---

COMPETITION TRIBUNAL

REPUBLIC OF SOUTH AFRICA

Case No: 67/LM/Dec01

In the large merger between:

Iscor Limited

and

Saldanha Steel (Pty) Ltd

_______________________________________________________________

Reasons for decision

_______________________________________________________________

Decision

1. The Industrial Development Corporation of South Africa Limited (‘IDC’) and Iscor Limited (‘Iscor’) currently own Saldanha Steel (Pty) Ltd (‘ Saldanha’) in equal shares. In terms of this transaction, Iscor seeks to purchase the IDC’s half share so that Saldanha becomes a wholly owned subsidiary of Iscor.

2. The merging parties have asked for the merger to be approved unconditionally. In its recommendation to us the Commission also asked for unconditional approval, but at the hearing it asked for the merger to be approved with a condition relating to supply to a downstream customer of Saldanha’s, Duferco.

3. The Commission and the parties have advanced different reasons for why the merger should be approved.

4. The merging parties argue that as the merger amounts to a change from joint control to sole control, there is no lessening of competition. Hence the merger should be regarded as neutral in substance with no effect on the competitive fabric of the market.

5. The Commission does not examine this argument in any depth in its recommendation to us, but finds that the two firms were never competitors in South Africa as Saldanha only supplied offshore customers. Since the Commission defines the market as a national one, the two firms were not competitors.

6. We find neither theory entirely satisfactory for the reasons we outline below. Nevertheless we have decided to approve the merger with conditions. Briefly our basis for approving the merger are-

1. Saldanha is a failing firm within the meaning of section 12A (2) (g) of the Act

2. Prohibiting the merger would in all probability lead to the closure of the Saldanha plant an outcome that cannot be justified on substantial public interest grounds because of the adverse effect on the Saldanha region.

3. The horizontal anticompetitive effects of the merger are hard to quantify. Firstly, because Saldanha never competed in the local market and secondly because of its disadvantageous location. The vertical anti-competitive effects can be cured by the imposition of conditions.

Background

7. Saldanha Bay according to one popular theory was named in error¹. The explorer in whose honour it was named had never been to the bay. If that is so, it was not the last miscalculation to plague the area.

8. The first record of a European explorer arriving at Saldanha was that of an English explorer in 1612. Yet although the bay’s potential was known of since that time, little development took place until the 20^th century. The area was proclaimed a municipality in 1954, but by the mid – 1960’s its population did not exceed 2500.

9. The town relied on fishing and the presence of a naval dockyard for its livelihood. During the Second World War Iscor officials had visited Saldanha to consider its possibilities. They were attracted by its favourable location in relation to international shipping lanes and the physical attributes² of the harbour. They believed that with the proper development it could be an industrially competitive export zone.

10. Their visit had no immediate consequences and it was not until Iscor required an outlet for its iron ore mine situated at Sishen, which is near Kimberley that their proposals were reconsidered.

11. Although Sishen was initially intended to supply Iscor’s steel factories domestic needs, it was soon realised that the mine’s iron ore production potential exceeded the needs of the local steel industry. Hence Iscor wanted to enter the export market. Given Sishen’s inauspicious location there was no obvious harbour outlet or a favourable transport link. The solution planners thought lay in Saldanha.

12. A railway line was constructed in the 1970’s to link Sishen to Saldanha a distance of 861 kilometers. The project also entailed expansions to the mine and the development of the harbour so it could serve as an iron ore loading port.

13. This project proved a great success. Today 76% of Sishen’s production is exported through Saldanha Bay.³

14. In the 1970’s planners came up with the idea that if iron ore could be beneficiated before it was exported this would yield even greater returns. They proposed that a steel factory be constructed at Saldanha Bay, that would receive the Sishen ore, beneficiate it and then export the resultant product through the port.

15. It is not hard to see why the scheme was so seductive for government. A large factory would boost the other ailing industrial efforts then at Saldanha, create new jobs and make South African exports competitive in a strategic industry. It was also foreseeable that a successful plant could lead to other downstream activity in the region.

16. For various reasons, which are not relevant to our enquiry, the idea conceived of in the seventies was not implemented until the mid-nineties.⁴

17. In 1994 the change in South Africa’s political dispensation was to have profound effects on Iscor. The company had previously enjoyed the protection of high tariff barriers and confined itself to a domestic market. Now the new government was committed to both lowering tariff barriers and to ensuring that its companies competed in export markets. In 1995 the tariff on steel was reduced from 30% to an effective 5%. At the same time the General Export Incentive scheme was phased out.

18. Iscor responded to these changes in 1994 by embarking on a five-year transformation program to transform its business to become internationally competitive.

19. The following year, in 1995, Iscor and the Industrial Development Corporation entered into a joint venture agreement to establish Saldanha Steel with each party owning 50% of the company. With the opening up of economic opportunities previously denied to apartheid South Africa, the object of the company was the:

“development of a robust competitive business which will compete principally in world markets..”⁵

20. The factory was only built in 1997 and production only started towards the middle of 1998, 18 months later than anticipated.

21. The delay cost Saldanha dearly. In the first place it led to massive cost overruns, and secondly, and perhaps crucially, it meant that Saldanha entered the market at the most inopportune moment for a nascent steel producer. In 1998 world steel prices had plunged due to the crisis in Asian economies.

22. Once the Asian crisis was behind them world steel producers believed there would be an international recovery in demand because the US and European economies remained buoyant. In eager anticipation plants that had been mothballed during the Asian crisis were brought back into production. The result of this optimism was that world steel production had grown by 5,1% in 1999 and by 8,2% in 2000 far exceeding economic growth.⁶

23. Saldanha came onto the market just as steel prices has plummeted internationally. This volatility remains to the present day. As a result revenue projections in its initial budgets have proved woefully unrealistic.

24. To compound the problems expenses were also underestimated. Bernard Smith, a former Managing Director, told the Steyn Commission in 1995 that the project would cost R4,7 billion⁷. Today the merging parties tell us the final costs were R8,7 billion.

