---

COMPETITION TRIBUNAL OF SOUTH AFRICA

Case No: 23/AM/May10

In the matter between:

Bedrock Mining Support (Pty) Ltd …..........................................................................Applicant

and

The Competition Commission of South Africa …..................................................Respondent

In consideration of the intermediate merger between:

Bedrock Mining Support (Pty) Ltd …......................................................Primary acquiring firm

and

The Letaba, Numbi and De Kaap businesses of

Mondi Ltd …..............................................................................................Primary target firms

Panel : N Manoim (Presiding Member)

A Ndoni (Tribunal Member)

A Wessels (Tribunal Member)

Heard on : 19 and 20 July 2010

Order issued on : 21 July 2010

Reasons issued on : 21 October 2010

Reasons for Decision

CONDITIONAL APPROVAL

1. On 21 July 2010, the Competition Tribunal (“Tribunal”) in terms of section 16(2)(b) of the Competition Act, 1998 (Act No. 89 of 1998, as amended) (“the Act”) conditionally approved the acquisition by Bedrock Mining Support (Pty) Ltd of the Letaba, Numbi and De Kaap businesses of Mondi Ltd. The reasons for the conditional approval of the proposed acquisition of these businesses follow below.

THE PARTIES AND THEIR ACTIVITIES

2. The primary acquiring firm is Bedrock Mining Support (Pty) Ltd¹ (“Bedrock”), a company incorporated in accordance with the laws of the Republic of South Africa. Bedrock is owned by Zelpy 2245 (33.746%); Platau Investment Holdings (25.1%); the P Els Family Trust (16.873%); the C Els Family Trust (16.873%); and Zelpy 2245 (Nominee) (7.498%). Bedrock owns, apart from certain dormant firms², the following subsidiaries in South Africa:

• Makhado Sawmill (Pty) Ltd; and

• Emakhosini Sawmill (Pty) Ltd.

3. Bedrock’s main business activities comprise the manufacture and supply of underground elongates³ and support packs⁴ for deep level gold and platinum mines in the Free State, Gauteng, Limpopo and North West provinces. These products support the structure of the mining shafts in order for them to operate. Bedrock manufactures primarily timber-based mining supports but also makes use of steel and concrete in combination with timber. In addition to this it provides certain value added services to these mining customers through chemical treatment of products, logistics and underground technical services.

4. Pre-merger Bedrock does not own any hardwood plantations, but owns a number of timber sawmills of which those likely to be affected by this transaction are:⁵

• Glenthorpe (located at Barberton in the De Kaap area);

• Makhado (located at Louis Trichardt in the Letaba area);

• Rutlands⁶ (located at White River in the Numbi area); and

• Venus (located in the Numbi area). Third parties have alleged that the Venus mill is being closed down due to timber supply shortages in this geographic region (also see paragraph below).

5. The primary target businesses are the Letaba, Numbi and De Kaap businesses of Mondi Ltd (“Mondi”)⁷. Mondi has substantial interests in the South African forestry, pulp, paper and packaging sectors. The target businesses are all timber plantations containing mainly hardwoods i.e. Eucalyptus and are active in the production and sale of mining timber, pulpwood and agricultural poles. The three plantations are located in the Limpopo and Mpumalanga provinces and also sell timber into these geographic areas.⁸

6. Only two current customers of the target businesses use the sourced hardwood timber for the production of mining support products, namely Bedrock (the primary acquiring firm) and Reatile Timrite (Pty) Ltd (“Reatile”) (a competitor of Bedrock; see paragraph below). Other than with Reatile the target businesses do not premerger have any timber supply agreements.

THE TRANSACTION

7. In terms of the Sale of Business Agreement⁹, Bedrock will wholly acquire the Letaba, Numbi and De Kaap businesses from Mondi as going concerns. In purchasing the target businesses Bedrock will also acquire certain sale assets and the rights of Mondi arising on or after the effective date under the contracts relating to the businesses and specified in the sale agreement. Upon completion of the transaction the target businesses will become wholly owned Bedrock subsidiaries.

8. The proposed acquisition of the target businesses thus constitutes a merger in terms of section 12(1) of the Act. The merging parties meet the thresholds prescribed in terms of the Act for an ‘intermediate’ merger (also see paragraph below).

RATIONALE FOR THE TRANSACTION

9. The merging parties submit that the transaction is congruent with Bedrock’s vision to become a fully integrated timber based mine support company with its own fibre resources¹⁰ required to service world class mining companies on a sustainable basis. The acquisition of the target businesses is said to enhance the profile of Bedrock and allow it access to land and plantations that previously were the domain of major listed pulp companies.

10. According to Mondi it wishes to dispose of the target businesses as they have no strategic fit with the rest of its assets and operations which centre around the Richards Bay complex and Merebank (Durban) in KwaZulu-Natal.¹¹

BACKGROUND

11. During June 2009 Mondi issued a tender for the disposal of its Business Unit North businesses (“BU North”) which comprise the Letaba, Numbi, De Kaap, Satico and Sjonajona businesses. The intention of the tender was for BU North to be disposed of as a whole or in any combination – but in each case as a going concern. Bedrock was the successful bidder in the tender for the Letaba, Numbi and De Kaap businesses. Reatile was awarded the Satico plantation (also see paragraph Error: Reference source not found below) and Sappi Ltd the Sjonajona plantation.

12. As stated in paragraph above, the proposed transaction is classified in terms of the Act as an intermediate merger and therefore the merging parties notified the proposed deal to the Competition Commission (“Commission”) for adjudication.

13. During the Commission’s investigation, two third parties raised concerns in regard to the proposed transaction namely (i) Tshifhire Timbers (Pty) Ltd t/a Shefeera (“Shefeera”) and (ii) Reatile.

