COMPETITION TRIBUNAL
REPUBLIC OF SOUTH AFRICA
Case No: 60/LM/Aug02
In the large merger between:
JD Group Limited
and
Profurn Limited
_______________________________________________________________________
_
Reasons for decision
_______________________________________________________________________
_
CONDITIONAL APPROVAL
1. The merger between the JD Group Limited (“JD”) and Profurn Limited
(“Profurn”) was approved subject to conditions in an order issued on the
12 December 2002. The reasons for this decision follow.
THE TRANSACTION
2. FirstRandBank is currently the controlling shareholder of Profurn with a
78,8% shareholding. 1 It has entered into an agreement with the Profurn
Group in terms of which the JD Group or one of its nominees will acquire
from Profurn all the subsidiaries of Profurn, or all the businesses of the
subsidiaries as going concerns including their assets and liabilities, or a
combination of the subsidiaries, assets and liabilities. At the time of the
hearing the parties had not elected which of these mechanisms they
1 FirstRandBank acquired a shareholding of this significance pursuant to a rights offer in 2002 in
which they assumed control over Profurn. This prior transaction was notified to the Tribunal and
approved unconditionally in decision FirstRandBank/Profurn Limited case no.32/LM/May02.
1
would follow but since the exact nature of the structuring is neutral on the
merger’s effects in terms of the Act, it is not relevant for us to require them
to make an election.
3. What is clear is that in return for the sale FirstRand Bank would acquire
shares in the JD Group as consideration and that in turn it would on sell
5/6ths of these shares to Daun and Cie a foreign investor.
4. Subsequent to the transaction FirstRand and Daun & Cie would hold
approximately 25% of the shares in the JD Group.
5. Presently Old Mutual SA is the largest shareholder in the JD Group with a
27,26% stake.
6. As part of the transaction the JD Group has entered into an interim
agreement with FirstRand and Profurn in terms of which the JD Group will
provide what it has described as “selected retail expertise and services to
Profurn. This agreement, which was approved of by the Commission
came into effect on June 28 2002 and was to last until the outcome of the
adjudication of this matter before the competition authorities.
7. On the 8 October 2002 the Commission recommended that this merger be
unconditionally approved.
The Parties
8. The primary acquiring firm, is JDG Trading (Pty) Limited, a wholly owned
subsidiary of the JD Group Limited.
9. The JD Group (whom we shall refer to hereafter as JD) boasts a
phenomenal history of growth and expansion and has earned a reputation
as the industry’s leader in technology and business management systems.
10. David Sussman, the present chairman, founded the group in 1983 when
he started the business with two Price ‘n Pride stores. Sussman’s
company was to grow spectacularly from its modest origins through
masterly timed acquisitions of troubled firms. In 1986 he acquired the then
ailing, but larger Joshua Doore chain from the Russels group. In 1988 he
ailing, but larger Joshua Doore chain from the Russels group. In 1988 he
acquired Bradlows and World from W&A and in 1993 he acquired the
Rusfurn Group.
11. The acquisitions were all successful and the group acquired a reputation
as a successful acquirer and integrator of rival businesses. According to
2
one analyst, furniture retailers were rerated after the Rusfurn acquistion
and, “in general, they generated superior earnings growth.” 2
12. The group’s rapid acquisition and successful integration of Joshua Doore,
Score Furnishers, World Furnishers and the Rusfurn group saw it become
the leading furniture retailer in South Africa. 3 In 2000 an attempt to
acquire its most established competitor, Ellerines, was however prohibited
by the Tribunal. 4
13. JD is structured as a holding company listed on the J.S.E. It owns the
following chains of stores that are located in South Africa: 5
i) Joshua Doore, Joshua Doore is the largest furniture, appliances and
home entertainment products discounter in South Africa and Namibia.
ii) Bradlows, sells branded appliances, home entertainment products and
furniture targeted at the LSM 68 group.
iii) Russells, sells furniture, branded appliances and home entertainment
products targeted at the LSM 46 group.
iv) Price ‘n Pride, sells furniture and appliances targeted at the targeted at
the LSM 35 group. It has recently positioned itself upwards and now
claims to target the LSM 46 group.
v) Electric Express, acquired in 1993, is primarily a discounter of
household electrical appliances and home entertainment products.
14. The primary target firm is Profurn Limited. The majority shareholder is
FirstRand Bank Limited, a wholly owned subsidiary of FirstRand Bank
Holdings Limited.
15. The Profurn Group emerged out of the liquidation of Supreme Holdings in
1992. In 1997 the group acquired Freedom Furniture a Cape based chain.
In 1998 it acquired the Morkels chain and in 1999, the appliance retailer Hi
Fi Corporation.
16. Between 1994 and 1999 Profurn grew its earnings by not less than 30 %
2 See UBS Warburg Report, Record page 984.
3 JD makes this claim itself. See its Annual report 2001
3 JD makes this claim itself. See its Annual report 2001
4 The Tribunal prohibited the merger between the JD Group Limited and Ellerine Holdings Limited in
August 2000. See case no. 78/LM/Jul00. We refer to this decision from now on as the Ellerines
decision to avoid confusion with the present matter.
5 JD owns businesses that operate overseas but we are not concerned with them here as they do not effect
the South African market.
3
per annum. This earnings growth was the product of an aggressive
expansion strategy that fueled an increasingly competitive market and
correspondingly a decline in margins, as well as consolidation in the
sector resulting from JD’ s acquisition of Rusfurn. 6 Because its margins
were so squeezed the ambitious group was unable to fund growth
internally and so increasingly relied on bank loans. This was to prove its
downfall in the next decade when market conditions soured for retailers
serving the lower end. In March 2002 the group’s chief executive Gavin
Walker, said to be at the helm of the growth strategy, resigned. A
committee comprising of the chairman, the chief financial officer, the chief
operating officer and Mr Theunie Lategan, CEO of FNB Corporate, has
since filled the embattled group’s CEO role.
17. In June 2002 the group, with the support of its principal banker, FirstRand,
undertook a rights offer in an attempt to reduce interest bearingdebt. The
poor response to the rights offer resulted in FirstRand, as underwriter of
the shares, becoming the controlling shareholder with 78,8% of the equity.
18. Profurn like JD is a holding company listed on the J.S.E. and owns the
following chains of stores that are located in South Africa and elsewhere:
i) Protea Furnishers, sells furniture, home audiovisual and household
appliances, targeted at LSM 35 although like Price ‘n Pride it has
ambitions of moving upwards to the LSM 46 market.
ii) Barnetts is part of the Protea division, focusing on the LSM 35 income
groups.
iii) Freedom, is also part of the Protea division and its target market
incorporates the LSM 46 income groups.
iv) Morkels, sells furniture and branded electrical products targeted at the
LSM 58.
v) Home & Electric, an urbanbased chain of specialist discount retailers,
LSM 58.
v) Home & Electric, an urbanbased chain of specialist discount retailers,
offering credit and focusing on key electrical, bedding and builtin
furniture products for the home.
vi) HIFI Corporation, is the largest audiovisual retailer in South Africa,
and offers a wide range of imported audiovisual equipment at discount
prices.
vii) Supreme Furnishers trades in 11 African countries outside the borders
of South Africa under the brand names Supreme, Barnetts, Protea,
and HiFi & Electric City. The first three chains are credit outlets and
the latter is a cash discounter dealing mainly in electronics.
6 UBS Warburg Report.
4
The Intervenors
19. A group of independent furniture manufacturers applied, in terms of
Section 53(1)(c) of the Act, read with Rule 46(1) of the Tribunal Rules to
participate in these merger proceedings. Although the merging parties
expressed some reservations, they did not oppose the manufacturers’
application, thus leave to participate was granted. 7
20. Prior to this these intervenors filed submissions detailing their opposition
to the merger, with the Commission. Their fundamental opposition to the
merger is founded on the ground of the vertical effects of the merger.
The Rationale
21. The fundamental rationale for the transaction is said to be the imminent
failure of Profurn. This in turn is a consequence of the general disarray
prevalent in the furniture retail industry, as well as the particular expansion
and credit granting strategies adopted by Profurn.
22. The parties aver that if not for this transaction, Profurn will exit the market,
approximately 7 000 jobs will be lost and a large number of suppliers and
services providers will in turn suffer financial losses that may eventually
see them exit the market as well.