25. This meant the two shareholders had to inject loans to keep the company financed so it was soon consuming more debt than iron ore. This led the auditors to remark in August 2001 that:

“ The ability of the company to continue as a going concern for the foreseeable future depends on the continued financial support by the shareholders. The shareholders have reconfirmed their intention to fund the company’s cash flow requirements for the next twelve months by way of interest free loans, as referred to in the Directors report..”⁸

26. At the end of June 2001 the firm had shown an after tax loss of R 3287 million dwarfing even the loss of the previous year of R 945 million.

27. In 2001 Saldanha was losing $ 91 per ton in production without even contributing to financial interest or depreciation.

28. Not surprisingly the shareholders insisted on action on all fronts. A report was commissioned by consulting firm McKinsey to assess whether the plant was viable in the long term. The consultants concluded it was, provided a number of recommendations were implemented .The recommendations heralded an improvement initiative that was sanguinely labelled Operation Rainbow.

29. Consultants together with Saldanha staff analysed the production process from start to finish in a bid to lower production costs and improve efficiencies. It is too early to determine whether these efforts have been successful because some significant repairs have still to be implemented but Saldanha claims some modest success to date.

30. But it was financial as opposed to production engineering that saved Saldanha from imminent failure at the end of 2001.

31. Since 2000 there had been an idea that Iscor assets were considerably undervalued because they combined both mining and steel assets in one listed company. The market it was thought would re-rate the firms if they were separated into a mining and steel company. The problem was what to do with the Saldanha debt in order to achieve this. If all this debt was to be inherited by the stand-alone steel company it would be very unattractive to investors. Eventually after months of tough negotiations, in which the IDC played a crucial role, an acceptable structure was arrived at, the crucial elements of which were that –

1. The Saldanha debt was apportioned between Kumba, the newly named company that inherited the mining assets and Iscor⁹;

2. Iscor was guaranteed supply of iron ore from Sishen on favourable terms; and

3. Iscor would acquire IDC’s 50% stake in Saldanha.

32. In effect debt was taken off the Saldanha book and bolstered by the injection of equity into Iscor and Kumba financed partly by the IDC and partly borne by the respective shareholders of those companies who given their re-rating by the market could not complain.

33. The listing of the two companies proved spectacular. After the unbundling Iscor ‘s shares started at R6 in November last year. At one stage they had reached R17 although they have appeared to have stabilised at around R 13,50 at the time of this decision.¹⁰

34. Then fortune at long last smiled on Saldanha. The spectacular crash of the rand at last made its export prices competitive. So it happened that for the month of December 2001 the company reported its first profit.

35. Despite these propitious events we still consider Saldanha to be a failing firm within the meaning of the Act, and we explain why below, but before that we need to determine whether the merger raises any competition concerns.

Competition Issues

1. Joint to sole control

36. In their competitiveness report, the parties main argument for the approval of the transaction was that the merger did not ‘de facto’ amount to a change of control, as Iscor had been effectively managing Saldanha Steel since January 2001. Although at the hearing the parties also invoked the failing firm doctrine, this initial argument was not abandoned and we must therefore consider it.¹¹

37. This argument was not fully developed beyond the parties’ observation that whilst there was a structural change as a result of the acquisition there would be no change in the competitive environment.

38. It is not self-evident that because a company goes from joint control to sole control, its competitive inclination remains unchanged. For this reason the onus is on the parties to the merger to establish that this is the case.

39. In Europe in the ICI /Tioxide case, the Commission had to decide the reverse situation, namely whether a move from sole control to joint control amounted to a concentration. In coming to their decision they had to evaluate the distinction between joint and sole control:

“…because decisive influence exercised solely is substantially different to decisive influence exercised jointly, since the latter has to take into account the potentially different interests of the other party or parties concerned.. By changing the quality of decisive influence exercised by ICI on Tioxide, the transaction will bring about a durable change of the structure of the concerned parties.”¹²

40. In this case the merging parties have done no more than make a bald assertion that Iscor enjoys ‘management control’ over Saldanha. ¹³We have not been provided with any agreement to that effect or any evidence to indicate that the IDC has been constrained from exercising any rights it enjoyed as a result of the 1995 shareholders agreement.

41. On the other hand there is evidence to suggest that the change of control would make a significant difference to the way Iscor relates to Saldanha Steel.

42. In the first place its incentives have changed. It has gone from a situation where it enjoyed only half of the benefits of an enterprise, which constituted a competitor, to being its sole owner and controller.

43. In the Saldanha shareholders agreement, entered into in 1995, Iscor insisted on a clause that would effectively preclude Saldanha from entering the local market as a competitor.¹⁴ A sole shareholder would not require such a clause as it could achieve the same end without a contract. Iscor required the clause because it did not have sole control and in this respect it knew that its interests and the IDC ‘s would not coincide.

44. Another fact that illustrates this, is the supply arrangement it has now secured for itself with Kumba in relation to Sishen. Prior to the restructuring, Iscor then owner of Sishen, supplied Saldanha on an arms length basis. Yet its mine at Thabazimbi supplied its Vanderbijlpark factory at cost plus a 3% management fee i.e. on more favourable terms. When Iscor was restructured it ensured its supply agreement with Sishen at cost plus a 3 % management fee i.e. on the same terms as Thabazimbi was supplying it. ¹⁵The fact that Iscor supplied Saldanha less favourably prior to the restructuring is again indicative of the fact that its incentives as a joint shareholder were different to that of a sole shareholder.

45. Furthermore in communicating with its shareholders Iscor has made much of the fact that having sole control of Saldanha would allow it to introduce changes beneficial to Iscor.

“Full control of Saldanha will allow the realisation of significant synergistic benefits for Iscor in the areas of product rationalisation, marketing, and future capital expenditure.”¹⁶

46. Presumably these changes would have been implemented some time ago if Iscor was able to behave as if it were a sole controller. The reality is that the IDC acted as a constraint. Despite the fact that at our hearings the IDC protested modestly that it was not in the business of making steel and was therefore content to let Iscor manage the business, the IDC was no mere passive investor or spectator at board level. It had risked an enormous amount in the undertaking and it is improbable that given its exposure it would not have kept the closest scrutiny over Saldanha’s affairs.

