SAFLII [Home] [Databases] [WorldLII] [Search] [Feedback]

South Africa: Competition Tribunal

You are here:  SAFLII >> Databases >> South Africa: Competition Tribunal >> 2005 >> [2005] ZACT 29

[Database Search] [Name Search] [Recent Decisions] [Noteup] [Help]

HARMONY GOLD MINING COMPANY LIMITED and GOLD FIELDS LIMITED (reasons) (93/LM/Nov04) [2005] ZACT 29 (18 May 2005)

.RTF of original document

COMPETITION TRIBUNAL
REPUBLIC OF SOUTH AFRICA

                                                              
Case no: 93/LM/NOV04


In the large merger between:

HARMONY GOLD MINING COMPANY LIMITED      Primary Acquiring Firm

and

GOLD FIELDS LIMITED                                          Primary Target Firm



Reasons for decision


Introduction
1.      
On 10 May 2005 the Tribunal approved the hostile take-over by Harmony Gold Mining Company Limited of Gold Fields Limited subject to the following conditions:

1.      
The following limitations shall be placed on the retrenchments at the merged entity –

a)      
There shall be no retrenchments of employees at the merged entity below the level of Patterson grade C or equivalent as a result of the merger;
b)      
The merged entity may retrench up to a maximum of 1000 employees at or above the level of Patterson grade C or equivalent as a result of the merger.

2.      
For purposes of paragraph 1 –

a)      
‘merged entity’ means Harmony Gold Mining Company Limited and its subsidiaries including Gold Fields Limited; and
b)      
‘employees’ includes contract labour.

3.      
The undertaking in paragraph 1 shall apply for a period of 24 months from the date of the Competition Tribunal order.

4.      
The Competition Commission must monitor the above conditions.

5.      
In order for the Competition Commission to properly monitor the above conditions the merged entity must adhere to the following procedures:

a.      
Provide the Commission with quarterly reports regarding the effects of the proposed transaction on employment.

b.      
Each report must include the following information:
(i.)    
the current levels of employment, per job category, at the merged entity;
(ii.)   
the number of actual retrenchments per job category in the quarter reported on;
(iii.)  
the reasons, per job category, for retrenchments;
(iv.)   
the number of planned further retrenchments per job category;
(v.)    
the status of further retrenchments; and
(vi.)   
the process upon which the retrenchments will take or have taken place.

c.      
The quarterly reports must be submitted for the period of this order. Should the retrenchment process in the merged entity, as a result of the proposed transaction, not be finalized within the period of this order, the merged entity shall be obliged to submit further quarterly reports until the entire retrenchment process has been finalized.

d.      
The quarterly reports must be submitted to the Competition Commission no later than one calendar month following the end of each quarter.


Transaction Background

6.      
In October 2004 Harmony Gold Mining Company Ltd (“Harmony”) launched a hostile takeover bid for rival mining house Gold Fields Ltd (“Gold Fields”). The bid was made in two stages, the first ‘the early settlement offer’ and the second stage, ‘the subsequent offer’.

7.      
The ‘early settlement offer’, which Harmony claimed was not subject to regulatory approval, was the subject of a whole set of procedural applications brought against Harmony and which resulted in the Competition Appeal Court ruling, on 26 November 2004, that ‘the early settlement offer’ and ‘the subsequent offer’ in substance formed part of a single transaction to acquire control of Gold Fields and, therefore, interdicted and restrained Harmony from voting its shares in the share capital of Gold Fields prior to the final determination of the merger by the Competition Tribunal in terms of section 16(2) or the Competition Appeal Court in terms of section 17 of the Act.

8.      
Subsequent to the Competition Appeal Court hearing Harmony acquired 11.8% of Gold Fields’ shares. On 8 November 2005, at the same time that Harmony announced its ‘early settlement offer’ it also filed its merger notification with the Competition Commission. Gold Fields responded on 15 December 2005, informing the Competition Commission that it is a hostile take-over.

