Stellen- bosch X X X X X X Orbit Motors Boland Worcester X X X X X X Table 3: Gauteng Dealer- Ship Location New PC MB New CV MB New Chrysler & Jeep New Pajero New Colt Use d PC Used CV Parts & Service Fore- court Sandown Motors Village Close Sandton X X X X X Mitsubis hi Motors Sandton Bryanston X X Randbur g Pajero Randburg X X X Sandown Truck Centre Kelvin, Sandton X X X Ellenby Motors Hatfield, Pretoria X X X X X Mitsubis hi Motors Centurion, Pretoria X X SMH therefore conducts all of the above activities to a limited extent throughout its 9 dealerships. However, its core activity remains the sale of new and used passenger and commercial vehicles. More detail on each type of service will be provided in the analysis section. The merger transaction This is essentially a vertical merger. DCSA, a manufacturer and supplier, is acquiring one of its retailers, the SMH dealership. Accordingly, an upstream manufacturer and supplier is integrating with a downstream retailer to sell motor vehicles to the end consumer. DCSA does not own any dealership at present, however SMH is an exclusive dealer, in that it sells only DCSA motor vehicles. 4 DCSA is a cquiring 75% of the shareholding in SMH. The remaining 25% will be held by Mr Roy McAllister who is the current Managing Director of the company. Background : Relationship with Dealerships There are approximately 1,400 motor vehicle dealerships across South Africa. 4 DCSA distributes and sells its cars through a network of franchised dealerships. These dealers are appointed to sell new, as opposed to used, vehicles. The dealers fall into two categories: 1. Mutli-franchise dealers who distribute DCSA vehicles as well as the vehicles of other motor manufacturers including competitors of DCSA. The multi- franchised dealers are large retail concerns as appears more fully from Table 6 below. There are three firms in this category. They are Barloworld; Imperial (trading as Cargo Motors and Mecurius Motors) and McCarthy. 2. DCSA also distributes its vehicles through various individual dealerships. The latter category which includes SMH, comprises exclusive dealerships. Table 4: Exclusive and Non-Exclusive DCSA Dealerships Dealer Total No. Dealers No. DCSA dealers No. dealers selling other brands (Non-Exclusive) Barloworld 56 9 47 Imperial 99 15 84 McCarthy 87 13 74 SMH 9 9 0 Other (Individual) 23 44 0 TOTAL 274 90 285 Source: Competitiveness Report Dealerships can be further distinguished between those selling passenger cars and those selling commercial vehicles: Passenger Cars: Total DCSA dealerships 81 SMH 8 4 As advised by Gary McCraw, the Retail Motor Industry Association (“RMI”) representative at the hearing. 5 Commercial Vehicles Total DCSA dealerships 58 SMH 3 SMH therefore constitutes 9 out of 90 DCSA dealers on a national basis. All dealers, whether exclusive or multi-franchise, operate through franchise agreements with DCSA. Each SMH outlet has separate franchise agreements, rather than one collective agreement in respect of the dealership group. There are also separate franchise agreements for the Chrysler and Jeep brands on the one hand and the Mercedes and Colt brands on the other hand. Franchise Agreements are presently of one year duration and notice of termination of the franchise agreements is six months before the end of every year. The current agreements are due to lapse at the end of the year, with new ones being drawn up for 2002. In terms of the franchise agreement, dealers are required not to take up agencies for any other m otor vehicles which are detrimental to the image of, or which are in conflict with any products or services offered by MBSA without the prior written consent of MBSA. The franchise agreements are presently undergoing a process of review, in accordance with a request from the National Dealers’ Association (“NADA”). One such amendment is to amend the manner in which vehicles are sourced through implementation of a central pooling system, or New Distribution Policy (“NDP”) system. In terms of this system, implemented last year, dealers no longer hold stock of motor vehicles on the shop floor. Instead orders placed by customers at dealer outlets are electronically transmitted to DCSA. This then places the required vehicle into the distribution and delivery process and delivery takes place on a first ordered-first received basis. The parties maintain that this system ensures that all dealers have equal access to DCSA vehicle stock and that no discrimination between dealerships takes place. In addition, the parties are in the process of negotiating a new dealer network strategy with both the major dealer groups and smaller, independent dealers. This strategy seeks to enhance brand focus and be more customer oriented, as well as adding value by focusing on separate brands, in order to develop a brand identity. This will be facilitated by the separation of DCSA brands into so-called dedicated “brand centres”, “hubs” and “spokes” within its existing network. This strategy envisages firstly creating a single franchise facility in metro areas for all dealerships, so as to concentrate on each and every brand separately, as opposed to the existing multi-brand facilities, where brand strength is diluted. 5 Although the parties haven’t stated this specifically, presumably the merger with Chrysler and the arrangement with Mitsubishi have lead to the group selling a 5 In DCSA jargon, these are referred to as “multi-franchise”. We use the phrase “multi-brand” since we have used “multi-franchise” to describe the non-exclusive dealers. 6 multiplicity of brands in different market segments. 