COMPETITION TRIBUNAL
REPUBLIC OF SOUTH AFRICA
Case No: 55/LM/Aug02
In the large merger between:
Rustenburg Platinum Mines and Eastern Platinum Mines Ltd “Pandora
Joint Venture”
and
Rustenburg Platinum Mines Ltd
Reasons for Decision
______________________________________________________________
APPROVAL
On 16 October 2002 we unconditionally approved the joint venture between
Rustenburg Platinum Mines and Eastern Platinum Mines Ltd “Pandora Joint
Venture and Rustenburg Platinum Mines Ltd. The reasons for our decision
follow.
Background
The Transaction
1. RPM and EPL are to establish a joint venture, (the “Pandora Joint
Venture” or “JV”) which will exploit the UG2 Reef on portions of 3 farms
near Brits in the North West Province. Both parties will have 45% each
in the J.V. while the two black empowerment companies, Newco and
Northam, will have 5% each.
The parties
2. The acquiring firm is a joint venture, (the ”Pandora Joint Venture or
J.V.”), between Eastern Platinum Mines Limited (“EPL”) and
Rustenberg Platinum Mines Limited (“RPM”), who will jointly control the
J.V. The Pandora J.V. is being formed to mine and concentrate
platinum group metals (“PGM’s”) – namely Palladium, platinum, iridium,
osmium, rhodium and ruthenium.
3. EPL is ultimately controlled by Lonmin Plc 1 and engaged in smelting,
base metal refining and precious metal refining.
4. Rustenberg Platinum Mines Limited (“RPM”), is ultimately controlled by
Anglo Platinum, (“AngloPlat”) the world’s leading producer of Platinum.
RPM is engaged in mining, smelting, base metal refining and precious
metal refining.
5. Northam is 22.5% owned by Anglo Platinum and 22.5% held by
Mvelaphanda Resources, a major black economic empowerment
group. It mines and mills ore and is responsible for smelting and base
metal refining up to PGM concentrate state. 2
6. Newco, likewise an empowerment company, represents the
communities of some 20 000 people surrounding EPL and the
proposed Pandora mine. They are members of the Bapo Ba Mogale
(“Bapo”) tribe. The company they formed is known as the Bapo
Bamagale Mining Company (Pty) Limited, and the shareholders of that
company will be the Bapo tribe.
7. RPM and EPL are committing assets into the joint venture and as such
they may be regarded as the target firms.
Rationale for Transaction
8. The transaction will facilitate the establishment of a large mine
(Pandora Mine) to mine the entire area of the UG2 Reef more
efficiently than having to mine each area separately. The parties state
that it is not feasible for RPM (which owns the mineral rights to two of
the mines and has secured mining rights of the third farm from the
state) to mine the entire area on its own, as it would require much
infrastructure investment. EPL already has shafts adjacent to the three
farms and it is more economical to upgrade existing EPL infrastructure,
rather than erecting an entirely new plant. In addition to having the
requisite infrastructure, EPL also has the experience to manage this
requisite infrastructure, EPL also has the experience to manage this
project. Generally, the acquisition would permit better utilization of
RPM’s mineral assets – leading to lower costs in extraction and in
management.
The relevant market
1 Along with its sister company, WPL, through LSA (UK) Limited, a wholly owned subsidiary
of Lonmin Plc.
2 Northam have a base metal refinery, not a precious metal refinery. The parties stated at the
hearing that Implats has a shareholding in Northam too although it is not clear what stake
they actually do have.
9. Platinum group metals – PGMs – comprise platinum, palladium,
rhodium, ruthenium, iridium and osmium. The properties of this group
of metals are such that substitution of PGMs with metals outside of this
group is not commercially or technically viable over an important range
of uses. There is a certain degree of substitutability between the
members of the PGM. However the 1996 European Commission report
on the proposed merger of the platinum interests of Gencor (viz.
Implats) and Lonrho (viz. LPD) (henceforth ‘the GencorLonrho report’)
found that PGMs do not constitute a single relevant market but rather
six relevant markets, each comprising the various members of the
platinum group of metals. 3 Although subsequent developments may
indicate a greater degree of substitutability between platinum and
palladium in the manufacture of autocatalysts than that suggested in
the EC report, we are confident that the relevant markets identified by
the European Commission remain valid. 4
10. Although the range of PGMs is implicated in this transaction we will, by
and large, restrict our comments to platinum itself. It is here that the
parties to the transaction – RPM, Lonmin and Northam Platinum
Limited are most active. Indeed South African PGM ores are
particularly richly endowed in platinum. Hence while Russian producers
have a strong presence in the mining of PGMs generally, they appear,
because of the character of their ore bodies, to hold a particularly
strong position in palladium. Stillwater, the US PGM mining and
refining company active in the US, also mines ores richly endowed in
palladium.5 Hence, of the six relevant markets identified above, it is the
platinum market that is particularly implicated by this transaction.
platinum market that is particularly implicated by this transaction.
