---

COMPETITION TRIBUNAL

REPUBLIC OF SOUTH AFRICA

Case No: 56/LM/Aug04

In the large merger between:

Ellerine Holdings Ltd

and

Relyant Retail Ltd

______________________________________________________________

Reasons

______________________________________________________________

INTRODUCTION

1. The Tribunal has been asked to approve the merger between Ellerines Holdings Ltd and Relyant Retail Ltd.

2. The hearing took place on 7, 8 and 9 March 2005. The following witnesses were called:

1. Mr Jeff Dritz, Executive Director of Ellerines

2. Mr Bruce Sinclair, Chief Financial Officer of Relyant

3. Mr Peter Squires, Chief Executive Officer of Ellerines

4. Mr Aubrey Karp, Divisional Manager OK Furniture and House and Home

5. Mr Peter Griffiths, Managing Director Steinhoff Africa: Furniture Division

6. Mr James Moore, Managing Director of Poco International

7. Mr Alan Schlesinger, Chief Executive Officer of Relyant

3. We have decided to approve the transaction. The reasons for our decision are set out below.

THE TRANSACTION

4. This is a horizontal merger. Both merging parties are active in the retail market selling furniture and home appliances. They each trade through a number of retail brands directed at various income-based segments of the market. In addition, a number of vertical issues are raised by the transaction. These relate to the relationship of Steinhoff, a major supplier of furniture, to the merging parties.

5. The primary acquiring firm is Ellerine Holdings Ltd (‘Ellerines’). It is a public company listed on the JSE Securities Exchange. Ellerine’s shares are widely held and the company is not controlled by any single shareholder. Ellerines trades through five primary branded retail chains:

• Ellerines

• FurnCity

• Town Talk

• Wetherlys

• Osiers

6. Ellerines, the oldest and strongest brand in the Ellerines Group, and Town Talk target the lower LSM categories. FurnCity targets the middle LSM categories. Wetherleys and Osiers are targeted at the top end of the market. There are 310 Ellerines stores, 156 Town Talk stores, 137 FurnCity stores, 11 Wetherley stores and 6 Osiers stores.

7. Ellerines is well positioned in the bottom and high end of the markets with strong, well-known brands. Its weakness, however, lies in the middle segment – as we shall elaborate below, it avers that although FurnCity is targeted at the middle segment of the market it has never succeeded in differentiating itself from the more down-market Ellerines chains.¹

8. The primary target firm is Relyant Retail Ltd (‘Relyant’), a public company listed on the JSE Securities Exchange. Relyant’s largest single shareholder is Poco International, a German-based furniture retailer. However, a consortium of banks controls a very large bloc of Relyant’s shares. Relyant’s shareholding structure is as follows:

Poco International 40,54%

First National Bank 26,28%

Absa Bank 15,67%

Standard Bank 5,08%

Investec Bank 5,00%

9. Relyant operates through three primary divisions. These are referred to as the furniture chain division, comprising a number of retail chains directed at the lower and middle segments of the market, a value retail division directed at somewhat higher market segments and a services division. The distinction between the furniture chain division and the value chain division essentially corresponds to the distinction between, on the one hand, customers who purchase furniture on credit extended to them by the retailers and, on the other hand, cash customers.² Relyant trades through the following brands:

• Geen & Richards

• Beares

• Lubners

• Savells/Fairdeal

• Furniture City

• Glicks

• Dial a Bed

• Mattress Factory

10. Ellerines intends to acquire Relyant. This is a share-based transaction with Relyant shareholders receiving 10 shares in Ellerines for every 225 shares in Relyant. Upon successful implementation of the proposed transaction Ellerines will control Relyant. The share exchange will result in Poco International, currently the largest shareholder in Relyant, becoming the largest shareholder in the merged entity.

11. The largest shareholders in the merged entity will be:

• Poco International 14.3%

• First National Bank 10.11%

• Public Investment Commissioners 9.98%

• Investec 7.31%

• Standard Bank 4.22%

• RMB 4.12%

12. The post-merger Board of Directors will comprise seven non-executive and four executive members. Two of the non-executive directors will be appointed by Poco International. Mr P.J.B. Pohlmann, the current chairman of Relyant and of Poco International, will serve as the non-executive chairman of the merged entity and the current chairman of Ellerines, Mr D.S. McGlashan, will serve as deputy chairman of the merged entity. Mr P. Squires, the current CEO of Ellerines, will retain this position in the merged entity.

THE BACKGROUND TO THIS TRANSACTION

13. The competition authorities have reviewed several transactions in the furniture retail sector. In July 2000 the Tribunal prohibited a proposed merger of the JD Group and Ellerines.³ In essence, the Tribunal held that the merger would substantially lessen competition in that segment of the furniture retail market that served low-income consumers. However, since the prohibition of the JD/Ellerines transaction, the competition authorities have approved five important mergers in this sector, these being the Massmart/Furnex transaction,⁴ the First Rand/Profurn transaction,⁵ the JD/Profurn transaction,⁶ the Ellerines/Wetherlys transaction⁷ and the Lewis/Lifestyle Living transaction.⁸

14. Most of these mergers were approved on the basis that the merging parties were not active in the same market segment. The exception is the JD/Profurn merger where the Tribunal was extremely concerned about the horizontal overlap between the merging parties. However, imprudent credit granting by Profurn, aided and abetted by their largest shareholder, First Rand Bank, compounded by an economic downturn, had brought Profurn to the brink of demise. The Tribunal was persuaded that, absent a merger, a significant part of Profurn’s capacity would have exited the market. It accordingly approved the merger subject to conditions designed to address certain vertical problems.

