32004D0284

2004/284/EC: Commission Decision of 29 September 1999 declaring a concentration compatible with the common market and the EEA Agreement (Case No IV/M.1383 — Exxon/Mobil) (Text with EEA relevance) (notified under document number C(1999) 3093)

Official Journal L 103 , 07/04/2004 P. 0001 - 0136


Commission Decision

of 29 September 1999

declaring a concentration compatible with the common market and the EEA Agreement

(Case No IV/M.1383 - Exxon/Mobil)

(notified under document number C(1999) 3093)

(Only the English text is authentic)

(Text with EEA relevance)

(2004/284/EC)

THE COMMISSION OF THE EUROPEAN COMMUNITIES,

Having regard to the Treaty establishing the European Community,

Having regard to the Agreement on the European Economic Area, and in particular Article 57 thereof,

Having regard to Council Regulation (EEC) No 4064/89 of 21 December 1989 on the control of concentrations between undertakings(1), as last amended by Regulation (EC) No 1310/97(2), and in particular Article 8(2) thereof,

Having regard to the Commission decision of 9 June 1999 to initiate proceedings in this case,

Having given the undertakings concerned the opportunity to make known their views on the objections raised by the Commission,

Having regard to the opinion of the Advisory Committee on Concentrations(3),

Whereas:

(1) On 3 May 1999, the Commission received notification of a proposed concentration by which the US undertakings Exxon Corporation and Mobil Corporation were to merge their activities worldwide.

(2) On 2 June 1999, the United Kingdom notified the Commission in accordance with Article 9(2)(b) of Regulation (EEC) No 4064/89 (hereinafter referred to as the Merger Regulation) that it considered that the concentration affected competition in the northwest of Scotland in the retail motor fuel sector. On 26 July 1999, the Commission issued a statement of objections identifying, inter alia, competition concerns in the market for motor fuel retailing in the whole of the United Kingdom. On 10 August 1999 the Commission informed the United Kingdom that in those circumstances and in accordance with Article 9(4)(b) of the Merger Regulation there were no grounds for referring the case to its authorities.

(3) By decision dated 9 June 1999, the Commission found that the notified operation raised serious doubts as to its compatibility with the common market. The Commission accordingly initiated proceedings in this case pursuant to Article 6(1)(c) of the Merger Regulation.

I. THE PARTIES AND THE OPERATION

(4) Exxon Corporation (Exxon) is a diversified company active worldwide in the exploration, development, production and sale of crude oil and natural gas, the refining and sale of refined petroleum products the development, production and sale of various chemical products, the production and sale of coal and minerals and power generation. Mobil Corporation (Mobil) is a diversified company active worldwide in the exploration, development, production, and sale of crude oil and natural gas, the refining and sale of refined petroleum products and the development, production and sale of various chemical products. Mobil and British Petroleum (BP) regrouped all their refining and retailing activities in Europe into a jointly controlled joint venture, BP/Mobil(4).

(5) This notification concerns a full merger within the meaning of Article 3(1)(a) of the Merger Regulation. Exxon and Mobil signed an agreement and plan of merger on 1 December 1998. Pursuant to this agreement, Mobil will merge with a wholly owned subsidiary of Exxon, with Mobil as the surviving corporation. As a result, Exxon will hold 100 percent of Mobil's issued and outstanding voting securities. Holders of Mobil common stock will receive 1,32015 shares of Exxon common stock for each share of Mobil common stock that they own at the time of closing. Exxon shareholders will own approximately 70 % of the combined Exxon Mobil entity, while Mobil shareholders will own approximately 30 %.

II. COMMUNITY DIMENSION

(6) As the Parties' combined aggregate worldwide turnover(5) exceeds EUR 5 billion (Exxon: EUR 121 billion; Mobil: EUR 47 billion), each of Exxon's and Mobil's Community-wide turnovers exceeds EUR 250 million (Exxon: EUR [...](6) billion; Mobil: EUR [...]* billion) and neither Exxon nor Mobil achieves more than two thirds of its aggregate Community-wide turnover within one and the same Member State, the concentration has a Community dimension pursuant to Article 1(2) of the Merger Regulation. The turnover of each of Exxon and Mobil exceeds EUR 250 million in the EFTA territory (Exxon: EUR [...]* billion; Mobil: EUR [...]* million). Therefore, the case is an EFTA cooperation case.

III. COMPETITIVE ASSESSMENT

(7) The oil industry is generally divided into the upstream oil and gas sector (i.e. exploration, development and production of crude oil and natural gas), the downstream sector (refining and marketing of fuels, distribution of natural gas, manufacture of lubricants) and the various petrochemical activities.

(8) The concentration does not raise competition concerns in the markets for exploration, development and production of crude oil and natural gas and with regard to the various petrochemical activities where the parties' businesses overlap. Both parties are active in the exploration, development and production of crude oil and natural gas and the Commission had expressed in the Article 6(1)(c) decision serious doubts on these markets. In the Article 6(1)(c) decision, the Commission had also expressed serious doubts on the "gas-to-liquid" market. However, there are, for the reasons indicated in sections A and B below, no competition concerns on these markets.

(9) For the reasons indicated below, the concentration would have given rise to the creation or strengthening of a dominant position in the following markets:

- wholesale transmission of natural gas in the Netherlands (Section C),

- long-distance wholesale transmission of natural gas in Germany (Section C),

- underground storage facilities for natural gas servicing the south of Germany (Section C),

- group I base oils in the EEA (Section D),

- motor fuel retailing in Austria, Germany, Luxembourg, Netherlands and the United Kingdom (Section E),

- motor fuel retailing on toll motorways in France (Section E),

- aviation lubricants worldwide (Section F),

- aviation fuels at Gatwick Airport (Section G).

A. EXPLORATION, DEVELOPMENT AND PRODUCTION PROCESS

(10) Upstream activities comprise three types of commercial activity: the finding of new reserves, the development and the commercial exploitation of those reserves. The finding of new reserves is generally described as "exploration". Development concerns the setting up of adequate infrastructure for future production (oil platforms, pipelines, terminals, etc.). The exploitation of reserves is called "production and sales". Previous Commission decisions(7) focused primarily on the impact of the then notified transactions on the production and sales segment.

(11) Exploration and development is a time and capital-intensive activity, which includes a number of successive steps. Firstly, countries which believe that hydrocarbon reserves could be discovered on their territory (such a country is called the "host country") organise bids to grant exploration licences. The bidding processes vary from public auctions in the USA to situations where bidders are pre-selected by the host country. This pre-selection can be based to varying extents on technical capability, financial strength or other considerations (lobbying, cultural affinities, etc.).

(12) Often, companies bid together in order to spread risks or to bring complementary skills. In some circumstances, host countries may require that some bidders join together in order to get the licence for exploration. Host countries may also require that their national oil companies be granted a share in any discovery made in the block under licence. In all cases when the licence is granted jointly to a number of companies, an operator for the block is named. The role of the operator is to manage technically and financially the exploration and possibly the development and production phases of the project. Most of the important decisions require unanimity from all the partners in the project.

(13) The total time between the granting of a licence and actual production typically ranges from five years to 15 years. Costs of exploration and development projects can reach the region of EUR 7 billion for projects in the "frontier areas". The split of expenditures between the exploration stage and the development stage is typically respectively 15 and 85 %.

(14) It must also be mentioned that when production is started in a new area, infrastructure needs to be developed. This includes notably oil platforms, pipelines and terminals. Smaller fields, which would not be economical to develop on their own, can then be exploited through using the infrastructure already put in place for the bigger field. Such smaller fields are sometimes called "satellite fields". Exploration in the North Sea has typically followed that process.

RELEVANT PRODUCT MARKETS

(15) It is often argued that exploration, development and production and sale are too closely inter-linked for more than one distinct product market to be defined. It may however be argued that exploration constitutes a separate product market. Indeed, it could be argued that a firm engaged in exploration has two categories of clients: the host country, to which it has generally undertaken to provide a future flow of resources in case oil or gas is found, and the subsequent producers and sellers of oil and gas that may buy it out or join in the development and production. The transfer of rights over blocks is a widespread practice in the industry.

(16) Since the possible contents of the underground are not known at the time of exploration, the Commission considers that it is not justified to make a distinction between exploration for oil and exploration for natural gas. As regards the development, production and sales product markets, since gas and crude oil have differing applications and are subject to varying pricing behaviour as well as cost constraints, the Commission considered it appropriate at the stage of the decision initiating proceedings to define one relevant product market for the development, production and sales of crude oil and another relevant product market for the development, production and sales of natural gas.

(17) The parties have contested the existence of an exploration market arguing that it is too closely linked to the subsequent steps of the production process to make it a possible product market. As will be explained in the section on effects of the merger on competition, it is not necessary for the appraisal of the effects on competition of the notified operation to define precisely the relevant product markets according to the various stages of the exploration, development, production and sales of crude oil and natural gas, respectively.

RELEVANT GEOGRAPHIC MARKETS

(18) The Commission considers that the exploration market is worldwide in scope and agrees with the parties that from a European-demand perspective the relevant geographic markets are worldwide in scope for development, production and sales of crude oil and probably include the EEA, Algeria and Russia for gas.

(19) The Commission examined whether smaller geographic areas could constitute relevant markets. Some purchasers of gas make, for security of supply reasons, a distinction in their purchasing policy as to the origin of the gas on the basis of the perceived political risks and their influence on the security of supply with regard to Russian and Algerian gas (Russia accounted for approximately 17 % of total western European supply and Algeria for 12 %). The most explicit example of this is the Spanish legislation ordering that not more than 60 % of the natural gas demand can come from the same country. It is, however, unlikely that a price increase of gas produced in the EEA would not be defeated by a small increase in the proportion of gas purchased from Russia and Algeria. In any case, the precise definition of the relevant geographic market can be left open, as this does not alter the conclusion of the competitive assessment.

MARKET PLAYERS

(20) Traditionally, companies involved in the exploration, development and production and sales business were grouped in three categories with different profiles, strategic objectives and future exploration and development capabilities. The categories are: (i) the State-owned (mainly OPEC and some OECD-based companies such as Statoil) producers, (ii) the "majors", that is, vertically integrated oil companies with international activities and (iii) a multitude of substantially smaller oil companies, most of which are non-integrated upstream explorers and/or producers.

(21) The competition concerns identified by the Commission were raised mainly by the fact that, as a result of the notified transaction, as well as of that notified by BP Amoco and Arco, a distinct (fourth) tier of players might prevail on the exploration and development activities. This group of "super majors" would be composed of Exxon-Mobil, BP Amoco-Arco and Shell.

POSITION OF THE PARTIES IN THE MARKET

Exploration

(22) There is no widely available or accepted indicator of market power for the exploration market. Measurement of market power could be carried out on the basis of the number of blocks where licences have been granted, the number of blocks where a given company is the operator, net acreage, total acreage under operatorship, capital expenditure in exploration, etc. There is wide agreement that proxies based on number of blocks or acreage would not be reliable as they cover widely varying situations.

(23) On the basis of capital expenditures, Exxon-Mobil, BP Amoco-Arco and Shell would represent together, with roughly equivalent shares, between 30 and 40 % of the total capital expenditures in exploration, development and production scheduled for 1999. The next largest player would represent one third of the average size of the three main players.

(24) The parties have contested the adequacy of this measure of market power for the following reasons. First, capital expenditures vary according to the sites explored. This biases downwards the market share of national oil companies in low-cost OPEC areas. Secondly, capital expenditures often do not reflect the correct share of ownership of the licence or of the output. This is because State companies often pay a share which is disproportionately lower than their licence interest or take of crude oil output. Thirdly, capital expenditures often include expenditure on investments such as gas liquefaction and infrastructure rather than sums invested purely in exploration.

(25) Another possible way to measure market power would be to calculate the share of expected production of these companies (since current exploration and purchase of third party rights will translate into future production). On the basis of the market investigation, the combined market share of the three super majors of the non-OPEC production, in some 10 years from now, would again be in the 30 to 40 % range.

Production and sales

(26) Exxon and Mobil would have together around [0 to 10]* % and [10 to 20]* % of, respectively, the worldwide production crude oil and the European production of natural gas in 1998. These numbers would be 3,7 % and 8,4 % for a combined BP Amoco-Arco and 3,3 % and 13,9 % for Shell. These numbers have not materially changed over the years. Of the total natural gas production in the EEA, the three super majors account for almost [30 to 40]* %. However, of the natural gas consumed in the EEA, the three super majors account for less than [20 to 30]* %.

(27) The industry often refers to proven reserves as a proxy for market power. On this basis, the parties would hold approximately [0 to 10]* % and [10 to 20]* % of, respectively, the worldwide proven reserves crude oil and European wide proven reserves of natural gas. However, it may be argued that proven reserves are not a relevant indicator of future production. Proven reserves as stated by the oil companies have just a stock function and, as any other private company, private oil companies need to limit their stocks as much as possible. By means of its proven reserves, a company indicates that it is (or is not) able to replace the foreseeable depletion of existing oil fields and thereby to remain a viable player on the market. However, national oil companies do not need to justify further exploration and development investments to their shareholders since they benefit from the national reserves, which usually largely exceed their needs. Therefore, such an indicator involves an upward bias in favour of national companies.

(28) Currently and for the foreseeable future, the OPEC producers collectively enjoy substantial market power in crude oil production. They control a substantial amount of production and proven reserves (approximately 40 % and 75 %, respectively, whereas Saudi reserves account for more than one third of OPEC reserves). OPEC has been able in a number of instances in the past to influence(8) crude oil prices up to a certain level.

(29) Gazprom (17 %) and Sonatrach (11 %), the national Russian and Algerian suppliers, account for approximately 30 % of EEA sales of natural gas and their reserves account for 88 % (Gazprom 81,5 %) of the combined proven reserves that can be sold in the EEA.

EFFECTS OF THE MERGER ON COMPETITION

(30) The Commission raised the following serious doubts as to the effect on competition in the markets for exploration, development, production and sales of crude oil and natural gas. The notified acquisition together with the BP Amoco-Arco merger could lead to the creation of a fourth tier of competitor in the industry composed of Exxon-Mobil, BP Amoco-Arco and Shell. Whatever the parameter retained (market capitalisation, oil and gas production, proven reserves, etc.) there is now a material gap between these three super majors and the remaining majors.

(31) It was considered whether it would be possible that the three super majors would in the future be in a privileged position to find and develop new important reserves compared to the other players in the market. This could come from their increased financial strength that would now enable them to hold a portfolio of risks of increased size. On the contrary the other competitors would have to pick and choose the areas where they intend to become active. Because of their less favourable access to capital and capacity to spread risks, smaller companies would, if they wanted to continue developing important fields, have to become smaller partners to the super majors to benefit from their combination of resources. Their only other alternative would be to become niche players focused on exploration. This may result in the super majors controlling the access by other explorers to the new frontier areas first at the time of the granting of the first licences and then for further surrounding blocks since the super majors will have built the necessary infrastructure.

(32) Since the typical time lapse between the first stages of exploration and the start of production varies between five and 15 years, in this scenario there would be a possibility that within some 10 years, new non-OPEC reserves and production would be influenced to a significant extent by the super majors.

(33) This could affect the production and sales markets in the following way. The competitive constraints limiting OPEC's ability to function as a cartel would be lessened. These three companies would have the same interests as OPEC and would be likely to align themselves on the OPEC decisions by limiting output to a certain level without fearing that others would take advantage of this. This would translate into the strengthening of OPEC's dominant position in the crude oil market through the creation of an oligopolistic structure combining both OPEC and the three super majors, thus creating incentives for alignment on OPEC's commercial strategies. OPEC would then be in a position to raise and sustain prices at the highest price threshold above which new exploration would be triggered.

(34) On the basis of the foregoing, the Commission concluded that the operation raised serious doubts as to its compatibility with the common market and the functioning of the EEA Agreement, which should be further explored in an in-depth investigation.

(35) The parties contested the Commission's serious doubts on mainly two grounds. First, the super majors would still be facing competitive constraints from smaller oil companies. Secondly, host countries control oil and gas production and in any case would have no incentives to let oil companies restrict production.

(36) The market investigation confirmed that small explorers do not seem to believe that the emergence of a new class of super majors would threaten their position. Because of size differences they would not compete for the same type of exploration rights and they would not be dependent on the bigger explorers to sell their oil. In addition, majors such as Chevron, Texaco, Elf or Total would all seem to have unchanged capacity to explore and develop fields anywhere in the world. The parties have provided numerous examples of the involvement of majors in current projects all over the world.

(37) Second, the market investigation confirmed that typical concession contracts between governments and explorers-producers prohibit the latter from limiting their output (and leave such a decision to the discretion of the government).

(38) As for natural gas, it is unlikely that the parties together with the other super majors could control EEA production and therefore that competition would be affected. Even if such a market did exist (in view of the perceived political risk attached to the gas production in Russia and Algeria), no collective dominance between the super-majors can be argued in view of the strong position of Norwegian gas delivered by competitors. In addition, producers of natural gas face very concentrated demand in the hand of the national wholesale gas transmission companies such as SNAM (Italy) or Transgas (Portugal).

