Santam Limited and Guardian National Ins

The judgment

The court's own file

IN THE COMPETITION TRIBUNAL REPUBLIC OF SOUTH AFRICA Case No: 14/LM/Feb00 In the large merger between: Santam Limited and Guardian National Insurance Company Limited Reasons for Competition Tribunal’s Decision Approval 1. We approved the merger between Santam Limited (“Santam”) and Guardian National Insurance Company Limited (“Guardian National”) without conditions on 4 April 2000. The reasons for our decision follow below. The Merger Transaction 2. The transaction involved Santam acquiring the entire issued share capital of Guardian National from Guardian National’s shareholders, which included Liberty Life Association of Africa Ltd (“Liberty”) and GRE South Africa Holdings (Pty) Ltd (“Gresa”). Prior to this transaction Liberty and Gresa held 40.8% and 52.56% of Guardian National’s shares. The holding company of Gresa, AXA S.A., was a party to the merger agreement and a participant in the merger hearing. Evaluation of the Merger The Relevant Market 3. Both Santam and Guardian National are registered short-term insurers in terms of the Short-Term Insurance Act 53 of 1998 and are authorized in terms of the Act to carry on short-term insurance business under all the classes of policies provided for in Section 67 the Act 1 . As a result, there is a direct product overlap between 1 The Act currently provides for eight classes of policies: accident and health, engineering, guarantee, liability, miscellaneous, motor, property, and transportation. 2 the businesses of the two firms. However, the two firms concentrate on different classes of products: Santam focuses on personal insurance products, while Guardian National focuses more on insurance for corporates and group schemes 2 4. There are at least three approaches to defining the relevant product market in this merger: (a) Defining a separate market for each type of short-term insurance product; (b) Defining a single market for all short-term insurance products; and (c) Defining a separate market for different ‘clusters’ of short-term insurance products. 5. The Commission’s analysis of the merger concentrated on the first two of these approaches, although their report did in passing mention the possibility of a relevant market based on ‘clusters’. 6. The first approach is consistent with international practice in respect of mergers in the short-term insurance industry. This practice is to define a separate relevant market for each type of risk covered 3 . Such a definition recognizes that insurance cover for a particular risk is a distinct product, which is not substitutable from the customer’s point of view for cover in respect of any other risk – for instance fire cover cannot be substituted for burglary cover. 7. In the present case, the Commission used the product classes provided for under the repealed Insurance Act, 1943 as the basis for defining a product market based on this concept of a separate market for each risk category. The repealed Act recognized six product classes, as opposed to the eight established under the new Act, to which we referred in paragraph 3 above. The six product classes under the repealed Act are fire, marine, motor, personal accident, guarantee and miscellaneous. The Commission appear to have chosen the product classification of the repealed Act rather than the classification established under the new Act in order to use historical (1998) information on market shares in their analysis. 8. The broader definition based on a single market for all short-term insurance products may nevertheless also be justifiable if sufficient supply-side substitution between the various product classes is possible – i.e. if insurers that only participate in certain of the classes are able to switch from providing one type of cover to providing another type of cover or are able to extend their product lines to include other categories of cover. Although we have not been presented with sufficient information to evaluate the extent of supply-side substitution in the industry i n any detail, the evidence on entry conditions, to which we will refer in 2 See Figure 2 at p. 40 of the Competition Commission’s report. 3 See for example the approach of the European Commission in Allianz/AGF (Case no. IV/M.1082); Commerial Union/General Accident (Case no. IV/M.1142); CU Italia/Banca Delle Marche/JV (Case no. IV/M.1627). 3 more detail below, suggests that the supply-side structure of the market may indeed support a product definition based on a single product market for all short- term insurance products. 9. The remaining market definition, which was not pursued by the Commission, is based on the general idea that a separate product market exists for various combinations of the distinct product categories. This would be the case if consumers of short-term insurance products showed a preference for insurance policies that covered more than one type of risk. The Commission advised us that some of the brokers that they interviewed had confirmed that consumers do in fact normally seek combination insurance – i.e. cover for a number of different risks under a single policy. Defining the relevant market in terms of specific clusters would, however, lead to a myriad of relevant markets due to the large number of different combinations of risk that can be incorporated under a single policy. Due to limited information, we have defined a surrogate cluster market instead of defining a separate market for each cluster permutation. Our surrogate market focuses on the supply-side of the market and includes all insurers that are in a position to offer clustered products. In this case, we have included all insurers that are registered in all or most of the product classes provided for in the repealed Act. 10. We do not have sufficient information to determine which of these approaches to defining the relevant product market is the most appropriate for analyzing this merger. We have therefore considered the merger’s effect on competition based on all three approaches. In any event, as is clear from our analysis below, the choice of product market definition is not determinative of our decision in this merger. 