15. The JD/Profurn transaction has proved to be something of a watershed in this sector. JD’s timing, whether a result of good fortune or prescience, was brilliant. Aided by a significant economic upturn, JD’s considerable managerial resources have managed, more rapidly and effectively than most observers believed possible, to turn around the struggling Profurn brands. In the process, JD’s risk taking has been rewarded by its successful establishment of a portfolio of brands across the entire spectrum of LSM profiles with particular strength in the middle-income segment of the market, generally designated as that segment of the market serving consumers in the LSM 4-7 category. This, as we shall elaborate below, is particularly significant because, it appears, it is precisely consumption in this segment of the market that has fuelled the strong economic upturn of recent years.

16. This confluence of factors – the acquisition and rapid turnaround of Profurn, the particular strength acquired through this transaction in the middle segment of the market, the economic upturn led precisely by consumers in this segment - has enabled the JD Group to assume an increasingly powerful position in the furniture retail market. In short then, JD’s success, it is argued, lies in its ability to serve, ‘from cradle to grave’, the increasingly large swathe of consumers who make up LSM 4-7. In other words, in its ability to service, its customers through their purchasing life-cycle. In this consumer life-cycle, a powerful position that is focused on the lowest consumer segments – LSM 3-4 – has diminished in importance because, it appears, a number of factors, notably, easier access to cheaper credit, has enabled a significant grouping of low to middle-income consumers to enter the consumer life cycle at a somewhat higher level than was hitherto possible and then to proceed somewhat more rapidly up the consumer ladder. Hence from a consumer life-cycle perspective – or, expressed otherwise, from the critically important perspective of ‘re-serving’ customers - it is the ability to satisfy consumers in the LSM 4-7 category that is crucial.

17. The JD group is not alone in possessing this attribute. Certainly the Lewis group appears uniquely capable of capturing this large grouping of consumers within the umbrella of a single brand, and the aggressive OK furniture chain is a growing presence in this market.

18. However it is argued that the venerable Ellerines group is left vulnerable in the face of these market dynamics. Although it continues to outrank its competitors on most operational and financial measures, Ellerines’ future growth is threatened by its weakness in the LSM 4-7 market. It pioneered the sale of furniture on credit to low income black consumers and this is where it is, for the most part, mired as its customers and their progeny steadily migrate to stores in competitor chains that serve higher incomes and aspirations. It has, through the acquisition of Wetherley’s, successfully located itself in the upper echelons of the furniture market but its attempts, through the Furncity chain, to establish itself in the crucial middle segment of the market have met with only limited success. Ellerines’ predicament is graphically described by Mr. Jeff Dritz, an Ellerines executive director, who testified at the hearings:

“Ellerines sits like a body with legs and arms without the body in the middle..”⁹

19. The target company, Relyant is, for its part, a group with different problems and prospects. The group was formed as a result of the merger of the Amrel and Beares groups, a merger that apparently bequeathed the new group a mixed legacy comprising a family of some solid and well-established brands and a number of less salubrious relatives. The group attempted to solve its inherited problem by closing down or repositioning some brands complemented by a programme of acquisitions designed to plug up gaps in the new group. The upshot is that Relyant has, throughout its lifetime, been characterised by an excessive number of brands in its stable and by what appears to be an almost permanent state of restructuring.

20. In fact at the time of the JD/Profurn merger, in 2002, Relyant had embarked on a strategic review of its major brands. The upshot of this review was inevitably a new strategy, the implementation of which required substantial capital. Shortly after the review process was embarked on, Poco, a German-based furniture retail company looking for investments outside Germany, invested in Relyant.¹⁰ The implementation of the results of the strategic review came to a halt while Poco, which had acquired a 40% share of Relyant and the chairmanship of the Board, attempted to negotiate more funding for the cash strapped Relyant with the banks, who collectively held 49% of the shares. However, in late 2003 these negotiations, between Poco and the Banks, broke down. It appears that the banks were able to hold up the restructuring of Poco both through their position as sources of loan capital as well as through their equity shareholding and board membership. It appears that Poco offered to purchase the banks equity with the intention of de-listing the company but this offer was rejected.

21. In early 2003 the Relyant board had also appointed a new CEO, Mr Schlesinger, who was given the task of stabilising and turning around the Relyant business. He was immediately tasked with reviewing the turnaround strategy that had been completed some time earlier. In February 2004 management advised the Board that the identified strategy remained viable. They also suggested that a rights issue for approximately R150 million should proceed to fund the process. Although the Board was not ad idem on the restructuring strategy in view of Relyant’s financial constraints, in March 2004 the Board approved the process and tasked Schlesinger and Sinclair with seeking underwriting for the rights issue. However, they could not persuade the Banks to support a rights issue.