47. Finally the IDC is a significant player in the steel industry apart from its stake in Saldanha. Its joint ownership of Duferco Steel Processing (DSP) a customer and a potential competitor of Iscor are illustrative of the difference in interests. Iscor as a sole shareholder in Saldanha has every incentive to squeeze DSP, but the IDC, because of its interest in DSP, does not.

48. In our view the argument that the change from joint control to sole control leaves the competitive environment unchanged, is without substance.

49. For this reason we need to go on to analyse the competitive effect of the merger.

b) Horizontal effects

50. Both firms are involved in the manufacture of steel products. Since steel products have a variety of different applications they are not all substitutes. In this case both the Commission and the parties are in agreement on the product market but some prefatory remarks are necessary before we consider this.

51. Steel is manufactured in an integrated process. Raw materials such as iron ore, coal and dolomite are charged to blast furnaces where they are converted to liquid iron. The liquid iron is refined in basic oxygen furnaces and electric arc furnaces to produce liquid steel. The liquid steel is cast into slabs, which are hot rolled into heavy plate in a plate mill or coils in a strip mill. The coil is either sold directly as hot rolled strip or processed further into cold rolled and coated products.

52. Both Iscor and Saldanha produce Hot Rolled Coil (HRC), which is also the only area of product overlap between the parties. However, Saldanha has been specifically equipped to roll to a thinner gauge of between 1mm to 1.6 mm in thickness, called Ultra Thin Hot Rolled Coil (UTHRC), which is thinner than the HRC that is produced by Iscor and other competitors such as Highveld Steel. Saldanha currently only exports this product.

53. HRC is a basic material used in the manufacture of general engineering products such as containers, pipe, wheel rims, agricultural implements and gas cylinders. Gauges as thin as UTHRC are traditionally produced as Cold Rolled Coils (CRC), which are used in automotive components, tubular products mainly in furniture applications and in high precision tubing for conveyance of gases and fluids. Since CRC tends to be smoother and more pliable than UTHRC, UTHRC can only be used as a substitute for CRC in those applications where surface conditions and annealing strength (i.e. brittleness) are suitable for hot rolled material. Currently no customers in South Africa utilize UTHRC. In order to convert from CRC to UTHRC technical changes to welding and forming equipment needs to be made by potential customers.

54. However UTHRC presently only constitutes 35% of Saldanha ‘s production and thus 65% of its current production comprises product, which could be sold in the domestic market.

55. Iscor in response to this have argued that Saldanha’s location in relation to downstream industry make it an unlikely entrant into the local market. The bulk of the downstream manufacturing capacity is located in the Gauteng region with only one major customer in the Western Cape ¹⁷ the nearest market to Saldanha.

56. The evidence of the parties is that the South African market consumes approximately 2,4 million tons of flat steel products annually. Of this approximately 900 000 tonnes is in the form of HRC.

57. The evidence of the parties as to transport costs is that for a firm located in the industrial heartland transport costs per ton from Saldanha would be in the vicinity of R240 to R 270 per ton. Iscor’s transport costs from Vanderbijlpark to the central market would be R 54 per ton. The parties argue that if the same firm were to import product from Europe, which would be landed at Durban, transport costs would be approximately R140 per ton. The parties argue convincingly, that based on these figures Saldanha’s transport cost disadvantage make it an unlikely entrant to the inland market.¹⁸

58. Iscor argues that had it been intended that Saldanha would compete in the local market it would not have been situated where it is.

59. However this bold assertion by Iscor is belied by its own conduct. As we mentioned earlier, Iscor in 1995 insisted on a clause in the shareholders agreement that provided that it would be responsible for the sale of Saldanha’s products in the domestic market.

60. Despite the euphemistic language, the effect of this clause is to exclude Saldanha from the domestic market except with Iscor ‘s consent. Without taking too conclusive view of it, the clause amounts to a market division between competitors and thus might well be pro no scripta on the grounds that it amounts to an unlawful horizontal restraint. In fact Saldanha had no domestic customers except for DSP.

61. The fact that Iscor insisted on this clause is a clear indication that it feared that Saldanha might enter the domestic market as a rival, despite its claims to the contrary that such rivalry is infeasible.

62. Post merger of course the clause becomes academic, but it cannot be used to justify the fact that Saldanha was never a competitor of Iscor’s. We do not know if but for the restraint imposed on it, Saldanha might not have entered the local market.

63. That is not to say that the market for HRC in South Africa is characterised by intense rivalry. Indeed, as the table below suggests, this is largely a two firm market with Iscor having achieved an overwhelmingly dominant share, more than 80 %, and its only local rival Highveld holding to a relatively constant 15 % share. Competition from imports is slight, less than 3 % at present and the trend shows this dropping in the past three years.

Table 1: HOT ROLLED PRODUCTS

S.A. MARKET SHARE

									INDEX
MARKET SHARE	100%
	90%								Iscor
	80%					80.4%	82.1%
	70%	79.3%	79.2%	72.7%	71.9%
	60%
	50%
									Highveld
	40%
	30%
	20%	13.9%	15.2%	16.1%	15.0%	14.3%	15.1%
									Imports
	10%
	0%	6.8%	5.6%	11.2%	13.1%	5.3%	2.7%
		95/96	96/97	97/98	98/99	99/00	Jul00-Mar01

64. The parties nevertheless contend that the market is an international one. The Commission argue for a national one. There is no doubt that international steel prices play a function in determining the price on the domestic market as Iscor has a policy of import parity pricing ¹⁹. According to Iscor’s Revised Listing Particulars:

“Iscor’s steel pricing policy in the domestic market is based on import parity principles as its main competition in most of its product ranges is represented by imports. The exceptions to this principle are:

- in the case of products where local competition exists and where over capacity results in prices that are lower than import parity, as it is the case with certain lower quality long products; and

- where Iscor offers lower price incentives (negotiated on a case by case basis ) with local manufacturers to encourage secondary exports.