6 In this context concerns about brand dilution are understandable. Secondly, country/rural areas will constitute multi-brand facilities (“hubs”). There is a possibility that sales centres will be split from service centres (“spokes”), the number of outlets in the former being reduced in the metro areas, while service centres will be more prevalent. The parties stressed that the exact delineation and composition of each type of centre, as well as their geographical distribution has yet to be fleshed out with the dealer networks as part of an ongoing, long- term negotiation process. We will return to the relevance of this possible new strategy later in our decision. Rationale for the Transaction At the hearing, Andreas Hiller, Divisional Manager of DCSA, cited four reasons for this transaction. 1. To enable DCSA to get closer to its customer base, by ensuring it h as greater access to customers, as well as understanding their needs and expectations. 2. To identify and address problem areas. This is expected to further complement the general trend at DCSA towards a seamless chain of supply to the customer as well as assisting DCSA to prepare for future e-commerce initiatives. 3. To give them a platform to test and improve systems and rollout various projects, by demonstrating that if the particular system can be implemented successfully throughout the SMH network, it is a viable option for the rest of the dealership network. 4. To realize greater profit opportunities from high retail margins and revenues, as well as the promise of greater return on investment. The Objectors There were some objections to the merger from v arious retailers. There were general concerns amongst larger dealerships about a dealer integrating downstream. Retailers fear that this transaction is the first step in implementation of an overall plan to reduce the number of DCSA dealerships across the country, i.e. the “mega dealership” concept. All the objectors who expressed concerns had in common reservations around the uncertainty and lack of transparency with regard to DCSA’s future strategy. The first objecting party who initially sought to participate in the proceedings was Georgeson’s Motors. They operate as a dealer of Mercedes-Benz motor vehicles in Estcourt, KwaZulu-Natal. The essence of their objection centred around the cancellation of their franchise agreement with DCSA. They alleged this was cancelled because it represented a threat to the larger, white-owned franchises in the urban areas, facing a 6 On 1 April 2001, DCSA took over the Mitsubishi brands, namely Mitsubishi Pajero and the Mitsubishi L300 ranges, for both passenger cars and light commercial vehicles from Ford Motor Company of Southern Africa. Effective this date, DCSA’s dealer network assumed the sales and service responsibility for these brands. (Mitsubishi Motors website/archived news) 7 shrinking market for Mercedes-Benz cars. They maintained that the merger is part of the overall “hub and spoke” strategy whereby non-strategic spoke dealers would have to source parts and supplies from hub dealers, paying them a percentage. The result would be the consolidation of those dealers favoured by DCSA, to the exclusion of the non- strategic dealers. Georgeson’s Motors was invited to make submissions at the hearing but elected, via their legal representatives, to stand by their written submissions. DCSA’s response was that the deteriorating relationship with Georgeson’s Motors was a result of a commercial dispute and not a function of anti-competitive behaviour. The second objecting party was the Retail Motor Industry Organisation (“RMI”), a non- profit employers association, whose members comprise principally motor dealers representing 95% of all franchised motor dealers around the country. Their concerns revolved around the DCSA’s overall strategic plans with regard to the dealerships. These were sparked by recent developments in the UK where dealers’ franchise contracts have been cancelled in order to facilitate the rollout of brand centres, giving these dealers the option to reapply for new franchise contracts. Local dealers’ impressions have been that DCSA were acquiring a strategically placed dealership to ultimately implement the brand centre business model in South Africa. This prospect has sparked concern of the entire RMI membership base (not just DCSA dealers) and the RMI’s mandate was to intervene to attempt to level the playing field for dealers by ensuring they were consulted in a transparent manner and given fair representation and, if necessary, compensation. 7 During the hearing it emerged that after various meetings with DCSA, a memorandum of understanding was reached whereby DCSA undertook to maintain transparency and to consult with the RMI with regard to all aspects of its new strategy in an appropriate forum. Any further concerns with regard to the rollout of its long-term network strategy would also be addressed in this forum. The RMI accordingly formally withdrew their objection to the merger at the hearing, but were nevertheless invited to make submissions with regard to various aspects of the transaction, as well as to the nature of the industry in general. The Tribunal has to decide whether these undertakings by DCSA vis-à-vis the RMI is sufficient to allay the general industry concerns surrounding the merger. ANALYSIS/EVALUATING THE MERGER Theoretical Approach Competition authorities have a permissive attitude to vertical mergers. Such mergers are generally thought to raise fewer competitive concerns than horizontal mergers. This is because vertical mergers are generally regarded as procompetitive insofar as they enable 7 Both Georgeson’s Motors and the RMI have lodged formal complaints with the Commission concerning the onerous terms of franchise dealer agreements across the industry. Both objectors agreed that these formed part of a separate, independent process to the instant transaction. 