(39) Therefore, the Commission concludes that the concentration will not lead to the creation or strengthening of a dominant position on the markets for the exploration, development, production and sale of crude oil and natural gas.

B. GAS-TO-LIQUIDS TECHNOLOGY

(40) The Commission expressed serious doubts in its Article 6(1)(c) decision that the notified transaction creates a dominant position with regard to gas-to-liquids technology.

(41) Gas-to-liquids technology (GTL) is a process of conversion of natural gas into finished fuels. GTL technology enables the conversion of natural gas into liquid form at the site of exploration and production, and thus facilitates the commercial development of gas reserves and transportation of natural gas from remote locations that are not accessible by pipelines. In addition, it has the advantage of producing high-purity products.

(42) At present GTL processes cannot produce economic substitution products for conventional refining products. Customers' demands for GTL products are expected to come first from the lubricants and petrochemical industries, due to the high purity and molecular arrangement of feedstock generated by GTL technology. Subsequent demands may include the other petroleum products.

(43) There are two GTL pathways at present, that is Fischer-Tropsh (F-T) and methanol, or methanol-to-gasoline process (MTG), which Exxon and Mobil have each developed respectively. Information collected in the market indicated that Exxon has developed the single largest patent portfolio directed to the F-T process, and holds foreign and US patents directed essentially to every significant step in the pathway. Industry sources stated that [...]*.

(44) Other oil and gas companies have limited patent positions in one or the other, or sometimes both, processes. However, none of these companies have positions that cover all of the essential steps of either pathway.

(45) The patent positions of Exxon and Mobil currently represent alternative, although not substitutable, positions. Post-merger, the strong patent position of the parties would create disincentives to any potential new entrant. Finally, the substantial financial resources necessary to compete in GTL (for example: currently a F-T plant costs [between USD 1 and 3]* billion to build) would prevent both expansion by incumbents and entry by potential competitors.

(46) In response to the Commission's serious doubts, the parties explained that GTL technology did not constitute a market in itself and that the quantities of petroleum products that such a technology could enable to produce in the next decade would be minute compared to those produced through the traditional refining process. In addition, other companies would hold patent portfolios sufficient to enable them to develop the technology.

(47) The Commission agrees with the parties that the notified concentration, even assuming that they would hold an essential portfolio of GTL patents, will not give rise to any dominant position given the small impact it would have on petroleum products markets.

C. NATURAL GAS

INTRODUCTION

Description of the "gas column"

(48) The natural gas that streams from the well contains elements that need to be removed before the gas reaches pipeline quality, composed almost entirely of methane and ethane, and can be injected into the high-pressure distribution pipelines. This gas is sold to the wholesale transmission companies that, in turn, sell the gas to the end consumers

(49) There are two levels of gas wholesale transmission companies. There are the "long-distance" wholesale transmission companies and the "short-distance" wholesale transmission companies. Long-distance wholesale transmission companies purchase gas from gas producers and transport this gas over long-distances through high-pressure pipelines. They then either sell to other more local wholesalers, namely the short-distance wholesale transmission companies, or directly to end customers or provide international long-distance transit services to other long-distance wholesale transmission companies located in other countries.

(50) In addition to the supply of gas, the services offered to the end customers include:

- the provision of "swing": consumption of gas is not evenly spread around the year. This is particularly important for the local distribution companies (LDCs) that supply via their low-pressure pipeline grid the households who require, for their heating purposes, a lot of gas in the winter and much less in the summer. A key element to provide swing is access to storage, mostly in the form of underground storage facilities,

- the provision of other services such as quality control (ensuring that the gas provided is within the agreed specifications), balancing (ensuring that the input of gas in the pipelines is in line with the required output), back-up (agreements with other wholesalers and/or producers interconnected with the own pipeline network in which they guarantee supply if the proprietary supply falls out) and metering.

(51) The end customers can be grouped in three categories: power generators, (firm or interruptible) industrial customers and LDCs. These customers differ in the following way:

(52) There are important price differentials between the different customer categories, even when their gas demand is the same. This is so because prices are traditionally based upon the value of the customer's alternative energy source (Anlegbarkeitsprinzip). This refers to the alternative energy source to which a customer might switch: heating oil for households, heavy fuel for industrial users and, to a certain extent, nuclear energy or coal for power generators.

(53) There are also qualitative differences when considering the supply to the different customer categories. In this respect, reference is made to the" swing" requirement for supplies to LDCs and the difference between firm (their gas supply cannot be interrupted) and interruptible industrial consumers. These aspects equally have an impact on the price.

(54) The last difference is the legal framework resulting from Directive 98/30/EC of the European Parliament and of the Council(9) (hereinafter "the Gas Directive") and national implementation measures (liberalisation). The Gas Directive requires Member States to open up their markets before 10 August 2000 so that by then gas-fired power generators, irrespective of their annual consumption level, and final customers consuming more than 25 million cubic metres (MCM) per year on a consumption site basis, are free to contract for natural gas ("eligible customers"). The total opening of the market must be at least 20 %, increasing to 28 % in 2005 and 33 % in 2010. The threshold for eligible customers will be reduced to 15 MCM in 2005 and 5 MCM in 2010.

(55) For the organisation of access to the system, the Member States may choose between two procedures which must operate in accordance with objective, transparent and non-discriminatory criteria. The first procedure concerns the "negotiated access" whereby natural gas undertakings and eligible customers negotiate voluntary commercial agreements and whereby the relevant gas undertakings have to publish the terms of use of their system before 10 August 2001 and on an annual basis thereafter. The second procedure is the "regulated access" whereby natural gas undertakings must give access on the basis of published tariffs set by a relevant authority in the Member State.

(56) The Directive allows for an important conditional derogation procedure whereby the Commission can accept, under certain conditions, that gas undertakings can refuse access if this would cause them serious economic and financial difficulties because of take-or-pay commitments accepted in one or more gas purchase contracts.

(57) In the Netherlands, the recently introduced proposal for a new gas law goes considerably further than the minimum requirements of the Gas Directive. Consumers with an annual consumption of 10 MCM (accounting for 44 % of total annual Netherlands consumption) are already free to purchase gas. As from January 2002, clients with an annual consumption of between 170000 m3 up to 10 MCM will be free. As from January 2007 all consumers (including households) will be able to source their gas freely. There is no special regime for power generators. The Netherlands has opted for the "negotiated access" and Gasunie has published its TPA tariffs in August 1998.

(58) In Germany, the wholesale transmission companies traditionally had regional monopolies and had concluded regional demarcation agreements, i.e. agreements whereby the wholesale transmission companies agreed not to undertake wholesale activities in each other's designated territory. This is forbidden since 29 April 1998, the date of the entry into force of the Energy Industry Act. The Act makes 100 % of the German market open, i.e. all customers are "eligible customers". The act also stipulates that municipalities have to make available rights of way for the construction of pipeline grids. The new competition act stipulates that network operators must not deny third parties access to networks without justified, objective reasons. In case of a refusal, the matter can be submitted to the competition authority or directly before the administrative court. Since February 1999 the industry associations are negotiating the "Verbändevereinbarung Gas" which should lay down the criteria whereby (negotiated) third-party access is going to be given, together with indicative prices.

Contractual framework

(59) The European gas industry has operated on the basis of long-term contractual arrangements between buyers and sellers. The parties have indicated that in order to achieve the finance necessary for the investment in the production and distribution infrastructure, the gas producers had to conclude long-term agreements with their buyers. A gas field would only be taken in production if a buyer for the gas, typically one or more national wholesale transmission companies, has been found.

(60) On the basis of such agreements, investments would also take place in the necessary distribution infrastructure (terminals and pipelines). Contracts are concluded sufficiently in advance of the actual production start to allow the necessary infrastructure to be built. In turn, the wholesale transmission companies conclude contracts with their clients whereby they endeavour to match their aggregated buying volume with long-term contracts with their clients. It can be said that traditionally the wholesale transmission companies balanced the demand with the supply.

(61) Typical features of the long-term contracts are the take and/or pay commitments for the volumes. The contractual price on each level is typically constructed as a netback price minus an agreed margin for the client. A contract will also provide for three-yearly price-reopeners whereby the price might be reviewed if there is a material change in the market of the buyer (gas-to-gas competition, change in alternative energies and so forth). The contracts also contain a parallel sales clause whereby the buyer is entitled to lower his take obligations with the volume that the seller sells directly in the buyers' territory. The gas companies [i.e. the buyers]* argue that this is necessary to avoid the producers selling their gas twice (once to the wholesale transmission company and a second time directly to the wholesale transmission company's client). On the other hand, such agreements are an important element of stability in the gas distribution chain as it maintains a certain hierarchy of producer - wholesale transmission company - end consumer.

(62) The policy to price gas by means of the alternative energies of the buyer's clients, also called the market value principle, is in the view of the European gas companies the key element to explain the building and expansion of a European gas infrastructure and, at the same time, allows the consumers to benefit from competitive prices. Maximising the income stream from gas sales in this way and maintaining the acceptance by all gas players of this principle is thus a key concern for the incumbent gas operating companies.

(63) [...]*

Brief overview of the gas column in the Netherlands and Germany

(64) The Netherlands is one of the largest gas-producing countries in the EEA. It is a net exporter of gas. The largest equity producers in the Netherlands are NAM, a joint venture (JV) between Exxon and Shell, and EBN, the Dutch State's oil and gas company. Exxon and Shell and the Dutch State (via EBN and directly) own Gasunie, the wholesale distribution company to which almost all Dutch gas is sold. Gasunie sells [80 to 90]* % of the gas consumed in the Netherlands and owns the total Dutch gas wholesale infrastructure (with the exception of one pipeline). Gasunie has concluded long term exclusive supply agreements with all the regional distribution companies concerning sales to clients with an annual demand [below a certain level]*.

(65) There is gas production in Germany. This accounts for approximately 20 % of the German gas demand. The main producers are BEB, the German production and distribution JV between Exxon and Shell, and Mobil. Around 80 % of the gas consumption is imported from Norway, Russia and the Netherlands. The most important importer and wholesale transmission company in Germany is Ruhrgas. Exxon, Shell, Mobil and BP together with German coal and steel producers are the owners of Ruhrgas. Other importers are BEB, Thyssengas (where Exxon and Shell each have a 25 % interest and RWE-DEA has 50 %; together these companies control Thyssengas) and VNG (the eastern German wholesale transmission company, where BEB and Ruhrgas are shareholders). Mobil and EWE, a company in which the parties do not have an interest, also import some gas for German consumption. All of the above are the traditional long-distance wholesale transmission companies whose position has been challenged by Wingas, a JV between BASF and Gazprom (the Russian gas producer), by means of their newly constructed pipeline network.

(66) Gas is sold by the German producers and importing wholesale transmission companies (referred to as "long-distance wholesale transmission companies") to a series of regional wholesale transmission companies (referred to as "short-distance wholesale transmission companies"). The around 19 long and short-distance wholesale transmission companies sell on the gas to around 700 LDCs who in turn supply the households.

WHOLESALE TRANSMISSION MARKETS

Relevant product market

(67) The parties consider that the distribution and marketing of natural gas can be divided into a wholesale transmission segment and a local distribution segment. Wholesale transmission companies aggregate gas from producers. They then sell it to a mix of a second tier of wholesale companies (the German short-distance wholesale transmission companies), local distributors and large end-users. Transmission companies utilise high-pressure pipeline systems and storage facilities to deliver gas at the location and volumes required by their customers. Local distributors sell on the gas to end-users (such as households).

(68) The parties consider that the suppliers at the wholesale and local distribution levels of trade tend to be different. The parties have no direct presence in the local distribution segment.

(69) In addition to the distribution and marketing of natural gas, the parties qualify in the notification "natural gas transportation - pipelines and LNG shipping" and "natural gas storage" to be separate markets. The Commission agrees with the parties that natural gas transportation and natural gas storage constitute distinct product markets.

(a) Wholesale transmission versus local distribution market

(70) Long-distance and short-distance wholesale transmission companies sell gas via high-pressure pipelines to industrial customers, power generators and local distribution companies. Local distribution companies sell on the gas to end-users such as households and residential customers via their extensive (street-by-street) low-pressure pipeline network grid. Local distribution companies buy from the wholesale companies the swing and other related services. The Commission therefore agrees with the parties that local distribution constitutes a distinct product market. As the parties are not directly active on this market, this product market is not further assessed in the scope of the present procedure.

(b) Long-distance versus short-distance wholesale transmission market: long-distance wholesale is a distinct product market where entry depends on the goodwill of the incumbents or on a very high sunk-cost investment

(71) As explained above, the German gas wholesaling industry is characterised by a two-tier structure: long-distance transmission companies sell gas that they have imported directly from foreign producers and transported over long-distance pipelines to the short-distance companies. It needs to be assessed whether this type of service could be easily substituted by the short-distance companies in case of a hypothetical increase in the prices charged to short-distance companies. In such circumstances, the only way short-distance companies could defeat a price increase would be to get directly the gas from the producers, i.e. to integrate upwards in the vertical chain. To do so, short-distance companies (or possibly the end customers) would need to get similar pricing conditions from gas producers to those enjoyed by long-distance companies and be able to transport the gas over long-distance. From its market investigation, the Commission concludes that even in the face of a price increase the short-distance companies would lack the incentives to by-pass long-distance companies.

(72) [...]*

(i) The parties' view: everybody can import gas

(73) The parties consider that the relevant market consists of the sales of German and imported gas by the around 19 long-distance and short-distance wholesale transmission companies to industrial customers, power generators and local distribution companies. The parties consider that the specific activities of the long-distance transmission companies is nothing more than a technical complication as everybody is able to bring gas to the German border.

(74) First, the parties indicate that imports of small volumes are possible. They refer to a couple of small-volume import contracts such the ones [...]*.

(75) Second, the parties consider that the cost of entering into the business by means of investment in reception facilities, pipelines and storage is a natural feature of a long-term capital intensive industry and does not constitute a barrier to entry. Specific reference is made to the investment made by Wingas to illustrate that it remains profitable to invest in German gas infrastructure. The parties also point out that investments are not needed to import gas [...]*.

(76) Third, the parties refer to the provisions of the new German Energy Law and the legal obligation for pipeline owners to offer spare capacity to third parties. Therefore, sales to end consumers in Germany are a realistic possibility and reference is made to the decisions of a certain number of local distribution companies to purchase directly from the producers and the fact that industrial groups purchase centrally.

(ii) This is not corroborated by the Commission's market investigation

(77) The market investigation has not corroborated the parties' opinion that importing gas is merely a "technical complication". The following statements have been made by some of the 11 German short-distance wholesalers:

- "To achieve interesting prices by way of an import contract, one is required at present to buy very big quantities with buying obligations as well as, amongst others, own investments or financial participations in projects or companies. Such an activity is hence typical for companies that bundle a big demand and have large financial means to bear the import-related cost risk. When importing, the supply security by way of diversification of the supply sources plays an important role."

- "To import, one would require a parallel pipeline network to the producers (or to the transmission points at the border) or TPA. The short-distance wholesaler could not estimate the necessary costs and time involved."

- "It is difficult to sell additional import quantities in Germany to the clients as these quantities can only go to regional distributors. Marketing of gas directly to end consumers is only possible with considerable efforts and proves to be unrealistic at present."

(78) In their reply to the statement of objections (the reply), the parties have re-stated that purchasing small volumes is possible, that the costs for importing are readily available (by means of own research or via available consultant firms), that the costs for import infrastructure are typically less than [0 to 5]* % and that supplies and customers are accessible (demand is growing and existing contracts expire). The Commission recognises that the costs for importing can be calculated. The other points will be dealt with further below.

(79) It can be noted that the Bundeskartellamt has in a recent procedure(10) concluded that the long-distance wholesale transmission market constitutes a distinct relevant product market.

(iii) German gas cannot challenge a price increase of imported gas

(80) German gas is all sold in Germany, where it competes with imported gas. However, German gas production accounts currently for around 20 % of total consumption and is expected to decrease to [less than 15]* % by 2010. Wholesale transmission companies that want to satisfy their demand therefore require increasing access to imported gas, i.e. to the services of the long-distance wholesale transmission companies.

(81) In addition, German gas is committed to a very large extent to the long-distance wholesale transmission companies so that, in general, the availability of German gas for other companies is too limited to challenge a price increase of imported gas. German gas production and reserves are held for around [70 to 80]* % by the long-distance wholesale transmission companies BEB ([40 to 50]* %) and Mobil ([20 to 30]* %). The remaining [20 to 30]* % of German production (thereby accounting for about [0 to 10]* % of German demand) is done by RWE-DEA with around 7,5 %, Wintershall and Preussag with each around 5,5 % and Erdöl-Erdgas Gommern with around 4,5 %. RWE-DEA commit to sell 60 % of their annual production to Ruhrgas and Preussag commits to sell 45 % of its annual production to Ruhrgas. The largest part of Erdöl-Erdgas Gommern's production is sold via VNG, another long-distance wholesale transmission company.