The hearing
23. A prehearing was held on the 18 October 2002 and hearing was held on
the following days:
25 November 2002
29 November 2002
03 December 2002
04 December 2002
24. The Commission did not call any witnesses.
25. The merging parties called three witnesses:
7 Although papers were filed collectively on behalf of the group, the following individual manufacturers
were allowed to participate in the proceedings: Contour Bedding & Lounge CC; General Manufacturing
Company; Confurn Sales (Pty) Ltd t/a Imperial Furniture; Khwaja Lounge; Kiran Sales (Pty) Ltd t/a Lylax
Bedding; Oaktree Products (Pty) Ltd t/a Pilot Furniture; Serta Bedding (Pty) Ltd; Taurus FBW Furniture
(Pty) Ltd; Sheppe Five CC t/a Firwood Furniture; Friedlein Company (Pty) Ltd t/a Friedlein Furniture.
5
Mr David Sussman, executive chairman JD Group
Mr Derrick Minnie, nonexecutive chair of Profurn
Mr Mias Strauss, MD JD Group
26. The intervenors called six witnesses:
Mr Roy Pritchard, Oaktree
Mr Eshu Seevnarayan, Lylax Bedding
Mr Khan, Khwaja Lounge
Mr A Seevnarayan, Imperial Furniture
Mr Sewpersad, General Manufacturing
Mr H Salligram, FBW Taurus
27. The Tribunal called three witnesses:
Mr Kevin Boyers, CFO Profurn
Mr Peter Squires, CEO Ellerines
Mr Theunie Lategan, CEO FNB Corporate
Background
28. The background to this merger is the dramatic change in fortunes for the
furniture retail groups since 1999/2000. There are a number of reasons for
this, but they are most succinctly stated in the Profurn Group’s 2000
Annual report, where ironically, its erstwhile executive chairman, Gavin
Walker, was to show great prescience about the state of the market, little
knowing that his firm was soon to become one of the more spectacular
victims:
“The retail industry in South Africa in 2002 has undoubtedly
experienced one of its toughest years ever. A spate of profit
warnings, SARS investigations, consumer spending on cell phones,
lottery tickets and gambling, and petrol price increases have all
taken their toll on the sector”
“ This sentiment not only affects the capital markets but also the ability of retail
companies to fund their businesses with debt. The major South African banks
are at their prudential limits for retail exposure.”
29. Walker’s pessimistic assessment is written even before the general crisis
experienced in the microfinancing sector, which comes about in 2001/2
through the collapse of Saambou and the adverse impact of the troubles
at Unifer, the microloan business acquired by ABSA. These setbacks
6
further served to alarm banks about their over exposure to retailers who
operate as extenders of credit to consumers in the lower end of the
market. It is not surprising that they became the catalyst for changes that
were to occur in this market and that ultimately led to this merger.
30. In 2002 both Profurn and the Relyant Groups were compelled to re
capitalise through rights issues, while the JD Group had to reconsider its
presence in the lower segment of the market.
31. The prospects remain gloomy as a recent JD Report suggests: 8
“During the period under review, market conditions in the durable credit
retail market remained under pressure and the overtraded environment
continued to impact negatively on product margin. Huge demands on the
disposable income of the Group’s target market have further exacerbated
the situation.”
32. While this is not the first time the Tribunal has had to consider a
merger in this industry 9, it is patently obvious that the prevailing
industry conditions are vastly different from what they were in 2000.
In fact in August 2000 Profurn was depicted as a vigorous competitor
of both JD and Ellerines, two years later, it is presented to us as the
industry’s ugly duckling. So much so we are told that all attempts to
secure a buyer, other than JD, for the entire group, and not just its
more attractive assets, have been in vain. 10
Competitive Evaluation
33. We will be considering the merger in two respects
a) The horizontal effects since it involves the merger of two competitors; and
b) The vertical effects insofar as they effect the merged firm’s buying power
in the market for the purchase of furniture from manufacturers
Horizontal effects
8 See the JD Group’s Reviewed Interim results for the six months ended 28 February 2002, Record
Annexure 5 page 137.
Annexure 5 page 137.
9See of course the Ellerine decision, the Massmart acquisition of Furnex (Sociedad Investments
Pty Ltd /Furnex) case no 09/LM/Feb02 and FNB’s acquistion of Profurn case no 32/LM/May02.
10 For details of alternative transactions, see the Commission’s recommendation p 910.
7
34. As we have seen from our earlier description, the two merging firms are
both holding companies that own subsidiary businesses, which compete in
the retail market selling furniture and appliances.
35. Their businesses range from dedicated appliance discounters to furniture
chains that sell a variety of goods and are aimed at distinct niches in the
market.
36. There is no dispute that the merging parties subsidiaries compete with one
another. There is, however, some dispute as to the boundaries of the
market in which they compete.
37. To understand the differences in approach to defining the market in this
case it is first necessary to understand how the retail furniture and
appliance sector is structured and then to locate relevant markets for the
purpose of competition analysis.
38. Retailers who sell furniture and appliances have, not surprisingly, found
many routes to the market.
39. That all compete are in the business of selling furniture is not in doubt.
What is in dispute is whether consumers consider these different routes to
the market as equivalent offerings, and hence as substitutes for one
another.
40. Outlets differ from one another not only in terms of the range of products
they sell and their level of prices, but which customers they target,
whether they offer credit and where they are located, to name but a few
distinctions.
41. Like JD and Profurn, certain retailers have organized themselves in chains
not only to afford themselves a wider geographic exposure, but also so
that they encompass, within a single corporate structure, a range of routes
to market.
42. One type of store is referred to in the industry as the ‘mass discounter’.
Mass discounters are stores whose competitive advantage is on price.
They trade in high volumes and can be price competitive because of low
They trade in high volumes and can be price competitive because of low
margins. In order to operate in this type of market a mass discounter must
be able to turn over a high volume of product in a short time, have a
regular cash flow and offer consumers a wide choice of product. For this
reason mass discounters operate from large stores usually located in
urban areas where customer volume densities are more assured, sell on a
8
cash only basis and offer minimum service.
43. The sale of appliances as opposed to furniture is the dominant product
offering of the mass discounter.
44. The next type of store is the furniture retail outlet. This type of store is, in
the hands of the chains, organized like a consumer product according to a
carefully marketed brand aimed at a targeted group of consumers. Within
the JD and Profurn groups we see this as well, as we observed above in
the description of the parties. These retail outlets are usually targeted at a
range of consumer Living Standard Measure (LSM) categories, typically a
brand will target between 3 to 4 LSM bands. 11 The chains’ carefully
research consumer behaviour to ensure that the particular brand remains
aligned to its targeted LSM groupings. This entails different product
offerings, price ranges, store location and medium of advertising. These
stores not only sell furniture, but also appliances and cell phone services.
45. These stores are smaller in size than the mass discounters but are for that
reason more ubiquitous or, in the language of retailers, have a wider
footprint. The chains refer in their reports to their roll out of stores in a
particular chain. What they mean is that they are continually on the
lookout for new locations for their stores whilst at the same time
reassessing the location of existing branches. Like skilful chess players
the chains’ managements are continually moving their pieces to outflank
their opposition or to avoid being removed from the board. Hence this type
of store is less rooted to any one location than the mass discounter.
46. The management challenge is to match the brand to its targeted market in
product, price and location.
47. Within this universe of furniture chains at least three broad segments exist
a mass market, middle market and mid/ upper market.
a mass market, middle market and mid/ upper market.
48. What distinguishes the furniture chains inter alia from the mass
discounters is that the former sell to the consumer on credit, a fact
particularly important for consumers who are ineligible for credit cards. As
they have with targeting their market in terms of product choices, so the
chains have developed the granting of credit into an art. They spend vast
resources on creating models of likely risk and use sophisticated
information technology to assess and reassess their credit policies.
11 Living Standard Measures or LSM’s refer to a method of segmenting consumers into profiles so that
marketers can accurately identify their target markets. Retailers use this information to form a picture
of their target customers and so to provide for them accordingly.
9
49. The JD Group has indicated that it has undergone significant changes to
its credit granting policies as a result of the decline in the lower end of the
market.
They are able to do so through sophisticated IT systems that enable them
to develop and continually update sophisticated risk profiles.
50. In a recent report this is described as follows:
“This process has been successful in predicting the payment behaviour of debtors
with a particular status in a specific geographical area. This has resulted in a
reduction of risk allowing for greater levels of predictability at store level.” 12
51. The final type of player in the market is the ‘independent’ store, so called,
because they are not part of any chain or group and are usually managed
by their owners. The independents are typically single outlet businesses
with the largest type of business in this category not exceeding four
outlets.13 Independent stores vary from being specialist outlets e.g.
retailers of expensive high tech equipment to general dealers who sell a
bit of everything.