11. We agree with the Commission’s recommendation that the relevant geographic market for short-term insurance products in South Africa is a national market. Unlike most retail products, insurance products are not sold to consumers through retail outlets at specific locations. Rather, consumers rely on a nationwide network of brokers who source insurance cover from insurers nationwide. The geographic market does not e xtend beyond the national boundaries because legislation requires insurers who operate within the country to be licensed here. Consumers can accordingly not source short-term insurance internationally. Market Concentration 12. Table 1 contains the p re-merger market shares of firms in the six relevant markets based on product classes. The column on the extreme right gives the market shares in the single market for all short-term insurance products. The post-merger market share of the merged entity in each of these markets is shown in Table 2. 4 Table 1 Insurer Fire Marine Motor Personal Accident Guarantee Miscel. Total Mut. & Fed. 12 13 16 8 1 13 12.9 Santam 8 18 21 6 1 12 14 GNI 19 16 12 20 4 8 12.2 CGU 9 21 11 5 3 10 9.5 SA Eagle 6 10 10 2 1 6 7.1 Others (a) 46 23 30 60 90 51 44.2 Total 100 100 100 100 100 100 100 Source: Fitch IBCA Statistical Report (August 1999) and draft 1998 FSB Report (Table 4 in the Competition Commission’s recommendation) (a) The “other’ category comprised more than sixty small firms with mainly very small markets shares. Table 2 Fire Marine Motor Personal Accident Guarantee Miscel. Total Short-term insurance Market Share 27.3 32.6 34.9 22.6 5.73 19 25.6 Source: FSB, Registrar of Short-Term Insurance, Annual Report, 1998 (Table 7 of Competition Commission’s recommendation) 13. The market shares in Tables 1 and 2 translate into low to moderate pre-merger concentration levels in most of the product markets, as reflected in Tables 3 and 4. However, the increase in concentration in most of the markets following the merger will be substantial. Table 3 Fire Marine Motor Personal Accident Guarantee Miscel. Concentration Measure Pre Post Pre Post Pre Post Pre Post Pre Post Pre Post HHI (a) 880 1188 1029 1560 1165 1746 1005 1186 2390 2402 697 877 ? in HHI 308 531 581 181 12 180 Source: FSB, Registrar of Short-Term Insurance, Annual Report 1998. (Table 7 of Competition Commission’s recommendation) (a) Herfindahl-Hirschman Index 4 4 According to the US Merger Guidelines (1992, as revised in 1997) a market with post-merger HHI of between 0 and 1000 points is considered to be unconcentrated; a market with post-merger HHI of between 1000 and 1800 points is considered to be moderately concentrated; a market with post-merger HHI above 1800 points is considered to be highly concentrated. 5 Table 4 Source: Competition Commission Recommendation, Table 10 at p. 28 14. Market concentration in the surrogate market for clusters, which we defined in paragraph 9 above, can be roughly estimated with reference to the market shares of those insurers registered in all or most of the six policy classes. According to Financial Services Board records, as at 30 June 1999 more than two thirds of the 90 registered short-term insurers were registered in all six product classes, and many of the remaining insurers were registered in three or more of those classes. The market for clusters should therefore not be much narrower than the market based on all short-term insurance products. This mean that the level of concentration in this market, and increase in concentration as a result of the merger, is likely to be only slightly higher than in the market based on all short- term insurance products. 15. In summary, while market concentration after the merger will be moderately high in most of the markets considered, the increase in concentration will be relatively high 5 . The only market that will be highly concentrated after the merger is the narrow market for guarantee insurance. However, this market was highly concentrated before the merger as well and the increase in concentration as a result of the merger is quite small. Effect on Competition 16. Despite the relatively high increase in concentration in most of the relevant markets, the structural and dynamic characteristics of the short-term i nsurance industry in South Africa suggest that the merger is unlikely to significantly restrict competition in these markets. 