Shefeera

14. Shefeera, a BEE firm located at Makhado¹², is involved in the treatment and supply of poles. Although Shefeera is not a rival of Bedrock in the supply of mining support products, it purchases the same type of timber product as Bedrock for use in its said pole business. Its complaints in regard to the proposed deal concern Bedrock’s acquisition of the Letaba plantation and relate to input foreclosure and public interest¹³ given its said BEE status. Shefeera averred that it is dependent on the Letaba plantation for timber for its supply of poles to industry and was concerned that Bedrock post merger might cease supplying it and utilise all the Letaba timber for its mining support products business.

Reatile

15. As stated in paragraph Error: Reference source not found above, Reatile is a competitor of Bedrock in the supply of timber supports to the mining industry. Reatile’s complaints in regard to the proposed deal related to alleged:

1. Post merger input foreclosure of hardwood timber

Reatile had expressed concerns that once Bedrock controlled the Numbi plantation its access to hardwood timber supply would be impaired.¹⁴ These input foreclosure concerns are limited to the Numbi plantation; Reatile has confirmed that it has made alternative sourcing arrangements in regard to volumes of timber previously sourced from Mondi’s Letaba plantation.¹⁵ We therefore in regard to Reatile’s input foreclosure concerns do not below any further deal with the Letaba plantation.

2. Post merger coordination between it and Bedrock

Reatile’s concerns of post merger coordinated conduct between it and Bedrock as horizontal competitors in the supply of mining support products and services relate to Mondi’s cession and assignment to Bedrock of its rights and obligations under the ‘Restated Timer Supply- and Settlement Agreement’¹⁶ entered into between Mondi and Reatile in respect of the Numbi plantation (hereinafter referred to as the “Numbi Supply Agreement”). Post merger this timber supply agreement would involve Bedrock and Reatile engaging in price negotiations on the primary input for the manufacture of mining support products (i.e. hardwood timber) at least twice for the duration of the contract.

According to Reatile, post merger, this agreement would require it to provide competitively sensitive information to Bedrock in regard to not only its purchase prices for timber supply from the Numbi plantation but also its timber purchase prices paid to suppliers other than Bedrock, as well the timber volumes that Reatile purchases from the Numbi plantation and from suppliers other than Bedrock (also see paragraphs and Error: Reference source not found below).

16. The concerns of Shefeera and Reatile respectively are dealt with in detail below under the competitive assessment (in regard to Shefeera’s input foreclosure/ public interest concerns see paragraphs to and in regard to Reatile’s input foreclosure concerns see paragraphs to and paragraphs to in regard to the coordination concerns).

17. The Commission, given certain identified competition concerns, had proposed that Bedrock divest the Numbi plantation. The merging parties however were not willing to do so and hence the Commission on 21 April 2010 issued a Merger Prohibition Certificate.

18. Consequently the merging parties on 05 May 2010 requested the Tribunal in terms of section 16(1)(a) of the Act to consider the Commission’s prohibition. This appeal process is known as a re-consideration. A prehearing took place on 25 May 2010 and the hearing was scheduled for 19 to 28 July 2010. Both the Commission and the merging parties intended to call several witnesses to testify at this hearing, including economic experts.

19. On 25 June 2010 the merging parties however requested the Commission to consider a proposed behavioural remedy which the parties believed addressed the Commission’s concerns relating to potential post merger coordination (see paragraph above).¹⁷ In brief, this proposal involved Bedrock’s willingness to post merger forgo of the contractually stipulated price negation process between it and Reatile (in terms of the Numbi Supply Agreement) by having the timber price that Reatile would pay determined by an independent expert for the duration of the contract. However, the Commission on 13 July 2010 rejected the merging parties’ proposal stating that the offer “in its current form, does not address all concerns raised”.¹⁸

20. On the morning of 19 July 2010 (the first day of the hearing of this matter) counsel for the Commission however informed the Tribunal that the Commission is now considering the merging parties’ proposed behavioural remedy and that the Tribunal would be informed of any progress in this regard in the course of the hearing. The Tribunal thus commenced hearing the matter: the Commission and the merging parties respectively delivered their opening addresses and the Commission’s first factual witness, Mr John Peter Lewington (“Lewington”), the commodity manager of the Mining Procurement Division of Gold Fields Ltd, testified.

21. On the morning of 20 July 2010 the respective counsels of the Commission and merging parties advised the Tribunal that they had in principle reached agreement in regard to certain behavioural remedies and that they might be able to settle the matter to meet both the concerns of the Commission and the aforementioned third parties, Shefeera and Reatile (see paragraphs and above).

22. By midday 20 July 2010 the Commission and merging parties proposed a set of behavioural remedies which included timber supply agreements concluded between Bedrock and respectively Shefeera and Reatile. From the Commission’s perspective these agreements adequately addressed its initial competition concerns.

23. The hearing consequently proceeded on this (narrowed) basis. The Tribunal gave the representatives of respectively Shefeera and Reatile the opportunity to respond to the proposed remedies and requested them to specifically indicate whether or not the concluded timber supply agreements adequately addressed their competition/public interest concerns. Mr. Albert Mulaudzi (“Mulaudzi”), the managing director of Shefeera, responded on behalf of Shefeera and Mr. Len O’Haughey (“O’Haughey”), the managing director of Reatile, responded on behalf of Reatile. These responses are dealt with below under the competitive assessment.

VERTICAL MERGER

24. From the description of the merging parties’ activities in paragraphs to above, it is evident there is no horizontal overlap between the activities of Bedrock and the target businesses. The proposed deal however has a vertical dimension since Bedrock will secure ownership of the stated three hardwood timber plantations from which it currently purchases timber for the manufacture of mining support solutions.