22. During the latter part of 2003 Mr. Theunie Lategan, a First Rand Bank employee and a member of the Relyant board of directors, introduced Mr. Peter Squires, CEO of Ellerines, to Mr. James Moore, with a view to exploring an arrangement that would allow Ellerines to piggyback on Relyant’s Triad credit control hubs. It was one of Ellerines stated objectives to improve its IT system. This came to nought largely because Ellerines was anxious to ensure the confidentiality of its credit information and customer data. However, although this process failed, the meetings did pave the way for Relyant and Ellerines to canvass a possible merger between the two competitors. It is this transaction that is before us today.¹¹

THE RATIONALE FOR THE TRANSACTION

23. This then is the rationale from Ellerines perspective for this transaction: it is principally interested in acquiring key Relyant brands in the LSM 4-7 category. It believes that in the absence of this transaction Ellerines will be unable to penetrate this critical market segment and that it will continue losing ground to its better placed competitors, notably, although not exclusively, the JD Group. It has attempted to penetrate this segment of the market by developing its own brands. Its chosen vehicle for this task was the FurnCity brand. However although FurnCity is by now a large and commercially successful chain, it has not effectively distinguished itself from the Ellerines brand, it has not, in other words, succeeding in penetrating its target market. Ellerines argues that the power of its flagship brand and its inextricable relationship to the lowest income consumer ‘taints’ all efforts to penetrate higher income segments of the market – it suggests implicitly that to aspire to move into a higher consumer class has become synonymous with moving away from those brands associated with Ellerines.

24. Relyant, although, on every measure, the weaker of the two parties, does not suffer the same structural shortcomings as Ellerines. It is reasonably represented across the LSM range and it has several strong brands in the all-important middle segment of the market. Under the leadership of its current CEO and with the active participation of Poco, its largest shareholder, it has made enormous strides from its parlous position of recent times although it has also been assisted by a favourable economic environment. But, while, unlike Profurn, it is careful to disavow any prospect of failure, it insists that its turnaround has not been secured, that it remains vulnerable to changes in the external economic environment and to several internal shortcomings. It believes that the turnaround and rationalisation of its brands will be secured by access to Ellerines’ financial and managerial resources and its IT infrastructure.

25. Moore describes the situation in which Ellerines and Relyant respectively find themselves as follows:

“At the moment Ellerines can’t do that because they don’t go high enough and we can’t do it now because we’re not strong enough. … We haven’t got enough money…” ¹²

COMPETITION ANALYSIS

Relevant market

26. As already suggested, in earlier transactions in this sector, notably the Ellerines-JD and the JD-Profurn transactions, the relevant markets were determined by a three-fold segmentation of furniture consumers into a low-income category (LSM 3-5), a middle-income segment (LSM4-7) and an upper-income segment (LSM8). For reasons which are exhaustively elaborated in these decisions, it was determined that the geographic markets were national – that is, that these three relevant markets comprised national chains of furniture shops and did not include locally-based ‘independents’. Without attempting to present the large body of evidence supporting the conclusion on the relevant geographic market, it suffices to say that we were satisfied that the pricing and other competitive strategies of the national chains were not responsive to the competitive behaviour of the independents.

27. The parties to this merger have argued for a single product market, that being the retail market for furniture and appliances.

28. As regards the geographic market, the parties insist that these are local and that the independents be incorporated into those local markets in which they are active. The merging parties aver that the chains have regard to the prices of the independent furniture stores, which act to constrain the market behaviour of the chains. They argue that the merchandise supplied by the independents is the same as that supplied by the national chains, that they are located in the same part of town and that many offer credit on terms much the same as the chains.

29. In its assessment of the current transaction, the Commission argues that the dynamics of the furniture retail market lend support to a new definition of the relevant product market. The Commission found that although certain practical indicia pointed towards a lower, middle and upper segment in the furniture market, it is difficult to clearly and accurately distinguish between the different LSM groups targeted by the various market participants. It found that certain branded stores targeted both the traditional lower segment and middle segment customers. Accordingly, it defined a somewhat narrower relevant product market than the parties, as furniture shops directed at credit sales excluding independent furniture shops and mass discounters.

30. On the geographic market the Commission recommends maintaining the Tribunal’s earlier view that this is national. Although this is opposed by the parties who continue to insist on local markets, the Commission points out that the very arguments advanced by the parties in favour of this transaction – in particular the argument that scale economies represent a critical competitive advantage in the retail furniture trade – emphasise the independents’ inability to compete effectively with the national chains.