52. As mentioned above we have previously had to deal with a merger in this
industry in the Ellerines case. In that decision we adopted the following
market definition:
“In summary then we conclude that the relevant market is composed of
furniture shops (with a product mix of furniture and appliances) directed at
credit sales to consumers in the LSM 3 –5 category)”
53. The merging parties have urged us not to adopt the same market
definition in this merger and have argued instead for a relevant market
that is composed of all furniture and appliance retailers aggregated
together into a national geographic market.
54. Using the Ellerines decision as their point of departure, they go on to
consider the possible ways of segmenting this market. These according to
them might be based upon:
a) Price and Quality differences
them might be based upon:
a) Price and Quality differences
b) Furniture store – cluster or product cluster
12 See JD Group’s Reviewed Interim results for the six months ended 28 February 2002,
record Annexure 5 page 138.
13 See Parties Competitiveness Report, record page 536 paragraph 11.4.17.5.
10
c) Credit sales
d) LSM customer grouping
55. They conclude after an examination of each of these possibilities that
there is no valid basis for finding narrower markets by applying any of
these criteria.
56. As to the geographic market they make the contention, whose logic is not
easy to follow, that:
“The geographic market is local, includes independents and can, for reason of
practicability, be aggregated into a national geographic market.” 14
57. The Commission has followed this approach to the market definition
basing its analysis largely on the parties’ submissions. For this reason and
given that the Commission has not opposed the merger, we shall confine
our analysis to the parties submissions.
58. The parties have, as is their want in mergers of this kind, sought to adopt
the most expansive view of the market. The reason for this is that they
feared that if we were to follow approach in the Ellerines case,
concentration levels would again, as in that case, exceed thresholds for
concern. On the other hand, if the universe of the market is expanded by
the addition of the mass discounters and the independents (both were
excluded from the market on the Ellerines definition) and not segmented
by LSM bands, this would have the effect of diluting concentration levels.
59. The parties’ supposition about the consequences of the contending
market definitions for concentration levels turns out to be correct when we
examine these below. If the Ellerines definition is adhered to, then the
merger would result in an already highly concentrated market becoming
more so – if the parties’ definition is followed the market is moderately
concentrated post merger.
60. It seems to us that the issues that need to be decided upon in order to
delineate a relevant market in this merger are:
delineate a relevant market in this merger are:
a) whether mass discounters are part of the same market as furniture
retailers if they sell some overlapping products such as appliances;
b) whether the chains compete in a furniture retail market that includes the
independent furniture retailers; and
14 See Parties’ Competitiveness Report, Record page 478 paragraph 2.7
11
c) whether the furniture retail market is a universal one or a series of markets
comprising bands of stores aimed at particular groups of consumers
approximating LSM classifications.
61. The mass discounters as we observed above are distinguishable from the
furniture chains in that they
(i) are primarily sellers of appliances not furniture;
(ii) primarily sell on a cash basis;
(iii) have larger stores in fewer locations and hence are less
ubiquitous than the chain stores;
(iv) offer a greater breadth and depth of those products they sell;
(v) offer better prices but less frills including reduced service;
(vi) are targeted at LSM groupings that are higher than the range of
many of the chains. Thus according to Perry and Associates the
Massmart chains have found in a survey that 88% of their
customers in the case of Makro fell into LSM 78 and in Dions
and Game, LSM 68.
62. Whilst the fact that they do not offer credit may not be a disincentive to
purchasers of inexpensive items they constitute a bar to more expensive
items to lowincome consumers who have no easy access to credit. This
was an important conclusion that we came to in the Ellerines decision and
despite the parties contentions they have not convinced us otherwise.
63. For instance in the marketing report of Barnetts, one of the Profurn chains
which is targeted at males in the LSM 4 – 6 category it states:
“Credit usage and the availability of various credit options are essential in this
target market.” 15
64. They go on to say :
“All retail furniture groups which are located in the black rural areas, are
considered competitors. Stores like Ellerines, Town Talk and Savells,
Faideals and OK Furniture, all of which offer the customers easier terms
and purchase facilities such as ‘ No deposits and Cash Back Offers”
compete and share the same customers base.”
compete and share the same customers base.”
65. Nevertheless unlike in the Ellerines merger, the merging firms in this case
will also have a strong footprint in the mass discount market due to the
combination of Hi Fi Corporation and Giddys Electrical Express.
15 See Barnetts marketing plan for 2002/3, Record page 1102
12
66. The next element is how to regard the presence of independents. The
parties devoted considerable attention to this issue recalling that in the
Ellerines case we had excluded independents from the relevant market.
Since independents may account for up to 17 % of the market according
to the parties and may consist of up to 10 000 stores, their presence in the
market not only dilutes concentration levels, but also suggests the
existence of firms well situated to expand their market shares if the merger
leads to unilateral or coordinated effects that result in higher prices.
67. To this end the parties produced lengthy spreadsheets of the names of
these firms and their location. Their conclusion was that these stores were
plentiful and ubiquitous. The parties also argued that the independents
have a competitive advantage in that they have lower overheads (no mass
advertising, marketing and head office expenses) and to the extent that
they have size disadvantages in extending credit to customers and getting
rebates from furniture manufacturers, this has been cured for some
independents who can become members of buying groups like Furnex
and The Shaw Group and can get credit for their customers through credit
groups such as Consumer Credit. 16
68. The parties acknowledge that both merging firms set prices for their
chains nationally. This is because it is impractical for groups with such an
extensive network of stores and which engage in national advertising
campaigns to leave pricing to be determined at a local level. This of
course is perfectly rational. What is harder to accept is that groups which
set prices nationally can then be said to compete on price at a local level
with independents when their local managers lack the authority to respond
with independents when their local managers lack the authority to respond
to pricing of their immediate competitors. Rather what seems to happen is
that the chains determine the success of a local operation on an ongoing
basis against a nationally determined return. If a store is under performing
the response is to either close the store, relocate it or to rebrand it so that
it is targeted at a different LSM group, as the minutes of both JD and
Profurn show. The annual statements contain statistics of all three
occurring not infrequently. It is this type of move that is the response to
competition in the local market not price competition.
69. On the other hand the documents that we have suggest that the chains do
respond to the pricing moves of rival chains, which they appear to monitor.
Independents may be perceived by local managers as competitive, but
16 The parties adduced evidence that a random sample of consumers who had had credit
applications rejected by the JD system qualified for credit according to the credit check of
Consumer Credit, who provide credit verification for some independent stores.
13
they do not operate to constrain price increases by groups which set
prices in accordance with national not local dictates. The firms that will
compete on price with chains are other chains not the independents.
70. Another example of the fact that chains to respond to one another
rather than the behaviour of mass discounters or independents was
the “industry’s free gift giveaway madness”, alluded to by Mr
Squires of Ellerines in his testimony, referring to the aggressive
campaigns virtually every furniture chain store embarked on, hot on
the heels of rival stores. 17 Other than the consecutive sequence in
which each chain launched its giveaway campaign, the similarities
in the marketing and in the actual giveaway products, are clear
evidence that these campaigns were directed at customers of the
opposition furniture stores, and not at those of either the
discounters or the independent furniture retailers.
71. The parties were not able to produce any contemporaneous evidence of
the manner in which they compete with the independents. All the evidence
was produced specifically for the hearing and when challenged by the
intervenors was open to some doubt.
72. Roy Pritchard from Oak Tree testified that he had attempted to verify a
sample of this evidence and he found that certain stores listed were no
longer in business whilst the extent of outlets owned by other stores was
overstated. Using the sample and the error data he extrapolated this on a
national basis to come to his own estimate of the number of independent
stores. He also attempted to get from the independent buying groups their
best guess of the number of independents and came up with a number
that accorded with his own. Thus on Pritchard’s evidence the size of the
independent retail market is vastly reduced. In his estimation there are a
independent retail market is vastly reduced. In his estimation there are a
maximum of 400 “real” independent furniture stores. 18
73. The parties’ riposte was to send a photographer around to Pritchard’s
sample area, who purportedly found far more independent stores than
Pritchard. Pritchard did not have time to verify this new evidence, but a
perusal of some of the photos shows that judging by their names and
appearance, some stores claims to being classified as furniture shops was
open to serious doubt. 19
17 See transcript of hearings, 29 November 2002, p1518.
18 As stated in Mr Pritchard’s witness statement page 5, paragraph 9.3.
19 For instance some appeared to sell only music. Stores listed included “Audio Den”; “Franks
Radio”, and “Selglo Radio and TV”.