9.      
On 11 February 2005 the Competition Commission recommended that the proposed merger be approved subject to the following conditions:


1.      
The following limitations shall be placed on the retrenchments at the merged entity –

a.      
There shall be zero retrenchments at the merged entity below the level of corporate, management and supervisory positions as a result of the merger;
b.      
The merged entity may retrench up to a maximum of 1500 employees in corporate, management and supervisory positions as a result of the merger.

2.      
Corporate, management and supervisory positions shall mean positions from shift boss level up to the chief executive.

3.      
The moratorium mentioned in 1 above shall apply for a period of 24 months from the date of the Competition Tribunal order.

Monitoring of the recommended conditional approval

The following procedures must be adhered to in order for the Commission to properly monitor the abovementioned proposed conditions:

1. The merged entity must:

a)      
Provide the Commission with quarterly reports regarding the effects of the proposed transaction on employment.

b)      
Each report must include the following information:
(i.)    
the current levels of employment, per job category, at the merged entity;
(ii.)   
the number of actual retrenchments per job category in the quarter reported on;
(iii.)  
the reasons, per job category, for retrenchments;
(iv.)   
the number of planned further retrenchments per job category;
(v.)    
the status of further retrenchments; and
(vi.)   
the process upon which the retrenchments will take or have taken place.

c)      
The quarterly reports must be submitted for the period of this order. Should the retrenchment process in the merged entity, as a result of the proposed transaction, not be finalized within the period of this order, the merged entity shall be obliged to submit further quarterly reports until the entire retrenchment process has been finalized.

d)      
The quarterly reports must be submitted to the Competition Commission no later than one calendar month following the end of each quarter.

10.     
A pre-hearing was held on 25 February 2005 during which intervenors were identified, a time-table for filing submissions was agreed on and discovery and confidentiality issues addressed. A second pre-hearing date was set for 20 April 2005 to discuss the final logistics of the case and the hearing dates were set down for 3 to 6 May 2005.

11.     
On 30 March and 8,19 and 20 April 2005 applications to intervene, confidentiality applications and discovery were heard. During the pre-hearing, held on 20 April it became apparent to the Tribunal that it would need additional hearing dates since the since the parties had called 13 witnesses in total. The hearing days were thus extended to 7,8 and 9 May.

12.     
Three parties intervened, Stitch Wise (Pty) Ltd, Paragon Textiles (Pty) Ltd and Knee’d’em (Pty) Ltd. The intervenors called one witness, Ms N. Killasy, who is a Director of all three intervenors.

13.     
The following witnesses were called by Harmony and Gold Fields:

Gold Field witnesses:

1)      
Mr. T. P Goodlace        Senior Vice President, Strategic Planning, Gold Fields
2)      
Mr. J McLuskie   Expert Witness on deep level mining
3)      
Dr N.S. Segal             Independent consultant
4)      
Prof. H Bhorat            Director, Development Policy Research Unit, UCT
5)      
Mr. J Hodge               Engagement Manager: Competition and regulation practice, Genesis Analytics
6)      
Mr. N. Goodwin   Gold Analyst, TSEC Securities
7)      
Mr. M.J. Mitchley        Senior Manager Gold Fields      
8)      
Prof. SA du Plessis      Associate Professor in macro-economics University of Stellenbosch


Harmony witnesses:

1)      
Dr. C. Caffarra  Economist and Director of Lexecon Ltd
2)      
Mr. B.M Saunders         Executive: Investor Relations Harmony
3)      
Mr. A. Clay               Director of Venmyn Rand (Pty) Ltd
4)      
Prof. S. Roberts         Associate Professor of Economics, University of Witwatersrand
5)      
Mr. Z.B. Swanepoel       Chief Executive Officer of Harmony


14.     
Although all the Unions that represent the mineworkers were informed of the hearing none were represented before the Tribunal. UASA and Solidarity attended the pre-hearing on 25 February 2005 but did not submit any further submissions nor attended any further hearings in this regard. The National Union of Mine Workers, on 26 April 2005, requested an opportunity to address the Tribunal at the hearing but never showed up.