(iv) Access to foreign gas at competitive conditions is not and will not be feasible post merger: the potential entrants in the long-distance wholesale transmission market depend on the services of the incumbent companies

(82) The day-to-day business of any short-distance wholesaler depends, to a significant extent, on the pipeline network, storage facilities and other services provided by the long-distance transmission companies. To upset such essential suppliers through by-passing them for part of their services would be a risky decision in view of the scope for retaliation from the long-distance wholesale transmission suppliers.

(83) However, there is one short-distance wholesale transmission company that does not depend on the wholesale transmission services of long-distance transmission companies. This is Erdgas Münster (EGM), a company that sells German gas produced by its shareholders, the most important German gas producers: BEB ([20 to 30]* %), Mobil ([20 to 30]* %), Wintershall ([20 to 30]* %), Preussag ([10 to 20]* %) and RWE DEA ([0 to 10]* %). The shareholders are obliged to sell their whole production in a specific part of Northwest Germany via EGM (accounting for [less than 10 %]* of total German gas consumption) and, in this way, the company is assured of its gas supplies [...]*.

(84) [...]*, BEB and Mobil would, after the concentration, control the strategic decisions of EGM. Such decisions include the position of EGM in a future liberalised market environment (see further below). The parties claim that they would not control EGM as [...]*. This reasoning cannot be accepted. Exxon exercises joint control in the sense of the Merger Regulation over BEB and, in addition, it would make economic sense for Shell to agree to [...]* as, thereby, Shell would also benefit from the advantages of being able to control EGM.

(85) Although both Mobil and BEB are long-distance wholesale transmission companies, Mobil's position on that market is substantially smaller than that of BEB. Therefore, Mobil would have an interest, pre-merger, to maintain the possibility to use EGM as an instrument in a future liberalised market environment. There is, therefore, not necessarily an alignment of interest between BEB and Mobil. This ability to be a potential competitor would be suppressed post-merger through the establishment of a formal control on EGM.

(86) The parties have also made some references to direct calls for tenders to foreign suppliers by a couple of relatively minor local distribution companies. In the call for tender, the local distribution companies, all of them Wingas customers, request quotes for supplies to the German border and indicate that they would organise the transport within Germany themselves. Another example mentioned by the parties is the decision of Stadtwerk (SW) Viernheim to purchase directly. However, it results from the Commission investigation that these calls for tender were not successful.

(87) The fact that there may be sporadic calls for tenders does not mean that the local distribution companies consider that it is feasible to transport themselves the gas within Germany. It therefore does not mean that they consider that they act in competition with the incumbents.

(88) The producers that would want to enter the market do not only rely on the German companies to sell their gas in Germany, but they also rely to a varying extent on the incumbents for their transit of gas to other export markets. For instance, Dutch gas transits the pipelines of the German incumbents to reach Switzerland and Italy (and possibly even Poland), Norwegian gas transits Germany to reach Switzerland and Russian gas transits Germany to reach France.

It is unlikely that smaller companies could get interesting prices or volumes or that traditional suppliers would by-pass the incumbents

(89) The long-term (10, 20 or 30 years) take and/or pay contracts whereby the off-take obligation of the wholesale transmission company is reduced by the amount that the producer sells directly (or via other companies) in the area of the transmission company (also called "parallel sales clauses"), increases the barriers for other companies to obtain supplies.

(90) Furthermore, the suppliers have no interest in by-passing their traditional clients if this would lead to gas-to-gas-competition (GTGC) reducing the prices obtainable in the market. And if the supplier is not willing to offer lower prices, there is no reason for the customer to switch. GTGC and price reduction is therefore the most likely outcome. As the producer receives net-back prices, such a price reduction would affect all gas sales. Selling to newcomers therefore puts the producers at risk of possible damaging price competition. [...]*

(91) According to [...]*, parallel sales clauses stimulate the wholesale transmission company and the supplier to compete for the same end-user and are thus pro-competitive. However, such competition is only likely to happen if the profit made by the supplier on its direct sales more than compensates both the short-term loss of net-back value on the totality of the remaining contracted sales volume to the wholesale transmission company and the long-term loss of risking to lose sales to the wholesale transmission company when the existing contract ends. This is unlikely to happen so that parallel sales clauses also act as a barrier to entry.

(92) The Commission also notes that three of the most important German long-distance wholesalers have purchased a substantial part of their non-German origin gas from Norwegian, Danish and Russian producers in common (Ruhrgas [...]* %, BEB [...]* % and Thyssengas [...]* %). By combining their demand, the consortia should be able to get, via long-term contracts, better prices than any other company that would negotiate the contract on a stand-alone basis. The parties disagree with this and indicate that purchase prices are unrelated to volume, but are set by alternative fuels' prices. The Commission cannot agree to this argument as, if it were taken to its logical extreme, it would imply that no price competition is possible within the European gas industry as everybody would charge the same price, namely the alternative fuels' price. This is clearly not correct as parties have a wide freedom to assess the parameters of the gas price, even using links to alternative fuel indexes.

(93) It also has to be noted that buying in common reduces the incentive of the producer to by-pass the wholesale transmission companies in selling directly or indirectly to others in the their territory. This is so because, when several companies combine their purchases, the producer risks losing all of his/her clients even if he/she would only by-passes one of them.

(94) And even if, as the parties argue, other incumbents such as VNG and Mobil can purchase gas directly and individually at similar prices, this does not imply that the same may be possible for newcomers. The incumbents will endeavour to use their position and influence vis-à-vis the producers to push them not to supply to competitors. Wingas has indicated that it had concluded a number of years ago a supply agreement for Norwegian gas, but that the GFU (the Norwegian gas sales organisation) had refused this. In addition, [...]* has stated that "[...]* will continue to try to prevent that [...]* is able to buy [...]* gas."

(95) Even if new volumes might be contracted, the long-distance wholesale transmission companies have concluded long-term (take or pay) supply agreements with the most important customers (short-distance supply companies, local distribution companies and so forth). It may therefore be difficult to sell the volume so that these long-term agreements form an additional barrier to entry in the long-distance market. The Commission recognises that these contracts all end one day and that gas demand is growing by 2 to 3 % per year. These contracts are, therefore, not an insurmountable barrier to entry, but they undeniably make entry more difficult than in a market where there are no long-term commitments between buyers and sellers.

(96) Finally, the German long-distance wholesale transmission companies have spent a lot of money in recent years and will probably continue to do so in downstream vertical integration by buying equity participations in short-distance wholesale transmission companies and in local distribution companies. [The Commission considers that]* this is aimed at maintaining the current supply structure. This will further increase the barriers to entry into the market.

(v) Access to the infrastructure or duplication are no options

Duplication of infrastructure would involve excessive sunk costs

(97) The important capital investments of the incumbents in reception facilities, pipelines, storage and quality adjustment cannot be used for any other purpose although some gas pipelines might be converted against a substantial cost in oil pipelines. The investment can therefore be considered to be, for a very substantial part, a sunk cost.

(98) The parties indicate that the cost of entering the gas business cannot be reasonably construed as a barrier to entry as the cost of the import infrastructure is typically less than [...]* of the value of the gas purchased. However, the import infrastructure is not the only infrastructure required (the most important part might be the long-distance pipeline network) and it remains that the important capital investment is, for a very substantial part, a sunk cost.

(99) In addition, it is always cheaper and quicker for the incumbents to increase their capacity of pipeline transport by adding compressors than it is for a newcomer to build a new pipeline.

(100) It is, therefore, not expected that a newcomer to the market will start constructing a new, major high-pressure pipeline network in Germany. In other words, the "Wingas experience" is not likely to be repeated. Wingas is a sort of a "lucky combination" between a very big (probably even the largest) German industrial gas consumer that was not happy with the prevailing prices and a very big Russian producer that had a conflict of interest with the incumbents. This conflict of interest will be further explained in the competitive assessment part.

(101) In addition, the net-back that the supplier can expect to achieve from renewing the contract with the original wholesale transmission company is greater than the expected net-back from a newcomer who has to make new investments (and therefore requires a higher margin). Furthermore, such a newcomer would have to "fight" to get his volumes sold, which will lead to price reductions for all players on the market and this is again reflected in a reduced net-back price for the producer.

The industry-proposed third-party access will have a limited impact on the ability of customers to directly import gas in competition with the long-distance transmission companies

(102) For the purpose of this decision, the Commission is not satisfied that the possible conclusion of the Verbändevereinbarung Gas (VV Gas) will lead to market entry by means of third-party access. This conclusion is based on the current state of play, namely the proposal by the German gas industry within the VV Gas negotiations.

(103) The VV Gas is an initiative of four trade federations: (a) BGW, the association of the German gas and water industry; (b) VKU, the association of the local distribution companies; (c) BDI, the German industry association (industrial consumers) and (d) VIK, the German Power Generation Association. Germany has chosen, in the context of the Gas Directive, for a "negotiated TPA" and the VV Gas wants to set the guidelines for such a negotiated TPA. The German competition authority and/or the Commission will have to assess the VV Gas when adopted. The aim of the VV Gas is, as indicated in a common letter of the four federations to the German Minister for Economic Affairs, "to set a signal to the political environment that no further regulation for natural gas is required. The VV Gas shall fulfil the criteria of market-oriented, competitive, fair, transparent and simple and thereby it has to help in the development of a natural gas stock exchange. The VV Gas shall contain the basic principles and rules for access to the network as well as a simple calculation of the tariffs. The VV Gas shall be concluded before the end of 1999."

(104) Despite this common declaration, it does not seem, on the basis of the "reference case" developed by the gas industry, possible to conclude that by means of this type of TPA it will be possible to import gas in competition with the incumbents.

(105) The first element is that the reference case stipulates that TPA is only possible if pipeline capacity is available and that long-term take or pay contracts shall be taken into consideration. It results from the market investigation that it is difficult to assess whether capacity is available and the incumbents have contracted up to 100 % of their demand by means of long-term take or pay contracts. The VV Gas will therefore not fundamentally endanger the situation of the incumbents.

(106) The second element is that the reference case indicates that the third-party gas has to be compatible with the gas specification in the relevant pipeline network of the pipeline owner. This is defined in such a way that "the specification of the gas is so that it can be taken in the pipeline without adaptation by the transporter; that it has a pressure that enables to take the gas without adaptation by the transporter and that it has a specification that allows usage together with other gas in the network." These might not be easy criteria to fulfil in view of the different gas specifications currently traded in Germany (see the discussion below on LCV/HCV gas and the fact that both LCV and HCV gas is currently sold in Germany within several distinct specifications). Natural gas is thus not a fully homogeneous product and the nature and practicability of TPA for electricity and telecom is different from TPA for gas. Or, as indicated in a document submitted by the parties, "a main difference with the VV Electricity is that for gas a distinct agreement is necessary with each pipeline operator and that the different gas specifications and the pressure differences at the starting and ending point complicate the concrete organisation."

(107) A third element is that the duration of a transport agreement has to be at least a year, with two starting dates every year (1 October and 1 April). It can be noted that, therefore, the TPA contracts will always include the winter peak utilisation so that there is all the more chance that the pipelines will be running at full capacity. In addition, it is the express aim of this requirement to prevent the development of a spot market. The relevance of this is that the development of a spot market would unbalance the traditional system. The role of the wholesale transmission companies would be especially endangered as a spot market would facilitate a direct contact between the end consumer and the producer with reduced margin for the wholesaler as a consequence. [...]* Nothing of this kind is foreseen by the VV Gas.

(108) A fourth element is that "[...]*".

(109) Finally, the VV Gas does not provide for access to storage, for access to quality adjustment services (conversion of LCV/HCV) nor for back-up.

(110) The parties point out in the reply that the final agreement on the VV Gas will require the support of end-users, the German cartel office and the German Government. With so many stakeholders involved in the process, the parties fail to understand the Commission's speculation that the end result will not facilitate both TPA and competition. The Commission has to make its assessment on the basis of the information that it has available. This information is the gas industry's reference case. Hence, to base an assessment on anything else would be speculation.

Conclusion

(111) The Commission concludes that the German long-distance wholesale transmission market constitutes a separate product market.

Low calorific value and high calorific value gas

(112) During the course of the investigation, several third parties have argued that in assessing the impact of the operation on the European gas markets a distinction should be made between "low calorific value" (LCV) and "high calorific value "(HCV) gas. The Commission agrees with the parties that LCV and HCV gas is part of an overall natural gas market. The Commission has reached this conclusion of the basis of demand-side substitution. In other words, a relative price increase of LCV gas of between 5 to 10 % would become unprofitable in view of the fact that consumers in the relevant area in Germany(11) could switch to HCV gas.

(113) The distinction between LCV and HCV gas has its origin in the discovery (1959) and exploitation (1963) of the massive LCV Groningen gas field in the Netherlands, the largest gas field in Europe and one of the largest in the world. This field meant the start of the natural gas industry in Europe. The whole consumption of natural gas was thus oriented on gas having the specifications of the Groningen field. However, the classification of LCV gas contains a broader spectrum than below or up to the Groningen calorific value. Based upon the definition for low calorific gas used in Germany (DVGW-code of practice G 260/I - Gas Quality) LCV gas is defined as having a Wobbe Index of up to 13 kWh/m3.

(114) The specifications of Groningen or LCV gas have an impact on the total downstream "gas column". High-pressure pipeline transmission (carrying gas over long-distances), underground storage facilities, low-pressure pipeline transmission (enabling the gas to be supplied at the appropriate low pressure for the burners used by households) and the installation by which the gas was consumed by the final consumer have been designed and can only run either uniquely for LCV or HCV gas.

(115) Later gas discoveries were of a different quality with a higher calorific value as the gas contains less inert nitrogen. In fact, LCV gas was only found in other onshore fields in the Netherlands and northwest Germany. Today, there are still specific LCV gas columns in the Netherlands, northwest Germany, a part of Belgium (in the provinces Antwerp and Limburg) and a part of France (Nord - Pas de Calais).

Demand-side substitutability: switch of the gas column from LCV compatible to HCV compatible

(a) A relative permanent price increase of LCV gas will speed up the ongoing substitution

(116) The parties have provided charts illustrating the continuous ongoing substitution of LCV gas consumption by HCV gas in Germany since 1975 based on industry statistics indicating both the absolute and relative volume of the different types of gas (HCV, LCV, coke-oven and other gas).

(117) Against this background, the question to be investigated is what the reaction of a wholesale transmission company would be to a price increase by LCV gas producers. The reaction of the wholesale transmission company will be influenced by its possibility to switch itself to an HCV producer and by the incentives coming from the possibility for its customers to switch to another wholesale transmission company that might supply cheaper HCV gas.

(118) Whereas it is technically easy to switch a whole LCV pipeline network to HCV, the main constraint on a wholesale transmission company to change its infrastructure is underground storage. This poses the problem that one needs to change the LCV gas in the storage to HCV gas. According to the parties, the conversion is simple and straightforward. According to Wingas, such a conversion takes time (typically three storage cycles and thus two years for the relevant facilities) and is costly (in view of the varying calorific content of the extracted gas when HCV gas is injected in a LCV facility).

(119) However, already before the end of the third storage cycle the extracted gas may be of HCV quality. Moreover, all wholesale transmission companies have the knowledge of "blending gas" (adding LCV gas to HCV gas to obtain a HCV gas of lower calorific value but within the HCV gas range, or vice versa) to achieve the desired specification. It may therefore be relatively easy for them to adjust the gas that they extract from their storage facilities during the conversion period by means of simple blending to keep it within the LCV or HCV specification. One can imagine that at the start they would be able to add LCV gas to keep the extracted gas within the LCV range, and that gradually they would add HCV gas to bring it to HCV specification. This would probably reduce the costs of the conversion considerably.

(120) On the basis of the above information, the Commission concludes that faced with a permanent relative price increase of LCV gas compared to HCV gas, the technical restrictions faced by the German LCV storage owners are not such that this would stop them switching. Therefore, the relative price increase contributes to the acceleration of the conversion that will have to take place sooner or later.

(b) The customers of the LCV gas supplying wholesale transmission companies have alternatives in the relevant German area to switch to HCV supplies by Wingas

Industrial customers and power generators switching

(121) The market investigation has confirmed that the switching costs for most industrial customers and power generators to adjust their burners are low as a percentage of their annual gas bill. It could, therefore, be economically feasible for an important industrial consumer or power generator to have a new specifically dedicated pipeline built linking the customer to a nearby (up to 50 km at most) HCV high-pressure pipeline. The question therefore is whether there are HCV pipelines in the neighbourhood that are owned by other companies than those that have an interest in LCV distribution. If this is the case, the risk of losing such clients to another wholesale transmission company will make the "incumbent" LCV wholesale transmission company more inclined to undertake the switch itself (if the price was increased by the producer) or would make a price increase by the wholesale transmission company unprofitable.