8 a firm to produce an improved or lower priced product or service, or to distribute it through the value chain in a more efficient way. A vertical merger will only be objectionable where one or both of the merging parties dominates the respective market in which each operates. In such a case, the anticompetitive effects could entail the prospect of increased entry barriers, as well as the possibility of market foreclosure and the related ability to raise rival’s costs. 8 What is clear from the antitrust literature is that each vertical transaction must be examined on the basis of its own, peculiar facts. The primary issue in this case is the effect on intra-brand competition, that is competition among retailers or distributors of the same brand. This may occur on price or non-price terms. It is generally accepted that where inter-brand competition (competition between different brands of product) is strong, the requirement to regulate the vertical relationship between a supplier and his distributor which primarily affects only intra-brand competition, is diminished. The rationale for this is that where inter and intra-brand competition is weak, there is less pressure within the market to deter prices increases. As stated in the EU Commission Guidelines on Vertical Restraints 9 : “Vertical restraints which reduce inter-brand competition are generally more harmful than vertical restraints that reduce intra-brand competition. For instance, non-compete obligations are likely to have more net negative effects than exclusive distribution. The former, by possibly foreclosing the market to other brands, may prevent those brands from reaching the market. T he latter, while limiting intra-brand competition, does not prevent goods from reaching the final consumer.” Similarly, in Continental TV Inc. v GTE Sylvania Inc. 433 US 36 (1977), 54 to 57, in a dispute between a manufacturer of colour television sets, S ylvania, and one of its franchised distributors, Continental, the United States Supreme Court recognised that although non-price vertical restrictions reduce intra-brand competition, they promote inter-brand competition by allowing manufacturers to achieve certain efficiencies. 10 8 See Schumann Sasol (SA) (Pty) (Ltd) and Price’s Daelite (Pty) (Ltd) merger, 23/LM/ May 01 9 2000/C 291/01 OJ Para 119(2) 10 97 S. Ct. 2560 9 “Although intra-brand competition may be reduced, the ability of retailers to exploit the resulting market may be limited both by the ability of consumers to travel to other franchised locations and, perhaps more importantly, to purchase the competing products of other manufacturers “ In a more recent decision, Graphic Products Distributors Inc v Itek Corp 11 the United States Eleventh Circuit Court of Appeals upheld this view, stating as follows: “We note first that a vertical restraint on trade, almost by definition, involves some reduction in intrabrand competition. When a manufacturer restricts a dealer to selling only within a certain territory, or only to certain customers, or only from certain locations, it is necessarily restraining intrabrand competition. However, this may or may not have a negative effect on the welfare of the consumer ... The effects of a restraint of intrabrand competition on consumer welfare cannot be viewed in isolation from the interbrand market structure. A restriction of intrabrand competition may depending on the interbrand market structure either enhance or diminish overall competition, and hence consumer welfare… Moreover, if enhanced dealer services to the consumer result from the restraint, interbrand competition should be sharpened. ” 12 The relevant product market Since this transaction involves the merger of a manufacturer and its dealer, the Commission identified the market for the manufacture and supply of motor vehicles on the one hand, and the market for the distribution and sale of motor vehicles to the final consumer, that is, via the dealerships, on the other. DCSA competes with a number of motor vehicle manufacturers in both the passenger and commercial vehicle markets in South Africa in respect of the manufacture and wholesale supply of motor vehicles. Such competitors include BMW, Ford, Nissan, Renault, Alfa Romeo and Toyota, to name a few. In the dealership market, dealers sell motor vehicles to the end consumer, at the same time offering a variety of different services. SMH provides such services in differing combinations across its 9 branches, as follows:- 11 717 F 2d 1560 (1983). This principle was also relied on in the Australian case of Melway Publishing (Pty) Ltd v Robert Hicks (Pty) Ltd 2001 AHC 13 12 Note that court in this case went on to affirm the viability of a § 1 Sherman Act claim based solely on alleged anti-competitive effects on intrabrand competition since the conflicting evidence did not indicate that Intek’s restraints were pro-competitive in purpose and effect. The same principle as expressed in Intek was followed in subsequent cases. See K.M.B. Warehouse Distributors Inc. v. Walker Mfg. Co. 61 F.3d 123 1995;Orson Inc. v. Miramax Film Corp. 79 F.3d 1358 1996 10 Sale of New Passenger and Commercial Motor Vehicles SMH sells all new DCSA motor vehicles to its main customers, primarily dealers who on sell it to the final consumer. The market for commercial vehicles can be subdivided into: v Light commercial vehicles v Medium commercial vehicles v Heavy commercial vehicles v Buses and coaches over ten tons. Sale of Used Passenger Vehicles SMH, along with most dealers, sells used vehicles to consumers. These vehicles are sourced as trade-in vehicles from customers; as outright purchases; from DCSA auctions or demonstration stock. There is sufficient competition in this market since all dealerships can buy second-hand vehicles and resell them to the end customer without