31. This view of the relevant geographic market may not hold for the up-market chains, those like Wetherlys and Osiers and possibly Glicks, who serve consumers in the LSM range 8-10. While it appears that the upmarket chains are also centrally managed, these generally comprise a significantly smaller number of larger stores than in the case of their counterpart chains serving the lower LSMs. Accordingly this segment of the market is not characterised to the same extent by the simultaneous existence of massive country-wide chains and locally based independents that is a feature of the lower segment. This is consistent with the evidence that suggests that in the upper LSMs the independents have a significant larger market share than do the independents competing with the giant chains in the lower LSMs.¹³

32. We do not accept the Commission’s broad market definition. Clearly the distinction between the LSM categories remain an important element in the furniture chains’ branding strategies. The fact that Ellerines acknowledges that it needs to compete in the middle market, in order not to lose its entry-level customers when they migrate to higher LSM categories, evidences the continuing segmentation of the furniture market and of the orientation of the retail brands.¹⁴ In fact Mr. Squires’ explanation for Ellerines’ failure to penetrate the ‘middle market’ – and Ellerines consequent inability to achieve the status of what Squires terms a “universal chain” – rests on the continued salience of LSM-based segmentation.¹⁵ Mr. Dritz also refers to the different LSM categories that exist when he notes that: “low-end consumers aspire to brands targeting higher LSM groups”.¹⁶

33. However we do accept that the lower and middle-income segments of the furniture market have drawn closer together. The greater congruence of these previously segmented markets has been promoted by no less significant a structural change than the emergence, for the first time in the country’s history, of a black ‘middle class’. As already noted, the evidence suggests that this has been facilitated by the significantly lower rates of inflation that have characterised recent years, the lowering of interest rates and thus the effective decline in the cost of credit, and the emergence of new sources of credit notably the extension of credit card facilities to the broad middle mass of furniture consumers. This is, undoubtedly, underpinned by the de-racialisation of mid- and higher- level employment opportunities in both the private and, particularly, the public sectors and is likely complemented by a greater sense of confidence amongst black consumers in the sustainability of these gains and, hence, in a greater willingness to assume debt.¹⁷ But whatever the underlying cause of this break down in the historic segmentation of furniture consumers, certainly, a major imperative of the furniture retail chains is to ensure that their stores provide for the possibility of a relatively seamless transition of their customers from LSM 3 to LSM 7 or, ‘from cradle to grave’.

34. We, therefore, find that this transaction covers two national product markets these being, firstly, the market for sale of furniture to consumers in the LSM 3-7 and, secondly, the market for the sale of furniture to consumers in the LSM 8-10 categories.

Market Shares

35. This is a horizontal merger, although there are certain vertical dimensions that will be considered. We have concluded that there are two relevant furniture markets, these being a segment whose principal target market is LSM 3-7 and then an upper segment focused on LSM 8 and above. Our analysis of the competition implications of the transaction will focus on the former segment, on the sale of furniture to consumers in LSM 3-7. The evidence presented does not point to likely competition concerns in the upper segment – LSM 8-10 - where, as already elaborated, the market shares of the merging chains are relatively low and where the increment that accrues in consequence of the merger is small.

36. Turning then to the lower segment – the retail furniture market serving customers in LSM 3-7 - we note that although the principal target market for this segment is LSM 4-7, key chains attached to the national furniture groups actively contesting this segment do focus on LSM 3-4 (Ellerines, Savells, Fairdeal). However it appears increasingly that the overriding value of this exposure to the lowest reaches of the broad furniture retail market is to establish a bridge from there into the broad middle segment. As already elaborated, Relyant’s weakness at the lower end does not provide it with a sufficient base of entry-level consumers to migrate into its strong mid-level brands. Ellerines, on the other hand, sees its powerful position at the lower end of the market dissipate as its entry-level customers migrate to the mid-level brands of rival chains.

37. The market shares of competitors in the middle segment market, LSM 3-7, before the merger are: ¹⁸

Group	Brand name	% market share	% market share per group before the transaction
Relyant	Savells/Fairdeal	4.69	16.46
	Bears	4.83
	Geen&Richards	1.97
	Lubners	2.61
	Furniture City	3.20
Ellerines	Ellerines	5.33	11.25
	Town Talk	3.18
	Furn City	2.74
JDGroup	Bradlows	4.76	45.50
	Barnetts	4.44
	Price & Pride	5.34
	Joshua Doore	9.17
	Morkels	6.27
	Russels	12.72
Lewis	Lewis	13.96	15.17
	Electric Express	1.21
Shoprite	House & Home	6.39	11.62
	OK Furniture	5.22
Total		100.00	100.00

38. The merged entity’s market share post the transaction will be 27.71%. It will be the second largest player after the JD Group, which will remain the largest player with a market share of 45.50%.

Impact of the transaction on competition in the relevant market

39. At the risk of repetition, we note that we have identified two relevant markets, the first of these being the market populated by national retail furniture chains that serve customers broadly categorised as belonging to LSM categories 3-7. The relevant geographic market is national.

40. The second relevant market identified is that serving purchasers of furniture in the LSM 8-10 categories. As already noted, we are of the view that the merger does not threaten competition in this market. Relyant has a weak presence in this market and hence accretions to market share resulting from the merger are relatively slight. Again, we have already noted that there are conspicuously fewer chains serving this segment of the market, with most having started their still relatively short lives as single stores in one or other of the large cities where they, by and large, remain. Although the development of small chains of these relatively upmarket stores has resulted in a significant degree of centralisation of purchasing and pricing decisions, it appears that competition from independent stores is significant and probably justifies finding that the relevant geographic market for this segment is regional. We will not examine this market further.