(122) With regard to industrial demand for gas, a distinction has to be made between Nordrhein-Westfalen (NRW) and Lower Saxony. In NRW, the gas sales per square kilometre are around three times higher than in Lower Saxony and the number of industrial clients per square kilometre is 2,5 times higher. In NRW, the Wingas "Wedal" and "Midal" pipelines are therefore probably in reach of a sufficient number of clients to make a switch possible. In addition, Wingas' underground storage facility is situated in Rehden, on the NRW/Lower Saxony border, thereby enabling the company to provide the necessary swing to its clients in NRW/Lower Saxony. In Lower Saxony, Wingas also has its HCV high-pressure pipeline network. It seems that the only important area that might not be in reach of this network is the Hannover area. However, the available scope is probably enough to make a price increase unprofitable.

(123) It remains, however, that a specifically dedicated pipeline needs to be built from the Wingas high-pressure HCV pipeline to the industrial customer. This takes time and costs money and has, in practice, only proven profitable if the client agrees to an exclusive supply arrangement with a duration of between five and 10 years. The incumbent supplier of the LCV gas is therefore in a position to increase his price over the prevailing HCV price by a margin up to the incremental costs of the new infrastructure. However, such agreements are not unusual in the industry so that the cost of the additional pipeline can be offset profitably as has proven the Wingas experience in other regions.

(124) Wingas argues that in Lower Saxony only a few customers with significant demand are located within reasonable distance of its high-pressure HCV pipeline so that its experience in other regions does not apply in what Wingas considers to be a distinct geographic market, namely the northern German LCV gas area.

(125) The Commission acknowledges that switching costs are not always low for industrial customers. [A company]* has indicated that industrial consumers that use natural gas as a feedstock for a chemical process may face costs of more than EUR 20 million. However, the Commission is not aware of such clients in the relevant German area.

(126) Overall, the Commission concludes that a sufficient number of German LCV industrial customers and power generators have the possibility to switch their demand to Wingas when faced with a permanent price increase of LCV gas to contribute to make such a price increase unprofitable.

LDCs switching

(127) LDCs supply households and other residential customers. They therefore require suppliers that are able to offer them high swing. They would only consider a switch if their new supplier is able to offer them HCV high swing gas. Wingas has its Rehden storage facility in the area and is, therefore, able to supply the swing.

(128) It remains that the distributor and/or the supplier must also absorb the costs involved in switching the households. These costs are estimated to be around EUR 60 per household on the basis of a detailed cost estimate provided, via the parties, by an experienced "household switching" company. On the basis of a typical household consumption of 80 MBTU/year, a local distributor price of EUR 3,4/MBTU and a typical household gas price of EUR 8/MBTU, a price increase of 5 % for LCV gas supplied by the LCV wholesale transmission company reduces the margin of the local distributor by EUR 13,5. A price increase of 10 % reduces the local distributor's margin by EUR 27. Therefore, the switching costs would be totally recouped by the local distributor within two to four years. As the typical duration of a supply contract between a wholesale transmission and a local distribution company is between 10 and 20 years, the 5 to 10 % price increase would be made unprofitable by LDCs switching to an alternative HCV gas supplier. Furthermore, once a household has been switched to HCV, it can safely run on the LCV gas until HCV enters the local distribution grid. The consequence is a temporary and marginal loss of efficiency; however, if this switch takes place in the summer months, the efficiency loss will be minimised.

(129) It has been argued that even if a switch from LCV to HCV gas would be possible, a LDC would not switch if it would become, by doing so, an "HCV island" in an LCV area. This would have important security-of-supply consequences that the customer might not be willing to take. Wingas considers this as the main reason why, despite offering discounts of 5 to 8 % (information from the parties) compared to the existing price level, it persuaded fewer than five local distributors (Stadtwerke) in Germany to switch from LCV to the HCV.

(130) This argument cannot be accepted in the present situation of a permanent relative price increase. In such a situation, the incentives for all the clients to switch would be the same so that, in the end, there would be no island situation anymore.

(c) The alternatives of the incumbent wholesale transmission company's customers will further increase the incentives for the incumbent to switch

(131) When facing a price increase of LCV gas relative to HCV gas, it is normal rational behaviour for a wholesale transmission company that concludes that a substantial part of its customers have the possibility to switch to a competing wholesale transmission company to defend its position. The company will therefore ensure that it itself is also able to offer the cheaper HCV gas. Therefore, the "incumbent" wholesale transmission company will speed up the switch itself.

Conclusion with regard to demand-side substitutability

(132) For the purposes of this procedure, the Commission concludes that a permanent relative price increase of LCV gas would be unprofitable in view of the fact that (a) sooner or later the LCV gas chain has to be converted on HCV gas, (b) that such a price increase will speed up the conversion, (c) that there are alternatives in the relevant German area in the form of HCV supplies by Wingas and (d) that this knowledge will further increase the incentives for the LCV wholesale transmission companies to switch themselves their consumption area.

(133) However, it remains that the costs of switching represent a competitive advantage of the transmission company that is able to offer the prevailing specification as it increases the threshold to change to another supplier.

Relevant geographic market

(134) The parties consider that, in general, the wholesale transmission markets are national. They indicate that distribution has traditionally been organised largely on a national basis through national or regional operators. However, the parties consider that the trend towards cross-border purchases following the advent of the Gas Directive cannot be ignored. The parties consider that attempts to increase post-merger prices in individual Member States would be defeated not only by competition at the national level, but also by increased cross-border purchases by wholesale transmission companies, local distribution companies, power generators and large industrial users.

(135) As the ownership of the infrastructure is still organised on a largely national basis and as the conditions for third-party access will remain essentially fixed on a national basis, the Commission considers that the markets will remain in the near future at most national.

(a) Netherlands

(136) For the reasons indicated in the previous point, the Commission agrees with the parties that the Dutch wholesale transmission market is national.

(b) German long-distance wholesale transmission market(s)

(137) With regard to Germany, there are indications that, at present, the markets may still be regional. The regions would be defined by (a) the previous existing, but now illegal, demarcation agreements such as those that Ruhrgas had concluded with BEB and Thyssengas, and (b) the ownership of the long-distance high-pressure transmission pipelines. Despite the fact that the demarcation agreements are illegal since April 1998, only two agreements with a German client outside the long-distance transmission company's own regional territory have been concluded, both by Ruhrgas for annual supplies of [...]* MCM each and via the construction of a new pipeline. [...]*

The old demarcation regions are still relevant in view of the fear of retaliation

(138) The internal documents of [...]*, [...]* and [...]* confirm the continuing relevance of the old demarcation regions in view of the fear of retaliation.

(139) [...]*

(140) [...]*(12)

(141) [...]*

(142) [...]*

(143) [...]*

(144) [...]*(13)

(145) [...]*

(146) [...]*

Definition of the regions

(147) The regions would thus consist of the "Thyssengas"-region (western part of Northrhine-Westphalia along the Belgian/Dutch border), "BEB"-region (Hamburg, Schleswig-Holstein and parts of Bremen and Lower Saxony), "VNG"-region (Mecklenburg-Vorpommern, Brandenburg, parts of Saxony and Saxony-Anhalt - a region where Ruhrgas does not have its own pipeline network) and the "Ruhrgas"-territory (the remainder of the "Alte Bündesländer" and the parts of Saxony and Saxony-Anhalt serviced by a short-distance wholesale transmission company that is jointly owned by Ruhrgas and VNG).

Position of some other long-distance wholesale transmission companies

(148) Mobil has never had an "own region" nor high-pressure pipelines, but was able to reach its customers by way of third-party access agreements with [...]*.

(149) Wingas competes, by means of its own pipeline network, in each of these territories and most short-distance wholesale companies are in reach of its pipeline network (corridor of at most 50 km). However, on that basis, it cannot be concluded that the competitive conditions within the whole country are equal. [...]* the respective long- and short-distance wholesalers differentiate their prices depending on the fact whether or not the customer has an alternative in the form of supplies by Wingas.

The proposed German industry TPA regime (VV Gas reference case) will not influence the geographic setting of competitive conditions

(150) It results from the description of the VV Gas in the product market section that the Commission does not consider that the VV Gas will make it possible to import gas in competition with the incumbents.

(151) For the reason indicated in that section with regard to the built-in limitations of the VV Gas concept and the reference case of the gas industry such as (a) the general difficulty to assess whether capacity is available, (b) the fact that with a minimum duration of one year the winter peak is always taken into account when assessing the free capacity in a pipeline, (c) the difficulties resulting from the specifications of the gas (caloric value, pressure and so forth) and (d) the limited scope of the services offered, the Commission considers that the VV Gas is not sufficient to consider, from a prospective analysis point of view, that the market is national. In addition, the Commission takes into account that in case of a dispute between a party that requests TPA and a party that refuses TPA, the procedures before the competition authorities and/or the courts might take several years before a "precedent" case is settled.

Conclusion

(152) On the basis of the above, the Commission considers that the wholesale transmission markets in Germany might still be regional, following the former demarcation lines. However, it may not be necessary, for the purpose of the present decision, to decide the geographic definition of the long-distance wholesale transmission market within Germany as the concentration also leads to the strengthening of a (collective) dominant position if the long-distance wholesale transmission market were national.

(153) The parties have stated in the reply that the market is national, not regional. They indicate that gas prices are determined by competing fuel, not regional location. However, this is in contradiction with the information obtained by the Commission in the market investigation that prices differ also on the basis whether or not a client has an alternative supplier (gas to gas competition). The parties also state that a forward-looking analysis must take account of the recent elimination of demarcation and the impending industry agreement on negotiated third-party access. However, there remain too many uncertainties to conclude that the final agreement would lead to a national market where competitive conditions are the same.

Position of the market participants

(a) The Dutch wholesale transmission market

1. Relationship between Dutch gas producers and Gasunie

(154) In 1947, Exxon and Shell formed a 50/50 JV, called Nederlandse Aardolie Maatschappij (NAM), with the object of exploring and producing hydrocarbons within the Netherlands. Following NAM's discovery of the large Groningen gas field in 1959, the Netherlands Government organised the Dutch gas industry to reflect the Groningen field's strategic importance for the Netherlands economy. To this end, the Netherlands Government in 1963 entered into an agreement with Exxon, Shell and NAM, under which NAM is assigned the responsibility to develop and produce the Groningen concession for the account of the Groningen partnership in which NAM has 60 % and the Netherlands Government, via EBN, 40 %. In addition, Exxon, Shell and the Netherlands Government formed a company called Gasunie, which was entrusted with the marketing of all Groningen gas. The Netherlands Government holds a 50 % equity interest in Gasunie (10 % direct - 40 % via EBN), while Exxon and Shell each hold 25 %. Finally, the Netherlands Government entered into a private law agreement with Gasunie under which Gasunie must seek the Government's approval for certain commercial decisions.

(155) Since 1959 NAM has obtained production licences for other Dutch on and off-shore gas fields and roughly 50 % of its current production comes from these other "small" fields (compared to Groningen). The parties have indicated that most of these concessions are also produced for the benefit of partnerships in which EBN generally has a 40 to 50 % participation. The gas of these fields has been sold to Gasunie. In addition, all future discoveries by NAM must be sold to Gasunie as a consequence of [...(14)]*. In total, NAM accounts for 75 to 80 % of total Dutch gas production (in view of EBN's equity interest in NAM's production, NAM accounts for about 45 % of production on a net equity basis). EBN has also consistently marketed all its equity interests via Gasunie. [...]*.

(156) Until 1994, Dutch producers were obliged to sell all gas that was destined for the Dutch market to Gasunie. Dutch producers could sell themselves to the export market, although the prices were and are still controlled by the Government. Furthermore, Gasunie is obliged to buy gas from the small fields prior to production at financial conditions fixed at the start if the producer has chosen to offer the gas to Gasunie. [...]*

(157) It appears that, in general, the Netherlands Government would not give its approval on the price in an export contract of Dutch gas if this price were below the Gasunie "net-back" price, called the "norm inkoop prijs" (NIP or normative buying price). In this way, the Government ensures that it receives its share of the "gas cake".

(158) [...]*(15)

(159) NAM owns two of the three Dutch underground storage facilities (depleted gas fields) which it has leased for 100 % to Gasunie. The third facility, operated by BP Amoco, is also leased totally to Gasunie. Therefore, only Gasunie is able to offer underground storage facilities in the Netherlands. Whereas there are a couple of depleted Dutch gas fields that could be turned into underground storage facilities servicing the Dutch market, it is not likely that these facilities will be developed by others in view of the risks involved. Such risks relate to the required investment and to the fact that Gasunie could always lower its tariff, thereby making such investment unprofitable. It is also noted that [...]*.

2. Gasunie's corporate governance structure

(160) Gasunie is owned 50 % by the Dutch State and 25 % each by Exxon and Shell. Decisions are taken with [...]* majority voting. There are no specific shareholder agreements. Therefore Exxon has no de jure control over Gasunie within the meaning of the Merger Regulation.

(161) As to de facto control, the parties argue that there is no common interest between Exxon and Shell. [...]*. However, for the reasons indicated below, it is not necessary to decide whether or not Exxon has de facto control over Gasunie within the meaning of the Merger Regulation.

3. Gasunie's gas selling position in the Dutch market

(162) Gasunie is the dominant Dutch transmission company with around [...]* % of total sales. Gasunie explicitly recognises its dominant position in the introduction of its new tariff system. [...]*(16). Furthermore, according to Gasunie's "Plan of Gas sales 1999", Gasunie will supply [...]*.

(163) Gasunie owns or has exclusive rights to (a) the whole Dutch high-pressure pipeline grid (with the exception of one line in the south of the Netherlands by which British gas is imported), (b) all processing and blending facilities, (c) all (underground and Liquid Natural Gas) storage capacity.

(164) It also appears from several [documents]* that Gasunie faces and will continue to face little competition. These documents, quoted in the following points, support Gasunie's internal assessment that it is a company with a dominant position.

(165) [...]*

(166) [...]*

(167) [...]*

(168) [...]*

(169) [...]*

4. Mobil's position on the Dutch market

(170) Mobil is one of three actual competitors. It has, according to the parties, a [less than 5]* % share on the total market. Statoil (or the Norwegian joint negotiation committee that markets all Norwegian gas) has an estimated 4 % market share by means of its supply contract to an important power generating plant. EnTrade, a trading company of the Dutch RDC Pnem/Mega, is the other competitor with a 4 % market share resulting from the sale of British gas to industrial consumers and a power generator within the reach of its partly owned Zebra HCV pipeline in the South of the Netherlands. It has a TPA transport agreement with Gasunie for the supply to one customer.

(171) Mobil's position in the Netherlands stems from its European gas sales strategy [...]*. This European strategy will be described in the following paragraphs. This description is also of relevance for the assessment of the position of Mobil on the German market

Mobil's European gas sales strategy

(172) [...]*

(173) [...]*

(174) [...]*

(175) [...]*

(176) [...]*

Mobil in the Netherlands

(177) [...]*(17)

(178) [...]*

(179) [...]*

(180) [...]*

(b) The German long-distance wholesale transmission market(s)

Market participants

(181) In the product market definition section, the Commission has reached the conclusion that the long-distance wholesale transmission market is a separate product market where the entry depends on the goodwill of the incumbents. It has also been concluded that new entry by means of a new own infrastructure investment is not likely to occur.

(182) The incumbents are those companies that (a) purchase gas from foreign producers, (b) are able to transport this to their customers by means of a high-pressure pipeline network, (c) are able to provide swing and (d) are able to provide other related services. They either perform these services on the basis of their own investments or because they have received for parts of them third-party access of the others.

(183) Ruhrgas, BEB, VNG, Thyssengas and Wingas are the most important players on the market. They all perform the above services by means of their own investment.

(184) Exxon is an active player on the German market by means of its long-standing 50/50 JV with Shell, BEB. Exxon acknowledges that it has joint control together with Shell over BEB within the meaning of the Merger Regulation. Exxon is also active on the relevant market by way of its 25 % share of Thyssengas. The remainder of the shares is held by Shell (25 %) and RWE (50 %). Exxon acknowledges that it has joint control with RWE and Shell over Thyssengas within the meaning of the Merger Regulation.

(185) Ruhrgas is the most important German long-distance transmission company. Both Exxon, through BEB, and Mobil are shareholders of Ruhrgas. BEB's shareholding accounts for 29,5 %. The largest part of the voting rights attached to this shareholding, (25 %) is voted directly; the remainder (4,5 %) is part of the "Schubert pool" within Ruhrgas. The Schubert pool accounts for 15 % of the votes in Ruhrgas. Members of the pool are Mobil (7,4 % share), BEB (4,5 %), BPAmoco (0,5 %) and Preussag (2,6 %). The votes must be expressed as one block and the Schubert pool rules stipulate that no member can have more than [...] % of the voting rights within the pool. It is claimed by the parties that, as a consequence of these rules, the current voting rights of BEB in the Schubert pool would be attributed, post-merger, to BPAmoco and Preussag. It has to be noted that the Schubert shareholders have contracted to supply [...]* % of their proven domestic reserves additions to Ruhrgas. These supplies to RG are transported by [...]*.