requiring an agreement with DCSA. Thus barriers to entry are minimal and numerous firms operate in this market. This market accordingly, need not occupy our competitive assessment any further. After-Sales Servicing of Motor Vehicles SMH’s workshop engages in routine servicing in terms of the maintenance plan accompanying new vehicle sales; repairs and maintenance of vehicles; outsourcing of specialised repair work and pre-delivery inspection. There are no competition concerns with regard to this area of activity arising out of the transaction, since this is not a core area of activity. A customer is not tied to the dealer from whom the vehicle is purchased for the purpose of service .Therefore this, too, need not delay us any further. Sale of Spare Parts SMH and other DCSA dealers source spare parts from DCSA. The parties maintain that approximately 35% of these spare parts are utilised in the dealers’ own workshops to serve as replacement parts for vehicles being repaired. The remainder of these spare parts are sold to non-dealers, comprising fleets, the repair industry and independent workshops. All dealers have equal access to spare parts, drawing from the factory on equivalent terms, when the need arises. We accept the parties’ submissions that this market is irrelevant for the purpose of this analysis. Notwithstanding these diverse services provided by SMH, their core activity relates to the sale of new and used passenger and commercial vehicles. Furthermore, it is only in respect of retail sales of new motor vehicles where competition concerns arise and it is accordingly this market that will be the subject of further analysis. 11 Passenger vehicles v Commercial vehicles DCSA sell both passenger and commercial motor vehicles. The distribution of commercial vehicles differs to that of passenger vehicles since in the case of medium and heavy commercial vehicles, manufacturers usually sell and distribute directly t o the end customers according to their specifications. Alternate means of distribution are through tenders and via commercial dealerships. 13 We are not concerned with the commercial vehicle market since this is not a core area of focus for SMH, as is apparent from the number of SMH dealerships engaged in the sale of commercial vehicles, by comparison to those selling passenger vehicles: Table 5: Commercial and Passenger Vehicle Dealer Outlets Total dealership CV’s SMH CV’s Total dealership PV’s SMH PV’s 58 3 81 8 Source: Competitiveness Report page 32 We will accordingly confine our analysis to the sale of new passenger vehicles. The dealers sell a wide range of DCSA passenger vehicles. The primary passenger car brands include Mercedes-Benz, Chrysler/Jeep, Mitsubishi Colt and Pajero. 14 Within the market for new passenger vehicles, further classification can be effected into niche segments. The European Commission has previously held that it is possible to delineate the passenger car market on the basis of a number of objective criteria, such as engine size or length of car. However, a final definition is seldom required and so the question has largely been left open. 15 This is because boundaries between these segments are not rigid, but are essentially based on customer perceptions and preferences. Nevertheless, the industry itself has traditionally utilised these fields of categorisation of motor vehicles and it is regarded as an accepted framework to determine the position of cars in the market. Generally-speaking, the European Commission has previously defined the passenger car market narrowly into the following segments: v Mini cars v Small cars v Medium cars v Large cars v Executive cars v Luxury cars 13 See page 72 of the Record. 14 See Table 1. 15 See inter alia, BMW/Rover IV/M.416 1994, Ford/Mazda IV/M.741 1996, Chrysler/Distributors IV/M.1036 1997, Daimler-Benz/Chrysler IV/M.1204 1998 12 v Sport coupes v Multi purpose cars v Sport utility cars (including off-road vehicles) The parties maintained that the appropriate market definition should be that of passenger vehicles in general. However, as pointed out by the Commission, not all passenger cars will compete with each other. A customer who wishes to buy a Mercedes-Benz would not likely substitute to a Toyota Corolla model. As in the EU, South African customer choice will be based on price, technical specifications and aesthetic appeal. Brand awareness also plays an important role. Furthermore, as will appear below, DCSA does not compete in all motor vehicle segments, such as within the entry level cars, middle and large car segments. We are therefore looking at competition between segments, not competition for the overall passenger market. According to internal marketing data they submitted, the parties in practice adopt an approach which parallels that of the EU, although they provide for more segments. They classify their motor vehicles as follows 16 : v Entry-level cars v Small Cars v Lower Middle Cars v Upper Middle Cars v Large Cars v Lower Luxury Cars v Upper Luxury Cars v Lower Speciality Cars v Upper Speciality Cars v Small Utility v Lower Middle Utility v Upper Middle Utility v Small minivans v Minivans MARKET FOR THE SUPPLY OF MOTOR VEHICLES The parties advised that i n the market for the wholesale supply of motor vehicles, the following volumes were sold directly to wholesale customers. Unfortunately only figures in respect of total units (passenger and commercial vehicles) were provided and are in respect of a national market: 16 See Table 1. 