41. Turning then to the first of these relevant markets – the national retail furniture market serving customers in the LSM 3-7 categories - we note that although the dynamic features of the market mentioned above have served to elide the distinction between the lower (LSM 3-4) and upper (LSM 4-7) segments of this market, there nevertheless remain two discernible segments. Although we accept that the precise location of consumers in this broad category is more fluid – that is, they are more upwardly mobile - than at any other time in South Africa’s economic history and, hence, that the LSM 4-7 categories have assumed overriding importance, there are many South Africans mired in the lowest LSM categories. In other words the lower segment of the market does not merely serve as a point of entry and transmission belt to the higher segment of this market. Many South African furniture consumers will not ascend into the higher LSM categories and so we are at pains to ensure that this transaction does not portend a substantial lessening of competition in this market.

42. However were we to segment the relevant market that we have found and then to refocus on the lower and upper categories of this market, we would find that this is a merger between one company, Ellerines, that holds a very powerful position in the lower segment and a fairly weak position in the upper segment, and another, Relyant, whose profile is precisely the reverse – that is, its chains that are directed at the lower segment are weak and do not much enhance Ellerines position in this segment while its chains in the upper segment are impressive. In other words by segmenting the relevant market we limit the horizontal overlap between the merging parties. It is precisely because we believe that this segmentation has eroded significantly – albeit not disappeared – that we are concerned about competition in a market that is focused on LSM 4-7, but one in which a strong presence in LSM 3-4 is important precisely because of the transmission belt that a strong position below LSM 4-7 provides to that core market.

43. Section 12A(2) of the Act provides that:

(2) When determining whether or not a merger is likely to substantially prevent or lessen competition, the Competition Commission or Competition Tribunal must assess the strength of competition in the relevant market, and the probability that the firms in the market after the merger will behave competitively or co-operatively, taking into account any factor that is relevant to competition in that market, including-

1. the actual and potential level of import competition in the market;

2. the ease of entry into the market, including tariff and regulatory barriers;

3. the level and trends of concentration, and history of collusion, in the market;

4. the degree of countervailing power in the market;

5. the dynamic characteristics of the market, including growth, innovation, and product differentiation;

6. the nature and extent of vertical integration in the market;

7. whether the business or part of the business of a party to the merger or proposed merger has failed or is likely to fail; and

8. whether the merger will result in the removal of an effective competitor.

44. It is our view that the relevant market is characterised by reasonably intense competition. The JD group is a very powerful competitor in each segment of this market as is Lewis. OK, too, is a powerful and, it appears, rapidly growing participant in the relevant market. While, as elaborated above, Ellerines and Relyant, have problems with their representation across the various segments, they remain significant competitors – Ellerines, certainly, is a very powerful brand and its position in the lower segment of the relevant market is particularly significant. However detailed consideration of the relevant factors that we are obliged to take into account when assessing the future of competition in this market, undoubtedly give rise to concern.

45. Barriers to entry (Section 12(2)(b)) are clearly significant. The large furniture retail chains are characterised by the longevity of their brands and it is difficult to imagine the successful establishment of a new chain in anything approximating the short or even medium term. The limited success of FurnCity even with the support of Ellerines’ infrastructure, managerial experience and brand awareness (although, as elaborated above, the brand is argued to be a mixed blessing) is noteworthy.

46. OK seems to have grown its presence in the furniture industry reasonably rapidly although it is undoubtedly assisted by its brand, by the deep pockets of its parent and by key aspects of the group infrastructure, particularly attractive locational advantages. However, even while enjoying these advantages, Mr. Karp, the witness from OK, estimated that he required 6 years to establish the brand in the middle market.¹⁹

47. We have seen that although the structure of the market is reasonably competitive, that the trend, over the past five years at least, is in the direction of consolidation and greater concentration. This process was given considerable impetus by the absorption of the Profurn brands into the JD Group. While we have no evidence of collusion in this market, it is very difficult for potential customers to engage in price comparisons – credit terms differ, ‘specials’ are ubiquitous, and comparisons between the slightly differing specifications of the products sold by the various chains are difficult to make.

48. The upshot of this is that much puffery combined with credit terms that are difficult to reduce to a common price may provide an appearance of competition but may equally serve to camouflage effective co-operation – certainly it is extremely difficult for customers to evaluate the merits of alternative offerings. As we will outline below, despite the appearance of product differentiation, we believe that the increasing homogeneity of the product offerings – and therefore the susceptibility of the market to collusion -is promoted by the increasingly powerful position of furniture manufacturers, notably Steinhoff, in the upstream market.

49. There is clearly no import competition (Section 12(2)(a)) or countervailing power (Section 12(2)(d)) in this market. We deal more fully with the question of vertical integration (Section 12(2)(f)) below.

50. The dynamic characteristics of the market (Section 12(2)(e)) have, as we have already shown, exerted a powerful influence on our identification of the relevant market. In summary, a number of critical factors – rooted in political change that is reflected in the changing demographic character of the labour market and in a macro- and micro-economic policy that has dampened product price increases and cheapened the cost of credit – has served to blur, although clearly not eliminate, the distinction between low and middle income consumers. However the impact of these dynamic features on the definition of the relevant market aside, the furniture market is not distinguished by its dynamism – in particular, as already suggested above, we are concerned that the structure of the market for the production of furniture and the interplay between the dominant player in the upstream market and certain of the retail chains, notably the JD Group, may have served to stunt innovation and product differentiation. We note however that import competition in the furniture market itself, may facilitate differentiation in the downstream retail market.