65. Many suggest that if a firm can set prices at import parity that fact alone is indicative of the fact that it has market power. We do not need to decide however whether the market is an international one or a domestic one as that assessment would not alter the outcome of our decision. We have assumed however that the market is a national one.²⁰

66. Nevertheless we conclude that although Saldanha poses potential competition to Iscor, its geographic disadvantage from the largest customer base in the interior, the absence of a customer base of significance in the Western Cape, coupled with an absence of a market for the product it is most competitive in (UTHRC), all suggest that its potential to compete successfully with Iscor is remote.

c) Vertical issues

67. Saldanha has only one local customer and that is Duferco Steel Processing (Pty Ltd) (‘DSP’) a company that is jointly owned in equal shares by the IDC and Duferco a Swiss company.²¹

68. DSP was established as part of the broader Saldanha industrial development strategy to process HRC. Saldanha Steel and DSP entered into an agreement in 1997 to set out the terms on which Saldanha would supply HRC to DSP for a 10-year period.

69. In terms of the agreement DSP was –

1. obliged to purchase exclusively from Saldanha a specified amount of HRC per annum; ²² and

2. was further prohibited from reselling the HRC that it had purchased from Saldanha and processed in markets in Southern Africa, without the consent of Saldanha.²³

70. The first clause was presumably inserted to guarantee DSP as a customer and from DSP ‘ point of view to ensure that it had a secure source of supply. The prohibition on resale was inserted again as with clause 16 of the Saldanha shareholders agreement for the benefit of Iscor which also processes HRC in the manner that DSP does. Again both clauses served Iscor by ensuring that potential rivals were confined to the export market.

71. The agreement was premised on the assumption that Saldanha would produce 1,2 million tons of HRC per year and that DSP would be able to absorb 450 000 tons of this output per annum. As we have outlined above, Saldanha’s production ambitions were not realised and it was not able to supply DSP on the original contract terms. An arrangement was entered into in 2000, in terms of which Iscor undertook to supply 50% of DSP’s needs.

72. The new arrangement has not been formalised into an agreement yet. The status of the 1997 agreement is unclear. Iscor alleges that the “agreement has effectively been abrogated by disuse.” Confusingly, they also in the same breath state that “if the merger is approved, Iscor proposes to take the business out of Saldanha Steel, and therefore there will be a need to be a cession from Saldanha Steel to Iscor” ²⁴ This suggests they regard the agreement as still binding if it was not there would be no need for a cession. Yet later on they state that part of their future strategy for Saldanha is to develop a steel grade for thin HRC for supply to DSP for cold reduction to dimensions suitable for West Africa.²⁵

73. In a submission to the Competition Commission, DSP expressed anxiety about its conditions of supply given its present lack of contractual security.²⁶

74. DSP points out that with the merger it would only have one supplier for it most critical raw material, which accounts for 70% of its production costs. This single supplier relationship exposes it to a potential squeeze from Iscor both in relation to price, quality and quantity. Since DSP is a potential competitor of Iscor, if it were to supply the local market, its vulnerability to a supply squeeze is well founded. DSP states that it does not have any alternative supplier to Iscor and when in the past it found alternative sources it was faced with the prospect of anti-dumping duties being imposed.

75. At the hearing we raised these concerns with the merging parties. In response to the proposal by the Commission that the vertical concerns could be obviated by the Tribunal imposing appropriate conditions, the merging parties raised no serious objections.

76. We have accordingly framed conditions, which are designed to ensure that whatever contractual arrangements the parties may enter into, DSP has no constraints on its source of supply and that it is free to compete in the domestic market.

d) Failing firm

1. Theoretical

77. The jurisprudence of competition law recognises what has become known as the ‘failing firm defence’ to sanitise a merger that might otherwise raise competition concerns.

78. The defence first emerged in US law when in a 1930 decision, International Shoe v FTC ²⁷the Supreme Court recognised that the acquisition of a failing firm did not violate section 7 of the Clayton Act.

79. What is surprising is that notwithstanding amendments to the Clayton Act subsequent to this case, and discussion of the failing firm defence in Congressional debates preceding the amendments, the language never found its way into the statute. ²⁸ Nevertheless the defence is well recognised in the case law and has been invoked on numerous occasions.²⁹

80. The defence is also recognised in the U.S. Horizontal Merger Guidelines which provide for the following ;

“A merger is not likely to create or enhance market power or facilitate its exercise if the following circumstances are met:

1. the allegedly failing firm would be unable to meet its financial obligations in the near future;

2. it would not be able to reorganize successfully under Chapter II of the Bankruptcy Act;

3. it has made unsuccessful good-faith efforts to elicit reasonable alternative offers of acquisition of the assets of the failing firm that would both keep its tangible and intangible assets in the relevant market and pose a less severe danger to competition than does the proposed merger; and

4. absent the acquisition, the assets of the failing firm would exit the relevant market.”³⁰

81. Influenced by the US approach the defence has also found its way into European law albeit in an altered form. As in US, the European Merger Regulation contains no express recognition of the doctrine, but it has been recognised now in case law and the practice of the Commission. In the judgement of the European Court of Justice (the ‘ECJ’) in the case of France v Commission³¹ the ECJ accepted the failing firm test applied by the Commission, in approving the merger between the only two German producers of potash.³²

82. The Commission ‘s test requires that the following conditions be satisfied:

1. the acquired firm would have withdrawn from the market if not taken over by the other firm;

2. the acquirer would gain the market share of the acquired firm if the latter were to exit the market; and

3. no alternatives were available that were less anticompetitive.

83. It has been pointed out that the second condition is a European innovation and is more stringent than the U.S. test, which has no such requirement.³³

84. In Australia, the Merger Guidelines indicate that the Australian Consumer and Competition Commission ( the ‘ACCC’) will recognise the failing firm notion provided certain conditions are satisfied. Their approach is instructive and worth quoting at length:³⁴

“Briefly, the Commission [the ACCC] will need to be convinced that the firm cannot be successfully re-organised and there is no other viable buyer for the business, and no likelihood of such a buyer emerging, such that the firm’s resources are likely to exit the market absent the merger and so cease to represent an actual or potential constraint on the market.