11. As noted above there is a degree of substitutability within the PGM
range of metals. For example, both platinum and palladium are
extensively used in the manufacture of autocatalysts, an important
3 Commission Decision of the 24 April 1996 declaring a concentration to be incompatible with the
common market and the functioning of the EEA Agreement (Case No IV/M.619 – Gencor/Lonrho).
Note that we shall refer to this report at other points in this decision. The EC report is particularly
apposite because both analyses deal with the same geographical market populated by the same
participants.
4 The parties pointed out that in 2001, the autocatalyst sector accounted for approx. 40% of platinum
demand and 71% of palladium demand. They remarked that this demand substitutability signified that
platinum and palladium could indeed constitute a single product market, especially as fabricators
become more price sensitive.
5 See GencorLonrho Para 80: ‘The individual PGM metals are produced in fixed ratios, determined by
nature, which depends on the particular ore body. Indications from official sources are that the ratio of
platinum/palladium/rhodium is about 100:42:21 at the Merensky reef and 100:83:54 at the UG2 reef.
In other countries palladium occurs in higher concentrations relative to platinum. In the main Russian
mine, in Noril’sk, the ratio is about 100:284:16, at the American mine in Stillwater 100:350:73 and at
the Canadian mine in Sudbury 100:110:73. This production structure often results in some stocking or
oversupply of the minor metals. It also means that palladium makes up a larger part of Russian
production than platinum. However, in the South Africa mines platinum is by far the most important
metal which accounts for more than 80% of the sales revenue.’
market for PGMs, although this substitutability does not extend to other
important markets, for example jewellery. 6 We repeat then the six
relevant product markets correspond to the members of the PGM and
our analysis will focus on one of these – the market for platinum. 7
12. The GencorLonrho report includes a careful analysis of the platinum
market. It’s conclusions are worth citing at length and speak for
themselves:
‘The portion of platinum demand accounted for by industrial
processes and autocatalysts is priceinelastic, probably with a
very low price elasticity, since there are basically no substitutes
for platinum for these purposes, apart from limited substitution
possibilities between platinum and palladium for certain types of
autocatalysts. The price elasticity for jewellery demand on the
Japanese market was found to be priceinelastic with an
elasticity of –0,6. Since autocatalysts and industrial processes
account for about 51% of the market, and the Japanese
jewellery market for about 34%, this means that the price
elasticity of 85% of the global platinum market is highly inelastic.
The remaining 15% of demand is for jewellery outside Japan
(5%) and investment (10%). The jewellery market outside
Japan is likely to have an inelastic demand, since platinum
jewellery is a special, upmarket product. Furthermore, the
effect of investment demand, on overall price elasticity, is
limited. All in all, it can therefore be concluded that the price
elasticity for the total market is inelastic (numerically smaller
than 1).’ 8
13. We recognise that these conclusions are based on research concluded
some 9 years ago. However, little has changed since then certainly,
the overall composition of demand is unchanged. As noted the degree
the overall composition of demand is unchanged. As noted the degree
of substitutability of palladium for platinum in certain autocatalysts
appears somewhat greater than predicted in the GencorLonrho report
6 The parties insisted on the substitutability of gold with platinum in the jewellery market. This,
together with the substitutability of platinum and palladium in the autocatalyst market would,
according to them, decrease any possible market power on behalf of Anglo Platinum . See paragraph 12
below.
7 Note an August 2001 Schroder Salomon Smith Barney report on precious metals: ‘On the question of
substitution potential within the PGM group, in simplistic terms, autocatalysts address three different
emissions – carbon monoxide (platinum best for that); hydrocarbons (palladium best for that); and nox
(rhodium best for that). Future autocatalysts are therefore likely to need all three in different
combinations. Mix will be affected to some degree by current prices ( i.e. if PD becomes very
competitive) but automakers are also concerned about longerterm supply issues and price volatility.
The reality is that for technical reasons, platinum is the only possible choice for diesel catalysts. So a
number of industry players are now suggesting demand within five years could more closely replicate
the production profile of the SA mines (i.e. 60%Pt; 30% PD; 10% Rhodium).’