13 Table 6: Wholesale Supply of Vehicles Customer` Units (Jan-May 2001) % of total DCSA Barloworld 1413 8.7 Imperial 2327 14.32 McCarthy 2250 13.85 SMH 1984 12.21 Debis Fleet Management 2970 18.28 Other Dealers 4186 25.77 Wholesale (volume) 17 437 2.69 Other Wholesales 587 3.61 Total 16 245 100 Source: Parties Competitiveness Report Accordingly 12% of DCSA product is channeled through SMH. Based on NAAMSA data, the parties submitted market share information in respect of their own internal segmentation of the market as follows: Table 7: National DCSA Market Shares according to vehicle segments Category Chrysler Mercedes-Benz Pajero Total Entry-level cars 0 0 0 Small Cars 3.7% 4.4% 8.1% Lower Middle Cars 0 0 0 Upper Middle Cars 0 0 0 Large Cars 0 0 0 Lower Luxury Cars 0 29.67% 0 29.67% Upper Luxury Cars 0 37.96% 0 37.96% Lower Speciality Cars 0 0 0 Upper Speciality Cars 0 24.79% 0 24.79% Small Utility 1.66% 0 0 1.66% Lower Middle Utility 19.13% 0 0 19.13% Upper Mi ddle Utility 16.04% 15.16% 31.2% Small minivans 12.64% 0 0 12.64% Minivans 60.16% 0 0 60.16% Source: DaimlerChrysler Market Analysis 17 Refers to direct sales to big private customers buying directly from DCSA 14 It is accordingly apparent that DCSA has a strong position in the luxury, utility, minivan and speciality car segments. However, the parties suggested at the hearing that the two luxury segments should more appropriately be combined into one segment. The parties maintained that there is a large degree of interchangeability between lower & upper luxury as well as lower & upper middle utilities in that customers might well switch between these categories, within the premium market. 18 We accept that it is possible that the segmentation in table 7 is too narrow, and if we were to take this to the next level, by combining two similar categories into one, the market shares would be lower. In the luxury segment therefore, it would be 31.3%, while a similar exercise for the minivan segment, incorporating the small minivan category, would yield a market share of 23.4%. However, these shares are calculated on the premise this is a national market, a view the Tribunal does not hold, as we discuss more fully below. DEALERSHIPS In broad terms, SMH’s comparative share of retail unit sales of Mercedes-Benz and Chrysler Jeep passenger and commercial vehicles throughout the national market is as follows 19 : Table 8: Retail unit sales across Dealer Groups-Mercedes-Benz & Chrysler brands* Dealer PC’s CV’s Sales % Sales % Barloworld 1584 10.5 412 13.6 Imperial 2580 17.1 359 11.9 McCarthy 1959 12.9 688 22.7 Other Dealers 6746 44.7 1323 43.8 SMH 2222 14.7 239 7.9 Total 15091 100 3021 Source: Competitiveness Report Accordingly, in broad terms, DCSA is acquiring access to passenger vehicle sales amounting to approximately 8% of the DCSA commercial vehicle market and 15% of the DCSA passenger vehicle market. When one looks at the dealer network in terms of sales by brand, the following emerges: 18 As submitted at the hearing, ref. Transcript, page 87 19 Data relating to commercial vehicle sales is set out for information purposes, though, as already indicated, we are not concerned with this market. 15 Table 9: Dealer Market Shares according to Brand - 2000* Dealer Chrysler % Mercedes- Benz % SMH 848 19.8 1374 12.7 Barloworld 430 10 1154 10.7 Imperial 694 16 1886 17.5 McCarthy 444 10.4 1515 14 Total Dealers 2416 5929 Total DCSA 4287 10804 *excluding Mitsubishi/Colt & Pajero Table 10: Mitsubishi Motors Retail Sales (Colt and Pajero) May-Aug 2001 Dealer Sales % SMH 367 22 Imperial 255 15.5 Barloworld 207 12.5 McCarthy 263 16 Total Groups 1092 Other 559 39 Total 1651 100 Source: Parties documents Accordingly, SMH has the highest market share out of all the dealer groups in respect of Mitsubishi, as well as Chrysler brands across the national market. When one refers to table 1 above , these brands occupy the utility and minivan segments of the market. Accordingly, this confirms that the relevant market is the sale of new vehicles in the luxury, utility, specialty and minivan categories. Geographical Market The parties maintained that the market is national. They contend that DCSA manufactures vehicles for sale throughout South Africa as do its c ompetitors. They further submitted that prices for motor vehicles are determined on a national, rather than a regional level. SMH sets guidelines on a recommended retail price, based on market conditions nationally and it is these prices that are applied, subject to minor flexibility on the part of the dealerships. At the hearing, they argued that there is no legal or other impediment from dealers selling to customers outside their region. The parties further maintained that the ability of consumers to order motor vehicles telephonically or via the 16 internet, as well as the implementation of the new NDP system, adds credence to the argument for a national market. The Commission endorsed these views. The Tribunal does not agree with the assessment of the geographical market. The Tribunal was presented with data prior to the hearing reflecting SMH’s customer concentrations per region. As seen on the attached annexure, when this data was logged graphically, it is apparent that for each of three SMH dealerships, there is a highly marked reduction in sales the further the customer resides from the immediate outlet. 20 This data is presumably representative of all the franchises, across all segments, therefore highly indicative of a local market. Secondly, the parties’ own internal new retail and network strategy document, produced just prior to the hearing is itself based on the premise that the markets are local. The document reveals that the new retail strategy is clearly centred around the main hubs of consumer purchasing activity, namely Johannesburg, Pretoria, Cape Town. 21 If the information contained in the tables reflecting SMH’s customer base is indicative of consumer behavior more broadly, and there is no reason to assume it is not, then customers in Gauteng are more likely to purchase a motor vehicle from a Gauteng dealer, than a dealer situated elsewhere in the country. The parties conceded this at the hearing and seemed to accept that it is logical that customers are most concentrated in the immediate vicinity of their nearest dealer. 