If the target firm is considered to be failing, the Commission will consider the likely effect of the acquisition on competition compared to the effect of the target’s assets exiting the market. Under the latter circumstances, the distribution of the target’s customer base among the remaining market participants would be determined by market forces, whereas an acquisition would tend to deliver those customers to the acquiring firm. If the competitive strength of the remaining participants is evenly matched, the level of competition in the market may be better served by allowing the firm to fail. However, the loss of capacity will tend to reduce competitive pressures in the market. Depending on the effect of the acquisition on the relative strength of remaining participants, retaining the failed firm’s capacity in the market may still be pro-competitive.”

85. In Canada the failing firm doctrine is given statutory recognition as one of the factors to be taken into account in determining whether a merger prevents or lessens competition. In this respect our Act is identical to the Canadian Competition Act.³⁵

86. There is thus no doubt that the failing firm doctrine is widely recognised in competition law jurisprudence, and regardless of whether the doctrine has become part of the case law or enjoys an express statutory recognition, has been applied with a degree of uniformity.³⁶ What is more difficult to discern, and which is relevant to an assessment of how much of their jurisprudence to adopt, is the exact rationale for the defence.

87. Writing in the context of European merger regulation, Bishop and Walker argue that if other firms will buy the assets after a firm has failed, then the failing firm defence is not valid³⁷. The defence is only reasonable if productive assets would otherwise exit the market.³⁸

88. A similar theme is echoed in the US Horizontal Merger Guidelines, where it is argued that if imminent failure would mean that the assets of the firm would exit the market, then post merger performance if the merger is approved, may be no worse than if the merger had been blocked and the assets left the market. ³⁹

89. The rationale expressed here seems to be that if the status quo remains unchanged if the failing firm’s assets exit or are acquired one would favour their acquisition. That still begs the question of the underlying rationale of invoking the defence when the post merger state of the market is less competitive than before because the acquiring firm would enjoy a greater market share on acquisition of the failing firm than it would have if it just exited.⁴⁰ Here case law is not that consistent.

90. Areeda alludes to this difficulty in his comment on the International Shoe case:

“International Shoe suggests two differing and perhaps inconsistent grounds for a failing firm defence (1) Acquisition of a failing company may have no significant adverse effects on competition; and (2) the interests of stockholders, creditors, employees, and others affected by the fate of the failing company deserve recognition. These propositions are hardly sufficient grounds for making the defence absolute in all circumstances; nor do they point to any unified or simple way of identifying the defence’s proper domain.”⁴¹

91. Areeda goes on to say that:

“as a general proposition the cases have not attempted to determine whether an acquisition benefits competition more than failure would do harm.”

92. Areeda is critical of the one leg of the International Shoe test namely the adverse impact on shareholders, employees or creditors. He points out that antitrust is traditionally unsympathetic to these concerns when they are asserted as a justification for an anticompetitive outcome. Thus a loss to shareholders is no panacea to firms involved in a price fixing cartel. Furthermore the private interests are not always reconciled if a merger goes ahead. A firm may be willing to purchase only the assets of a failing firm and thus satisfy shareholders to some extent but not employees.

93. Areeda does however acknowledge that there are circumstances where private interests may be worth recognising. Although he gives several examples ⁴²one is particularly relevant to the circumstances of this case. He argues that new entry and investment are encouraged the more the costs of failure are reduced. Expressed differently, if firms taking investment decisions know that competition law might increase the costs of failure by harshly applying merger control in the event of their exit, they would be less likely to assume the risk.

94. Areeda then looks at the other rationale for the failing firm defence which is to ensure that productive assets are reassigned to more efficient users. He points out correctly that a number of these defences are more appropriately classified as efficiency defences and need not be used to broaden the failing firm defence beyond what it need be.

95. This is particularly apposite in interpreting our statute where we have an express efficiency defence, which is absent in the Clayton Act.⁴³

96. Areeda however goes on to identify efficiency defences which arise in a failing firm context and which do not qualify for acceptance in the typical economies defence and he argues for their recognition as part of the failing firm defence. Into this category he would include the following :

1. an acquisition would preclude the administrative costs of bankruptcy

2. it would keep in the market resources that might otherwise be taken out of production, even though they are not yet depleted

3. it would assign productive assets to managers capable of using them more efficiently.⁴⁴

97. If we follow the Areeda approach and look at the failing firm defence in the context of our Act, then there may be other aspects of a failing firm defence that can more appropriately be assumed under other sections provided expressly for by the Act. Thus as is the case with the Kali und Salz merger, although the transaction led to unacceptably high concentration, the failing firm defence was successfully raised in part because the merger obviated a large loss of jobs in Eastern Germany.⁴⁵

98. In our statute there would be no need to invoke the failing firm doctrine to such a situation when the adjudicator can have regard to the employment effect in terms of the public interest criteria.⁴⁶

99. Similarly, where the failing firm defence is in reality an efficiency defence, the merging parties should rely on the efficiency defence in section 12 (A) (1)( a)( i). This is not to say that alternatives cannot be alleged, but it is important that we avoid conceptual confusion. The danger of following some overseas jurisdictions uncritically, is that they may recognise under a failing firm defence, factors that we, with our more extensive merger regime, might recognise under another classification, employment loss being a good example.

100. This perhaps leads us to an important contextual difference between the way we should apply the failing firm doctrine and the way it is applied in other jurisdictions.

101. In our Act, the failing firm doctrine is not used as a ‘defence’ to a merger that has been found on an initial market analysis to be anticompetitive. Rather it is recognised as one of list of ‘factors’ that one takes into account before one can determine whether a merger is anti-competitive.

102. Thus in some jurisdictions, such as the United States, the adjudicator would first determine whether a merger was anti-competitive and then only if it was, consider the failing firm doctrine as a defence.