(186) Ruhrgas is formally controlled by the "Bergemann pool", a shareholder pool representing 59,76 % of the votes. The pool groups the 25 % shareholding of BPAmoco together with the votes held by a group of German coal and steel producers (that are also very often active in electricity production and distribution). These complicated rules must prevent that Ruhrgas would be controlled by multinational oil and gas groups.

(187) [...]*(18)

(188) Furthermore, in view of the shareholders agreement between BEB and Ruhrgas, a [...]* % vote is required if a question is "essential to BEB". This gives BEB therefore a veto for all questions that are essential to it. The parties have indicated that this clause has never been invoked. [...]*

(189) [...]*. It therefore seems that, despite the absence of de jure control, the oil companies have de facto influence on Ruhrgas.

(190) Verbundnetz Gas (VNG) is a long-distance wholesale transmission company active in the new German Länder. It was founded in the 1960s and privatised after the German re-unification. Its current shareholders are Ruhrgas (36,8 %), Wintershall (15,8 %), local communes (15,8 %) and BEB (10,5 %). In addition, Statoil, Erdöl-Erdgas Gommern, PreussenElektra and Gazprom each hold 5,2 %. It follows that all VNG's shareholders with the exception of the local communes are gas producers and/or long-distance wholesale transmission companies.

(191) Wingas is a 65/35 % JV between Wintershall (a BASF subsidiary) and Gazprom.

(192) [...]*

(193) EWE is de facto a short-distance wholesale transmission company as it is not in the business of re-selling imported or own-produced gas to other companies operating a high-pressure grid (long and short-distance wholesale transmission companies). However, it imports directly from Gasunie a substantial part of its gas sales via its own high-pressure (LCV) pipeline. It is, therefore, considered to be a long-distance wholesale transmission company.

Market position

(a) Local

(194) BEB has indicated that it has a market share of [40 to 50]* % in northern Germany (Schleswig-Holstein, Hamburg, Bremen and Lower Saxony. However, this "Northern Germany" region is larger than the BEB demarcation region that only comprises parts of Bremen and Lower Saxony. The remainder is "Ruhrgas territory", whereby Ruhrgas had concluded in turn a demarcation agreement with Erdgas Münster. It therefore results that the BEB market share in its demarcation region is in excess of [40 to 50]* %.

(195) [...]*

(196) [...]*

(197) RG's strategy paper from October 1998 indicates that as a consequence of GTGC, it cannot maintain its 100 % share in its area. Although no precise data are available, the fact that RG accounts for around [+ 50 %]*of the supplies to the short-distance wholesale transmission companies operating in its area is the best available proxy for its market share.

(198) VNG has, even according to the parties' reply to an Article 11 request, a [9 to 100]* % market share in 1997. TG has according to the parties [70 to 80]* % in 1997. [...]*

(b) On the German market

(199) All sales data in this section refer to 1997. Total gas imports in Germany account for 75,6 billion cubic metres (BCM).

(200) BEB produces half of the German gas accounting for [...]* BCM of production. It also purchases another [...]* BCM from Gasunie, Gazprom, DONG (the Danish state's oil and gas company), from the GFU (the Norwegian gas sales entity) and a small share of other German producers. Thyssengas imports [...]* BCM. Mobil imports [...]* BCM of its Dutch equity production and from Norway. Mobil is also a German gas producer (some [...]* % of German production, accounting for [...]* BCM). Ruhrgas imports [...]* BCM from the Netherlands, Norway, Denmark and Russia. It purchases [...]* BCM of German-produced gas. VNG imports [...]* BCM. Wingas imported 6,9 BCM. EWE imports, as indicated above, gas from Gasunie. This volume accounts for [...]* BCM.

(201) The abovementioned import figures give the long-distance wholesale transmission companies the following market shares: RG [50-60]* %, BEB [10 to 20]* %, VNG [10 to 20]* %, Wingas [0 to 10]* %, TG [0 to 10]* %, Mobil [0-10]* % and EWE [0 to 10]* %.

Mobil's German strategy

(202) Mobil has described in its European strategy paper its position in Germany as follows:

[...]*

(203) [...]*

(204) [...]*

Competitive assessment

(a) Dutch wholesale transmission market

The parties and Gasunie acknowledge in their internal documents that Gasunie is dominant and will remain so

(205) On the basis of the documents described in the section on the position of the parties in the market, it can be concluded that Gasunie is dominant and will remain so (its current market share - [80 to 90]* % - is likely to remain at the same level for the next [...]* years).

The competition pressure of the other actual competitors remains limited

(206) Apart from Mobil, there are two other actual competitors, Statoil and Entrade. Statoil has indicated that "its policy in the Dutch (and German) gas wholesale distribution markets can in short be characterised as a 'wait and see' approach." [...]*.

(207) The "direct sales" by Statoil to the Eems plant is the only sales contract ever done directly by Statoil so that, on the basis of the above, the Commission considers that, overall, Statoil cannot be considered as a competitor that will remain active on the market. [...]*.

(208) The other actual competitor is Entrade. Entrade has contracted [...]*. Entrade's sales potential along its pipeline seems to be more or less exhausted. The potential to increase further the sales via TPA with Gasunie seems to be limited [...]*.

(209) Mobil has annual gas sales of [...]* under contract. Around [...]* is of British origin, the remainder from Mobil's north Friesland concession (Anjum). One of these contracts is a [...]*. Mobil has indicated that it has no more uncommitted high swing of own production.

Potential competition

(210) [...]*

(211) The parties refer to the proximity of the existing Zeebrugge LNG terminal with spare capacity and connecting pipelines that make immediate entry of LNG possible into the Netherlands and indicate, in their reply, that Sonatrach, the Algerian gas company, could thereby enter the Dutch market. The parties did not provide data concerning the existing spare capacity in the Zeebrugge terminal and connecting pipelines. However, new LNG gas supplies are much more expensive (cost of 16 to 19 ct (HFL)/m3) than pipeline supplies and there is no indication of vertical integration by Sonatrach in any other European gas market. Therefore, potential entry of Sonatrach in the Netherlands is not considered to be very likely.

(212) The parties also indicate that the recent TPA agreement that Gazprom secured with Ruhrgas facilitates the possible entry of Gazprom into the Dutch market. Whereas this agreement might facilitate the possible entry, it does not make such an entry more likely. [...]*. In the reply, the parties have indicated that the potential relief from the parallel sales agreement is only for direct marketing by Gazprom to an existing Gasunie customer. However, all customers except those of Statoil, Entrade and Mobil (representing [80 to 90]* % of the market) are Gasunie's existing customers, so that the protection is substantial.

(213) In the reply, the parties indicate that large continental wholesale transmission companies possess gas at various points along the Dutch border. Distrigaz, Thyssengas, Wingas, SNAM, Gaz de France, BEB and Ruhrgas all have supplies which can be potentially sold into the Dutch market, and no parallel sales agreements exist between Gasunie and these companies for such sales in the Dutch market. It is true to say that all these companies possess gas at various points along the Dutch border as all of them except Wingas purchase gas from Gasunie. Even if the contract would allow these companies to re-sell the Gasunie gas in the Netherlands or even if they have access to nearby gas from other sources to sell in the Netherlands, it is unlikely that these companies would "attack" their supplier in view of the risk for retaliation by Gasunie [...]*.

(214) Even if it cannot be excluded that one or more of the above companies or any other companies might enter the Dutch market, the Commission can conclude that the scope for development of potential competition is quite minimal as Gasunie itself does not expect to lose much more market share anyway [...]*. This estimate was made in the full knowledge of the own TPA tariff structure and the Netherlands Government's liberalisation plans.

(215) With regard to the Gasunie tariff, the Commission recognises that all possible services necessary to operate a gas business in the Netherlands are included in the tariff. In this way, the scope of the tariff contrasts with the rather limited "pipeline transport-only" approach taken by the German gas industry in the ongoing VV Gas. However, these tariffs do not make such an entry more likely. The tariff is based upon the customer's alternative cost to provide the service itself and is thus calculated on Gasunie's replacement costs for the infrastructure. Whereas the Commission recognises that the tariff passes Gasunie's economies of scale on to the customer, [...]*.

(216) [...]*

(217) [...]*

Mobil is in a unique position

(218) [...]*

(219) The parties argue that Mobil has no unique strengths to compete in the Dutch market and that any new entrant can replicate and exceed Mobil's Dutch market position. The parties indicate that any Dutch producer with uncommitted reserves has several marketing options: export, sell to a trader, to a Dutch end-user, an RDC or to Gasunie. Whilst this is true, it has to be noted that all current reserves are and have been committed to Gasunie except a part of the reserves of Mobil. Mobil has [...]* BCM of proven reserves in the Netherlands of which [...]* BCM are not yet committed. In addition, Mobil has another [...]* BCM of probable reserves. [...]*. Again, this makes Mobil "unique".

(220) In addition, reference is made to Mobil's strategy which clearly indicates [...]*.

The merger will eliminate Mobil as a competitor

(221) A merger between Mobil and Exxon would in normal circumstances lead to a greatly diminished incentive for Mobil to compete against Gasunie. Exxon has an important and lucrative stake in Gasunie and would suffer from lower returns resulting from price competition between two of its ventures. However, in view of the contractual arrangements on which Gasunie has been founded, Mobil will not be allowed to continue competing against Gasunie.

(222) Exxon's 1963 agreement with Shell and the Netherlands Government requires the sale of all Exxon's Dutch gas to Gasunie. Therefore, Exxon would require agreement from both Shell and the Netherlands Government prior to competing directly with Gasunie. Even if one assumed that the Netherlands Government would give its agreement to Mobil competing with Gasunie in view of the Government's support for the liberalisation of the gas market, it would not be in the economic interest of Shell to give such an agreement in view of the enormous profit loss for Gasunie resulting from price competition. [...]*

(223) The concentration therefore has as an effect that one of only two "active" actual competitors of Gasunie will be eliminated from the market. The Commission therefore considers that the concentration strengthens the dominant position of Gasunie as a result of which effective competition would be significantly impeded on the Dutch gas wholesale transmission market.

(224) In the reply, the parties indicate that Gasunie's market conduct has not been altered by Mobil's presence as a competitor nor will it change if Mobil and Exxon merge because gas pricing is fixed by reference to alternative fuels and services are priced comparably with services provided elsewhere in Europe. If this reasoning is accepted, it becomes tantamount to indicating the competition in the gas industry is irrelevant. This is clearly not true, even on the basis of the internal documents of Gasunie itself.

The Merger Regulation applies to the strengthening of a dominant position that is not held by the parties to the concentration

(225) It can be noted that Article 2(3) of the Merger Regulation states that a concentration which creates or strengthens a dominant position is to be declared incompatible with the common market. Article 2(3) is thus, in terms, not limited to the prohibition of an operation which will give rise to the creation or reinforcement of a dominant position on the part of one or more of the parties to the concentration - see also recitals 1, 5 and 7 in the preamble to the Regulation.

(226) The reasoning by the Court of Justice in the Kali und Salz(19) judgment and of the Court of First Instance in Gencor(20) does lend indirect confirmation to the above conclusion. The Commission acknowledges that the support is indirect as both cases deal with oligopolistic dominant positions.

(227) However, the Court, in point 171 of the Kali und Salz judgment, confirms in general terms that the dominant position that is created or strengthened does not need to be one of the parties to the concentration:"A concentration which creates or strengthens a dominant position of the parties to the concentration with an entity not involved in the concentration is liable to prove incompatible with the system of undistorted competition laid down by the Treaty. Consequently, if it were accepted that only concentrations creating or strengthening a dominant position on the part of the parties to the concentration were covered by the Regulation, its purpose as indicated by the abovementioned recitals would be partially frustrated. The Regulation would thus be deprived of a not insignificant aspect of its effectiveness, without that being necessary from the perspective of the general structure of the Community system of control of concentrations."

(228) The Commission considers that the creation or reinforcement of a dominant position by a third party is not excluded from the scope of application of Article 2(3) of the Merger Regulation. In addition, it should be born in mind that Gasunie is a JV in which one of the parties to the concentration has a substantial interest.

(229) In the reply, the parties indicate that there is no precedent for the above application of Article 2(3). This is correct, but irrelevant. In addition, the parties claim that it is not necessary to apply the Merger Regulation to such cases as European and national antitrust authorities have other means at their disposal to curb abuses of the dominant position of a third party. However, the Merger Regulation provides for a structural control, "not allowing the creation or strengthening of a dominant position", rather than an abuse-control system.

(b) Long-distance wholesale transmission market(s) in Germany

Uncompetitive German gas market

(230) As indicated before in the geographic product market section, there are two ways to approach the long-distance wholesale transmission activity in Germany. It is possible to conclude that the markets are still regional whereby the regions are defined along the lines of the "old demarcation regions". However, one could also consider that the market is national. The Commission has examined the long-distance wholesale transmission market from both approaches and considers that, as a consequence of the notified operation, the existing single (regional market) and/or collective (national market) dominant positions are strengthened.

Barriers to entry are very high in the German long-distance wholesale transmission market.

(231) As indicated in the relevant product market section, the barriers result from the sunk cost investment nature of the important capital expenditure necessary to have an own long-distance wholesale transmission infrastructure, the long-term take-or-pay contracts between producers and long-distance wholesale transmission companies, the parallel sales clauses in those contracts and the long-term agreements between long-distance wholesale transmission companies and their clients. This makes the industry very stable. In Germany, the stability was until May 1998 regulated in private arrangements (the demarcation agreements). It was also indicated in that section that on the basis of the proposed German industry TPA regime (reference case VV Gas), it is not likely that on the basis of this TPA the position of the long-distance wholesale transmission companies can be challenged.

Supply concentration

(232) RG, BEB, TG and VNG account for [80 to 90]* % of gas imports in Germany. This makes the supply very concentrated and the documents cited in the geographic market section (regional markets) show that because of the common fear of retaliation, the stable relationship amongst the four companies will continue.

Links among market players further limit the incentives to compete among Ruhrgas, BEB and Thyssengas

(233) The six long-distance wholesale transmission companies already all had, prior to the merger, some form of equity link with each other. BEB and Thyssengas are jointly controlled by Exxon and Shell. BEB and Mobil are shareholders in Ruhrgas. Ruhrgas, BEB and the two shareholders of Wingas (Gazprom and Wintershall) are all shareholders in VNG. It results from the above that the links between Wingas and the other companies are less strong than those between BEB, Thyssengas and Ruhrgas.

(234) Ruhrgas, BEB and Thyssengas buy in common a substantial part of their non-German origin gas (RG [...]* %, BEB [...]* % and TG [...]* %). This, therefore, results in symmetry of costs at the "border level". In addition, these companies also transport a substantial part of their gas via pipelines with common ownership(21). They therefore also have an insight into this part of the cost.

Transparency of the market

(235) Finally, the market is transparent in so far that each of the long-distance wholesale transmission companies would know immediately when one of their clients changed supplier and they would know immediately who the other supplier is.

No incentives to compete for other long-distance wholesale companies

(236) None of the market participants apart from Wingas would have an interest in entering the region of another market participant. VNG cannot be expected to endeavour to do this. This is so in view of the minority equity interests of Ruhrgas and BEB in VNG, the short-distance wholesale transmission JV between RG and VNG in certain of the "new" Bundesländer and the fact that VNG has no own high-pressure pipelines nor storage facilities outside its core territory in the "new" Bundesländer. Furthermore, the other shareholders would have no interest in the company challenging RG and BEB in their area (the TG area is too far away) and risk retaliation. The shareholders would prefer the company to return some dividends as soon as possible from its core business in the "new" Bundesländer.

(237) EWE might have a certain potential to increase its share of the imports for its own re-sale in the "Ems Weser" part of its sales area by substituting purchases from Gasunie for purchases from BEB. However, sales by EWE outside its "Ems Weser" sales area make the company vulnerable to retaliation by BEB and RG. And as it is much more vulnerable to retaliation by BEB and RG than BEB and RG for retaliation by EWE, it is unlikely that EWE will expand its position on the long-distance wholesale transmission market by buying gas from Gasunie for sales outside its "Ems Weser" area.

Limited ability of Wingas to further increase its market share

(238) Wingas is, in view of its historic position as an outspoken challenger, quite clearly not a member of the oligopoly. It was the first and so far the only company to bring gas-to-gas competition in Germany by building and currently further expanding an own high-pressure HCV pipeline grid. It has never concluded a demarcation agreement with Ruhrgas, BEB or Thyssengas. However, the Commission's investigation has resulted in a number of elements that support a conclusion that Wingas will not be in a position to challenge the oligopoly much more.