COMPETITION EVALUATION

The Gold Market – competition implications

15.     
It is common cause between the Commission and the acquiring and target firms that this merger presents no competition problems in the gold market. Although the merged entity will, by most relevant measures, be the largest gold producer in South Africa and in the world, it’s share of world gold production will still only be 9,5%. In short the structure of the market for the production of gold is characterised by its high degree of fragmentation. Moreover daily prices are fixed internationally through a relatively transparent mechanism, which appears relatively impervious to direct producer influence. The gold market has been analysed in previous transactions and this transaction does not change the conclusions reached in these earlier decisions. Accordingly the gold market will not be examined further.


Markets for the supply of inputs to gold producers – competition implications

16.     
Gold Fields alleges that the merger will lead to a substantial lessening of competition in the markets for the supply of inputs to the gold mining sector. Gold fields argues that because the merger will reduce the number of South African gold producing majors from three to two, competition will be substantially lessened in many of the markets in which inputs are sold to the gold mining sector – the merger will, in other words, create oligopsonistic or buyer market power. The consequence, asserts Gold Fields, will be manifest in the ability of the buyer to force input prices to sub-competitive price and output levels. Gold Fields also argued that, in addition to these static allocative inefficiencies, the accretion to buying power will give rise to dynamic consequences insofar as the suppliers’ incentive to invest and further develop their products will be dampened. It is not clear whether Gold Fields contended that these dynamic consequences would be generated by a change in the structure on the demand side of the market or in consequence of the particular modality employed by Harmony to procure supplies. It was alleged that Harmony’s approach to procurement is dictated by considerations of price alone whereas the Gold Fields’ approach is allegedly more sensitive to quality considerations. Gold Fields argues that the latter mode of procurement is more conducive to supplier investment in product improvement.

17.     
Gold Fields attempted to adduce evidence in support of its contentions. This evidence was gathered by a survey conducted of its suppliers and a meeting it convened of some 200 suppliers. On this basis Gold Fields concluded that 39% of their suppliers would be forced out of business and 56% expected a material downscaling in business volumes.

18.     
We are sceptical of evidence gathered in the manner. Gold Fields’ suppliers would, under these circumstances, understandably be inclined to provide answers supportive of an important customer’s clearly expressed standpoint – indeed the survey result may well reflect the power of Gold Fields vis a vis its suppliers rather than genuine apprehension of the merger. We should add that the presentation to the suppliers has compromised the value of this evidence. Highly disputed figures of the employment loss predicted by Gold Fields are presented as fact. The tone of the communication is decidedly alarmist and manifestly designed to strike fear into the hearts of the suppliers and their employees. It appears, moreover, that the presentation, on behalf of Gold Fields, was made by its legal advisers, Edward Nathan – the presentation slides certainly carry Edward Nathan’s explicit imprimatur. Not only does this provide a veneer of independence, but it may have imbued certain important statements with a legal authority which they should not have enjoyed. For example the analysis presented to the suppliers and the dire impact predicted is predicated on an analysis of the ‘extensive efficiencies’ promised by Harmony, ‘promises’ which, Edward Nathan baldly and quite incorrectly asserts (in bold type), ‘Harmony is not entitled to abandon’. The reasonable reader of this document would assume that Harmony was legally obliged to achieve these efficiencies and hence been more susceptible to come to panic induced conclusions. We attach no weight to evidence gathered under these circumstances.

19.     
Our scepticism of the evidence presented by Gold Fields is deepened by the conspicuous failure of any suppliers to make submissions to the Tribunal. Indeed a significant number of suppliers had initially made submissions. However it appears that when they realised that they may be obliged to repeat their allegations in an open enquiry and subject to cross-examination they all withdrew their statements and objections. This may well indicate that they feared victimisation in the event of a successful Harmony acquisition (although as the Commission points out, the mere fact that they feared victimisation from Harmony post-merger indicated that they envisaged competing for the merged entity’s custom – that is, contrary to Gold Fields’ evidence, its suppliers had not resigned themselves to going out of business). However their reticence to defend their claims may also indicate that they were largely designed to please an increasingly anxious Gold Fields’ management and they feared that they would not stand up to scrutiny. In the event, the only supplier who made submissions to the Tribunal was a representative of what is best understood as a corporate social investment project and she, indeed, made her representations in terms of the public interest.