103. Our Act does not divide the inquiry into two discrete stages. The way section 12A(2) is drafted, the consideration of whether a merger is anti-competitive involves as part of that inquiry a consideration of a non-exhaustive list of factors, one of which is whether a party to the merger is a failing firm. In contrast to this, our ‘efficiency defence’⁴⁷ is a real defence to an anti-competitive merger and for this reason is located in a self-standing sub-paragraph of section 12A, as a reading of section 12A(1)(a)(i) makes clear. To invoke the efficiency defence one first has to conclude that the merger is anti-competitive. This one does by first conducting the inquiry into the factors contained in section 12A(2). Then, and only if that inquiry leads to the conclusion that the merger is anti-competitive, does one go on to enquire whether the efficiencies gained outweigh the anti-competitive effects.

104. The distinction might appear to be a mere lawyers quibble with no practical significance, but this is not so. If the failing firm doctrine was a “true defence”, then if all the elements of the defence could be established, an anti-competitive merger would have to be approved. However if it is just one of a list of factors to be taken into account, in the list set out in section 12A(2), then even if it was established, that the firm was failing, by say the application of the US or EU tests, one might nevertheless still find a merger to be anti-competitive because of the way one balances this failing firm factor in relation to all the others. It is thus possible in terms of the Act, that a credible failing firm theory may nevertheless not save a merger, which raises serious competition concerns.

105. Conversely, where the competitive loss is low, then one may be less exacting in requiring a showing of all the elements of the traditional failing firm defence. If the failing firm concept was a defence, in the sense that the efficiency defence is, then this type of flexibility would be impermissible and one would have to satisfy all the elements of a test that the legislature had provided before it could be invoked.

106. Whilst this approach may risk making the concept more nebulous, it does allow the adjudicator the flexibility to achieve real interest balancing, as opposed to the application of rigid formulas.

107. A flexible approach, allows one for instance to have regard to some of the rationale for the failing firm defence even though they do not constitute elements of a traditional defence. Thus if one rationale, as Areeda puts it, is to avoid inhibiting investment decisions because we regard investment in general terms as pro-competitive, then evidence that not allowing the acquirer to rescue a failing firm might inhibit entrepreneurial or investor risk, would be relevant and pertinent despite the fact that they are not elements of a traditional test. This distinction is important to bear in mind before one imports en bloc the jurisprudence on this doctrine from other jurisdictions.

108. This is not to say that we would abandon all reference to these tests. Rather we are saying that some of the elements of the defences would be more sacrosanct than others. If the merging parties had not established the requirement that there were no other viable less anti-competitive outcomes, the failing firm doctrine would not avail them. This is the approach which we took in the Sasol/Schumann⁴⁸ merger, where although we applied the test set out in the US Guidelines the failing firm doctrine was not found to have been established on this ground.

109. Other controversies in foreign case law about the failing firm doctrine are partially answered for us by the language of the Act. For instance it is not necessary to show that a firm has failed only that it is likely to fail. Nor does failure need to equate to insolvency, otherwise the legislature would presumably have used that language. As Areeda has expressed it “ insolvency is not a necessary ingredient of failure:

“Low revenues may dictate that a firm withdraw from a market rather than reinvest capital, even though it is continuously able to pay its maturing debts. Although proof of such a case poses more serious problems, the opportunity to make it cannot reasonably be foreclosed.”

110. In summary we are saying the following:

1. A failing firm defence should not be invoked if it amounts in substance to another factor or defence which the Act already provides. In particular we draw attention to the efficiency defence and the public interest criteria.

2. The merger criteria for a failing firm set out in the tests of other jurisdictions will carry serious weight in our assessment. Organising ones evidence on the basis of these criteria would thus be useful and instructive to the Tribunal.

3. A merger would not be regarded as lessening competition if the conditions laid out in the more stringent EU test can be satisfied.

4. A party falling short of the “market share would have gone to us” requirement, but that could satisfy the other elements of the test or the standard in the US test, would have a reasonable possibility of success depending on the degree of the anti-competitive sting. Thus where the anti-competitive effects of the merger are otherwise slight, then the Tribunal might be less stringent in the application of some of the criteria. Here the party should have regard to evidence that establishes some rationale for the existence of the failing firm doctrine. We have referred to some of these in our discussion although we do not suggest that this is an exhaustive list.

5. Evidence of the extent of failure or its imminence, would be weighed up against the evidence of the anti-competitive effect. The greater the anticompetitive threat the greater the showing that failure is imminent

6. No leniency would be afforded to the requirement that there be evidence that there is no less anticompetitive alternative.

7. The onus is on the merging firms to establish the evidence necessary to invoke the doctrine of the failing firm.⁴⁹

ii) Applying the failing firm doctrine to the facts of this case

111. At the time that the Saldanha project was first conceptualised in the 1970’s, Iscor was still a parastatal. In 1989 Iscor was privatised and then listed. At that stage the company was involved in both mining and steel manufacture although they were operated as separate businesses. When Iscor listed, the markets initial euphoria was soon tempered by the fact that analysts did not like the idea it was running two businesses and there were perceptions it lacked focus and would land up doing neither well. After listing the Iscor share price was to drop from R2 to 60 cents per share.

112. In 1994 Iscor brought in new management to run the separate businesses. According to an article in the Financial Mail, Iscor had not adapted quickly enough to the post ‘94 economic environment.

“Though it took advantage of export opportunities, especially in Asia, it did not restructure quickly enough to remain internationally competitive in steel.”⁵⁰

113. The idea was then to transform both businesses to make them internationally competitive. This theme of the international competitiveness of Iscor was to drive both the company’s and government policy thereafter and provided the context for the decision we have to arrive at in relation to this transaction.

114. In 1995, as we noted earlier, Iscor and the IDC entered into a shareholders agreement to establish Saldanha. The construction of the factory commenced in May 1996 , but the first coils were only produced in the fourth quarter of 1998.

115. The plant was bedevilled by numerous problems from the outset.

• Concerns about the environmental impact of the plant including its site led to delays in the commencement of construction.⁵¹

• The construction itself took longer than anticipated.

• When production commenced it was at levels substantially below those budgeted. The plant had been designed to produce 100 000 tons per month, but for most of 1998 and 1999 the production levels were considerably below this. The Corex smelter and gassifier was producing at only 70% of its capacity due to design flaws. It was only in October 2001 that production reached a peak of 90% and it is by no means clear that this is sustainable as the Corex will require repair later this year. The Conarc furnace is one of the few in existence in the world and thus there was limited experience in dealing with it when problems arose, as they did.