RELEVANT MARKETS

25. Since the Commission and the merging parties did not call any further witnesses after agreeing to a set of proposed remedies (see paragraphs to above), we shall in this matter not definitively conclude on the parameters of relevant markets in question. We shall however below describe and elaborate on the potential relevant markets to contextualise the potential theories of harm and our assessment of the proposed set of behavioural remedies.

26. We summarise below the relevant upstream and downstream product and geographic markets as identified by respectively the Commission¹⁹ and merging parties²⁰:

Upstream market: hardwood timber supply

27. It is common cause between the Commission and the merging parties that the upstream market is confined to the supply of hardwood²¹ timber. They however disagree on the narrower delineation of this product market.

28. The merging parties in their competitiveness report submit that the upstream market is the market for the supply of hardwood in Mpumalanga and Limpopo (the so-called Northern Regions). In regard to the geographic scope of this market the merging parties state “[t]he key elements of financial success in plantation forestry are growth rate and transportation costs to the market. In order to be successful it is important to acquire productive land close to the market”²² (also see paragraphs to below).

29. The Commission was of the view that the merging parties’ product market delineation is overly broad considering three aspects when observing the supply volume of hardwood suitable for mining timber usage, namely:

1. The portion of the total tree volume suitable for mining timber purposes and the required diameter of the timber.

The Commission’s market investigation found that not all parts of a hardwood tree are suitable for the production of timber-based mining support products. The Commission namely found that only hardwood with a diameter of between 9 and 26 cm²³ is suitable for the manufacture of mining timber and further that on average only approximately 65% of a hardwood tree can be usefully applied to produce timber mining supports - the balance of the tree being sold off as pulp and building and fencing poles.

2. Certain long term timber supply contracts and the internal supply of timber.

The Commission’s market investigation led to an understanding that certain long term contractual supply agreements effectively lock a potential source of hardwood timber out of the ‘open market’. One example of this is Mondi’s current supply contract with Reatile i.e. the Numbi Supply Agreement (see paragraph above). The Commission concluded that the above volumes as well as the tonnages of hardwood used exclusively for plantation owners’ internal supply purposes, for example Sappi’s supply from its hardwood plantations to its downstream paper production operations, are not contestable and therefore should be excluded from the relevant upstream market.

3. The existence of so-called ‘swap’ agreements.

The Commission’s market investigation identified certain current swap agreements between plantation owners/hardwood customers. These agreements in essence allow the parties involved to obtain the relevant parts of a tree for their respective businesses and purport mainly to reduce timber transport costs. A three-way timber swap agreement for example exists between Bedrock, Mondi and Sappi, ensuring that the Bedrock sawmills have certain quantities of mining timber in exchange for pulp timber.²⁴ The Commission concluded that the timber supply locked into these agreements is highly unlikely to be contested away by a third party and therefore must be excluded from the timber volumes that are available on the open market.

30. The above implies that one must first determine the total standing timber (i.e. planted hardwood hectares) and then determine the usable mining proportion that is contestable. We shall refer to this product market as defined by the Commission as the market for ‘contestable hardwood supply usable for mining support products’.

31. The Commission defined the relevant geographic market as the Limpopo province, Mpumalanga North and Mpumalanga North/South including the Piggs Peak area in the Northern tip of Swaziland.

Downstream market: underground mining support products

32. The merging parties in their competitiveness report defined the downstream market as the manufacturing and supply of underground mining support for deep level mines, including pure timber products; stope support products made from concrete and steel; and so-called composite products made of a combination of timber/concrete²⁵ or timber/steel, in the Limpopo, Free State, Gauteng and North West provinces.

33. The Commission defined this product market as “products that provide stand up support whether it be timber based or composite good including timber and either steel or cement, or just timber, steel or cement”²⁶ and defined the relevant geographic market as “multi-regional” comprising “only those provinces as listed by the merging parties” i.e. the Limpopo, Free State, Gauteng and North West provinces.

34. As stated in paragraph above, we shall not definitively conclude on the exact parameters of the relevant markets, save to make the following observations for future reference and to contextualise our latter competitive analysis:

Upstream market

35. In regard to the upstream market, we note that the Commission’s report and the merging parties’ competitiveness report do not explicitly deal with hardwood sawmilling²⁷ in a market delineation context, but conflates it into either the above-mentioned upstream or downstream markets. Be that as it may, hardwood sawmilling is an integral, unavoidable stage in the mining timber supply chain and must be taken into account in the overall assessment of the potential competitive effects of the proposed deal. In particular: the relevant stage of supply under assessment i.e. whether (i) hardwood supply for sawmilling (i.e. saw logs) or (ii) milled hardwood as input into the downstream production of mining support products, critically affects the parameters of the relevant geographic market, as explained below.

36. The submissions of various market participants suggest that a sawmill must be located relatively close to the plantation supplying it with logs in order to be economically viable, but that mining support customers can be viably supplied at comparatively further distances from the sawmill i.e. within a comparatively larger geographic area. Reatile for example submits that in order for a mining timber sawmill to be viable it must be located within [0-50] kilometres of a plantation.²⁸ The Commission in its conflated analysis considers the available hardwood in a radius of 100 to 150 kilometres from a specific plantation.