14
74. Perhaps the most interesting error on behalf of the parties, that Pritchard
identifies, is the fact that a firm called Ster Discount Furnishers which the
parties initially identified as having 36 stores in fact has only 4 stores of
which only 1 is specifically a furniture store. Had it had the 36 stores it
would have been the size of a small chain and dwarfed the other
independents. The error was acknowledged by JD who attributed the error
to one of their agents responsible for compiling the information. We have
no reason not to accept their bona fides on this. Nevertheless it illustrates
the point that we make above – had the independents been real
competitors of JD they would have known all about them. In the Ellerines
merger the Commission had identified exactly the same weaknesses in JD
lists’ of would be competitors. 20 These errors are repeated two years
later. Their recurrence proves that the only time independents appear on
the JD radar screen is when they are resuscitated as competitors to
massage market shares in an otherwise concentrated market. Every other
aspect of JD’s approach to this case has been meticulous and considered;
furthermore its executives demonstrate an encyclopedic knowledge of
their market – this anomaly, the apparent deficiency in their knowledge of
the independents, can only be accounted for because they do not, in the
minds of JD’s management, constitute a serious competitive threat.
75. In our view independents do not act to constrain the market behaviour of
the chains and their plurality, whether one accepts Pritchard’s estimates or
JD’s, should be excluded from any relevant market determination.
76. The last issue concerns the legitimacy of segmenting stores into LSM
bands, the approach we had taken in the Ellerines decision.
bands, the approach we had taken in the Ellerines decision.
77. The parties argue that surveys of customer profiles conducted by the
chains, suggest that even though a store may be targeted at a particular
segment of consumers, persons outside of the targeted LSM profile still do
business at the store. This they use as the basis to argue that there is no
such thing as a segmented market defined in terms of LSM groups, or
some other function of income or social status, but one universal market.
We certainly agree that the fact that a chain of stores is branded at a
particular LSM segment does not mean that it becomes hermetically
sealed from the ingress of the disavowed class, like the separate
20 See JD/Ellerines decision page 28 where the discrepancy between the figures presented by the parties
and those verified by the Commission was noted: “ The evidence gathered by the Commission regarding
the former grouping of independents is striking. The parties informed the Commission investigators that
there were 1251 independents in the 99 local markets in which both parties compete. A survey conducted
by the Commission of 202 of these independents in the Eastern Cape, Northern Province and the Free State
established that 93 were no longer in business, 12 were not in the relevant product market and four declined
to respond.”
15
entrances of an apartheid style segregated institution. The fact remains
that the marketing materials and the strategy of the stores suggest that
they remain focused on their core group of consumers who form the rump
of their business and in so doing they create a differentiated shopping
experience not wholly substitutable by one aimed at a different group of
consumers.
78. A selection of quotes from the marketing plans of certain of the chains
implicated in the merger illustrates this:
In the JD stable:
79. Price ‘n Pride, its chain that was previously targeted at the LSM 35 group
and now intends moving up market to focus on the LSM 46, describes
this move in its internal documents as positioning itself at “the top end of
the traditional black mass market but below that of the Universal
Chains”21
80. Other extracts indicate that the chain will embark on a new merchandise
range, which reflects the new market positioning. 22 They also state that
the advertising has been changed to reflect the new merchandise. 23
81. In relation to Bradlows the 2002/3 Business Plan states “Bradlows
remains the leading furniture retailer in the middle /upper end of the mass
market. Given the turmoil around the competition Bradlows is well
positioned to cement its position as the Number 1 furniture retailer in its
market segment. The customer profile is 85% black predominantly in the
LSM 68 Group” 24
21 See Price ‘n Pride launching of a new business, Record page 1060.
22 Record page 1066.
23 Record page 1065.where they state “ This process has been substantially successful with a clear change
in customer base moving from old LSM 35 to LSM 46.”
24 Record page 1090.
16
82. In relation to Russels it is stated that the longterm goal is to cater to LSM
57 [Russels present position is said to be 46]
83. In the Profurn stable we see the same approach in its internal plans.
83. In the Protea furnishers marketing plan for 2002/2003 Protea
targets LSM 3 to lower end LSM 5. Protea introduced a status store
concept in 2002 aiming at upper 5 to LSM 6 market. 25
84. Similar sentiments are expressed in the Barnetts material about its focus
on the LSM 46 category, which we quoted earlier. 26
85. In the Profurn Annual report Morkels is described as targeted at LSM 57
customers.
86. The approach in all these internal documents, which were prepared for the
purposes of preparing strategy not merger notification, is consistent in its
emphasis of the distinctiveness of each respective chain’s bands of
targeted customers.
87. As we observed in our decision in the Distell merger where we adopted a
similar approach to vertically segmenting a retail market:
“...the vast price differential within each of the various categories” (of spirits),
provides for market segmentation, so that a “multimillion rand Rolls Royce could
not sensibly be placed in the same market as a humble Opel Corsa.” 27
88. The existence of welldefined submarkets in a broader market is
something that has long been recognised in antitrust law. 28
25 Record page 1099. At page 1100 it says Protea says its main competitors are “ retail
furniture chains operating in the LSM 3 to lower 6 market, black consumer market, these
chains offer credit terms and similar purchase facilities also having similar merchandise.
Main competitors are Barnetts, Ellerines., Town Talk, Fair Deal, Savells, Price and Pride,
Lewis Stores.” “ Protea Status stores competitors are Furncity, Oxford, O.K. Furnishers,
Lewis Stores.” “ Protea Status stores competitors are Furncity, Oxford, O.K. Furnishers,
Russels and Lubners.” They go on to say that Markinor has identified Proteas main
competitors as Ellerines, Lewis and Joshua Doore. “ This will be focused on the best
learned from them, introduced and improved.”
26 Record page page 1102
27 Distillers Corporation SA Limited and Stellenbosch Farmers Winery Group Ltd , case no 08/LM/
Feb02 page 38
28 See Brown Shoe v United States, 370 US 294, 325 (1962)
17
89. In Federal Trade Commission vs Staples Inc. and Office Depot Inc 29
the Court recognised that
“The mere fact that two different classes of retail vendors both sell a
particular type of merchandise does not mean that they are in the same
product market.” 30
90. In support of the above the Court quoted from State of California v
American Stores 31, a case involving supermarkets:
“ …even if convenience stores competitively price a few food items
such as bread and milk, in direct competition with supermarkets, such
is not sufficient to justify inclusion of all retail grocery sales from
whatever outlet in the relevant product market.”
91. And later on the Staples Court follows a similar approach to the
evidence that we have when it concludes:
“The documents reviewed by the court show that the merging parties
evaluate their competition as the other office superstore firms, without
reference to other retailers, mail order firms, or independent stationers.”
92. In our view this case is an appropriate instance for defining a submarket
that constitutes a relevant market – the practice, language of the industry
and documents of the parties all support this conclusion. This does not
constitute gerrymandering by the Tribunal, these are submarkets carved
out by the practices of the industry’s players. 32
93. How then do we approach the issue of concentration in view of this
controversy? Most usefully, the parties have calculated market shares
using the various permutations that have been considered in the prior
analysis– one is a segmentation based on two possible LSM bands,
29 Case no. 1:97CV00701, pages 1820.
30 Page 19
31 State of California v American Stores 697 F Supp (1998) at 1129
30 Page 19
31 State of California v American Stores 697 F Supp (1998) at 1129
32 The phrase “gerrymandering” is borrowed from Gavil et al, who in their discussion of the Staples case
state “ If the loss of headtohead competition among merger partners would permit the firms to
raise price regardless of the response by other sellers in the market, the competitive concern
could be described equivalently as a unilateral competitive effect within a broad market, or as a
merger to near monopoly within a narrow submarket. Framing the case in the latter terms may
make it more appealing, at least if the submarket definition can be simply stated, so as not to
appear to reflect resultdriven gerrymandering. (See Antitrust Law in Perspective: Cases,
concepts and problems in Competition Policy. Andrew Gavil, William Kovacic and Jonathan
Baker, Thompson West, 2002, at page 522.
18
another separates out appliances from furniture and finally the market
shares that emerge based on their ‘total universe’ market.
94. They also then take furniture market shares that are provided in two
analysts reports and show, after some adjustments they deem
appropriate, that the combined shares post merger of the parties will, on
whichever market definition is adopted, be in a range between 25,5% to
27,5%.
95. Yet we need to approach these figures with caution. Unfortunately the
Commission in not opposing the merger did not go out to test this
information or find new sources. The parties’ working of the figures is
based on assumptions, estimates and calculated guesses. Thus figures
for Profurn are based on reworking done to provide for envisaged store
closures whilst OK/ Hyperama is considered still to have 70% of its
turnover in the LSM 35 range, based on the assumption that because this
was the approach adopted in 2000 in the Ellerines merger that assumption
is still valid today. Thus the parties’ figures show flexibility when adjusting
the merging firms shares down to meet current conditions, but their rivals’
shares receive no current reappraisal and are assessed on historical
assumptions.