22 The evidence provided by the parties suggests that prices vary marginally, within a recommended price band, between dealer outlets. Accordingly, if prices are raised by the merged entity, it is unlikely that a Gauteng customer would buy a car from a Cape Town outlet, even if it were marginally cheaper. Is there a reduction in intra-brand competition In order to answer this question, we must examine whether intra-brand competition existed prior to this transaction. The parties were somewhat equivocal at the hearing as to whether dealers competed on the basis of price. The RMI opined that there is no difference in pricing between dealers. Prices are recommended at what the vehicle should be sold – dealers have some flexibility to give discounts within the recommended price bands but, according the them, it would be “suicidal” to have carte blanche to give unlimited discounts since ultimately, dealers have to protect their margins. DCSA maintained that margins are low already, below 10%, making little room for price competition. They however did imply that there might be some degree of price competition since they admitted that they do not like to see their dealers competing on the 20 This exercise was done only in respect of the luxury market segment. 21 “Local” refers to the areas as described in the parties documentation where they refer to 6 metropolitan areas, which are Gauteng, Gauteng East, Gauteng North, Gauteng West, Cape Town and Durban, as well as their rural markets which are located in areas outside of the metropolitan areas. Whilst the parties make no attempt to sub-divide the rural market any further, there would presumably be a number of rural markets that can be differentiated on the basis of some relationship between the customer and their proximity to the dealership. It is not necessary for us to go into this as nothing turns on this classification. 22 See transcript, pages 46, 67. 17 basis of price, which could reduce the viability of their dealerships, but would prefer to see them compete on some other basis or with other brands, such as Audi, Volkswagen etc. As Mr Hiller explained at the hearing: “We would rather like to see a customer moving over from BMW to Mercedes Benz rather than from Sandown to McCarthy or vice versa…because what is happening at the moment is a kind of dealing with vehicles dealing with specs and destroying prices and whatever we’d rather like to see them focusing on our competitors.” 23 . It appears that dealers are likely to compete on a non-price basis by providing other value-added services, such as after-sales service, warranties, in-house motor vehicle financing, courtesy car offers or life insurance to cover the cost of the vehicle. 24 While we can accept that no manufacturer would like to see dealers resorting to an all out price war in respect of their products, this does not nevertheless mean that dealers will not attempt to out price each other, even if by a small margin. The parties conceded that there is some degree of price competition in the rural areas, where overheads are lower, therefore allowing these dealerships to reduce their margins. 25 This fact, together with the existence of non-price competition between dealers, leads us to the conclusion that there is some intra-brand competition between dealers presently in the market. The next question then is to what extent intra-brand competition would be reduced as a result of this transaction. The parties have maintained that the dealer outlets will not be closed, since this would be self-defeating. As pointed out by Mr Hiller the hearing on the question of terminating dealerships: “We want to grow daily as a wholesale organization. And therefore it is just not an option for me. And I understand the concern but it won’t be viable for us at all because we’re making good wholesale profits here and we want to continue and we want to grow with them.” 26 Even should they do so, it is common cause that countervailing power exists in the form of the three competing chains, McCarthy, Imperial and Barloworld, in all regional markets in which SMH operates. SMH does not dominate any particular region, there being a strong presence amongst dealers in each region where SMH operates, as is illustrated in Table 11. Accordingly, any anticompetitive behaviour is likely to be vigorously resisted by these three dealerships, who hold substantial bargaining power. If it did terminate franchises, DCSA would have no way to push through supply of its vehicles since the non-SMH 23 See transcript, page 12 24 As explained by Gary McCraw, representing the RMI at the hearing. 25 By way of example Mr McAllister, MD of SMH, testified of a dealer in, Vereeniging who was competing within his area ( Johannesburg North) for his customers on a price basis because he ( the Vereeniging dealer) had a lower cost structure and hence the flexibility to reduce his margins. 26 See transcript, page 34. 18 dealers are responsible f or distributing a not insignificant volume of DCSA vehicles, approximately 85%, if we refer to the data in table 8 and Table 11 below. DCSA would then have to reallocate this supply amongst its own 9 dealers, or alternatively invest heavily in new outlets. 