20.     
The Commission investigated the possibility of a substantial lessening of competition in the supplier markets. Parties notifying a merger are required to identify their 10 largest suppliers. The Commission telephonically contacted about 52 Gold Fields’ suppliers, including its 10 largest suppliers. The Commission summarises its efforts and some of its most important findings:

The suppliers contacted ranged suppliers of hydraulic pumps, gum boots, bearings, cleaning chemicals, pumps, rolling stock, skips, heat exchange for underground use, batteries, motors, valves, mechanical seals, hoses and fittings, conveyor belts, backfill bags, knee and arm guards, radiators, oil coolers and the construction of underground dams, pump stations and pipe installations, inter alia. Of the suppliers contacted none was wholly dependent on Gold Fields as a customer with many citing Gold Fields, Harmony and Anglo American-Ashanti and being their major customers. Some of the contractors located in the gold mining areas were branches of larger companies such as Rocla (part of the Murray and Roberts group), Builder’s Market (part of the Iliad Group with branches in Welkom, Klerksdorp, Rustenburg, Gauteng and Polokwane), Conway Johnson (part of the Inmins group listed on the JSE) and Alstom (an international company). Contractors and suppliers in the Carletonville area cited that, in addition to the gold mines, the platinum mines are important customers.

21.     
It appears that the Commission’s enquiries elicited a range of responses. Predictably, some amongst those Gold Fields’ suppliers that have no established relationship with Harmony expressed disquiet at the prospective merger. The Commission summarises its findings:

In summary it can be said that the majority of businesses who have expressed concerns about the merger operate in markets where competition takes place. The concerns of many of the company representatives with respect to the merger can be said to be related to their uncertainty about being able to secure business in a market where most projects go out on tender.

22.     
Counsel for Gold Fields made much of the Commission’s failure to go behind the responses gleaned from the telephone interviews and to identify the relevant markets and then conduct a detailed competition analysis of the impact of the transaction. The Commission reasonably responded that because its interviews with a significant sample of key suppliers indicated little if no competition concern with the merger there was no need for a time consuming and resource sapping examination of each of the many product and geographic markets in which a vast array of inputs are supplied to the merging parties.

23.     
Gold Fields also relied upon evidence and argument presented by one of its expert economic witnesses, Dr. S. du Plessis. Dr. du Plessis purported to measure concentration in these markets by using a measure dubbed the ‘dispersed HHI’. The Herfindahl Hirschman Index or HHI is widely and legitimately employed as a measure of market concentration in anti-trust analysis. However, a necessary prior step to calculating the HHI is the identification of the relevant market. This is more than a filter – it is the necessary preliminary step towards the construction of a filter and that filter is the HHI, which measures concentration in the relevant market. In other words, once the relevant markets have been identified the market shares of the various participants in these relevant markets are calculated and the index is applied in order to assess the change in market concentration that will arise from the merger of two of the participants. It is widely used by competition authorities, including the South African Competition Commission, as a measure of the increase in concentration resulting from a horizontal merger and, hence, as a first cut indicator of a possible accrual of market power that may arise from this increase in concentration. However, the HHI is never, on its own, construed as sufficient evidence of market power – this requires a detailed evaluation of a range of factors including barriers to entry, the prospect of import competition and the dynamics of the market in question.

24.     
However the dispersed HHI utilised by Dr. du Plessis does not serve the same useful filtering purpose as the HHI and accordingly is, du Plessis acknowledged and the Harmony expert, Dr. C. Caffarra, confirmed, rarely, if ever, utilised in anti-trust analysis. There is no substitute for defining the relevant market and the Commission’s interviews constituted precisely the appropriate first step in this direction. Had its interviews revealed any concern amongst the diverse suppliers that it contacted then it would have been obliged to conduct a deeper investigation in order to identify the relevant market. Thereafter it would have calculated market shares and then computed the HHI in order to measure the changes in concentration that would accrue in consequence of the merger of the two market participants, of the two buyers.