(239) The company has announced that on the basis of its current investment it would be able to reach a 15 % market share. According to the parties, it has around 12 % of the market under contract. It therefore comes close to its target. On the basis of internal documents from the parties, it is clear that the market was attaining stabilisation and that Wingas was already on the way to becoming established. For instance, [...]*

(240) [...]*

(241) [...]*

(242) This new cooperation with RG has led to RG taking a 4 % participation in Gazprom and the conclusion of a new long-term take-or-pay contract by means of which RG obliges itself to purchase [...]* of its German natural gas demand from Gazprom and whereby Gazprom is prepared to supply up to [...]* %. As part of this agreement, RG is discharged from its purchase obligations in so far as Gazprom/Gazexport sell gas in Germany in competition with Ruhrgas, [...]*(22). One of Wingas shareholders and its major gas supplier (Gazprom) has thus agreed with the main competitor (Ruhrgas) to limit the further growth of Wingas. The Commission can therefore reasonably conclude that the future potential of Wingas to attack the incumbents is very limited.

Indications of uncompetitive conditions

(243) Reference is made to the important ROCE of the German wholesale transmission companies. [...]*. Only VNG has so far returned minimal profits. With regard to VNG it can be noted that this is probably due to the enormous investments it had to make following the German re-unification. However, it also seems that the current shareholders have an interest to show a low value for the company as this decreases the risk that the BVS (former Treuhandanstalt) might ask for a high retroactive payment through a court decision.

(244) In 1996, prices to the final consumers in Germany are, on a weighted average, higher than in the United Kingdom, Netherlands and Belgium. They are a bit lower than in Italy. France is the most expensive market(23). For the period 1997 to May 1999, German prices to commercial customers, firm and interruptible industrial customers were continuously amongst the highest of the above six countries. This conclusion is valid for prices before and after tax.

Single dominance in old demarcated region

(245) In the section of relevance to the market participants, it was indicated that each of Ruhrgas, BEB, Thyssengas and VNG have estimated market shares in their respective regions in excess of 50 %. Although the old demarcation agreements are no longer valid, there may still be a very substantial protection of the own territory in place between the "traditional" long-distance wholesale companies. First of all in view of the common purchasing of the imported gas. As there are clear arrangements between the companies on what part of the common imported gas goes to which company, they are not going to "attack" each other with their share of the imported gas on the others' territory. Secondly, there are the "parallel sales letters". By means of such letters the sales of the one party into the territory of the other can be deducted from the purchase obligation that the "attacked" party has vis-à-vis the attacking party. [...]*. Thirdly, there is the fear of retaliation. On the basis of the above, Ruhrgas, BEB, Thyssengas and VNG are considered to have a dominant position in their own region.

Oligopolistic dominance

(246) Market growth is expected to remain steady at, at most, 2 to 3 % per year. Whereas the parties have indicated that as a result of such a growth there might be a supply gap [...]* onwards, such market growth cannot but be considered moderate at most.

(247) The price elasticity for gas demand is limited: it is rather price inelastic. There are numerous indications in the file that a "price war" would destroy the margin for everybody and that both producers and wholesale transmission companies (see the section on "the old demarcation regions are still relevant in view of the fear for retaliation" above) are aware of this and act accordingly.

(248) In the light of the uncompetitive nature of the market as discussed above, the limited growth of the market and the limited sensitivity of demand to price movements, the Commission considers that there is already pre-merger an oligopolistic dominant position between at least Ruhrgas, BEB and Thyssengas on the German long-distance wholesale transmission market.

(249) The best illustration of the existence of a collective dominant position is that BASF, probably the largest user of natural gas in Germany, considered it necessary to invest in an own pipeline network to have access to cheaper gas. It has used the leverage of its demand to help to create Wingas.

Result of the merger

Mobil - uniquely placed to be a protagonist in a more liberalised German gas market

(250) Mobil currently holds a [0 to 10]* % market share. This is however a significant underestimate of its competitive potential in the German market. [...]*.

(251) Mobil is in a somewhat atypical situation in Germany. Mobil produces a substantial part of the German gas ([...]* %) and thereby accounts for [...]* % of German gas consumption. Mobil also has an equity interest in Erdgas Münster and in Ruhrgas (Supervisory Board member). It is therefore a part of the German gas establishment. This is probably the reason why Mobil was able to import gas without having its own high-pressure pipeline network by means of TPA with especially BEB (their close partner in the German production). [...]*.

(252) [...]*

(253) The potential of Mobil bringing more competition into the German market thanks to liberalisation will completely disappear. Unlike Mobil individually, the merged entity has more to lose than to win from a liberalisation than the individual companies Exxon and Mobil before.

Reinforcement of equity link between Exxon/Shell and Ruhrgas

(254) As a consequence of the merger, the equity links between Exxon/Mobil and Shell with Ruhrgas will be further reinforced. The parties argue on the basis of the formal corporate governance structure of Ruhrgas ([...]*) that this will not increase the combined entity's impact on Ruhrgas, but rather decrease.

(255) [...]*. The Exxon and Shell interests in European gas production and in gas wholesaling in the Netherlands and Germany are all held in common. Consequently, their economic interests are parallel. This should, despite the distinct philosophy of Exxon and Shell with regard to certain elements of the wholesale transmission business as seen for instance with regard to the discussions on Gasunie's future role, provide a sound basis for a commonality of interest. This implies that, following the merger, only BP needs to be convinced to align itself with common Exxon/Shell interests, compared to BP and Mobil before the merger. This therefore substantially increases the scope of the merged entity to increase its de facto influence on Ruhrgas.

(256) Furthermore, even abstraction made from the possibility of the merged entity to increase its influence on Ruhrgas, a reinforcement of the equity links between companies on the same market increases the possibility for the use of market power. In order to appreciate the level of concentration in this market pre-merger and the impact of the merger, the Commission has estimated HHI indices that take into account the existence of cross shareholdings among most of the players in that market. This calculation was based on the work of Bresnahan and Salop(24). The assumption was made that BEB is identified with its owners, Exxon and Shell. Under this assumption the HHI would rise 139 points from 4243 to 4382. Such an increase at this level of concentration is a sign of the scope for a material increase of market power.

Control of Erdgas Münster, a potential competitor

(257) As a consequence of the merger, BEB and Mobil would control the corporate governance structure of EGM. Mobil would have an interest, pre-merger, in maintaining the possibility of developing the only short-distance wholesale transmission company that does not depend from the wholesale transmission services of the long-distance wholesale transmission companies into a potential competitor in a future liberalised market environment. This ability to be a potential competitor would be suppressed post-merger through the establishment of a formal control on EGM because, unlike Mobil individually, the merged entity has more to lose than to win from a liberalisation than the individual companies Exxon and Mobil before.

Conclusion

(258) It can, therefore, be concluded that the concentration strengthens (a) the individual dominant positions of each of Ruhrgas, BEB and Thyssengas within their region and/or (b) the collective dominant position between Ruhrgas, BEB and Thyssengas on the German long-distance wholesale transmission market as a result of which effective competition would be significantly impeded on these markets.

(259) In the reply, the parties indicate that on the basis of Mobil's mere [0 to 10 % maket share]*, [...]* and the fact that BEB and Thyssengas, the other entities controlled by the parties, have a market share of only [10 to 20]* %, the notified operation cannot impede, much less significantly impede, competition whatever the market definition. The parties argue that the Merger Regulation establishes a two-pronged test and that this principle has been accepted by the Commission and confirmed by the Community courts.

(260) However, as indicated above, the changes on the relevant market resulting from the concentration result not only from the current position of Mobil on the long-distance wholesale transmission market, but also from Mobil's potential [...]* and the reinforcement of the equity links between the incumbents as expressed in the HHI calculation.

UNDERGROUND STORAGE

Relevant product market

(261) Underground storage facilities are key facilities in the provision of "swing". This is particularly relevant for the gas demand of the local distribution companies. The Commission therefore agrees with the parties that this forms a separate product market.

Relevant geographic market

(262) There are two types of storage facilities: pore versus cavern (see LCV/HCV discussion previously). The economic radius for pore storage is less than 200 kilometres whereas for cavern, the economic distance from the end-user is limited to 50 kilometres.

(263) It is thus possible to delineate a distinct relevant geographic market of the pore storage facilities in the south of Germany, namely an area of around 200 m around Munich.

Underground storage facilities servicing the south of Germany

(264) There are for the moment five pore underground storage facilities located relatively close to Munich. In addition, taking into account the 200 km radius of a pore storage, one other facility (Eschenfelden) covers the entire region around Munich. These facilities serve the south of Germany, which is a substantial part of the common market.

(265) The following table gives the overview of these facilities, the storage volume and the maximum withdrawal rate as well as an indication of the ownership interests.

>TABLE>

(266) Of the above capacity, the [...]*.

(267) RG has the beneficial interest (ownership + lease) in [...]* % of the storage volume and the daily withdrawal potential. [...]*

(268) However, the parties have an interest in [...]* of the 16 depleted Bavarian reservoirs that can reasonably be assumed to be suitable for conversion into an underground storage facility serving the Munich region. According to the parties, the four depleted reservoirs in which they do not have an interest could be converted in storage facilities with a working gas volume of around [...]* MCM. However, according to the concession holders and other interested third parties, the potential of these reservoirs would be only around [...]* MCM and doubts have been expressed with regard to the suitability of some of the fields involved.

(269) On the other hand, the [...]* depleted reservoirs in which the parties have an interest have, according to the parties, a potential of around [...]* MCM working gas volume. In addition, the parties have an interest in the individual fields with the highest potential in working gas volume.

(270) [...]*

(271) However, as a consequence of the concentration, these opportunities in the Munich area will be evaluated taking into account the 36 % equity ownership of BEB and Mobil in Ruhrgas. Therefore, it is much more unlikely that (a) Exxon/Mobil will start competing themselves in the area and (b) that competitors of Ruhrgas will be given the opportunity to develop additional storage.

(272) It can, therefore, be concluded, that the barriers to entry for potential competitors in the market for swing capacity in the south of Germany will be increased as a consequence of the concentration. Therefore, the concentration strengthens the dominant position of Ruhrgas in the market for swing capacity in the south of Germany (Munich area).

Compendium of abbreviations

NATURAL GAS

BEB German gas production and long-distance wholesale transmission JV between Exxon and Shell

BCM billion cubic metres

DVGW Deutscher Verein des Gas- und Wasserfaches, German industry association deciding technical issues on gas

EBN Energie Beheer Nederland, the Netherlands State Oil and Gas Company

EGM Erdgas Münster

GTGC gas-to-gas-competition

HCV high calorific value

kWh kilowatt hours

LCV low calorific value

LDC local distribution company

LNG liquified natural gas

MBTU million british term units

MCM million cubic metres

NAM Nederlandse Aardolie Maatschappij, the Dutch Exxon/Shell gas production JV

NRW North Rhine-Wetphalia

RG Ruhrgas

SW Stadtwerk (German LDC)

TG Thyssengas

TPA third-party access

VNG eastern German long-distance wholesale transmission company

VV Gas Verbändevereinbarung Gas, the negotation between the German Industry Federations to set the basis for the liberalisation of the market

D. BASE OILS, ADDITIVES AND LUBRICANTS

INTRODUCTION

(273) The lubricants industry involves three different products with a vertical relationship: base oils, chemical additives and lubricants. Base oil is blended with chemical additives to produce lubricants.

(274) On the basis of market investigation, the Commission is of the opinion that the merger will lead to the creation of a dominant position held by Exxon and BP/Mobil on the EEA market for base oils.

DESCRIPTION AND FUNCTIONING OF THE MARKET

Base oils are the end product of crude oil refining

(275) The base oils used to formulate lubricants are, generally, mineral oils produced by oil refiners from the residue of crude oil left after separation of the other major oil products. A minor part of the base oil market is the result of chemical processes and are called Esters or PAO's (polyalfaolefins).

(276) Base oils have various performance capabilities due to their different chemical/physical compositions resulting from different crude feedstocks and refinery processes. Base oils are classified on the basis of their inherent characteristics. Independent bodies (API/ACEA/ATIEL) have defined standards by which the quality of base oils can be closely monitored(25).

Characteristics

(277) An individual base oil is defined by its viscosity, its viscosity index (VI), its volatility and a number of subsidiary specification points.

(278) Viscosity is the measurement of a fluid's resistance to flow. High-speed/low-temperature/low-pressure conditions call for a low-viscosity oil. Low speed/high-temperature/high-pressure conditions call for a high-viscosity oil. The viscosity is expressed in grades, but for better understanding, the Commission will refer to low-viscosity base oils (so called "light neutral", or LN) and high viscosity base oil (so called "heavy neutral", HN, and "solvent bright stock", SBS).

(279) The viscosity index is one of the most important characteristics of almost any lubrication product. It is the measure of the flowability at definite temperatures. All lubricants change viscosity with temperature changes (lubricants become thinner as temperature increases). Viscosity changes at varying rates depending on the base oil. This rate of change is measured by a "viscosity index" and a specific base oil is said to be of a certain "VI". The higher the VI of an oil, the lower its tendency to change viscosity with temperature, which is, especially for engine oils, crucial to ensure efficient lubrication. Base oils can be classified according to their viscosity index by the terms LVI, MVI, HVI, VHVI, and XHVI. Although these are marketing terms, and not recognised by Exxon and Mobil, they are widely used in the industry. The table below provides a definition for the terms which will be used in the document:

>TABLE>

(280) Volatility is an important factor in engine oils and their consumption rate. The more volatile a petroleum liquid, the higher its consumption rate. Volatility is related to the shape of the distillation curve and is measured in Noack.

(281) The pour point is the lowest temperature at which the oil will pour. This is, of course, very important for engine oils and other lubricants used at low temperature. The pour point is directly related to the type of crude and processing used.

Classification

(282) Depending on their performance, base oils are classified in five categories ranging from conventional base oil with viscosity index 80-120 and an acceptable level of impurities (group I) over base oils with a viscosity index of 120 or greater (group III)(26) up to PAO's (group IV) used in synthetic lubricants with a high viscosity index. Group V includes all other base oils such as naphthenics and very high quality ester base oils.

Crude feedstocks

(283) Base oils may be categorised by type of crude: paraffinic crude; naphthenic crude and mixed crude. Paraffinic crude is most commonly used for base oils although naphtenic crude has certain superior characteristics for some applications. Since the parties have stated that they do not produce naphthenic base oil anywhere in the world, the assessment will be limited to paraffinic base oil.

(284) The selection of the crude slate has an impact on the production capacity of a base oil refinery. The difference between a crude suitable for base oil production and a crude which is not is the concentration of those base oil molecules and the ease with which they are extracted from the crude. Arab Light is considered to be an excellent quality crude for base oil production, unlike for instance west African crudes.

(285) Since base oils refineries are usually integrated with fuels refineries and share many common facilities, the chosen crude slate is also determined by an economic analysis of the entire production slate of the refinery where the base oil plant is located. The crude necessary for fuels refinery can determine up to a certain extent the base oil production level (and vice versa).

Additives

(286) Additives are the chemicals compounded with base oils to help the oil perform a specific lubricating job. Lubricants use specific additives depending upon the applications and the base oil. Additives can be split into two categories: detergent inhibitor packages (or "DI packages") and viscosity index improvers (or "VI improvers"). In most automotive lubricants applications VI improvers have to be used with DI packages. It must be noted that all additives use base oils as a support.

(287) DI packages serve to suspend oil contaminants and combustion by-products as well as to prevent oxidation of the oil with the resultant formation of varnish and sludge deposits. A DI package will typically include approximately six individual additive components, the most important being detergents, dispersants, corrosion inhibitors and antioxidants. The combination of these components and their proportion varies according to the end-use and specification of the finished lubricant.

(288) VI improvers modify the viscometric characteristics of lubricants by reducing the rate of thinning with increasing temperature and the rate of thickening with low temperatures. VI improvers thereby provide enhanced performance at low and high temperatures. They can be considered as synthetic oil thickeners which are relatively inactive at lower temperatures but react with heat to counteract the natural tendency of the base oil to thin at higher temperatures. All multigrade engine lubricants contain VI improvers.

(289) The additives market is very concentrated. The five leading suppliers hold more than 90 % of the world market. Lubrizol is market leader and accounts for [30 to 40]* % of the market. The Infineum joint venture (combining the additive-interests of Exxon and Shell) takes the second place with [30 to 40]* %. Ethyl and Oronite (owned by Chevron) each have around [10 to 20]* %.

Lubricants

(290) Lubricants have several uses but their main function is to reduce friction and wear between moving parts. Lubricants have four general types of application. They are used in the automotive vehicles (automotive lubricants), for industrial applications (industrial lubricants), in ships (marine lubricants) and in jet engines (jet lubricants). Finished lubricants are produced by lubricant blenders through blending chemical additives into base oils to obtain the properties required.

(291) Roughly 50 % of lubricants consumption is for automotive applications, 48 % for industrial applications (including a significant part for "process oils" - 36 %). Marine and jet engine applications account for the remainder. Jet lubricants are dealt with in section F.