51. We must then examine Sections 12(2)(g) – the possibility of the failure of one of the parties to the merger – and 12(2)(h) – the removal of an effective competitor.

52. In contrast with the claims made and accepted during the JD-Profurn transaction, there is no claim made here that Relyant is threatened with failure. Indeed Mr. Schlesinger, the Relyant CEO, specifically insisted that there was no threat of failure on the part of Relyant. And nor would this be a convincing claim. Certainly, there was a time when Relyant’s future appeared to hang in the balance, however a combination of the introduction of a strong and committed shareholder in the shape of Poco and a competent management in the shape of Mr. Schlesinger and his team and a favourable turn in the economic environment, has served to place Relyant on a considerably sounder footing. Indeed there are claims made throughout the papers that Relyant has already taken the bitter medicine that the other furniture retail chains are yet to imbibe and that this will hold Relyant in good stead in the immediate future.²⁰

53. Equally, however, Mr. Schlesinger, insists that although Relyant is not threatened with imminent failure, nor is it quite out of the woods yet:

“ What I’ve been able to do is in a favourable environment, a favourable business environment give people maybe a greater degree of freedom, helping them in their thinking, in their planning. In their merchandise planning particularly we’ve been able to improve that. So there are some superficial changes, which have helped us to exploit the favourable economic times that we’ve been in. But I fear that in the event of the market becoming tight again, I think we will fight very, very hard not to slide back to being a poor, weak competitor… At the moment we are still vulnerable to down trend.” ²¹

54. From his evidence it is clear then that Mr. Schlesinger is particularly concerned at the prospect of an economic downturn – in that eventuality, he argues, the underlying vulnerabilities that continue to beset Relyant would be harshly exposed. We accept this evidence and argument. And although it does not amount to the invocation of a failing firm defence to the merger, it does go to the one remaining factor to be analysed, namely, the question of whether or not the merger results in the elimination of an effective competitor.

55. The merging parties have attempted to justify the merger on the grounds of JD’s increasingly powerful position in the market. As noted, its position has been strengthened by the timing of its acquisition of Profurn and by its demonstrated ability to rapidly absorb Profurn’s brands. The Profurn acquisition has also consolidated JD’s presence across the consumer categories, in particular in the two segments represented in the market relevant to this transaction, namely the retailing of furniture to LSM 3-7. The merging parties in this transaction argue that JD’s growing strength is borne out by its market share, which as the table above shows, has advanced from 43% in 2003 to 45.50% in 2004. This evidence is then deployed to support the contention that the merging of Ellerines and Profurn is necessary to provide an effective competitive counterpoint to JD.

56. The Commission has examined this argument – including relevant international jurisprudence – at some length.²² The Commission cites Areeda who insists that “..simply reducing size disparities without increasing efficiency is not likely to increase competition”. It also quotes from Areeda:

“When one starts with large disparities in an already concentrated market, however, reducing the number of firms and increasing the size of the market’s smaller firms makes non-competitive pricing more likely. Unless there are scale economies, the merged firm is less likely to compete on price than the unmerged firms previously were …. Mutual beneficial coordination of pricing is much easier in a four-firm market… [M]ere reduction of size disparities is not a persuasive reason for treating mergers in highly concentrated industries more liberally.”

57. The Commission is not convinced that the furniture market requires a further concentration to attain efficiencies. It is uncomfortable with the post merger market structure where two significant firms will be competing with two relatively small firms and is concerned with the effect that this will have on what is identifies as the two fringe firms, Lewis and OK.

58. We concur with what we interpret as the tenor of anti-trust jurisprudence on this question. That is, we are extremely reluctant to permit the merger of two effective competitors on the grounds that the merged entity will provide more robust competition to the largest player in the market concerned. However we are prepared to approve the merger on the following grounds:

• Given that the post-merger market structure will comprise four significant groups in JD, Ellerines-Relyant, Lewis and OK Furnishers, the merger will not result in a substantial lessening of competition – we would not follow the Commission and describe the Lewis and OK chains as ‘fringe’. Moreover, while, as noted above, we are not persuaded that the retail market is impervious to collusion, we do not think that the ability to collude is strengthened by the fact of this merger.

• While, again as noted above, Relyant is by no means a failing firm, we are not persuaded that it is in the short to medium term a particularly effective competitor. On the other hand, we are persuaded that the merger will strengthen Ellerines, which, though undoubtedly an effective competitor, appears to face significant obstacles in its attempts to establish a competitive position in the upper segment of our relevant market, that is, in the sale of furniture to LSM 4-7.