17. The role played by independent insurance brokers as intermediaries between consumers and insurers is a particularly significant c haracteristic of the industry from a competition perspective. According to the Commission’s report, more than 95% of insurance business is conducted through independent brokers. These brokers receive requests for insurance cover from customers and then shop around for the best products available from insurers in terms of price and product characteristics. Because brokers have extensive knowledge of the industry and are well informed about product choices and market conditions, this arrangement contributes to a competitive market for short-term insurance products. 5 The US Merger Guidelines consider an HHI increase of more than 100 points in a merger that leads to a moderate level of concentration as a relatively large increase in concentration. Short-term Insurance Pre-merger HHI ? in HHI Post-merger HHI 700 328 1028 6 18. The manner in which brokers are remunerated supports the pro-competitive role broker-intermediation plays in this industry. They are remunerated in a manner which both encourages consumers to use brokerage services to source insurance cover as well as encourages brokers to pursue the best interests of customers in doing so. Consumers are encouraged to use brokers as intermediaries because they do not themselves pay for the brokers’ services; the insurer whose product is eventually chosen by the consumer pays the broker a commission. Moreover, insurers are prohibited by legislation from paying brokers an incentive bonus. This measure seeks to ensure that the fact that brokers are remunerated by insurers does not encourage brokers to establish a ‘comfortable’ relationship with any given insurer and thus reduce competition between insurers. Instead, brokers’ incentives in recommending an insurer are directed at protecting their client-base through satisfactory customer service. 19. The broker’s role as intermediary between the customer and insurer effectively consolidates the buying power of customers and should therefore contribute significantly towards countervailing the potential market power established by moderate to high concentration levels on the supply side of the markets. This conclusion is consistent with the findings of the European Commission in insurance mergers 6 . 20. Furthermore, the increase in concentration in the markets arrived at on the basis of simply summing the market shares of the merging firms probably overestimates the true increase, because this approach does not take into account the “run-off” that could be expected as a result of the merger. The merging firms estimated that they stood to lose up to 15% of their combined market share after the merger. They attributed this to two factors: the practice by brokers of sourcing insurance products from a number of different insurers; and the effect of uncertainty after the merger on s ervice levels. Although the extent to which these factors will contribute to the merged entity losing market share is difficult to quantify, the practice by some brokers of using a ‘short-list’ of insurers with which they place their business supports the view that a certain amount of run-off will result from the merger – if both merging firms appeared on a broker’s short list before the merger, an additional firm would now be included on its list after implementation of the transaction, which could result in less business being placed with the merged firm than the combined amount of business placed with the two merging firms prior to the merger. 21. Another factor that enhances competition in the short-term insurance markets is the ease with w hich customers are able to move their business from one insurer to another. According to a number of brokers interviewed by the Commission, the existence of claim-reducing measures such as so-called “no-claim bonuses” does not restrict customers from moving between insurers since these bonuses are generally transferable between insurers. 6 See for example Allianz/AGF (Case no. IV/M. 1082) at par. 33. 7 22. Apart from regulatory requirements, there do not appear to be significant barriers to entry into any of the short-term insurance markets. The most significant regulatory r equirement is that registration in terms of the Short-Term Insurance Act, 53 of 1998 is required to gain access to the industry. The Act prescribes a number of requirements for registration mainly of a prudential nature, none of which are particularly onerous. The large number of smaller firms that have entered the market recently supports the view that barriers to entry to the short- term insurance markets are not significant 7 . 23. The existence of a highly competitive environment in the short-term insurance industry is verified by comments made in internal management documents of Guardian National, which were prepared for purposes unrelated to notification of the merger. A background document for a Business Review and Budget Meeting held on 2 November 1998 mentions the aggressive marketing campaigns of competitors who were engaging in fierce price competition, stating that “competition from new and existing players [was] making it increasingly difficult to retain clients”. 24. Based on the above, we conclude that this merger is unlikely to substantially prevent or lessen competition in any of the relevant markets. We therefore need not consider the long list of mainly unsubstantiated efficiency gains claimed by the merging firms. ________________ Date: 3 May 2000 N.M. Manoim D. H Lewis and S. Zilwa concurred. 7 See Table 13 at page 34 of the Commission’s report.