8 GencorLonrho Para 56.
but even in this limited area the tide – driven largely, it appears, by
technical considerations – seems to be turning platinum’s way once
again. The conclusions of the GencorLonrho report are appropriately
qualified by the observation that the demand for platinum is only price
inelastic over its current price range. Against that, though, it is clear
that, even in response to significant price swings, the possibilities for
substitution are highly limited even over the medium term and
particularly for platinum’s industrial applications. Nevertheless this may
seem an important qualifier in the case of a volatile commodity market
– however, as we shall elaborate below, the peculiar structure and
pattern of interactions in this market may be seen as part of a strategic
approach that is precisely intended to check price volatility through
control of supply.
14. South Africa, as already noted, is particularly richly endowed in PGMs
and South African ore bodies are particularly richly endowed in
platinum. This has enabled South African based companies – notably,
although not exclusively, Angloplats, though RPM, a party to this
transaction, and Implats – to assume a dominant position in the mining
and refining of platinum. Lonmin, through EPL, also a party to this
transaction, is a British owned participant in the PGM market which
also controls a significant share of the mining and refining of PGMs in
South Africa. Russia is the other area in which PGMs are extensively
mined. PGMs are also actively mined in the US and Canada.
15. PGMs are homogenous products with no apparent barriers to
international trade. 9 While it appears that mining activities are
generally served by refineries located in the countries in which the
generally served by refineries located in the countries in which the
PGMs are mined (this accounting for the strong position of South
African companies in the refining of PGMs) we are not aware of any
insurmountable barriers to exporting the raw material resource from
countries where no refining capacity is located to those countries in
which established refining capacity is located. Indeed, according to
Schroder Salamon Smith Barney, Implats – the only PGM company
engaged in refining mining output belonging to independent PGM
mining companies – refines, at its South African refinery, the metals of
twenty groups from five continents. 10
16. Accordingly the geographic market for the mining and refining of PGMs
is international. This view is supported, although not necessarily
determined, by the existence of internationally quoted prices for PGMs.
17. World shares of the platinum market are calculated by measuring
9 ‘ PGMs are fungible assets, are easily transported, are refined to the same purity standards throughout
the world and readily traded without tariff barriers. PGMs are sold on a worldwide basis either under
longterm contracts or on the metal market’ (GencorLonrho para 68).
10 In addition, Northam exports concentrate to Heraeus in Germany.
shares of the refined product. Data submitted by the parties reveals
the following market shares are attributable to the major producers:
Party Share of Refined World Production in 2001
Angloplats 38.4%
Implats 21.6%
Lonmin (Lonplats) 12.4%
18. Note however that Implats also owns a 27% share of Lonmin . Lonmin
Platinum11 is a UK registered company, whose South Africanbased
mining and refining activities accounts for 12.4% of world output.
Implats is party to a shareholder’s agreement which appears to give it
significant influence over this company. If Lonmin’s output is counted
as part of Implat’s output, its market share in 2001 would increase
somewhat. However, Implat’s stake in Lonmin is, in our view,
noteworthy, not because of its implications for Implats’ market share,
but because it evidences widespread cooperation amongst the
participants in the market for mining and refining PGMs. Note also that
Angloplats has a significant stake in Northam, one of the smaller South
African platinum producers and one of the black empowerment
partners in this project. 12
THE IMPACT ON COMPETITION
19. This transaction has both a horizontal and a vertical dimension.
Angloplats and Lonmin, both established miners and refiners of PGMs,
are entering into an agreement that will result in the exploitation of
PGM resources. As already noted this will involve mining on three
tracts of land. On two of these, RPM hold the mineral rights while the
state holds the mineral rights on the third. EPM is mining on territory
immediately adjacent. Effectively, mineral rights held by the state and
RPM are being acquired by a joint venture controlled by RPM and
EPM. This gives the transaction its horizontal dimension.
20. The transaction may also be viewed as an act of backward integration
20. The transaction may also be viewed as an act of backward integration
by refiners acquiring additional sources of input. This provides a
vertical dimension to the contract.
21. The market in question is highly concentrated. The two largest South
African producers – Angloplats and Implats stand astride the world
market, the more so if, for the purpose of evaluating this transaction,
the Implats and LPD market shares are consolidated. Moreover the
two largest companies are clearly taking steps to consolidate their
11 This is the trade name for Western Platinum Limited and Eastern Platinum Limited . Lonmin has
73% in EPL.
12 See above discussion in paragraph 2.
powerful position in the world market. Angloplats has a massive ore
reserve and is in the process of establishing a second refinery in South
Africa.