37. The available evidence further suggests that mining support producers either own their own sawmills (for example Bedrock and Reatile); enter into toll manufacturing agreements with sawmills; or source mining timber from third party sawmills. It is evident that although Bedrock and Reatile²⁹ are vertically integrated in sawmilling both these firms have in the past on a partial basis sourced timber products form third party sawmills.³⁰ For example, Bedrock makes use of the Scotmar and Spectrum sawmills³¹ and the Paul Pietersburg³² sawmill.³³ Producers of mining support products that are not vertically integrated into sawmilling such as Duraset³⁴ and Strocam Mining on a more extensive basis source mining timber from third party sawmills.

38. In conclusion: the relevant stage of the supply chain being assessed significantly affects the geographic scope of the relevant market. Hardwood sawmilling may constitute a separate relevant market given inter alia that certain providers of hardwood sawmilling services provide these services exclusively and are not vertically integrated into either plantations (unlike Bedrock post merger) or the downstream production of mining support products. Furthermore, the ability to potentially relocate sawmills and to source mining timber from third party sawmills may affect the assessment of any potential foreclosure effects (also see paragraphs and Error: Reference source not found below).

Downstream market

39. We provide below a synopsis of the available evidence in regard to the substitutability of various mining support solutions. This is of relevance to the competition analysis of both potential input foreclosure³⁵ and coordinated effects. The main issue here is if respectively pure steel-based and pure cement-based support products are substitutable with timber-based products or composite products form a pricing perspective.

40. The price-related evidence at our disposal indicates that it is highly unlikely, from a pricing perspective, that pure steel mining support solutions are substitutable with either pure timber solutions or possibly with composite products. In fact, the submissions of mining customers confirm that pure timber solutions are the most cost effective mining support. Lewington in this regard testified that “you are going to look at about a 35% increase from just a normal timber elongate to an engineered timber and steel elongate and from an engineered timber steel elongate to an engineered steel, predominantly steel prop is going to be in the region of a good 80% to 90% increase”. He further testified that, depending on the product height, a pure steel product could be as much as 120% more expensive than a pure timber product.

41. Furthermore, the substitutability between pure cement solutions and pure timber/composite solutions is also contested from a pricing perspective: Lewington submitted that the cement pack supplied by Duraset is “quite cost prohibitive” and AngloGold Ashanti Ltd (“AngloGold”) submitted that steel and cementitious products “are extremely costly and have never been considered a practical or alternative solution in totality due to the high costs”.³⁶

42. From a pricing perspective we therefore conclude that it is highly unlikely that respectively pure steel-based and pure cement-based mining support products are substitutable with either timber-based products or composites (i.e. mixtures of timber/steel or timber/cement). We therefore below confine our analysis of the competition issues to timber-based mining support products.

COMPETITION ANALYSIS

43. Two main theories of harm emanate from the above-mentioned concerns raised by third parties (see paragraphs and above) and the Commission’s findings, namely post merger (i) input foreclosure of hardwood timber to users thereof in downstream production processes; and (ii) horizontal coordination between Bedrock and Reatile (in the context of the Numbi Supply Agreement, see paragraph above) in the market for timber-based mining support products and services. We shall discuss below each of these theories in turn as well as the proposed behavioural remedies in regard to each.

1. Input foreclosure

44. The key issue raised by the Commission in an input foreclosure context is that Bedrock’s competitors in the supply of mining support products may post merger not have access to an essential input i.e. hardwood timber. The two main competitors in this market, Bedrock and Reatile, appear to have the same strategy, namely to vertically integrate into timber production (see paragraph above). As stated in paragraph above, Reatile raised this concern in relation to the Numbi plantation.

45. The merging parties acknowledge that the target businesses’ current hardwood timber customers that use this timber for purposes other than mining timber post merger may have to look to other sources for their requirements.³⁷ Shefeera is one such customer and, although it is not a direct competitor of Bedrock, it in regard to the Letaba plantation raised input foreclosure/public interest concerns regarding its pole manufacturing activities (see paragraph above).

46. We note that no third party raised concerns regarding the sale of the De Kaap plantation. We therefore below first deal with Shefeera’s concern regarding the Letaba plantation, followed by Reatile’s concern regarding the Numbi plantation.

Shefeera – Letaba plantation

47. As background to Shefeera’s BEE status, it is important to provide a brief history to its establishment: the business was originally a division of the Mondi Group and sold in 2003 to management, after which it was sold in 2009 to a consortium of BEE partners. Shefeera stated that the sale to management in 2003 was premised on an agreement that Mondi will continue to supply raw materials to it in order to ensure the survival of the new company and furthermore that the BEE consortium buyout in 2009 had taken this prior arrangement into account.³⁸

48. Shefeera further submitted that it has been purchasing hardwood timber from Mondi’s Letaba plantation on an ad hoc basis for a number of years. Mondi submits that this supply has been in the order of [0 – 10 000] tons per annum in the period 2007 to 2010.

49. Shefeera further alleged that a post merger lack of supply from Bedrock could seriously threaten its survival as BEE firm – a survival that, as explained above, Shefeera claims ought to be Mondi’s concern as seller of the Letaba plantation given the historic parental relationship between these firms and the fact that Shefeera’s BEE ownership structure was a Mondi initiative. Shefeera claims that therefore it should be the primary beneficiary of the Letaba timber supply source.

50. However, we shall not deal with Shefeera’s concerns in any further detail since Mulaudzi confirmed that Shefeera on 20 July 2010 signed a timber supply agreement with Bedrock which satisfactorily addresses Shefeera’s concerns.³⁹

51. The Tribunal has made adherence by Bedrock to this timber supply agreement with Shefeera a condition of its approval of this transaction. A copy of this agreement entered into on 20 July 2010 is attached to these reasons as Annexure B1. In our view Shefeera is in a position to effectively monitor Bedrock’s compliance with this condition of the merger.