• By the time Saldanha was ready to place its product on the market, as we noted above, steel prices had plunged and were continuing downwards. The project could not have come to market at a less opportune moment.

• Expertise in running a plant with technology that was new to the South African steel industry was not what it should be and production delays occasioned by technical problems were frequent.

• The delays in commencement and the sub-optimum returns on initial production had meant that the shareholders had to dig deep into their pockets to provide loans to continue the joint venture.

• In 2001 at then prices and production costs the firm was losing $ 91 per ton.

116. Despite the fact that the two shareholders are very large concerns with a considerable ability to raise capital their balance sheets started to take strain under the weight of the obligations that they started to incur. This was to lead to a strained relationship between the shareholders, and according to the Financial Mail, the IDC had at one stage contemplated taking over control of Iscor.⁵²

117. Iscor’s own performance was compromised by Saldanha’s difficulties. Iscor’s share of Saldanha Steel’s after tax loss excluding impairment charge for the year 2001 amounted to R 526 million compared to R 473 million for 2000.

118. Faced with these growing problems the shareholders engaged the services of consultants in 2001, who as part of their wider brief, considered the option of liquidating Saldanha steel or mothballing it for a period until international prices for HRC had improved.

119. The consultants concluded that neither of these options was attractive and might indeed result in a greater cost in the short term to the shareholders than continuing the operations.

120. The shareholders accepted the recommendation and the fortunes of Saldanha and its shareholders have now improved both as a result of the restructuring we describe earlier, the improvements to Saldanha’s production and the favourable rand dollar exchange rate.

121. This change in fortunes begs the question of whether Saldanha should still continue to be regarded as a failing firm. Indeed Saldanha reported in December 2001 that it was in the black for the first time for its figures of that month.

122. The parties have urged us not to place too much emphasis on these factors to come to a conclusion that the firm is no longer failing.

123. Saldanha’s recent results are not yet evidence that it has turned the corner and its revival in fortunes is largely the result of the decline of the rand which on a rand / dollar exchange gave a windfall to exporters at the end of last year.

124. The fact that Saldanha’s debt has been restructured does not mean it has been eliminated. If Saldanha was still carrying its pre-restructuring debt, the December results would have shown a considerable loss. It is the deep pockets of its shareholders that have kept Saldanha from exiting the market. The fact that they have concluded that the costs of its retention outweigh the costs of its closure does not make it any less than a failing firm. As we pointed our earlier in our theoretical discussion, our Act does not require that a firm must have failed only that it is likely to fail. Without the financial restructuring of Iscor and the IDC that we have referred to Saldanha would have failed - no independent firm would have provided the finance necessary to bail it out of its debt obligations given its past performance and its less than certain future.

125. Whilst there is no doubt that Iscor and the IDC remain confident that Saldanha has a future, this will only become clear in the long term. In the medium term the risks to the shareholders remain considerable.

126. The extent of the financial disaster is not in doubt and the numbers speak for themselves.⁵³ In a letter to the Tribunal, dated 15/02/02, by the Chief Executive Officer of ISCOR, Mr Louis van Niekerk, indicates that by 2001 Saldanha Steel had a debt burden of R5,8 billion consisting of the following:

Foreign Loans R 3.3 billion

South African Banks R 1.5 billion

IDC (through project loan) R 1.0 billion

TOTAL R 5.8 billion

127. He also said that both Iscor and the IDC would absorb R3.0 billion each of Saldanha’s debt i.e. re-capitalisation of Saldanha’s debt by shareholders.⁵⁴

128. The IDC, according to Mr Ngqula,⁵⁵ (the CEO of the IDC) stands to lose R 6.5 billion, Iscor R 2.7 billion and the local banks R 1.5 billion if the merger was not approved. This amounts to a total of R10.7 billion.⁵⁶ He also mentioned that there were foreign loans to the amount of R3 billion. Both Iscor and the IDC would absorb R3 billion of Saldanha’s debt, as shareholders.

129. According to the McKinsey Report, Saldanha Steel’s ultimate net present cost to shareholders would be R11.5 billion, should Saldanha continue business as pre the merger. Ceasing operations at Saldanha Steel will have a total cash impact of R5.8 billion for mothballing, R7.3 billion for liquidation and a further equity write off of R4.7 billion is to be expected in both scenarios. For both liquidation and mothballing shareholders are likely to be exposed to the full Saldanha Steel debt position of R 5.5 billion in addition to the R1.0 billion injected up to June 2001. Moreover, large contract cancellation costs of approximately R 7.0 billion must be faced.

130. Iscor has launched a rights issue for R1.67 billion to bring down its Saldanha-driven debt, while the IDC recorded the impairment loss of R2.6 billion in its interim results. Iscor’s share of Saldanha’s operating loss in the 6 months to December 2001 was R266m.⁵⁷

131. In the short term Saldanha will have to decrease its production levels as its main refinery Corex will be shut down for 45 days to undergo repairs in May 2002.

132. The evidence that the patient is in remission is too inconclusive. Saldanha must at the present moment still be regarded as a failing firm.

133. The next issue we must examine is whether even if the firm is failing there is not an alternative buyer whose purchase would not raise competition concerns.

134. The evidence of the merging parties is that they have approached various international concerns over the past few years. All it seems were approached on the basis that they would be offered a stake in Iscor. All apparently placed various conditions on the deal which included an insistence that Saldanha be integrated first into Iscor. None it appears would have been interested in purchasing Saldanha as a going concern and that is hardly surprising as given its debt levels the shareholders could never have got their price.⁵⁸ On the other hand if Saldanha is liquidated equipment is more likely to be sold off piecemeal to another firm based offshore and hence will not remedy any domestic competition concerns.

135. The fact that the shareholders had commissioned the consultants report, which had looked at all these issues in detail lends credibility to the fact that the present option is the only realistic scenario.