59. The players that remain in the market post the transaction are:

• The JD Group, which has managed to set up a portfolio of brands across the entire spectrum, with particular strength in the middle to upper sectors of the market, thereby successfully attracting and keeping customers throughout their lifecycle. It is a financially strong company and the largest customer of the biggest furniture supplier, Steinhoff.²³

• Lewis, which has its roots in the traditional white market with stores that are usually found in the more upmarket part of towns, is a strong brand that spans most of the LSM categories. It has also indicated in its prospectus that it is focussing on the rapidly growing LSM 4-7 categories. Although it conducts business with Steinhoff it has a policy of supporting independents suppliers.²⁴ It is also a large importer of furniture.²⁵

• The Shoprite Group comprising House & Home, focusing on LSM 7-10 and OK Furniture, focusing on LSM 5-7. It has the backup of a large shareholder, favourable locations and is something of a maverick in its pricing policy, focussing on the volumes of customers that visit its supermarkets, which are located adjacent to its furniture divisions.²⁶

• The merged Ellerines/Relyant Group. Ellerines dominates the bottom end of the LSM categories with very strong brands and is clearly intent upon maintaining a base of independent furniture suppliers.²⁷ Relyant, which has recently restructured its brands, is well represented in the middle market with Beares being its flagship brand. Relyant also boasts an advanced credit control system which is much more efficient than that of Ellerines.²⁸

60. We have already noted that while Relyant is clearly not a failing firm – and nor has it claimed to be – its recovery nevertheless remains delicately balanced. In particular the new Relyant has not been tested in an unfavourable economic climate. And there are significant competitive lacunae that remain – these range from their inadequate IT infrastructure through to deep problems in certain of the chains. Certainly the problems at the ‘city’ brands – Furniture City and Appliance City – as well as Glicks remain unresolved and, possibly more important, the sustainability of their brands in the lower segment of our relevant market – Savells/Fairdeal - is still open to question. Nor are we persuaded that these problems will be easily overcome. It is clear from the evidence that restructuring the problematic brands and upgrading the Relyant IT infrastructure will be costly.

61. Although Relyant has a committed and experienced shareholder in the shape of Poco, the latter’s positive influence has to be set off against the consortium of banks that together carry equal shareholder weight. In general the banks have shown themselves to be more than willing to use the furniture retailers as conduits for passing on credit to a part of the population whose direct custom they appear to eschew. However as both the Profurn and now the Relyant experience indicate, the banks are fair-weather friends, quick to withdraw their support as soon as the sale of furniture on credit appears to be under threat. The practice of credit-granting by the furniture retailers is undoubtedly confronting a somewhat more hostile environment than hitherto, not least because the banks are entering the LSM 4-7 market more aggressively themselves and would not necessarily want to be supporting a formidable rival – the retail furniture chains – in this competition for the provision of credit.²⁹

62. Seen against the background of these continuing problems at Relyant and Relyant’s stronger representation in the middle market, the merger with Ellerines offers a compelling synergistic-type fit. In particular Ellerines’ undisputed strength at the lower end of the relevant market will bolster Relyant’s considerably weaker brands at that level (or will allow for their consolidation into the Ellerines brand). Ellerines considerable IT capacity will be extended to its partner and its managerial and financial capacity will more easily allow the new group to confront some of the problems that Relyant face in respect of certain of its chains, notably the ‘cities’ chains.

63. The potential gains for Ellerines are also considerable. Although the old models of credit granting by the furniture retailers face challenges it is, however, decidedly premature to proclaim the death of furniture credit. The banks’ ability and willingness to sustain entry into this market is not proven. And certainly Ellerines itself sees much life yet in the credit side of its business. Ellerines pioneered the granting of mass consumer credit in South Africa and it appears to have lost little of its edge. It is, of course, also true that of the major furniture groups Ellerines is most exposed to the lower LSM segment of the market, that segment of the market most reliant upon credit and which the banks are least likely to penetrate. It is, however, likely that credit granting will become more customised, more closely directed at the particular recipient of credit. To this end Ellerines will require access to sophisticated credit monitoring IT, which appears to be a current strength of Relyant.

64. But IT can be bought and Ellerines undoubtedly has the capacity to make the necessary investments. However in order to enter successfully the upper segments of the relevant market – LSM 4-7 – it may well require a merger with Relyant. Several witnesses testified that the continued competitiveness of Ellerines is threatened by its inability to establish itself in this middle segment, which is increasingly regarded as the core segment of the broader furniture market.

65. As outlined earlier, its only exposure to the LSM 4-7 market is in Furncity. There are 137 stores in this chain that has, the various witnesses testify, never been able to shake off its association with the Ellerines brand, which, it is averred, is inextricably associated with the Ellerines core brand.³⁰ Mr. Squires testified that FurnCity gets only 1% to 2 % repeat business that it draws from the Ellerines and Town Talk brands.³¹ It appears that, for consumers buying furniture, to aspire upwards, is to aspire to leave behind the Ellerines brand in favour of a brand more strongly associated with serving somewhat higher income consumers. Nor, it appears, is this only a matter of consumer perception. Witnesses describe Furncity as a thinly disguised version of the Ellerines brand stores.³² Ellerines has been well known for the tight, highly regimented disciplines imposed on its staff. This has undoubtedly reaped a rich harvest for the company that is more than 50 years old. However, if the witnesses are to be believed, it appears also to have instilled a certain rigidity that is partly responsible for the firm’s inability to adapt to changing market conditions.

66. We now turn to the vertical aspects (section 12A(2)(f)) of the transaction. As indicated earlier a number of vertical issues arise, which relate to the merging parties’ relationship with Steinhoff. The increasing homogeneity between products offered by the retailers, Steinhoff’s growing market share and the continuing exit of independent furniture suppliers, even in these economic buoyant times, are just some off the concerns that have been raised in this transaction.