52. Furthermore this supply agreement affords Shefeera time to secure alternative sources of hardwood timber within an economic distance of its production facilities on termination of the said agreement with Bedrock. Premerger Shefeera procures the balance of its timber requirements from third parties including small timber plantations in the relevant area and it shall post merger continue to do so. According to the Commission’s report the suppliers of hardwood timber in the larger Letaba region⁴⁰ other than Mondi include NCT Forestry Co-operative Ltd, Northern Timbers (Hans Merensky), Tzaleba Timbers and numerous small farmers (see footnote 50 below).⁴¹

Reatile – Numbi plantation

53. In regard to Reatile’s concern of input foreclosure of timber volumes from the Numbi plantation, the Commission interestingly enough in its merger prohibition did not find ostensible post merger input foreclosure a likely outcome of this transaction. According to the Commission’s report Mondi’s Numbi plantation has a market share of approximately [20-30]% based on current contestable mining timber in the larger Numbi area⁴².⁴³

54. Bedrock submitted that it would post merger release a significant quantum of timber in the open market, namely approximately [0 - 100 000] tons (excluding the 60 000⁴⁴ tons supplied to Reatile in terms of the Numbi Supply Agreement). It further submitted that it will stop seeking to bid for/contest such volumes in the spot market in future.⁴⁵

55. The Commission was of the view that Bedrock and Reatile post merger are likely to “swop positions” in terms of their respective sources of hardwood timber supply since Reatile would be forced to source less hardwood timber from a single large supplier and more from smaller suppliers in the open market. The Commission assumed a post merger direct reversal of roles of these two competitors in regard to their timber procurement which it was of the view that Reatile “is in a position to cope with”.

56. The extent to which this assumption of the Commission in regard to a direct reversal of roles holds true depends inter alia on the intentions and ability of Bedrock to post merger expand its mining support business. We unfortunately have very scant information in regard to Bedrock’s future expansion plans, save for its statement that “at this stage” it “does not have plans to expand its current capacity post merger”.⁴⁶ However, Bedrock in regard to its Venus sawmill (which appears to be within geographic scope of the Numbi plantation, see paragraph Error: Reference source not found above) makes the noteworthy observation that it “... cannot say with certainty whether or not the sawmill will be fully operational post merger”.⁴⁷

57. The Tribunal is highly sceptical of the assumption that Bedrock will not, post merger, use the acquired Numbi plantation to expand its mining support business and therefore disagree with the notion of a post merger direct reversal of roles between Bedrock and Reatile in regard to their timber procurement.

58. Turning to Reatile’s views in regard to the Numbi plantation, O’Haughey confirmed that he, on behalf of Reatile, on 20 July 2010 signed a timber supply agreement with Bedrock⁴⁸ and that Reatile is satisfied with the terms thereof.⁴⁹ Thus, Reatile is satisfied that as a result of the contractually secured timber volumes from Bedrock it would not be significantly foreclosed by the instant deal in the short to medium term.

59. The Tribunal has made adherence by Bedrock to the supply agreement with Reatile a condition of its approval of this transaction. A copy of the agreement entered into between Bedrock and Reatile on 20 July 2010 is attached to these reasons as Annexure A1. In our view Reatile is in a position to effectively monitor Bedrock’s compliance with this condition.

60. O’Haughey further confirmed that said timber supply agreements give Reatile “a further period whereby we will be able to buy timber in the Numbi area”; that Reatile will continue to buy form private growers in the area, which together with the concluded agreement enables it “to operate in the Numbi area for 6 years in total”; that Reatile “would have to look at expanding sources, either from the Swaziland operation or in an area away from Numbi”; and that “[i]t will be challenging, but I’m confident that we will manage to do that.”

61. In regard to Reatile’s fate when these supply agreements with Bedrock expire, we note that Reatile will have another six years to explore any combination of a number of options, including:

1. the negotiation of supply agreements with private timber growers in the Numbi region - the bulk of the contestable mining timber in the larger Numbi area, apart from the Numbi plantation that is the subject matter of this merger, is made up of individual farmers previously selling collectively through NCT Forestry Co-operative⁵⁰;

2. moving one or more of its sawmills closer to an alternative timber source - this is not an uncommon phenomenon in this industry and Reatile has previously done so⁵¹;

3. the optimisation of volumes at Satico⁵² (see paragraph above);

4. expanding capacity at any other mill within viable distance of the mining customers where timber supply is more favourable; and

5. an innovative procurement policy such as making use of swap agreements (see paragraph above) could further assist in obtaining mining timber from further afield.

62. Furthermore, any potential foreclosure effects are diluted given that the timber volumes that Reatile sources from the Numbi plantation comprise a limited proportion of Reatile’s total timber requirements (see paragraph below).

2. Horizontal tacit coordination

The Commission’s findings

63. The Commission prohibited the proposed deal on the basis of it, in the context of the Numbi Supply Agreement, promoting coordinated conduct between the two most effective rivals, Bedrock and Reatile, in the market for the manufacturing and supply of mining support products and services. The Commission was of the view that the proposed deal would create a commercial supplier/customer relationship between said two competitors which did not exist pre-merger and which is likely to result in the exchange of commercially sensitive information through, in particular, the negotiation of timer prices – timber being the primary input in said downstream production process. The Commission, in particular, was concerned that these post merger price negotiations could be a forum for coordination and that this commercial interaction may act as a monitoring device for Bedrock.