Automotive lubricants and industrial lubricants

(292) Automotive lubricants include (passenger and commercial) engine car motor oils, automotive transmission fluids, brake oil, automotive gear oil, etc.

(293) Industrial lubricants include industrial gear oils, metal working fluids, turbine oils, transformer oils, compressor oils, hydraulic fluids and process oils/aromatic extracts (used as components in the production of, inter alia, tyres). Industrial oils need to have specific characteristics provided by the additives.

Development of lubricants

(294) Development of lubricants involves the interaction of four categories of actors: base oil producers, additive producers, lubricant blenders and the quality standard entity (either anoriginal equipment manufacturer - OEM - such as an automotive manufacturer or an industrial end customer). Lubricants are developed in two steps between the customer who defines the product specification, the lubricant blender who chooses the suppliers and the additive producer who will give special characteristics to the lubricant.

(295) The first step is between the lubricant blender and the OEM (or other end-use customer). Together they identify which performance requirements the new lubricant will have to meet.

(296) The second step is between the lubricant blender and the additive producer. The additive supplier has then to develop an additive package which together with the selected base oil is capable of giving the lubricant its required properties. The additive supplier can also take the initiative to ensure that its additives packages will work with the available base oils, often without the knowledge or the cooperation of the blender.

(297) The finished lubricant blender can specify the base oil to be used in combination with the additive package or rely on the base oils for which the additive supplier has already formulated an additives package. The additive package gives to the lubricant the required specification at the lowest cost possible. Once the blend has been composed, it has to be tested in order to prove its performance in industry recognised test programs.

(298) According to additives suppliers, the testing is financed either by the lubricant manufacturer or by the additive manufacturer depending on who initiates the development and the commercial agreement between both parties.

(299) Additive producers tend to develop also general additive packages ("off-the-shelf packages") which are suitable to be blended with a number of base oils(27). In such a case, the investment to meet the required specifications for these "off-the-shelf packages" is made by the additive supplier. The investment will depend on the existence of a "cover" which means that a number of base oils from different suppliers can be used as the effect of using base oils from different suppliers on the finished lubricant is known and documented. If such "cover" does not exist, an "interchange programme" will be undertaken leading to a documented base oil read-across. According to additives suppliers and lubricant blenders, the cost of using " off-the-shelf packages" is significantly higher than the cost of a base oil specific additive package. A generic additives package would need to be designed to work with the "weakest" base oil in respect of each performance bench test. As different base oils can have conflicting appetites, the weakest for any test might be different. In other words, some base oils would need to be over-treated with additives in order to equal the set performance requirements.

Quality requirements for industry and marine lubricants

(300) Industrial and marine lubricants are either approved as the result of commercial trials or else they meet a standard "product specification" whereby the product has the properties required (for instance ISO or DIN). Although the performance accreditation process is kept "in-house" (as opposed to automotive lubricants), the base oil/additive combination is important. Industrial customers buy products based on performance reputation, which must be protected. A base oil switch would require validation tests of a number of key products and critical performance parameters. Certain industrial tests verifying that the additives will not behave differently in a changed base oil are time-consuming (for example a 4000 hour oxidation test) preventing a rapid base oil switch. According to additives suppliers, the accreditation process for industrial and marine lubricants can take over one year to complete, but typically lasts six months. Accreditation costs typically fall in the range of USD 25000 to USD 50000 per accreditation.

Standard-setting and accreditation systems for automotive lubricants

(301) The European standard-setting system for automotive lubricants accords a prominent role to OEMs. This is because the demand for such products originates from individual customers who do not have the means or expertise to make sure that the lubricant they buy is the right one for their needs. That role is currently filled by the ACEA, which represents the interests of all European manufacturers of cars and trucks. In December 1995, ACEA introduced a new classification system consisting of nine different sequences to define engine oil quality for European automotive service fill applications. The system is based on a schedule of physical, chemical and engine tests similar to those used in the United States of America and set out by the API (American Petroleum Institute). The system consists of two codes of practice - one developed by ATC (the technical committee of petroleum additive manufacturers in Europe) and one by ATIEL (Association Technique de l'Industrie Européenne des Lubrifiants) - and defines the process for developing, testing and reporting the necessary performance data.

(302) Together, ATIEL and ATC regulate the conditions under which lubricant and additive suppliers can change the constituents of their lubricants or additive packages, respectively, while continuing to employ previous test data and to make the same performance claims. ATIEL, a group of lubricant suppliers (both pan-regional and national integrated), has established the nomenclature to identify lubricants that have met ACEA standards. It further regulates in a code of practice the conditions under which lubricant blenders may alter lubricant constituents (mainly by changing the VI improver or base oil) while still claiming to satisfy the performance standards for which the lubricant has been tested. ATC establishes a code of practice governing the tests that must be satisfied before a lubricant can make various performance claims, and the means of registering and conducting tests and sharing test data.

(303) ACEA/API accreditation is granted for the blend of a specific base oil with a specific additives package, through series of tests. The approval will explicitly specify the components used in the blended lubricant; the base oil from a specific manufacturer and the identified additive package are locked into the accredited formulation. Since accreditation is only valid for the base oil of a specific producer and not for base oils with comparable quality from other producers(28), an ACEA or API performance claim has to be supported for each and every formulation (i.e. combination of base oils and additive packages). If a lubricant blender wishes to switch one or more components with other components (base oil or additives), then the formulation needs to be tested according to rules set out by API or ACEA.

(304) According to additives suppliers, the accreditation process can take from one month to over a year to complete, but typically lasts six months. Accreditation costs range from USD 50000 to several million per accreditation, but typically falls in the range of USD 200000 to USD 400000 per accreditation (separate accreditation is needed for ACEA and API). The exact scope of investment (both in time and money) depends on many factors including performance level, previous testing history and degree of change required(29). The ease of switching will mainly depend on the extent to which the new supplier (base oil or additives) has already gained accreditation with the preferred base oil or additives. If this "cover" does not exist then the cost to complete an "interchange programme" would be USD 100000 (for a simple base oil read-across programme) to USD 1000000 (for a complex test protocol related to infrequently traded base oils), assuming the base oil/additives combination is technically capable of achieving the desired performance.

(305) In the event that a lubricants blender would be forced to change the base oil with an infrequently traded base oil, then a fuller, more fundamental programme of testing would become necessary, costing between USD 500000 to USD 1 million. Such a process would take 18 months or longer.

(306) All OEMs have adopted these ACEA (or API) specifications as minimum standards for lubricants. Some individual OEMs, such as Mercedes-Benz for automotive and Volvo for truck lubricants, impose more stringent limits, which require extra testing with other performance parameters and certification in their specific equipment. Achieving an OEM approval requires additional testing (both test benches and field trials) and triggers additional investments in both time and money. Each OEM retains a degree of freedom and flexibility in approving products and this includes the area of base stock interchange. Additive companies provide services to obtain these OEM approvals although some OEMs require testing to be performed by independent bodies.

(307) A lubricant blender has to store base oils that are used in lubricants having ACEA/API accreditation separately from base oils that do not have/need such accreditation. In order to economise on storage capacity, or to increase their buying power vis-à-vis the base oils supplier, blenders may not differentiate according to the lubricant application. In such cases they would tend to use base stock used in lubricants having ACEA/API accreditation for both automotive and industrial applications.

MARKET DEFINITION

Product markets

Base oils

(308) According to the parties, base oils are homogeneous products and therefore would form one single relevant product market whatever their application or group. As explained above there are many applications for lubricants and each of these applications require a different base oil with varying characteristics. It could therefore be argued that there are as many relevant product markets as there are specific requirements. There seems however to be a high level of supply-side substitutability whereby the same refinery could produce any quality of base oil within group I. Finally, group I and groups III/IV base oils do not appear to belong to the same relevant product market.

Demand-side substitutability: no possible distinction between automotive and industrial applications

(309) The following list shows the technical requirements for base oils according to the application the lubricant blender may use the base oil for:

- automotive crankcase lubricants: group I (HVI), group II (although not in Europe), group III (VHVI and XHVI), group IV (synthetic base stock) and group V,

- automotive gear and transmission oils: group I (MVI and HVI) and group III/IV,

- industrial lubricants and marine lubricants: group I (LVI, HVI and HVI) and group II,

- additive manufacturers: group I (LVI, HVI) and group III (XHVI),

- exports, waxes and process oils: group I: LVI, MVI and HVI,

- white oils: group I and III,

- jet engine lubricants and some industrial lubricants: group IV (synthetic base stock) and group V.

(310) It follows from this list that most group I base oils can be used for both automotive applications and industrial applications apart from a number of exceptions (notably the LVI grades which are used for rubber process oils, refrigerator oils, greases, metalworking fluids and shock absorbers). In addition, it results from the market investigation that there are only minor quality or price differences between group I base oils for industrial applications and automotive applications and no supply differentiation.

Supply-side substitution is possible within group I base oils

Viscosity index

(311) In Europe, more than 90 % of base oils are HVI. At the level of the viscosity index, base oils are not fungible products, as each type of lubricant needs a specific performance level determining viscosity index (VI). A base oil competitor has for example indicated that "HVI oils of 150 viscosity, low volatility and VI below 99 form an important subgroup for automotive applications". Segregation must be maintained between base oils in user facilities to maintain accredited product status.

(312) However, a refiner can obtain group I base oils with higher viscosity indexes through solvent extraction by more severe processing. This supply substitutability reduces the yield and has therefore an effect on the production capacity of the base oil plant. Throughput reductions may be substantial and have an impact on base oils production capacity (10 % or more depending on the crude and hardware configuration of the refinery). More severe processing can however not lead to obtain the very high VI's which are typical for the group III base oils. Additional additives used in the finished lubricant formulation can compensate for low base-oil VI. The output in terms of VI also depend on the refinery configuration and the lubricant feedstock.

Viscosity and other characteristics

(313) Group I base oils obtained through the normal solvent extraction separation process can be divided into two categories on the basis of viscosity: low-viscosity or light base oils (N150 for example) and high-viscosity or heavier base oils (500N and brightstocks for example). Lighter base oils are preferably used in areas with a more moderate climate and are driven by the automotive engine oil requirements asking for low-viscosity base stocks. They give better performances and have properties that allow better emission control and fuel consumption. Demand outside the EEA and North America is predominantly for the higher viscosity grades because of the warmer climate. Exxon, Shell, BP/Mobil, Total and Elf for instance export very significant volumes to their affiliates in Africa, Asia, and Latin America. Up to a certain degree, there is also a difference in base oil demand between the north of Europe (lighter base oils) and the south of Europe (heavier base oils).

(314) The parties argue that the relative volume split between light and heavy neutrals can be varied substantially, depending on crude quality, distillate capacity, demand requirements, and hardware limitations. Although there is a certain degree of supply substitutability from the production side, market investigation has shown that this is only possible up to a certain degree, as a specific feedstock can only assure a limited output of light grade base stock.

Group I base oils are not substitutable with group III and group IV base oils

(315) There are significant price differences between the various groups of base stock. Group III for example costs two to four times as much as group I HVI. In addition, group I and III base oils can be produced on the same refining hardware (although refinery tankage must be allocated to group III base oils to prevent commingling with group I base oils) but are not produced from the same feedstock. Group I base oils are produced from crude or crude fractions (reduced crude, vacuum gas oil, residue) while group III base oils are produced from the bottom stream that comes from certain hydrocrackers (hydrocracker residue) in fuels refineries. The group IV polyalfaoleofins (PAO's) are the result of a chemical process.

(316) Base oils belonging to group III and group IV are used for blending high value added lubricants. They are also sometimes blended with lower-viscosity index base oils in order to obtain finished lubricant with increased properties (for example, HVI can be blended into finished lubricants with synthetic base stock although such group III equivalent is more expensive to blend than it is to purchase the group III base oils on the market). Similarly, refiners can change the quality of crude or further refine their output in order to alter quality of their production of base oils for one specific group (although not to such extent that properties of a higher group can be obtained). Finally, additional amounts of additives can be blended with base stock in order to improve properties.

(317) In Europe, more than 95 % of base oils are group I, while group III quality actually accounts for less than 5 %. Market investigation has indicated that group III is likely to grow over the next few years by 8,5 % a year and group I decline by 0,5 % a year. Demand for group I is declining due to the evolution of engine technology, the higher requirements of OEM's on fuel consumption, performances, oil drain intervals and more rigid emission legislation.

(318) On the basis of the above, group III and group IV can be considered as belonging to different product markets from group I base oils. The economic drive is to maximise the use of group I base oils and to meet the lubricant requirements with a balance of base oils and additives based on cost considerations. The choice to use group III or IV in a finished lubricants formula is made when their specific properties (other than just their VI) compensates the higher cost in meeting the required performance. [...]*, it is not necessary to further delineate on group III and IV as separate product markets (group II is currently not used in the EEA).

Additives

(319) According to the Commission's Infineum JV decision, "VI improvers, DI packages for automotive applications and DI packages for industrial applications are separate markets"(30). The parties have identified only one affected market, relating to DI packages for automotive lubricants. As regards DI packages for engine and transmission applications, the results of the investigation suggest that from a demand-side point of view these products probably do not constitute one product market, but that there could be different markets depending on the application. However, as Mobil represents less than [...]* % of additives production, it is not necessary to further delineate the product market as regards DI packages for engine and transmission applications.

Lubricants

(320) The notified operation does not give rise to competition concerns in any of the possible lubricants markets (apart from aviation lubricants), so the definition of the relevant product markets can be left open.

Relevant geographic markets

Base oils

(321) It is the parties' opinion that the market for base oils is at least EEA-wide but is more likely worldwide. The parties base their contention on the following considerations. First, the same types of base oils are sold and used throughout the world. Secondly, base oils may be shipped economically over long-distances and there is already significant trade from the EEA. Thirdly, base oils prices and gross margins present some correlation. None of these contentions resist closer examination and the market is therefore EEA-wide.

(322) Base oils prices and margins in different regions of the world indicate that the EEA forms a distinct relevant geographic market

(323) Fob prices have been consistently different in the EEA, the US and Asia with Exxon's IABP prices in Singapore and the USA being in average greater than EEA prices by from [...]* % (for SN 150 in Singapore) to [...]* % (for SN 600 in the USA). This difference is even greater when taking into account the transportation costs. The parties explain that transportation from Sicily to the Middle East costs around USD [...]* per tonne. Taking this as an estimate of transportation costs from Asia to the EEA and intra-EEA transport costs at [...]* % of landed prices, that still leaves Exxon's Singapore prices at a higher level than EEA prices. Only 1998 and 1999 prices in Singapore would be at a level that could make imports into EEA profitable if EEA prices were hypothetically increased by [...]* %. However, these prices are the result of the crisis in south-east Asia and are not likely to remain on a lasting basis at such levels. [...]*(31).

(324) Another important element is the fact that export prices from the EEA have been typically lower than the prices charged in the EEA. The parties have estimated that difference in average at a level of USD [...]* per tonne. That is [...]* % below the current EEA prices. This is roughly equivalent to the freight costs and shows that no arbitrage is possible between the EEA and other areas of the world unless there are exceptional circumstances such as the Asian crisis. This is underlined by the fact that there is almost no trading activity.

(325) Europe has historically been the base oil producer with the largest excess capacity production, and is consequently the swing producer. Exports from Europe to Asia, Africa, the Far East and Latin America have been important in the past since these countries were in need of the heavier base oils.

(326) New production capacity plans have been announced or have emerged in these countries (mainly group II or III capacity through Exxon, Mobil or Shell projects), which were previously net importers of European base oils. The EEA surplus production is mostly composed of heavy grades for which there is limited demand in the EEA and sufficient demand outside the EEA. Heavy grades (for instance 600 solvent neutral) are therefore dumped on the spot export market. Surplus production for light grades (for instance solvent neutral 150) is sold in the EEA market. Exports are predominantly for heavy grades; on the basis of the figures provided by the nine most important competitors to Exxon and BP/Mobil, the breakdown into viscosity grades is 24,3 % for light neutrals, 12,63 % for medium neutrals and 63 % for heavy neutrals. Heavy grades are also used as process oils.

(327) Exxon has explained in its management documents that [...]*(32)(33)(34)(35)(36)(37)(38).

(328) If EEA prices were to increase by, say, 5 % relative to Singapore prices, imports into the EEA would not be able to compete readily with EEA production. Since lubricant formulations are bound by base oils interchange rules of ACEA and API, major investments would have to be made to obtain accreditation for these infrequently traded base stocks. Market investigation has indicated that none of the major additives companies have invested in accreditation programs for imported base stocks from companies other than Exxon, Mobil or Shell. Even if the base oils interchange rules issue were solved (which would be costly and take at least 18 months) this would only have an effect on the ACEA accreditation and not on API. Transportation and handling costs would in any case need to be added.