27 In Table 11 we see a regional breakdown of distribution amongst DCSA dealers. Table 11: Table DCSA Regional Dealer Network Dealer Johannesburg Pretoria Cape Town Durban Barlows 10% 0% 0% 37% Imperial 50% 0% 0% 0% McCarthy 4% 71% 63% 25% Sandown 28% 23% 25% 0% Others 8% 6% 12% 38% Source: DCSA Network Strategy documents The parties argued that on the basis of the figures in table 14, in a worst case scenario, SMH accounts for a maximum of 30% of DCSA vehicle sales in each region. The table further indicates that despite having a strong representation in all regions other than Durban, SMH does not dominate any particular region, there being sufficient competition amongst dealers in each region where SMH operates. There is no evidence that the merger will raise entry barriers in the new passenger car market, or any component thereof. Based on the evidence before us we cannot say conclusively that there will be a reduction in intra-brand competition. What is clear is that this merger is not changing the status quo significantly - post-merger, SMH will remain an exclusive retailer of SMH products. Only the relationship between the parties will change from a contractual one to one of ownership, entrenching a pre-existing vertical relationship. We have no reason to dispute this. What did cause some concern were remarks made in the marketing documents and in Mr Hiller’s comment that dealers were competing the price away. That might indicate that the merger could be utilized to inhibit price competition amongst dealers as it might make more credible a threat to terminate franchises if dealers were seen to be undercutting the normal price offered in the market. That being said there is insufficient evidence to suggest that even if DCSA were to embark on such a strategy, they need the merger in order to do so, when it appears that their real power over dealers is through existing vertical relationships. There is also little 27 It is however possible that inter-DCSA dealer competition could be eliminated if and when the network strategy is implemented but we accept that this strategy must be viewed as separate from the instant transaction and in any event, has not as yet materialized into anything concrete. 19 evidence that intra-brand price competition is vigorous amongst metro dealers who are the rivals most likely to constrain SMH. In any event, as enunciated in a previous section, all competition authorities around the globe take the view that any reduction in intra-brand competition can be offset by inter- brand competition. Assuming therefore for the moment that there is likely to be a reduction of intra-brand competition, we need to evaluate whether there is sufficient inter-brand competition to offset this. Is there sufficient inter-brand competition? In the overall market for passenger cars, DCSA market shares are high in the identified relevant market segments, that is, with respect to the luxury, utility, speciality and minivan segments, where they are in the region of 23-31%. These shares could well be higher when we consider the market on a local basis. It is important to note that there is no incremental increase in market shares as a result of this merger s ince the takeover of SMH will not alter DCSA’s market position. DCSA is not acquiring more brands. Despite the high market shares we nevertheless find that there is a sufficient degree of inter-brand competition in the market for various reasons: v There a re many new entrants into the motor vehicle market and market shares tend to shift. The data reflects that in the lower and upper luxury segments, DCSA has been losing market share over the last five years to competitors. For instance, in the upper luxury car market, Mercedes’ segment share has declined from 51% in 1996 to 38% to July 2001. The parties themselves concede that new entrants (such as in the minivan segment) have high market shares initially upon entry into the market. They also advised that their market share for the sale of minivans has been decreasing as customers switch to lower speciality cars. v All manufacturers import some products and therefore face the same barriers to entry. v Inter-brand competition is strong amongst the competing m otor manufacturers. There are many competing brands in each segment. By way of example, for each relevant segment, the competitive position is as follows:- 20 Table 12: Competition per Segments Segment Competitor Market Share Lower Luxury Audi 11.72 BMW 45.88 Upper Luxury Audi 11.93 BMW 39.16 Volvo 2.36 Alfa .40 Upper Speciality Audi 17.9 BMW 35.58 Small Utility Landrover 23.04 Toyota 66.78 Lower Middle Utility Isuzu 32.76 Landrover 21.95 Nissan 15.62 Upper Middle Utility BMW 11.92 Landrover 17.47 Minivans Volkswagen 34.31 Source: DCSA Market Analysis There appears to be a great degree of overlap between these defined segments from a customer perspective. For instance in Europe, minivans (multipurpose vehicles) are regarded as highly substitutable with estate cars. 28 The parties themselves maintained at the hearing that despite the narrow market segments submitted by them, for competition purposes, there is in actual fact a large degree of interchangeability between the lower & upper luxury and the lower & upper middle utility segments in that they all occupy a premium market. Were we to adopt the traditionally accepted categorization of cars in the EU, the parties’ market shares would probably be lower. However, in view o f the degree of existing and potential inter-brand competition in the relevant markets on the narrower level, it is unnecessary for us to go up a level and examine market shares in respect of broader categories of all luxury or all utility cars. Conclusion At the hearing the parties asserted that though there might be some collaboration with DCSA at a strategic level, from an operational perspective, SMH will remain a stand alone dealer and continue its business activities on this basis. There is no risk of foreclosure at the wholesale level of other suppliers, since the business being acquired sources only DCSA vehicles, in any event. 