Merging parties’ views

64. The merging parties had disputed this theory of harm alleging that the Commission failed to take into account a range of features of the relevant market that make post merger collusion improbable.⁵³ They argued that the Mondi - Kohler case cited by the Commission had far greater conformity to economic theory on coordinated effects since the deal inter alia enhanced two-way information flows on the pricing of upstream products in a market that was found to be conducive to collusion.⁵⁴

Background to relevant market

65. As background to the coordinated effects analysis we below sketch the basic features of the market for the production and supply of mining support products and services before providing a more in-depth analysis of certain key characteristics of this market (see paragraphs to Error: Reference source not found below).

66. The first issue of import in a coordinated conduct context is that Bedrock and Reatile appear to have symmetrical market positions in a highly concentrated relevant market. According to the Commission’s assessment the combined market share of Bedrock and Reatile in the said relevant market is at minimum [70-80]%, depending on the precise market parameters (see paragraphs to and to above in regard to market definition).⁵⁵

67. The market in question furthermore is characterised by seemingly high barriers to entry. Apart from the obvious need for mining timber as input, market participants and customers alike have highlighted as key constraints the technical aspects of mining support and customers’ requisite of suppliers having a reputation as reliable and full service providers. Regarding the technical aspects, the merging parties themselves have submitted that the legal requirements of the Mine Health and Safety Act and technical and quality requirements have affected the ability of small players to survive in this market.⁵⁶ Northam Platinum Ltd, a customer in the relevant market, points to the fact that it has “a production environment and [is] governed by strict safety regulations” and “cannot afford poor service”.⁵⁷

68. Furthermore, customers have stressed that the suppliers in the relevant market must provide full solutions to complex realities. Lewington confirmed this during his testimony: he stressed the importance of having a reputable supplier of mining support products, stating that factors such as required volumes, quality assurance programs and especially standards and sustainability of supply were imperative in the selection of a suitable supplier; he furthermore confirmed that “price would most definitely not be the only consideration”. He also provided a lucid perspective on the importance of security of mining support products supply in relation to potential mining production losses, or put differently, the risk profile associated with (unreliable) mining support:

“Our [Gold Fields’s] annual timber contract is … about R180 million. Now consider that if we cannot blast just on the west bit operation of Gold Fields and that is Kloof and Driefontein operations, we are most probably standing at R80 million a day of loss if we cannot blast. So in one day you nearly lose half of the total contract value, so if you are not circumspect in your view on this and you do give the contract away to a person that can jeopardise supply, you stand, in one day you stand to lose far more than what you could possibly make up by going to a cheaper alternative.”

69. Entry barriers may furthermore be enhanced by this transaction since potential future entrants may find it difficult to effectively compete with Bedrock if they are not similarly vertically integrated.⁵⁸ As stated in paragraph Error: Reference source not found above, we cannot simply assume that this transaction will result in a direct reversal of roles between Bedrock and Reatile in regard to timber procurement. Furthermore, the available evidence suggests that the current smaller niche players in the mining support market are not effective competitors of Bedrock and Reatile which suggests that scale may be a requirement for effective entry given mining customers’ needs (see paragraph below). We furthermore reiterate that non-timber based mining support products appear not to be price competitive with timber-based products (see paragraphs to above).

70. Regarding the players of various size in the relevant market Lewington stated that “Bedrock and Reatile are currently the only two reputable firms” capable of supplying “every need” of Gold Fields. He was of the view that competitors such as Strocam⁵⁹ and Duraset⁶⁰ were niche players in the mining support market only catering for “specific application” i.e. a specific mining condition or activity. He also confirmed that the latter firms have not approached Gold Fields with “total solutions”, i.e. with a full suite of mining support products, and that they provide a very limited service component. AngloGold echoed this view stating that the “specialist” producers cannot (i) provide the required quantities of product; (ii) consistently provide security of supply; (iii) ensure consistent quality; and (iv) supply products at competitive prices due to limited resources or specialist practices.⁶¹

Proposed remedy in regard to information exchange

Background

71. To contextualise the potential extent of post merger information exchange between Bedrock and Reatile, we note that the Numbi Supply Agreement (see paragraph above) makes provision for a comprehensive price-setting mechanism, inter alia for the calculation of the timber price by reference to “the market prices paid for similar products, ... and such other factors as may be relevant”.

72. O’Haughey, based on the historic price negotiation process between Reatile and Mondi in terms of the Numbi Supply Agreement, envisioned that the post merger price negotiation process would involve putting forward “what they [Bedrock] believe are the key issues in determining what the market price is for mining timber in that area and we [Reatile] would put forward views on what happened with other timber products, like what’s happened in the pulp industry, what the overall supply and demand is and then what the mining timber information is in the Numbi area.”

Commission and merging parties’ proposal

73. To address the above-mentioned issue of potential information exchange the Commission and the merging parties, as part of a proposed set of behavioural remedies, suggested that any post merger price negotiations between Bedrock and Reatile (in terms of the Numbi Supply Agreement and the additional supply agreement entered into on 20 July 2010) should be replaced by a mechanism of price determination by an independent expert, such expert preferably having both qualifications and experience in the field of forestry and finance.

The Tribunal’s assessment

74. The immediate concern with the above-mentioned post merger commercial supplier/customer agreement between Bedrock and Reatile is the creation of a platform or forum for the exchange of competitively sensitive information between the two main competitors in the mining support market. Coordination of commercial decisions may ultimately lead to anticompetitive outcomes such as higher prices than would prevail absent the interaction/merger. Such a forum theoretically could create the potential for either (i) explicit communication i.e. overt or express collusion (price fixing) or (ii) tacit (oligopolistic) coordination if the firms in question recognise the competitive interaction between themselves and adjust their mode of competitive behaviour accordingly.