96. A better proxy for market share might be using the number of stores in the
relevant 35 segment since store numbers are current and in the public
terrain. On the assumption that each store is of an equivalent size and has
a comparable turnover then market shares based on number of stores in
the LSM 35 category would look like this:
Participant No. of stores
in LSM 35
% of market
JD Group 123 8.5
Profurn 226 15.7
Relyant 140 9.7
Ellerines 399 27.7
Lewis 420 29.1
Other 134 9.3
TOTAL 1442 100
Lewis 420 29.1
Other 134 9.3
TOTAL 1442 100
97. On this basis the merged firms have 24,2% of the post merger market.
The HHI is 2113 premerger and 2380 post merger, yielding a change in
HHI of 267, suggestive of a concentrated market. 33
33 The HerfindahlHirschman Index ("HHI") of market concentration is calculated by
19
98. Approached from a different perspective the UBS Warburg analyst
suggests that Profurn, JD, Relyant and Ellerines collectively account for
87 % of sales to consumers in the LSM 3 – 5 market. 34
99. On the parties’ turnover figures, the HHI’s lead to a highly concentrated
market for the LSM 35 market, that excludes independents and mass
discounters, but is not concentrated using their universal definition i.e a
market spanning LSM 38 stores, independents and mass discounters. 35
100. However for reasons that we advanced above the universal market
definition is based on three fallacious assumptions and cannot be correct.
Even if the Ellerines definition is too narrow, it yields such high
concentrations that even if slightly diluted by stretching LSM boundaries
by another band, or allowing for some hypothetical independents, one
would still have a concentrated a market post merger.
101. It is not necessary for us to definitively conclude whether this will result in
a substantial lessening of competition because, as we explain below we
have concluded that it is probable that this concentration would have
occurred even absent the merger.
102. The conclusions about concentration are however indirectly pertinent to
the vertical effects of the merger which we deal with below.
103. Because we make no adverse finding against the parties in respect
summing the squares of the individual market shares of all the participants. The spectrum
of market concentration as measured by the HHI is divided into three regions that can be
broadly characterized as unconcentrated (HHI below 1000), moderately concentrated
(HHI between 1000 and 1800), and highly concentrated (HHI above 1800). A merger that
results in an increase in the HHI of more than 100 points is likely to raise competition
concerns.
concerns.
34 See UBS Warburg report, Record page 949. The CR 4 model for concentration levels suggests that a
market can be regarded as concentrated if the top four firms enjoy more than 75 % of the market and the
merged firm will supply at least 15% of the relevant market.
35 The market shares in the LSM 35 excluding independents are given as: Ellerines 26.1%; Relyant
15.67%; JD Group 9.35%; Profurn 12.35%; OK/Hyperama 25.6% and Lewis 10.36%. On this basis the
HHI is calculated at 2189 premerger and postmerger the HHI is 2446. On our calculation the HHI pre
merger is 1959 and post merger 2190. In the LSM 38 market (excluding discount stores) the premerger
HHI is calculated as 1093 and post merger 1356, including discount stores the pre merger HHI is 876 and
post merger the HHI is 1027. These market shares were adjusted by the parties to account for the planned
store closures and in LSM 35 the pre merger HHI is then calculated to be 2073 and post merger 2218. In
LSM 38 the pre merger HHI is 1066 and 1232 post merger (excluding discount stores). Including discount
stores in LSM 38 the pre merger HHI is 861 and post merger the HHI is 956.
20
of the horizontal aspects of the merger, it is unnecessary for us to
come to some definitive conclusion on the boundaries of the
relevant market. Whether we follow the definition we applied in the
Ellerines case or, the universal market favoured by the Commission
and the parties, or some permutation in between the two, the choice
of classification in the present case will not result in different legal
consequences. We would observe in passing that despite the
diligent efforts of the parties to the contrary we are not persuaded on
the present facts that the market definition adopted in the Ellerines
case does not still seem the best template to apply.
104. Causal link between the concentration and the merger.
105. Section 12A requires there to be a causal relationship between the merger
and the substantial lessening of competition in the implicated market. This
is because of the clear language which states that:
“…the Tribunal must initially determine whether or not the merger is likely to
substantially prevent or lessen competition …”
106. If the loss of competition would have occurred in any event due to other
factors, absent the merger, then the merger should not be enjoined
because the remedy would be pointless.
107. This approach has been followed by the European Court of Justice in
France v The Commission :36
“If a concentration is not the cause of the creation or
strengthening of a dominant position which has a significant
impact on the competitive situation on the relevant market, it must
be declared compatible with the Common Market.”
108. The parties have alleged that Profurn was a failing firm and rely on this
factor to sanitise the merger lest it be otherwise be found to be anti
competitive (see our decision in Iscor/ Saldanha where we held that the
failing firm is not a defence, as it is in some other jurisdictions, but a factor
failing firm is not a defence, as it is in some other jurisdictions, but a factor
amongst others to be taken into account in the assessment of a
36 France v EC Commission , cases C68/94 and C90/95 [1998] ECR I 1375, [1998] 4
CMLR 829. See also Whish, R. Competition Law, 4 th edition.
21
merger).37
109. If Profurn was failing, and its market share then went to the remaining
firms in the market, in some permutation leading to concentration levels in
the market and an increase in market share by the merged firm no worse
than they would be in the postmerger scenario if the merger goes ahead,
then at least on one version of the failing firm doctrine the merger should
not be enjoined because of its lack of causative effect. 38
110. Was Profurn a failing firm?
111. Profurn’s fall from grace in the past few years is attributable to two major
factors firstly, the deteriorating business environment faced by firms
extending credit to the mass market, and secondly a lack of strategic
management focus that made the firm particularly susceptible to
misfortune in an adverse environment.
112. We have referred to the change in the business environment in the
Background section and we need not repeat it again here. Profurn was ill
placed to weather the storm because its previous executive chairman
Gavin Walker had pursued an aggressive management strategy to expand
the group as rapidly as possible.
113. It is instructive to note that in 2000 when we considered the Ellerines
merger we quoted from Profurn’s 1999 Annual report where they made
the point that the quality of their debtors book was improving and that this
was also enhancing cash flow. Two years later this sanguine sentiment
has proved ill founded. The retail industries’ swan had now become its
ugly duckling.
114. Profurn’s cash flow had not improved to fund its growth largely because it
was a credit not a cash business. The roll out was financed by a steadily
rising mountain of debt extended by a consortium of banks principally,
First National Bank. The UBS Warburg report suggests that Profurn was
First National Bank. The UBS Warburg report suggests that Profurn was
growing at rates that exceeded the capacity of its balance sheet and at the
same time it was experiencing gross margin pressure. In five of the eight
years that they analysed they had negative to negligible operating free
cash flows. 39
37 ISCOR Limited / Saldanha Steel (Pty) Ltd case no. 67/LM/Dec01
38 This is the approach taken in the European Union. See the discussion in the I scor decision
supra.
39 See UBS Warburg report page 965 of the record (page 25 of actual report).
22
115. In the course of 2001, FNB alarmed at the decline, began to canvass
other firms that might be interested in purchasing Profurn. Buyers that
were interested wanted only parts of the business, and not the whole, and
the matter was not pursued.
116. In 2002 FNB decided to recapitalise the business by way of a rights offer
that led to FNB assuming control with a stake of 78,8%. Once it had
assumed control over the business, FNB resumed its search for a buyer.
This appears to have been known in the market and led, according to
FNB’s Theunie Lategan, to a surprising, unsolicited approach from JD to
purchase the entire business, something no one else had been willing to
do, in exchange for shares in JD.
117. Serendipitously, this overture was followed by an approach from a
prominent entrepreneur, Claus Daun of Daun et Cie who agreed that if the
deal with JD went ahead he would be willing to purchase five sixths of the
shares in JD that FNB acquired from the sale.
118. It is not difficult to see why the deal was so attractive to FNB. Not only was
the ugly duckling to be taken off its hands, but also it received shares in an
expanded more viable group in a less concentrated market. The Daun leg
had the added virtue of reducing its exposure to the furniture retail
industry.40
119. Subsequent to the deal being announced the parties approached the
Commission for permission to allow JD to act in a consultative capacity to
the Profurn board until such time as the merger had been cleared. The Act
does not allow parties to implement a merger prior to approval. The
Commission approved the plan. It appears to have been agreed that the
Commission would be provided all minutes of the Board to enable it to
monitor the decision making processes, presumably to ensure that JD
monitor the decision making processes, presumably to ensure that JD
had not precipitately assumed control. 41
120. When the merger was notified to the Commission the parties relied on the
fact that they considered Profurn to be a failing firm as part of their case.