25.     
The Gold Fields’ expert witness has made no attempt whatsoever to delineate the relevant markets. What he has done is to utilise highly aggregated (and dated) industrial statistics – specifically, the ‘Supply and Use’ tables for South Africa in 2000 – which have then been employed to identify 9 broadly defined sectors, or, more accurately, products, in which the gold mining sector accounts for more than 4% of total demand. Then, assuming that each purchasing sector constituted a single monopoly – an assumption patently at odds with reality - a dispersed HHI was calculated which purported, in the fashion of the HHI, to identify the market concentration of the sector as a purchaser of the output in question – he effectively insists that he has, in the manner of the HHI, provided an indicative measure of market concentration without defining the market. In this manner – and after several iterations with the Harmony expert in which he purported to modify his results in order to accommodate specific criticisms of his efforts – Dr. du Plessis identified three problematic groups of products, these being ‘other rubber products’, ‘pumps’ and ‘mining machinery’. These are product groups in respect of which the dispersed HHI exceeds 0.18. It appears that this threshold was chosen in order to provide an appearance of conformity with the HHI threshold utilised by the US Department of Justice.

26.     
Let us consider, by way of example, the product designated by the industrial statistics as ‘pumps’, one of the ‘markets’ in which Dr. du Plessis’ dispersed HHI indicates cause for concern. But the most casual observation tells one that this is no market at all. It is an aggregation of several diverse markets. It incorporates a range of non-substitutable products – pumps utilised in underground mines, pumps utilised in farm boreholes, pumps utilised in domestic swimming pools and, for all we and, indeed, Dr. du Plessis, know, pumps utilised in automobile engines. Had he even spoken to a single supplier of pumps to the gold mining sector he may have discovered that, as did the Commission on many occasions, the producer in question was unconcerned with the merger. Had he been of a mind to interrogate this response further, it may then have been revealed that these pumps are utilised across the mining sector and, hence, that the merger of even two large gold mining companies would, in the face of an attempt to exercise market power, not necessarily depress prices below the competitive level – the pump producers would simply turn to their other mining customers. He may also have found that the mining pump producers also actively supply international markets. He may have found that the mining pump supplier, faced by an exercise of buyer power on the part of his mining customers, is easily able to switch to supplying pumps to the agricultural sector. Or he may have discovered that the pump supplier enjoyed a monopoly in its market and, in consequence, that he felt relatively impervious to the change in the structure of his customers’ market.

27.     
We could go on in this fashion ad nauseam. The conclusion, though, is reasonably clear: anti-trust investigation does not easily lend itself to desktop research utilising highly aggregated industrial data. It is micro-economic research and there is, as the Commission has shown, no substitute for engaging with the actual producers and customers. This was not even attempted by the Gold Fields’ expert. Instead reliance was placed on a measure and a data set that yielded, at best, no results of consequence. At worst, the results may be downright misleading – just as Dr. du Plessis’ high dispersed HHI’s may, on relatively cursory examination of the relevant markets embedded in the product ranges, reveal no cognisable competition problems, so may some of his low dispersed HHI’s camouflage, on a proper definition of the relevant market, very definite problems. In other words, the dispersed HHI is susceptible to both Type 1 and Type 2 errors – it may signal problems where there are none; and it may signal the absence of problems where there are some. Dr. Caffarra, the Harmony expert, characterised the dispersed HHI:

‘..as a screening devise that doesn’t really screen because (it) is subject to type 2 errors. It can tell us that there are some sectors where a problem exists, but then we may find there isn’t one. It may tell us that there are some sectors where there isn’t a problem, because you are below the threshold and still there is a problem when you define the market correctly. That is to me the hallmark of a not very useful approach. You want to do other things instead. You want to look….that’s why surveys exist. You want to look…in a perfect world, you would want to calculate the elasticity of supply, but in a world where you don’t have this type of data it won’t be possible to do complicated econometrics and calculate the elasticity of supply or the supply curve.