75. Foremost we were concerned with the extent to which the commercial interaction between said parties could create a forum for two-way information flows on the key dimensions of competition (for example prices, volumes and input costs) which could provide an effective coordination mechanism. Therefore, in regard to the Commission and merging parties’ proposal regarding the price determination of timber supply by an expert (see paragraph above), the Tribunal indicated its concerns in respect to the lack of expressed terms in the suggested remedy effectively eliminating potential information flows between Bedrock and Reatile via the said expert.

76. After enhancement by the Tribunal the final set of imposed behavioural conditions:

1. eliminates the need for direct contact between Bedrock and Reatile and therefore addresses any concerns that the proposed merger through the post merger contractual supplier/customer relationship between these firms may create a platform for outright collusion; and

2. effectively restricts both the quantum and the quality of potential information exchange between Bedrock and Reatile, thus effectively preventing the flow of enhanced strategic and commercially valuable information. More specifically, it limits the degree of transparency to a one-way information conduit via the said expert - only the mining timber volumes that Reatile would purchase from Bedrock’s Numbi plantation and the final input price that Reatile would pay for this timber supply would be known to Bedrock via the expert.

77. The Tribunal also felt that the expert upon appointment should provide appropriate undertakings to the Commission regarding confidentiality and the prevention of any representations and information from Reatile being disclosed to Bedrock and vice versa. These enhancements form part of the Tribunal’s order (see paragraph below).

78. Since there is no evidence before the Tribunal in regard to existing premerger coordination in the market for the production of mining support products and services, and therefore of any strengthening or enhancing of such existing coordination through the proposed vertical merger, we shall below limit our assessment to the likelihood of coordination becoming probable, in the context of the aforementioned imposed behavioural remedy of an expert, as a result of this merger.⁶²

79. Despite the imposed behavioural condition, post merger Bedrock will still have the following information about Reatile’s business; it will know how much timber it is purchasing from Numbi and at what (input) price. However we do not believe that this will establish or enhance co-ordination for the reasons given below.

80. The relevant market in question clearly portrays some features that are considered conducive to likely tacit collusion inter alia high levels of concentration and seemingly high barriers to entry (see paragraphs to above), although any perceived enhancement of entry barriers as a result of this merger in our view does not significantly impact certain other key market features (see paragraphs to Error: Reference source not found below). Furthermore the outcomes of bidding or tender processes, i.e. so-called win/loss data, are known to the incumbent firms. We have also dealt with the post merger commercial relationship that will exist between Bedrock and Reatile and the imposed behavioural remedy to limit the scope of information exchange between these firms (see paragraphs to above).

81. We shall focus our analysis below on a closer scrutiny of certain other key conditions of the market under consideration.

Heterogeneous products

82. The available evidence indicates that the relevant market is by no means homogeneous but rather highly differentiated by performance and design given that mining support suppliers will provide different layout options for the required support in a mine which usually include a mix of products. Underground mining customers are ultimately concerned with their costs per metre mined and thus require a variety of products to address a number of required support solutions. A mining solutions supplier therefore must have the ability to recommend and supply products using the correct technical approach in the most cost efficient manner.

83. Lewington during his testimony explained that a number of factors determine the support system that would apply in a specific mining condition inter alia mining depth, seismicity and scoping width. He further explained that elongates, although more cost effective from an installation perspective, have limited yield ability compared to packs, especially compared to composite packs. A mining customer therefore at any one time requires a mix of support solutions including both elongates and packs.

84. The extent of product variety is further evident from the product offerings of both Bedrock and Reatile. For example: Bedrock provides the following mining support products: (i) brick/timber composite packs; (ii) timber chocks; (iii) matpacks (various configurations); (iv) timber composites; (v) saturn props (including pre stressing device); (vi) poles (various diameters and lengths); (vii) pencil props (various lengths and diameters); and (viii) terramacs (including pre stressing device). Reatile provides a comparable full range of support solutions. The available evidence furthermore shows that the price ranges of the various mining support product categories vary significantly. The prices of the above-mentioned products of Bedrock for example range from below R[0 – 10 000] to more than R[0 - 10 000] per cubic metre.⁶³

85. At the request of the Tribunal, mining customers submitted detailed tender data which further confirm the large number, mix and heterogeneous nature of the mining support products involved: for example, the tender data of Gold Fields involve circa 80 individual mining support products; AngloGold confirmed that it requires a mining support product range consisting of some 13 individual product categories⁶⁴; and an Impala Rustenburg mining support products contract involved more than 40 products consisting of mats, planks, wedges and wedgeprops of various lengths/diameters.⁶⁵

86. Furthermore, Lewington testified that the prices quoted for said individual products supplied to Gold Fields were inclusive of the required service component i.e. the service component cost is implicit in the individual product prices. These support services may include mine yard management, chemical treatment, logistics and underground technical services. This value added or service component also affects the ability of players to accurately assess a competitor’s costs based purely on timber-related input costs. Moreover, as stated in paragraph above, Lewington confirmed that from a customer perspective price is not the only consideration. Economic theory predicts that in industries where product or service quality is an important element of competition, tacit coordination becomes more difficult to reach and sustain since competition takes place on more than one parameter.

87. Differentiation on the above scale that extends to a vast array of products for different layout solutions combined with a significant service element, in our view to a large extent hinders successful tacit collusion. This is due to the difficulty and complexity in obtaining an accurate depiction of final layout costs and thus in reaching agreement on a cartelised final price and/or other supply conditions.⁶⁶

Lack of transparency of final prices and volumes

88. No evidence exists in the instant case of price transparency in the relevant market in regard to final selling prices. The merging parties have indicated that the incumbents do not use price lists and that there are no standard quoted prices. As concluded above, the relevant product market is essentially a complex non-homogeneous market involving both a large number of products and a significant service element. Furthermore, mining timber is used in different mixes as input in final mining support products which complicates the underlying cost structures of the final goods. Moreover, the supply of a total mining solution including a significant service element implies additional calculated margins to the base price of the individual products.