The supporting information was provided by Lategan, from FNB, who
although at that time a member of the Profurn Board, was a nonexecutive
40 FNB had also been heavily exposed at this time to Relyant, which was at the same time going through
reconstructive surgery.
41 That this undertaking seems to have been more honoured in the breach appears from two entries in the
minutes of the period. One indicates that the minutes need to be written so as to reflect that the Profurn
Board not JD is making the decisions (minutes of meeting held on 5 August 2002 at point 9). The other is
an observation that the minutes have not been sent to the Commission and that this must be rectified (board
minutes 19 September 2002).
23
director. This was later to become a controversial issue within Profurn
Board as Kevin Boyers, the financial director, vehemently denied Profurn
was a failing firm and made his views known at a Board meeting
questioning why the report had been made without input from either the
executive directors of Profurn or the Board. 42
121. A month later Boyers was given an opportunity by the Profurn board to
present his own conclusions as to why Profurn was not a failing firm. He
was not able to influence the board at this subsequent meeting to change
its submission.
122. It was for this reason that we called Boyers as a witness in our hearing.
Boyers testimony was cautious and inconclusive on the issue of Profurn
being a failing firm. This is not surprising given the invidious position in
which he now finds himself as the dissident opinion on a Board wholly
committed to the merger. He still appears to stick to his view that at the
time of the submission (2 August 2002) Profurn was not a failing firm
although at the time of the hearing (25 November 2002) he conceded that
it probably was.
123. One of the difficulties that also makes the assessment difficult, and which
is raised by Boyer’s evidence, is the change in accounting policy by
Profurn at the end of December 2001, when the company changed its
policy for reflecting debtors to an arrears basis and as a result it recorded
an additional provision of R477 million.
124. According to UBS Warburg:
“We believe that the substantial increases in debtors provisions by Profurn and
Relyant, are at least in part attributable to a change in their accounting policies
with regard to arrears.” 43
125. Although JD in its crossexamination sought to establish that Boyers was
influenced by his alleged animosity to the new owners over a
compensation issue and that as a director for obvious legal reasons he
compensation issue and that as a director for obvious legal reasons he
was disinclined to concede the firm was failing in a public filing, we have
no reason not to accept Boyers testimony; indeed it would have been
much easier for him like other members of the Board to accept uncritically
the tide of history and to support all steps to facilitate what appeared an
inevitability – the merger with JD.
42 See Board minutes of Profurn dated 5 August 2002.
43 See UBS Warburg report op cit, Record page 946.
24
126. This rendition of the facts is not intended to record a saga of corporate
infighting. The relevance of the dispute is how much value to attribute to
the claim that Profurn was a failing firm.
127. The reason for our concern is that the dominant influences on Profun
during the periods of this assessment were FNB and, later after the
Commision assented to the interim arrangement, JD and FNB. Both were
committed to the deal and doubtless wary in the light of the Ellerines
decision as to how the transaction would be viewed by the Competition
authorities. The failing firm assertion would importantly create the basis
not only for distinguishing the present notification, from that in Ellerines ,
but would also help to fast track it.
128. An examination of the board of directors during the period for which we
have access to the minutes, demonstrates the extent to which it had
become dysfunctional. The exit of the tarnished wonder boy Gavin Walker
as executive chairman saw no serious subsequent attempt to replace him
with anyone as CEO whilst the two Chairman who followed were all
interim appointments. The non execs, who included among them the
founder of the group, left as well and soon the company during the time of
its restructuring was effectively run by a board composed of FNB
appointees, executive directors of the pre FNB era who were either
disgruntled and disillusioned or time servers waiting for their moment to
jump ship. Little wonder that the FNB execs and the JD people, after the
commencement of the interim agreement, who whilst not on the board
attended the meetings that mattered and although in the side chairs
proclaimed on issues with all the authority of a crown prince in waiting,
proclaimed on issues with all the authority of a crown prince in waiting,
were able to mould Profurn in its last premerger months to suit their
desired end – the sale of Profurn to JD. It is not surprising that Profurn
never explored the alternative purchasers other than a sale to a rival
because with Walker’s exit they never acquired a leadership committed to
rescuing the group.
129. Noteworthy is the fact that on 20 May 2002, prior to the JD offer, the board
minutes of Profurn record that First Rand had agreed to the following:
• “reconstitute the board
• decide what to do with Profurn after they became the majority
shareholder
• appoint in consultation with the board a CEO”
130. Of these three undertakings the first was served largely by the
appointment of its own people to the board, rather than new blood
25
committed to turning the company around, the second was answered by
the sale to JD and the third was never fulfilled.
131. To what extent then is Profurn’s final demise a function of difficult market
conditions coinciding with historically ill considered management choices
or the desire of its final custodians to rid themselves of it and take no
steps short of sale to a rival to revive its fortunes.
132. We cannot answer this question with any certainty as to the Profurn
business as a whole. In all failing firm cases managements are in some
way responsible for their firm’s demise. In this case our concern is not
that, but whether the failing firm factual scenario, is either not credible on
the figures, in view of the dispute with Boyers, or if it was, has Profurn
been conveniently allowed to further wither into decline, unhindered by
any haste in remedial action from its custodian shareholder, FNB, which
showed great activity in selling the firm, but conversely only passivity in
trying to resurrect it.
133. Although we cannot resolve this dispute of fact on the evidence before us
we can say with more certainty that the businesses that Profurn had in the
lower end of the market were in terminal decline and even had Profurn
found another owner it is unlikely to have been a buyer for these
businesses. This is not only corroborated from the testimony of other
would be buyers such as Ellerines, but also the minutes of the board in
2002 and JD’s subsequent stance toward these stores.
134. Since these are the submarkets where we see competition being most
seriously impaired the inevitable exit of these brands means that the
merger is not the agent of their demise.
135. Perhaps the best exposition of this theory of the inevitable consolidation of
the industry comes in a 2002 report prepared by an analyst with UBS
the industry comes in a 2002 report prepared by an analyst with UBS
Warburg appropriately titled “Consolidation is coming” .
136. In the report the analyst observes what he terms the “dissaray in the
industry” and the fact that industry profitability and balance sheets are
under pressure. He notes that ;
“the excessively competitive environment resulting from Profurn’s aggressive
strategy has also eroded margins.” 44
137. Later on at 944 says:
44 Page 941
26
“ We believe that the key driver of this negative trend in cash flows was the
aggressive expansion strategy adopted by Profurn. The effect was to increase the
degree of competition within the industry, which, in our opinion led to
gross margin pressure.”
… “ We believe that the closure of these stores and bringing Morkels and Hi FI
Corporation under JD Groups management should reduce competition within the
industry and benefit all participants.”
138. The report which is written in September 2002, prior to the Commission’s
recommendation, goes on to state that whatever the Commission’s
decision:
“ Protea will be closed and so a reduction in competition in the lower end
of the market is unavoidable.”
139. According to the analyst consolidation would take place anyhow and the
only difference is how quickly it would come about. 45
Conclusion
140. For reasons we set out above we do not find that all the elements of the
failing firm doctrine have been established. Nevertheless we are
persuaded by the evidence that certain businesses of Profurn would have
exited the market if there had either not been a merger or if another firm
with a less competitively adverse profile had been the purchaser.
141. Although the exit of certain of the Profurn chains will lead to an increase in
concentration in the market in those segments of the market in which
those effects might be most pronounced namely the LSM 35 segment,
and where entry is less likely (this is conceded by the parties), the
inevitable exit of some of Profurn’s stores in this end of the market,
suggests this would have occurred in all probability absent the merger.
Consequently that market share would inevitably have gone to the
remaining chains, with Ellerines best placed to be the major beneficiary.
remaining chains, with Ellerines best placed to be the major beneficiary.
The merger is thus not the causative effect of the present state of the
market.
45 UBS Warburg conclude that Ellerine has the most store overlaps with Profurn in the
lower end of the market – for this reason its elimination will mean that they will be the
main beneficiary of the merger
27
142. The merger at worst shifts the calculus in favour of JD, which is smaller
than both Ellerines and Lewis in this segment. In the other segments
implicated by the merger namely the appliance sector and the mid/upper
furniture market, concentration is less pronounced, although entry
because of the higher margins and lower risks is more likely.
143. On this basis we find that the merger itself will not lead a
substantial lessening of competition hence we do not impose any
conditions in relation to the horizontal aspects of the merger.
144. Thus although the merger would increase concentration and reduce
competition in the market this would have occurred absent the merger and
the merger is thus not the causative effect of the present state of the
market.