136. Purchasing only Saldanha is highly unlikely to prove attractive to a third party purchaser. Firstly they would have to source supply from Sishen and it is highly unlikely that they would be able to procure supply on the same favourable terms as Iscor has, which it could only have obtained as a result of the window of opportunity afforded by the unbundling.

137. Without a competitive source of supply and given Saldanha’s as yet untested production process, it is hardly a jewel to be added to any steel makers crown. Nor would any firm want to enter into the domestic market to take on the dominant Iscor more favourably located next to the domestic customer base. The firm would then only enter to supply the international market and hence have no effect on our domestic market.

138. We conclude that it is unlikely that an alternative buyer for the firm would be found.

139. What distinguishes this case from the typical failing firm case, is that the buyer is not an arms length independent firm, but a 50% shareholder seeking to reduce the loss it has incurred to date. To prevent Iscor from protecting the investment it had made to date would be to discourage investment in new plant and innovation, which is what Iscor has done to date in Saldanha.

140. Iscor is not taking advantage of a competitors demise to secure a monopoly, the standard problem presented by a failing firm defence. Together with its partner the IDC it had embarked on a project to invest considerable capital in new and innovative productive capacity. The dream has run into trouble. It would not serve the cause of competition policy, which seeks to encourage firms to invest in plant and innovation, to apply the failing firm doctrine so rigidly that we inhibit such schemes, otherwise firms may become afraid to risk their capital.

141. Finally the transaction costs occasioned by Saldanha’s failure or exit either by liquidation or by temporarily mothballing would be excessive and exceed mere loss to the shareholders. Both Iscor and the IDC have absorbed debt by way of institutional loans to finance their respective stakes in Saldanha from both domestic and foreign firms. If the venture were allowed to fail this would according to McKinsey‘s report lead a re-rating of their credit status, increasing their costs of capital. Given that both firms continually raise capital in the market to finance development, the costs of such a failure would be felt throughout the economy.

142. Our conclusion is that the failing firm concerns outweigh the loss to potential competition that might otherwise be occasioned by the transaction and that accordingly the merger does not substantially lessen or prevent competition in the relevant market.

Public interest

143. If the merger is approved we have no evidence that this would be in conflict with any of the public interest grounds set out in section 12 A (3) . We have been informed that employment levels will remain unaffected by the transaction .

144. However if the merger is not approved this will have adverse public interest considerations. One of the factors that we have to look at in examining the public interest is the effect on the particular industrial sector or region.

145. There is evidence that the Saldanha Steel plant is a vital part of the town’s economic life. If the plant was to be shutdown or be mothballed for a period this would not only have a substantial impact on the employees of the plant who would be retrenched, but also on all the firms and individuals in the West coast region whose livelihoods are so dependent on the plants functioning.⁵⁹ The effect of Saldanha on the local economy is best expressed in a report by Wesgro ⁶⁰on the contribution that the project made to the gross regional product of the region during its construction phase – a figure of R 1,15 billion.

146. Were Saldanha a larger economic region the effect might be muted but given its dependency on a small number of industries the effect would be devastating. Saldanha Steel has to its credit been a good corporate citizen in the area and is involved in a number of important social programs that contribute to the upliftment of the area. Without the plant these projects would become less viable, if they survived at all.

147. The public interest thus strongly favours the approval of the merger.

Views of the Department of Trade and Industry

148. Government industrial policy has played and will continue to play a key part in the structure of our local steel industry. The Department of Trade and Industry has been involved with and has been a champion of the Saldana project for many years and for this reason we invited the DTI to make submission to us on the merger⁶¹. Regrettably these submissions were only received at the hearing itself in the form of oral testimony by Department representatives.

149. Nevertheless, the thrust of the DTI submission was that they were in favour of the merger, as it would enhance the competitiveness of Iscor internationally.

150. Iscor needs to compete in a global market against firms much larger than it. Unless it can expand its productive capacity it will falter in its attempts to secure international markets.

151. The DTI is also of the view that if Iscor can become a lower cost producer this will benefit the downstream steel industry in South Africa and thus Iscor’s customers will also become more competitive. The DTI also endorsed the public interest submissions made by the parties which we referred to earlier.

152. What is missing from the DTI’s approach and which we could not get a satisfactory answer for, was why Iscor should pass on these cost savings to its customers in the domestic market where it faces no serious local competition.

153. The fall in the rand means that Iscor can maintain its price at an import parity price and since it faces no domestic pressure retain the margin for itself. Whilst it’s true that our tariffs on steel imports are comparatively low at 5 % the authorities have not been reticent about imposing steep anti-dumping duties at the behest of Iscor. In 1999 the Board on Tariffs and Trade (the BTT) imposed anti-dumping duties of 81,7 % and 94,8%, respectively, on steel imported from Russia and the Ukraine.⁶²

154. The low figures of import penetration indicate that foreign competition does not impose a serious price constraint on Iscor. Within a few days of our order being handed down in this matter Iscor announced that it was to increase prices on its flat-steel products by 9,8% from May because, according to its Corporate Affairs Executive, Phaldie Kalam, the company’s flat-steel business has been “under severe pressure” because of dollar prices for steel being at a 30-year low. “In particular, margins at our Vanderbijlpark and Saldanha steel mills have been constrained,” he said.⁶³

155. Were Iscor faced by a strong domestic competitor it might be more constrained in the local market. We are concerned that downmarket firms may not see Iscor’s growth through the same rose tinted spectacles as the DTI, but that is not a subject that we need dwell on for the purpose of this decision for all the reasons given above.

Conclusion

156. We are of the view that the merger will have an anti-competitive effect both because of the removal of Saldanha as a potential competitor to Iscor and the vertical effects on DSP. The vertical problems can be cured by the conditions we have imposed. In respect of the horizontal effects, when we balance the loss of potential competition with the prospect of Saldanha failing we conclude that the merger will not substantially lessen or prevent competition. In addition, from a public interest perspective, we arrive at the same conclusion as the failure of the transaction would in all probability lead to a closure temporarily or permanent of the firm, and with that a devastating impact on the region.

157.The merger is approved subject to the conditions set out in the order attached hereto.

4 April 2002

N Manoim Date

Concurring: P.E. Maponya and M. Holden

---