22. We are enjoined by the Act to determine whether or not the transaction
substantially lessens competition. This transaction, on its own, does
not impact significantly on market shares. Angloplats already owns the
rights to most of the resources that will be exploited by the joint
venture. The additional market share that accrues to it by virtue of its
acquisition of the state’s rights is minimal. Indeed, in terms of market
share the largest gain is recorded by Lonmin, the smallest of the three
large PGM mining companies active in South Africa. Viewed
collectively, however, these small transactions are the mechanism that
account for steadily increasing concentration levels in the market for
PGMs. However, we cannot find that the horizontal dimension of this
single transaction is ‘likely to substantially prevent or lessen
competition’, the test provided for in the Act.
23. What of the vertical dimension? In Two Rivers 13 we examined the
vertical aspect of the transaction in some detail. We will not repeat this
analysis in detail here – suffice to say that the present transaction
confirms our concerns. In Two Rivers we showed that the steady
accretion, largely by Implats, of small pockets of ore reserves either for
the purpose of mining or simply refining, combined with Angloplats’
massive reserves, effectively foreclosed entry into the capital intensive
refining stage. Angloplats, Implats and Lonmin controlled sufficient of
the present ore output and reserves to deter any wouldbe entrant at
the refining stage. Hereaus, the German refiner, indicated that it was
considering establishing refining capacity in South Africa. The
considering establishing refining capacity in South Africa. The
Commission also indicated that several other companies were
considering setting up refining activities in South Africa, however we
consider this unlikely, indeed it is made all the more unlikely by the
steady accretion of mined and reserve ore in the hands of the
established refiners. Again, the present transaction on its own is too
small to constitute a foreclosure sufficiently significant to represent a
substantial lessening or prevention of competition.
24. However, from a competition perspective the most striking and
disturbing aspect of any examination of the PSG markets is the
extraordinary degree of cooperation between the various producers.
This transaction is a yet another case in point: it is a joint venture
between the largest (Angloplats) and third largest (Lonmin) platinum
producers in the country. The second largest producer (Implats) has a
very large stake in the Lonmin – indeed the agreement in place
between Lonmin’s shareholders would require that Implats approve the
13 Two Rivers Platinum Limited and Assmang Limited 54/LM/Sep01
JV. One of the black empowerment beneficiaries, Northam, is a small,
but growing player in this market. Angloplats owns a 22.5% stake in
Northam and, so, it appears, does Resources Limited.
25. We have been presented with evidence suggesting that there are
exceptional production efficiencies to be derived from the JV. We have
not found a substantial lessening of competition and hence there is no
reason for us to examine the claims for increased efficiency. However,
on the face of it the efficiencies to be derived from the JV do appear to
be unusually significant. Much has been made of the fact that this is a
production JV, and that refining and marketing will take place
independently under the control of each of the shareholders. Again, in
the ordinary course, this aspect of the arrangement is not without
significance in a competition assessment. However, neither the
efficiencies claim, nor the limited nature of the JV address the concern
that this transaction, and, it appears many others like it, further cement
the structural ability of the key players in the platinum market to engage
in anticompetitive cooperation.
26. The parties appear to concede that there is no price competition to
speak of. This, they aver, reflects the fact that all the producers are
‘price takers’ with price set by supply and demand conditions on the
international market, implicitly suggesting the existence of a perfectly
competitive market. However, reality is at odds with this suggestion –
far from a market characterised by large number of producers, each
incapable of influencing aggregate output, it is common cause that the
market is dominated by a small number of very large players.
27. In the Two Rivers matter we pointed to evidence that precisely
suggested tacit collaboration in influencing supply. The parties to that
suggested tacit collaboration in influencing supply. The parties to that
transaction denied that they cooperated in determining supply and
these denials have been reiterated here. We naturally made no finding
on this point in the Two Rivers decision and we will not do so here.
However we simply point out that the extraordinary web of interactions
between the key South African platinum producers provides wide
ranging opportunity for collaboration. Had these structures of potential
cooperation been established through this transaction, we may well
have concluded that competition had been substantially lessened.
However, the structures of the industry are well established and are not
unduly strengthened by this transaction. Our conclusion in the Two
Rivers matter is apposite and we emphatically restate it here:
It is now for the competition authorities in South Africa as well
as other jurisdictions to ensure that this anticompetitive market
structure is not abused, in particular to ensure that the
oligopolistic structure of this market does not permit its small
number of major participants to manipulate the supply, and
hence effectively set the price, of these important products.
CONCLUSION
We therefore conclude that the merger will not lead to a substantial lessening
of competition. The Tribunal therefore approves the transaction
unconditionally. There are no public interest concerns which would alter this
conclusion.
_____________ 6 December 2002
D. Lewis Date
Concurring: N. Manoim, M. Holden
For the merging parties: Deneys Reitz Attorneys
For the Commission: A.Coetzee, Competition Commission