145. This is perhaps the most distinguishing feature between this merger and
the JD/Ellerine merger, apart from the dramatic changes the industry has
weathered in the interim. 46
Vertical effects
146. Although several witnesses gave testimony on behalf of the intervenors
their evidence was largely consistent and can be approached
collectively.47
147. The core concern of the intervenors, is that the deepseated relationship
between the JD Group and Steinhoff, the countries largest furniture
manufacturer and JD largest supplier, will, in the post merger scenario
result in the independent manufacturers losing the custom of the Profurn
stores to Steinhoff. 148. In turn this customer foreclosure would see the
demise of the independent manufacturers, a concern echoed by the CEO
of Ellerines, Mr Peter Squires.
149. The shareholding of Mr Daun in Steinhoff immediately prior to this
transaction amplified these concerns, since this transaction would
transaction amplified these concerns, since this transaction would
46 We note however, that given Relyants’ precarious financial position the Ellerines
group remains the JD Group’s most effective rival. Analysts speculate about the appetite
of Lewis and OK Furniture for this end of the market. Had the Tribunal permitted its
merger with the JD Group two years ago, the demise of Profurn would still have taken
place and consumers at the low end of the market would be left with even less choice
than they have today.
47 The independent manufacturers claim they represent 46% of the total
furniture manufacturing industry.
28
see him acquire JD shares as well. However, as was highlighted by
the merging parties and widely reported in the media, Mr Daun
disposed of his shares in the Steinhoff Group almost simultaneously
with the notification of this transaction. 48 However, Mr Daun remains
at the date of this decision a director of Steinhoff.
150. The Commission investigated the submissions of the independent
manufacturers, opposing the merger, and found no evidence of the
alleged links between Steinhoff and the JD Group, hence the
conclusion that these allegations amounted in its opinion to no more
than a conspiracy theory. 49
151. This view of the close relationships at least between Steinhoff, Relyant
and JD is not the sole preserve of the independents. In an article for
Money Web, wellknown journalist David Gleason, had the following
observation about the takeover of Relyant by German group Poco:
“There is a little twist to the tale at this point, however. It seems that
Poco’s founder Peter Pohlman, is a close friend of one Bruno
Steinhoff whose company is, apparently, Relyants biggest single
creditor to the tune, I understand, of about R 400 million. So if you’re
cynical( I am), this could be interpreted as a deal structured, in part
at any rate, to protect Steinhoff’s interests. Indeed, and given the link
between Steinhoff and the JD Group, it looks as though Steinhoff is
seeking to establish an impregnable position in the furniture
industry.”50
152. Despite the lack of proof before us of any structural links between JD,
Steinhoff and other key players in the furniture industry, we probed the
vertical relationship between JD and Steinhoff, to examine the likely
impact of the merger on independent manufacturers who have historically
supplied Profurn.
impact of the merger on independent manufacturers who have historically
supplied Profurn.
48 Business Report, 1 October 2002. Mr Daun sold Steinhoff shares to finance an interest in Metz
Manufacturers and BCM Holdings, in the bedding components industry.
49 In their submission to the Commission the Intervenors refer to Steinhoff, Daun and David Sussman of
JD as a syndicate attempting to gain total control over the furniture industry in South Africa.( See
Annexure C to their affidavit in support of their application for intervention, Record page 1022. According
to the Commission’s report, the submissions they received, named Ellerines and Relyant as participants in a
collusive arrangement with JD and Steinhoff, aimed at controlling the furniture retail industry, by virtue of
unproven family relationships and shareholdings.
50 See article entitled “Relyant shareholders can rely on being told to take a hike.” Money Web 10 July
2002.
29
153. It is common cause that Steinhoff is the dominant firm in manufacturing of
furniture and bedding industry. According to the independents Steinhoff
has 54 % of this market. They segment this market into three categories
as follows:
LOUNGE WOODEN BEDDING TOT
AL
Steinhoff &
Associates 55 % 39 % 69 %
5
4 %
Independents
45 % 61 % 31 %
46 %
154. Founded in 1964 by Mr Bruno Steinhoff in Germany, Steinhoff is one
of the five largest furniture manufacuturers in Europe and the largest
in Africa. The group describes itself as manufacturing, warehousing
and distributing a “wide range of household goods (mainly
comprising bedding, case goods and lounge furniture) as well as the
raw materials used primarily in the manufacturing of household
goods.” 51
155. Steinhoff has grown to its preeminent position as a result of a series
of mergers of the past decade, which saw it acquire its major
rivals.52
156. The Steinhoff Annual Report of 2001 contains a list of superlatives
describing its extensive penetration of the local market:
Sealy is described as the “ largest bedding manufacturer in South
Africa” Edblo one of its other bedding divisions has a longterm
contract to manufacture beds under licence from the Relyant
Group.
Gomma Gommma is the “biggest producer of showwood
furniture in South Africa servicing a broad spectrum of lounge
51 This information comes from the website for Steinhoff International Holidings Limited.
52 In 1989 M Jooste and M Delport acquired Gomma, in 1991 Gomma acquired Bakker & Styger
and Wesco take out Delport, in 1994 Victoria Lewis is acquired, in 1995 Daun takes out Wesco,
in 1996 Gomma buys Loungefurn, in 1997 Victoria Lewis, Gomma, Bakker & Styger become
Steinhoff Africa, in 1998 Gomma acquires assets of Beechfurn, in 1998 Cornick acquire Afcol
Steinhoff Africa, in 1998 Gomma acquires assets of Beechfurn, in 1998 Cornick acquire Afcol
and in 1999 Steinhoff Africa merges with Cornick.
30
markets.”
Grafton Everest, which manufactures upholsetered lounge
suites in fabric and leather is described as “the largest lounge
factory in South Africa.”
157. Steinhoff has grown, largely due to its capacity to cater for the production
volumes required nationally by the large retail chains, which have
benefited from a shift in the buying power. Correspondingly the number of
viable independent manufacturers has declined alarmingly.
158. The decline in the independent furniture manufucturing sector and the
consolidation of the larger firms into Steinhoff has meant that there are an
insufficient number of local manufacturers capable of supplying the large
chains with their needs.
159. In fact Mr Squires, CEO of Ellerines testified that despite a policy of
spreading the ratio of purchases between the independent
manufacturers (80%) and Steinhoff (20%), Ellerines has difficulty in
limiting its purchases from Steinhoff. Despite a decrease in the past
year or two, Ellerines’ current purchases from Steinhoff are still in
excess of 25% 53. Retailers dependence on Steinhoff gives it the
opportunity to exercise market power.
160. Steinhoff acknowledges that it benefits from the consolidation taking place
in the SA furniture retail industry. Its managing director , Markus Jooste is
quoted in Business Report, dated Sept 10 2002, as saying;
“The consolidation of the local furniture retail sector is nearing completion,
which is expected to benefit the group.”
161. According to the reporter one of the deals he was referring to was JD/
Profurn.
162. During the hearing we were presented with details of the purchase
arrangements between JD and Steinhoff. According to Steinhoff, JD
accounts for 33.50 % of its sales making JD easily Steinhoff’s largest
customer.
163. Both JD and Steinhoff deny that there is any supply contract in
163. Both JD and Steinhoff deny that there is any supply contract in
existence between them so what evidence we have about their
53 See hearing transcript, 29 November 2002, p 78.
31
relationship has to be gleaned from the effect felt by other players in
the market at both manufacturing and retail level. 54
164. According to Ellerines’ Mr Squires, not only does JD enjoy volume rebates
from Steinhoff, but Steinhoff in turn benefits from a right of first refusal. It is
understood that the right of first refusal entitles Steinhoff to either equal or
better the cost price of a product. Although Ellerines purchases a
substantial percentage of its supplies from Steinhoff, it does not enjoy
similar arrangements.
165. Mr Roy Pritchard, one of the intervenors, testified that JD’s store
merchandisers informed him that they had been instructed to “achieve
targets with Steinhoff” and as a consequence “the bread and butter” lines
manufactured by independents were substituted with variations made by
Steinhoff.
166. Almost all of the intervenors testified that they had lost the custom of the
JD Group, presumably to Steinhoff. Profurn, in contrast to JD, was a chain
that purchased overwhelmingly from the independents. Their underlying
fear was therefore that post merger the merged entity would favour
Steinhoff at their expense given JD’s current purchasing policies.
167. Despite the parties’ insistence that in order to maintain the Profurn stores’
brand awareness, they would be compelled to continue sourcing supplies
from the independents, we are not convinced that the intervenors fears
are unfounded.