89. In the above regard O’Haughey opined that even though the price of timber “is the biggest single cost element and that cost element would be a significant item that would be driven in your overall price adjustment application you would make from a mine”, Bedrock post merger would not be able to determine from these input prices “a specific price for a specific product, because there would be different profit margins and other costs”.

90. From an impending coordination perspective it is important to relate the post merger quantum of timber supply by Bedrock to Reatile (in terms of the concluded supply agreements⁶⁷) to Reatile’s total timber requirements for the downstream production of mining support products. The post merger interaction between Bedrock and Reatile (through the expert determination of timber prices) would be limited to a maximum of 60 000⁶⁸ tons of timber supply per annum for no more than six⁶⁹ years. Moreover, this quantum of timber supply reduces significantly in years 5 and 6 of said concluded timber supply agreements. This supply volume constitutes a relatively small portion of Reatile’s total annual mining timber requirements, namely less than [0-30]% thereof.⁷⁰ Given this limited supply volume it would be difficult for Bedrock to determine Reatile’s final (output) prices to mining support customers on the basis of knowing only the input price of the mining timber via said expert.

91. Given the aforementioned imposed behavioural remedy and based on the available evidence, we conclude that Bedrock is unlikely to through this merger gain meaningful insight into the final (output) prices that Reatile would charge to its mining customers. One of the main premises for this conclusion is the above-mentioned confined, decidedly one-way, information exchange.

92. Therefore we conclude that in the context of the imposed behavioural remedy, the proposed deal is unlikely to lead to significantly enhanced transparency in the relevant market and the ability of competitors to agree (absent any direct communications) on collusive final output/selling prices or final volumes.⁷¹

Conclusion

93. Given (i) the one-way conduit of limited information via an independent expert in terms of said supply agreements between Bedrock and Reatile, and (ii) certain key relevant market conditions, considered collectively and which are not altered by the proposed transaction, we are of the view that successful post merger tacit collusion in the relevant market as a result of the proposed deal is unlikely.

94. In conclusion, given the behavioural remedy that prevents significant information sharing between Bedrock and Reatile as well as a lack of further compelling evidence that the market in question post merger on balance resonates with the economics regarding likely coordinative conduct, we conclude that a likely coordinated effects hypothesis - in this case - lacks substance. This assuages any coordination effects concerns as a result of this deal and therefore there is no need for us to assess this issue any further.

PUBLIC INTEREST

95. The concerns raised by BEE firm Shefeera have already been dealt with above, see paragraphs to .

96. In regard to the effect of the proposed deal on employment, the merging parties submitted that the target businesses will be sold as going concerns and that the employees will be transferred in terms of section 197 of the Labour Relations Act, 66 of 1995. The merging parties have further confirmed that no retrenchments are envisaged as a result of the proposed transaction.

97. No other public interest concerns arise from the proposed deal.

CONCLUSION

98. As stated above the Commission, after an initial prohibition of this proposed ‘intermediate’ merger, during the Tribunal’s re-consideration process proposed a set of behavioural remedies to address certain input foreclosure/public interest and coordination concerns. After certain further suggestions from the Tribunal and enhancement of the proposed set of remedies, terms were proposed by the Commission which the Tribunal found acceptable in the context of the transaction under consideration. This set of conditions form part of the Tribunal’s order.

99. The imposed behavioural conditions involve inter alia adherence by Bedrock to two commercial timber supply contracts: first, a supply agreement entered into between Bedrock and Shefeera on 20 July 2010 which addresses Shefeera’s input foreclosure/public interest concerns and, second, a supply agreement entered into between Bedrock and Reatile on 20 July 2010 (in addition to the Numbi Supply Agreement (see paragraph above)) which addresses Reatile’s input foreclosure and coordination concerns.

100. The conditions further provide for a post merger price setting mechanism designed to eliminate any significant information exchange between Bedrock and Reatile as a result of the aforementioned supply agreements. This mechanism provides for an independent expert to determine Bedrock’s timber supply prices to Reatile for the duration of said agreements. More specifically, Bedrock and Reatile must disclose their pricing negotiation information to an expert only and not to the other party. The independent expert is also precluded from disclosing any information submitted to him/her to the other party or any third party.

101. Therefore, further to the applicant’s request to the Tribunal for consideration of this intermediate merger in terms of section 16(1)(a) of the Act, the Tribunal approved the proposed deal subject to the following conditions:

1. The merging parties must comply with their contractual obligations in terms of the agreements concluded between Bedrock and Reatile, and Bedrock and Shefeera respectively, on 20 July 2010. These contractual obligations are set out in the attached agreements marked respectively as Annexure A1: ‘Timber supply agreement entered into by and between Bedrock and Reatile’, and Annexure B1: ‘Timber supply agreement entered into by and between Bedrock and Shefeera’.⁷²

2. The expert envisaged in clauses 6.2 and 8 of the supply agreement between Bedrock and Reatile must keep all information submitted by Bedrock confidential and not disclose it to Reatile or any third party, and must keep all information submitted by Reatile confidential and not disclose it to Bedrock or any third party. The said expert must upon appointment give undertakings to the Commission to this effect.

____________________ 21 October 2010

A Wessels Date

N Manoim and A Ndoni concurring

Tribunal Researcher : Mahashane Shabangu

For the merging parties : Adv F Snyckers and Adv M Le Roux instructed

by Bowman Gilfillan

For the Commission : Adv V Maleka and Adv G Ngcangisa instructed by the State Attorney

---