28 Daimler-Benz/Chrysler Case No. IV/M.1204, para 17 21 The parties maintained there is no relationship between its two initiatives, namely this transaction and its new retail network strategy. “… the opportunity to get involved in Sandown is a different opportunity where we’re saying there’s a downstream business opportunity. We have now an opportunity to learn more about our customers has for me nothing to do with the network strategy because there’s a re-alignment we need anyway in place, if we own something or not.” 29 We accept that the network strategy has no bearing on the instant transaction. This is an unconfirmed, long-term strategy in respect of which we cannot predict or speculate the likely anticompetitive effect. Although it is within the ambit of merger control to speculate into the future and to constrain anticompetitive market structures, to do so in this case would be extending our mandate into the realm of subjective speculation. There is no evidence to suggest that the merger is being proceeded with in order to enable DCSA to embark on a strategy of discriminating against its dealers. In the first place such a strategy appears irrational. They would not only end up causing dissatisfaction amongst their dealers to the detriment of their own distribution system, but also could render them uncompetitive in relation to the inter-brand competitors. Secondly, even if we were for a moment to assume it was a rational strategy, the merger is not required to facilitate it. As we have observed, the franchise agreement allows DCSA to terminate a dealer on six months notice. This leverage over dealers exists by virtue of a pre-existing vertical relationship upon which the merger has no impact, or at best, a de minimis one. The RMI was asked whether they felt that this transaction would increase DCSA’s leverage vis-à-vis the other dealerships by allowing them to gain a foothold in the retail market, enabling them to exploit this market. The RMI representative replied that if the merger took place the DCSA would have no added advantage in respect of the negotiations with the RMI since DCSA could open up their own or close down dealerships, in any event. 30 Amongst the reasons put forward by Mr Hiller in favour of the merger were some pro- competitive arguments. The transaction would make dealerships more aggressive vis-à- vis their competition. Developing closer ties with their customers would reinforce inter- brand competition, albeit at the expense of intra-brand competition. Similarly, rolling out pilot systems in their in-house dealerships would allow DCSA to assess risky types of marketing strategies, without necessarily exposing the other dealerships to those risks. This type of innovation would not occur if DCSA did not own dealerships. The larger dealers as we have seen are multi-franchise dealers who are not likely to finance a risky strategy in relation to DCSA brands since they are less committed to DCSA as an individual manufacturer. Thus it is entirely plausible that a DCSA-owned dealership 29 per Hiller, see transcript, page 49. 30 In a recent press release, it was revealed that VWSA intends to reduce the number of Volkswagen and Audi dealers, in line with international trends towards vertical downstream integration, illustrating that such is an acceptable market strategy and one not necessarily facilitated by acquisition or merger. 22 would be more committed to inter-brand competition and innovation than its multi- franchised counterparts. The more likely loss of intra-brand competition would be a loss of price competition amongst rival metro dealers. Yet as we have seen there is not much of that at the moment as dealers in the metropolis, such as McCarthy and SMH, are loathe to compromise their margins by competing on a price basis because their costs are high. Indeed it is more likely that the country dealers have been and will continue to be the source of intra-brand price competition, as they are faced with lower overheads than their metro rivals. The other possibility would be that DCSA would want to reduce the number of dealers in the metropolis that would compete with SMH. However since these dealers are largely the multi-franchise dealers this strategy seems less likely because they possess sufficient countervailing power in relation to DCSA. With their portfolios of dealer outlets DCSA would feel more constrained in terminating any one of them without concerns about its impact on their overall distribution capacity. The likelihood of DCSA utilizing this transaction to substantially eliminate intra-brand competition is too speculative to justify our intervention. In any event, the evidence reflects that there is sufficient inter-brand competition. We considered imposing conditions on our approval of this transaction, however, for the reasons expressed above, decided against this. Any conditions we imposed would be constraining conduct based on a fluid, unconfirmed strategy. The RMI presented us with the possibility of imposing a condition in terms of which their negotiations with DCSA were incorporated into our order. However, we decline from doing so since these understandings are best left to the parties to negotiate between themselves and we would be reluctant to enshrine this in an order relating to a Competition Act merger appraisal. We accept the Commission’s view that any future competition concerns would flow out of the franchise contracts, and any anticompetitive conduct on the part of either party with respect to the dealer network strategy will be more adequately addressed when it arises, under the auspices of the Competition Act’s restrictive practice legislation. Accordingly, we have approved the transaction without imposing conditions. _____________ 15 November 2001 N.M. Manoim Date Concurring: C. Qunta, F. Fourie