Conclusion on vertical effects
168. There is no doubt that the increased buying power of the merged entity
will have an effect on the manufacturing industry. The US horizontal
merger guidelines recognise the potential harm of this phenomenon:
“Market power also encompasses the ability of a single buyer (a
"monopsonist"), a coordinating group of buyers, or a single
"monopsonist"), a coordinating group of buyers, or a single
buyer, not a monopsonist, to depress the price paid for a
product to a level that is below the competitive price and
thereby depress output. The exercise of market power by buyers
("monopsony power") has adverse effects comparable to those
54 This denial was made by Steinhoffs secretary in response to questions we asked the Commision to put to
them (letter received from Commission dated 14 November 2002) and on behalf of JD by Mr Sussman
during his testimony.
32
associated with the exercise of market power by sellers.” 55
169. What emerges from an analysis of the evidence is the following
a) Steinhoff is the dominant supplier of bedding and furniture to
the South African retail industry. Although the extent of this
is disputed on either version this holds true.
b) Steinhoff enjoys particularly close supply relationships with at least
two of the chains, JD and Relyant.
c) Steinhoff has over the past decade grown dramatically through a
strategy of acquisitions of the major players in the industry as well
as securing its position through vertical integration.
d) At the same time the position of the independent manufacturers
has declined. Once they enjoyed a position of market power over
retailers but this has gone. There has been a decline in the number
of manufacturers in recent years and there is reason to believe that
this trend will continue in the foreseeable future.
e) The demise of the manufacturer has an impact on the retail
industry. It means that retailers have fewer manufacturers to look to
inhibiting their ability to secure best prices and product choices. In
this industry product variety is part of a firms competitive
advantage.56
f) Independent manufacturers will depend for their survival on
preserving the custom of the chains in order to achieve some
economies of scale to retain their competitiveness.
g) However the chains need to secure high volumes of product and
can only get this at the best prices by securing rebates from
Steinhoff. This has led to even a powerful group such as Ellerines
being more dependent on Steinhoff than it wishes to be.
h) Independents claim that the volume rebates offered by Steinhoff
are such that they act as a disincentive for large retailers to
purchase from them. The Ellerines experience appears to
substantiate this.
purchase from them. The Ellerines experience appears to
substantiate this.
i) Profurn is identified by the independents as being one of the chains
that had “championed the cause of independent manufacturers” 57
j) JD on the other hand has the closest relationship of the all the
chains to Steinhoff and is its major customer.
k) Within the merged group JD is likely to alter Profurn’s purchasing
55 US Horizontal merger guidelines, US Dept of Justice and Federal Trade Commission, April 1992
(revised 1997) p4.
56 Morkels for instance in its marketing report notes that one of its problems is to overcome consumer
views of its ‘sameness” with its competitors. Record page1107.
57 See Submission from Independent manufacturers to the Competition Commission entititeld Proposed JD
takeover of Profurn, Record page 1022.
33
policies away from independents in favour of Steinfhoff
l) This will be a further nail in the coffin of the independents and likely
to lead to further failure and exit thus further increasing
concentration in the manufacturing sector.
m) The demise of the independent manufacturer sector is likely to
raise barriers to entry to the increasingly concentrated retail sector.
New entrants will be faced with higher supply costs and less choice
for their customers. Whilst imports are an alternative they are less
attractive to firms that have smaller volumes and those that sell to
the lower income groups, precisely the area in which we find
competition to be most reduced.
n) The merger is likely to increase concentration in the furniture
manufacturing sector and accordingly increase barriers to entry in
the retail sector.
170. The adverse effects referred to above can be remedied by placing
conditions on the merger outright prohibition is not warranted. We discuss
these conditions in the next section after first considering the efficiencies
claimed by the parties and the public interest issues raised by the merger.
Efficiencies
171. The parties submit that the introduction of JD’s management system will
result in annual huge efficiency savings, and that additional savings will
flow from the store closures. In a report prepared by Deloitte & Touche it is
submitted that the annual efficiencies are calculated at R83.4 million.
172. Since we do not intend prohibiting the merger and the efficiencies claimed
remain unaffected by the conditions we are imposing, we do not need to
pursue the efficiency arguments any further.
Public interest issues
173 . Employment
174. As is required in terms of the Act the parties informed two representative
trade unions of the proposed transaction. SACCAWU furnished the
trade unions of the proposed transaction. SACCAWU furnished the
Commission with a comfort letter in which they indicated their support for
the merger. ECCAWUSA met with the Commission but did not provide
any written submissions.
175. Essentially the merger, although resulting in potential job losses of
34
1284, as a result of proposed store closures, is presented as being
less harmful than would be the case if Profurn were to exit the
market. The parties aver that the merger will save some 5716 jobs
(without the merger 7000 jobs would be lost) 58.
176. We were however, not completely convinced that that this claim could be
regarded as a substantial public interest ground on which the merger
should be approved. Even if Profurn were to exit the market, it is clear that
some of its better stores would be bought as going concerns, thus not all
of Profurn’s workforce would be without employment. On the other hand, it
is also not clear what the precise impact of JD’s stated rationalisation
plans will be. Thus it is extremely difficult to conclude that the merger will
result in fewer job losses than would the exit of Profurn.
177. However since we have approved the merger, it is again not necessary for
us to come to any conclusion on whether its effect on employment creates
a substantial public interest ground, which justifies the merger.
SMME’s / Historically Disadvantaged Persons
178. Although the independent manufacturers are not SMME’s, historically
disadvantaged persons control most of these businesses. We have,
however, considered and addressed the potential impact on these
businesses, in our conditions, in terms of the vertical integration issue, not
as a public interest concern, although its impact will also advance this
particular aspect of the public interest.
Remedies
179. The intervenors have argued that the only solution to their competitive
concerns would be to prohibit the merger.
180. We are not persuaded by this as a less drastic solution would in some way
ameliorate some of the concerns of the sector without being overly
ameliorate some of the concerns of the sector without being overly
solicitous of specific independents who enjoyed a relationship with Profurn
in the past.
181. The JD group claims it has a policy of encouraging independents and the
figures that it produced were offered as evidence of this. According to it, at
the present moment, it is procuring 30% of its lounge furniture, case goods
and bedding needs from independents the remaining 70% was from
58 Parties competitive report pages 603604; Commission’s recommendation pages 6869.
35
Steinhoff. However they attempted to show that between 1998 and 2002
the share of independents had grown by 57% whilst that of Steinhoff’ by
only 56%. A closer look at the way these figures were computed showed
that they compared Steinhoff only to independents who continue to supply
JD and excludes those who no longer do. The true comparison reveals
that in total independents supplied 35% of JD’s needs in 1998 whilst in
2002 they supply only 30%. In contrast Steinhoff’s proportion had gone up
from 65% to 70%. At best JD may have increased its purchases from
some independents over this period but that increment has been at the
expense of other independents not Steinhoff whose overall position
continued to improve.
182. Whilst the professed policy is less easy to discern from the numbers
produced to support it, and hence the concern voiced by the intervenors
has validity, it would seem that the imposition of conditions for a short
period that JD purchases from independents in a specified percentage
does not run counter to its currently expressed business objectives.
183. For this reason we have imposed conditions on the JD Group in relation to
its purchase of supplies from independent furniture manufacturers. The
various chains that make up the newly merged firm will be obliged to
purchase from independents, at least in the same proportion as they do
presently. The conditions will be in force for the next three years, after
which they fall away. During the period in which the conditions are in
force, independents will have an opportunity to compete for the JD
Group‘s custom whilst adjusting to the changes in the retail sector. The
order of the Tribunal obliges the JD Group to submit an annual report from
its auditors confirming that the retail chain is abiding by this condition.
its auditors confirming that the retail chain is abiding by this condition.
184. On the other hand the conditions do not oblige the JD Group to use any
particular independents, even those who have traditionally supplied the
Profurn group. This means that the JD Group will not be forced to support
inefficient suppliers and thus raise its costs in relation to those of its rivals.
185. The conditions will not require detailed monitoring by the Commission
beyond ensuring that the merged firm complies with its obligations by the
specified dates. Verification becomes the responsibility of the merged
firm’s auditors. Doubtless as well the vocal lobby of independents will
monitor compliance very closely. However this does not detract from the
Commission’s statutory powers and responsibilities, which it must use to
ensure overall compliance with the conditions as set out in the order.
These conditions are set out in Annexure A hereto.
36
29 April 2003
N. Manoim Date
Concurring: D. Lewis, U. Bhoola
For the merging parties: Adv. W Pretorius instructed by Feinsteins Attorneys.
For the Commission: M. Simelane, Legal Services Division, Competition
Commission.
For the Intervenors: Adv. R Meyer SC instructed by Edelstein Bosman Inc.
37