67. The competitors in the upstream market are Steinhoff, a major international supplier that has supply agreements with most of the major furniture retail groups, a number of much smaller independent local suppliers and more recently imports by retailers, mainly from the East.³³

68. Steinhoff is Relyant’s largest single supplier. Relyant is Steinhoff’s second largest customer and Ellerines its fifth largest.³⁴ Though both merging parties are Steinhoff customers, they, in fact, have very different relationships with Steinhoff.

69. Ellerines is recognised in the industry as a group that strongly supports independents. Schlesinger also alluded to this when he noted that the merger must be of some concern to Steinhoff:³⁵

“I mean they are suppliers to Ellerines, as I understand it, but Ellerines is on record as very actively supporting the small manufacturing sector.”

70. In the JD/Profurn transaction, in 2003, Mr. Squires had raised some concern about the effect that the transaction would have on independent suppliers since JD and Profurn both had close relationships with Steinhoff. It is now two years since that merger and Ellerines is still committed to supporting independent suppliers insisting that its policy of spreading the ratio of purchases between the independent manufacturers (80%) and Steinhoff (20%) remains. According to Squires the combined Ellerine’s Group currently buys 22.9% of all its merchandise from Steinhoff.³⁶

71. Relyant, however, has had a much stronger relationship with Steinhoff, with Steinhoff extending Relyant’s credit when it was in dire financial straits. As Mr. Schlesinger puts it:

“I think that during the period when Relyant was at its very lower ebb and in serious financial distress, I think Steinhoff probably went beyond the call of duty in terms of providing financial support, in terms of not demanding early payment at times when early payment would have embarrassed Relyant possibly to the point of failure.” ³⁷

72. Although Relyant’s purchases from Steinhoff have decreased during the past 2 years, more so in the last 6 months³⁸, it seems that the Relyant group currently sources some 42.5% of its merchandise from Steinhoff.³⁹

73. In order for furniture retail suppliers to remain competitive two factors are taken into account, pricing and product differentiation. The parties and also the Checkers/Shoprite furniture chain identify a lack of differentiation in the product offerings of the competing chains as a major problem and one that is rooted in Steinhoff’s dominant position in the supplier market.⁴⁰ Differentiation is achieved mainly through importing product or purchasing from independents or, for example through obtaining exclusive fabric ranges. Lewis, for example, is known by the industry as a large importer, it imports approximately 30%, and Geen and Richards and Glicks approximately 20%.

74. We are accordingly concerned at the impact of the transaction on the upstream furniture manufacturing market. As in the JD/Profurn transaction, our concern, in the context of the present transaction, is that were this merger to further consolidate the power of Steinhoff, or, conversely, weaken the position of the independents, it would raise already high barriers to entry into the furniture retail maket. Here we have a significant Steinhoff customer merging with a retailer that is significantly more supportive of the need to maintain an independent supply base. Would this tend to strengthen or weaken Steinhoff’s position?

75. Our view is that the merger is, on balance, not likely to strengthen Steinhoff’s position. The merged entity will, in effect, be managed by Ellerines and will, we expect, be guided by the policies that are core to that group, including its procurement policies.

76. Mr Squires responded as follows to the Tribunal’s questions regarding the merged entity’s procurement policies:

“I am a trader and I think we alluded to this on Monday when I was also under oath to give you a couple of givens and I think Ellerines history is a given and Ellerines history of 80/20 syndrome not more than 20% of our eggs in any basket has been a discipline that I have worked with for 25 years. It is a discipline if I carried on, as MD for 7 years and it is a discipline that I will continue, whilst I am chief executive officer of Ellerine Holdings. … Obviously, the Relyant methodology is going to take on the Ellerines methodology.” ⁴¹

77. Mr. Karp of the Shoprite/Checkers group also testified that as long as Ellerines controlled the merged entity, independent suppliers had nothing to fear.⁴² The Steinhoff board minutes which we scrutinised as well as the testimony of the witness from Steinhoff supports the view that, on balance and in marked contrast with the JD/Profurn merger, this is not a transaction which will favour Steinhoff’s narrow interest.

78. We had considered imposing a condition on our approval which would have obliged the merged entity to place a ceiling on its purchases from Steinhoff. However, we have found Mr. Squires to be a consistently reliable witness and this, combined with the actual policies and practices of Ellerines, have persuaded us that such a condition will not be necessary.

79. We note too that the dynamics of the furniture market have shifted since the JD/Profurn transaction. In particular the strengthened local currency combined with the retailers’ desire to promote differentiation has underpinned the increasing importation of furniture. Indeed this may, at this stage, constitute the greater threat to local independent manufacturers.

80. We accordingly find that neither the horizontal nor the vertical dimensions of this transaction are likely to lead to a substantial lessening of competition. There are no negative implications for the public interest. The merger is accordingly approved without conditions.

05 July 2005

D Lewis Date

Concurring: Y Carrim and U Bhoola

For Merging Parties: Arnold Subel instructed by Cliffe Dekker

For Relyant: Willem Pretorius

For the Commission: Maarten Van Hoven and Rudolph Labuschagne

---