(329) Finally, it must be noted that, contrary to the parties' assertions, there is no real spot market for base oils in Europe with any material liquidity. All traders that were questioned have responded that they have not imported base stock volume in the EEA in the past years. Traders play only a small role as intermediaries in export business. All but one competitor to the merging parties (apart from Repsol: 24 KT) have indicated that they do not use traders for exports or imports from outside of the EEA.

(330) From the discussion above it can be concluded that the production and sale of group I base oils in the EEA forms a relevant market. This is notably due to the current specific European consumption profile and to its quality requirements.

Additives

(331) The parties have argued that the relevant geographic market is at least the EEA. The determination of whether the market would be the EEA or wider does not alter the competitive assessment of this transaction and therefore the market definition can be left open.

Lubricants

(332) For all the possible lubricant markets (with the exception of ester-based turbine lubricants that is discussed later), their geographical scope can be left open since, on the basis of the above market definitions, the concentration will not create or strengthen a dominant position which will significantly impede competition.

COMPETITIVE ASSESSMENT

Base oils

Current market structure

Market positions for base oils; a two-tier competition

(333) The producers of base oils can be divided into four groups:

(i) the pan-regional companies: integrated oil companies who have base oil refineries worldwide with a strong EEA presence and production which exceeds their needs: Shell, Exxon, BP/Mobil;

(ii) the regional companies who have national base oil offer with production that for some of them exceeds their internal needs: Cepsa, Repsol, Agip, Fortum, Elf, Total, KPI, Petrogal and DEA;

(iii) producers of infrequently traded base oils like Iplom, MOH, SRS. These base oils are likely to be of a lower quality and are destined mainly for exports, high volume industrial lubricants and process oils. The four major additive suppliers have indicated that they have not undertaken accreditation for these base oils;

(iv) the unconventional base oils producers like Nynas (naphthenic base stock for speciality industrial lubricants) or Fortum (previously Neste) producing PAOs.

(334) Exxon supplies base oils throughout the EEA from three plants located in Fawley, United Kingdom, Port Jérôme, France and Augusta, Italy. BP/Mobil has five plants located respectively in France: Gravenchon and Dunkirk (the latter including a 40 % participation of ELF), the United Kingdom (Coryton), Germany (Neuhof) and in Spain where it operates, together with Cepsa, the base oils plant in Algeciras. Shell has four plants, in the Netherlands (Pernis), France (Petit Couronne), Germany (Hamburg) and the United Kingdom (Stanlow). Exxon, BP/Mobil and Shell control 12 of the 24 base oil refineries in Europe, among which eight of the 10 largest. In order to calculate the volume of the EEA market for group I base oils and the parties' market share the Commission has taken the following elements into consideration. The market shares were calculated according to three methodologies: first, on the basis of nameplate capacity as proposed by the parties, secondly, on the basis of effective production and thirdly on the basis of group I production.

Nameplate capacity and effective production

(335) All figures are based on 1998 data provided by Exxon, Mobil and their base oils competitors (third parties) for their respective plants. Where third parties had not presented the requested figures (as is the case for MOH, SRS and to a minor extent AGIP), the Commission used the data as presented by the parties in the form CO. As third parties have requested to consider the provided data on effective capacity, production and sales as confidential, only ranges of percentages have been retained in the table. The following base oil production and capacity table compares the effective capacity and output with the Enerfinance '96 nameplate capacity provided by the parties.

EEA production and capacity

>TABLE>

(336) Process oils are considered as base oils. Aromatics, parrafins, white oils and greases are not included in base oils sales as they are either a by-product or need further refining. White oils, paraffins and waxes depend very much on the raw materials output and competitive advantages and economies of scale lead to very large market shares. These products are, however, either the result of a further step in the refinery process of base oils or a by-product of base oils production. The assessment of the parties' competitive position therefore is linked to that of base oils.

Market shares based on sales to third parties

(337) Merchant market sales have been defined as sales to third parties both inside and outside the EEA including spot-market sales. Sales to affiliates are considered as captive sales. Sales to clients outside of the EEA has been kept in the calculation to reflect the parties position that exports could be redirected to satisfy the EEA demand. Sales to additive companies are considered as merchant market sales (except in the case of AGIP additives and sales by Shell and Exxon to their additive activities prior to the JV).

>TABLE>

(338) Exxon's ability to sell to the merchant market, due to investments in base oil capacity in the 1950s and 1960s, is extremely important: [...]* of its production is available for the merchant market while BP/Mobil's capacity for the merchant market comes to [...]*.

(339) On the basis of 1998 sales figures, the calculations indicate Exxon's significant share of the merchant market ([30 to 40]* %), with BP/Mobil taking second place with [10 to 20]* %. Shell is third with a stake close to BP/Mobil (although it has to be indicated that nearly half of the volumes Shell has available for the merchant market are sold as process oils). Total and Agip have market shares below 8 %. Smaller competitors, such as DEA, Cepsa, Repsol, KPI, Petrogal, SRS, Petrogal and MOH, have between 3 % and 5 % of the merchant market. Elf is to be considered as a net buyer on the market.

(340) Finally it must be stressed that the market shares indicated above have not materially varied in the past few years.

(341) Overall, it appears that whatever the methodology chosen, Exxon is the clear market leader in the base oil market both in terms of absolute production and in terms of production available to third parties. The main competitors to Exxon are BP/Mobil and Shell but they sell a smaller proportion of their output to third parties.

Barriers to entry and exit

(342) According to the parties, new capacity in the EEA is unlikely at current overcapacity and depressed prices, and if new capacity were to be installed, it is likely to be group III capacity for which increased demand is to be expected in the future. The parties have estimated that the construction of a new plant (group II/III) would require USD [...]* million for a [...]* KT/year capacity. Third parties have confirmed this by indicating that the construction of a new plant would take two to three years and involve investment of USD 400 million for a 400 KT/year plant.

(343) Converting existing group I capacity to group III is feasible and requires an investment of USD [...]* compared to a USD [...]* grassroot investment. [...]*(39).

(344) Closing a refinery entails significant costs, notably because of the need to rehabilitate the land. Nevertheless, three base oil refineries have closed over the past two years: Llandarcy (BP/Mobil UK: with a nameplate capacity of [...]* Kty.), Addinol (170 Kty in former East Germany) and one of the two DEA base oil plants in Germany. These base oil refineries were producing heavier grade base oils and were therefor hit by the drop in export prices since 1997. Although there is a significant cost, refiners take the decision to close when the break-even point for closing is reached. Other base oil plants are said in the industry to be likely to be closed in the near future because of low productivity.

Privileged role of Exxon, BP/Mobil and Shell with limited constraints exercised by smaller base oil producers

Base oils interchangeability for all refineries and limited accreditation constraints.

(345) API/ACEA accreditation is given for the finished lubricant blend of base oil and an additive after having passed the necessary tests. This accreditation precisely defines the producer of the base oil (accreditation for the base oil is given for all plants) and the specific additives package. If the origin of components is changed, re-accreditation is necessary (10 % group I interchange is accepted without testing required). The consequence of this is that pan regional base oil suppliers (that have invested(40) in obtaining interchange for their base oil production plants) have to pass accreditation of their base oil with a certain additive package only once. As they have several production units and multiple storage units with geographic coverage, they are in a better position to supply several blending plants. Regional base oil producers, who do not have this geographic production spread, either have to face higher transportation costs from their refinery to the customers premises or have to buy base oil on the market close to their customer (the blenders) in order to avoid higher transport costs but then have to pass accreditation for each base oil bought on the market.

(346) Competitors to Exxon and Mobil (both for base oil and for lubricant blending) have indicated that blending plants are very sensitive to the certainty of continuous supply and quality of base oils as the stocks they have available at their plants are limited. Supplying relatively small volumes from a single base oil plant to a pan-regional network of blending plants at regular intervals therefore imposes significant costs for even the largest of national integrated base oil companies with sufficient merchant market capacity.

Advantages derived from plant economies of scales

(347) The parties have acknowledged that economies of scale play a certain role in the production of group I base oils. Exxon, BP/Mobil and Shell control [...]* of the 24 base oil refineries in Europe among which [...]* of the 10 largest. That brings them a certain cost advantage over their smaller competitors because they can better spread fixed costs over the production. Only Agip in Italy and Total in France have plants of comparable size and could benefit from comparable economies of scale. As Mobil has indicated in its business plans(41), [...]*. Shell is considered by the parties as efficient and profitable thanks to a integrated (fuels and lubricants) view of business and profitable waxes/greases/white oils.

(348) The following table provides a ranking according to base oils plant dimension:

>TABLE>

(349) Additional benefits derived from dimension are (a) that large plants can optimise the crude intake, and (b) that free production capacity is created, which is important in optimising the production process as it offers more flexibility: this is important when producing several base stock categories.

Advantages derived from economies of logistics

(350) As indicated previously, Exxon, BP/Mobil and Shell have multiple refineries in the EEA. This Europe-wide spread is further supported by storage capacity at strategic places in Europe

(351) This contrasts with the smaller national producers who only have one refinery from which it is not as easy or economical to organise distribution to potential customers not located close to their production. This is obviously the case for those landlocked refineries such as SRS and DEA, which can supply local customers only by truck and train. These means of transport limit the maximum distance for which it is economically viable to transport the product, but also impose additional transport costs. Competitors have indicated that transport by land is limited to a radius of 500 km by truck and 1000 km by train. Transportation costs are not comparable due to the limited volume truck and train can transport and vary between 6 % and 10 % of the sales price of the product.

(352) Even single refineries based on the sea face higher average transport costs. The parties have estimated the average EEA transport cost at [...]* % of the landed price. They also explained that EEA transport costs comprise from [...]* % (United Kingdom to the Netherlands) to [...]* % (Italy to the Netherlands) of variable costs. Assuming that the fixed costs remain constant for the two possible journeys, this means that the cost of transport from Italy to the Netherlands is [...]* % higher than from the United Kingdom to the Netherlands(42). In other words if transport costs from the United Kingdom to the Netherlands were [...]* % of landed price, they would be [...]* % if the products were transported from Italy. The geographic spread, coastal location or connection to blending plants therefore determine the competitive position of a base oil refinery and its ability to compete on the EEA merchant market as opposed to local or outside EEA export market.

The dynamics of the base oil market reinforce Exxon, BP/Mobil and Shell's advantages

Ability to set the standards for future lubricant development

(353) Because of the need to obtain accreditation for blends of base oils and additives, size is a self-reinforcing competitive advantage in the base oil market. On the basis of market information, additive packages for new lubricant formulations are typically developed first for Exxon, BP/Mobil and Shell base oils. The major European automotive additive suppliers formulate additive packages for the base oils selected by the lubricant blender. In general, independent blenders select Exxon, BP/Mobil or Shell base stock, while national integrated oil companies with blending activities may select their internally produced base stock. Additive producers will independently invest in additional accreditations, but will only do so for the major traded base oils (principally Exxon, Shell and BP/Mobil), as for these base oils, the additional accreditation investment will be outweighed by the opportunity to secure extra sales. Other base oils than Exxon, BP/Mobil and Shell tend to be accredited in an ad hoc manner to meet customer demands if such appears to be technically possible and commercially viable.

(354) Additive suppliers have indicated that, on average, more than half of their additive sales are made of "off-the-shelves packages". The Commission has asked the four main additive suppliers to indicate with which base oils they have developed duly accredited "off-the-shelf DI packages", either for industrial or for automotive lubricants purposes. On average, more than 70 % of all the packages available are formulated on the basis of Exxon, BP/Mobil and Shell base stock. Apart from these companies, only Total and Elf (and to a minor extent AGIP and KPI) have accreditations with multiple additive suppliers. Exxon, BP/Mobil and Shell are therefore the reference base oil suppliers with few other smaller base oil suppliers being compatible. Additive suppliers have confirmed that they tend to propose and develop formulations for automotive applications only with base stock from the big three since they can expect to generate more sales for these base oils which are readily available on a pan-regional market in the necessary volumes and at constant quality.

(355) Finished products composed of Exxon, BP/Mobil and Shell base oil can be more rapidly adapted to new standards since base oil interchange ensures that the product can be purchased close to the blending plant without imposing additional accreditation. If a multi-regional blender would opt for purchasing base oils from different local producers, multiple testing and accreditation would be required for obtaining finished lubricant accreditation. That makes them less attractive for additives suppliers.

(356) According to third parties (both blenders and base oil producers), this inherent advantage of base oils produced by Exxon, Shell and BP/Mobil is of paramount importance in an industry that has witnessed over the recent years a shortening of the product life of lubricants. It is today in the range of three to five years for automotive lubricants. Moreover, OEM qualification requirements are becoming more severe. Lubricant manufacturers therefore constantly have to develop new formulations. Because of the shorter time available to develop new products and to amortise these costs, lubricant blenders tend to turn to the three major base oil producers, with whom additive manufacturers systematically develop adequate packages. Additive suppliers indicate that they first develop the new formulations with Exxon, BP/Mobil or Shell base oils since only these base oils can generate enough sales to recoup the investments made. OEMs, sometimes in collaboration with oil companies, steer the next generations of lubricants. Competitor lubricant blenders have indicated that OEMs will tend to discuss their requirements with Exxon or BP/Mobil and to a lesser extent with Shell (principle of one-stop shopping) as the only global players with presence on all levels of the supply chain. [...]*(43). Competitors have indicated that their limited ability to compete with Exxon, BP/Mobil and Shell will be further restrained by the ongoing shortening of the product life of lubricants.

(357) Industrial lubricant blending is equally affected by this trend, as blenders try to minimise storage and the diversity of base oils/additives. Lubricant blenders subscribing to high quality standards do not store base oil of different suppliers in the same base oil tank (not even for base oil with similar specifications). Due to limited storage, the blender therefore tends to use a "compliance base oil" that can be used for both automotive and industrial lubricant applications.

The base oils market is driven by the higher quality requirements

(358) Blending plant logistics also add to the cost of switching base oils as blending plants are normally optimised to keep the number of storage tanks needed for base oils and additives to a minimum (some of the very large blending plants may store over 30 grades of base oil). A "compliance base oil" is therefore often used. A compliance base oil is locked into particular specifications across a wide range of products. This means that the base oil used for top-tier products is also used for lower-tier products. Changing the formulation of a top-tier product leads to a wide span of formulation costs driven by the choice for the top-tier base oil. The alternative is to purchase new tankage capacity and carry greater stocks, which can be expensive, or to specialise in the production of either only automotive lubricants or only industrial lubricants (although the parties have indicated that this is a recent and growing trend, they have claimed that there is no separate base oil product market for industrial and automotive lubricants). The fact that base oil and additive combinations are locked in by accreditation also affects the blender-additives relationship as a lubricant blending company typically deals with additive suppliers on the basis of a whole basket of additives it needs to purchase. If the blender changes the lubricant formulation by switching base oils suppliers, then the additive supplier will need to agree on a new (joint) development programme. If this understanding is absent, the oil company would be forced to shop around for base oil-additives combinations that are already fully accredited, thus reducing his buying power vis-à-vis the additives producer.

(359) Virtually all branded automotive products are approved against standard industry specifications (API, ACEA) under tightly managed and documented test regimes. Achieving these minimum performance standards are qualifiers, merely allowing the product to be used for the particular application. Choice of approved product is then up to the channel customer or end-user (OEM). Major oil companies liaise with OEM's and contribute to their strategic thinking, which may direct their technical development to specific customer requirements. When an OEM changes its specifications, this will prevent any sales of a lubricant, whatever its brand, until the new specifications can be complied with. Pan-regional competitors (Exxon, Shell, BP/Mobil) are increasingly adopting a single pan-European product range (both for automotive as for maintenance lubricants).

(360) A lubricant blender cannot afford to see its competitive position endangered when the launch of a new product formulation might be delayed because of the time needed to acquire extensive testing documentation for a less frequently traded base oil. When an additive supplier has the choice of initiating a new product formulation, he would generally opt for Exxon, BP/Mobil or Shell base oils. Base oils interchange and finished lubricants accreditation limit the frequency of changes in basestock suppliers once a lubricant is formulated. [...]*(44).

(361) The degree of customer switching as provided by the parties indicate that national base oils producers compete for local blenders (both automotive and industrial), specialist industrial blenders, additive suppliers [...]*(45) and will occasionally win a contract for first fill or major blender in their home market. Exxon and to a minor extent BP/Mobil are focused on term contracts with the pan-regional blenders and OEM first fills.

(362) Exxon and BP/Mobil have lost a number of small blenders as clients to competitors, such as Total, Elf, KPI and Shell. Sometimes this was on the basis of quality and sometimes on the basis of price, but in all cases the winning competitor had production capacity close to the client. This does not, however, imply that Exxon and BP/Mobil do not compete on price; in a number of cases, the lost customer was recuperated the year after.

(363) Although Exxon and BP/Mobil have [...]*, they often lose out when competing with Petrogal, Repsol, Cepsa, KPI, Total, Elf or even small amounts of eastern European base oils. It is however to be noted that, as the parties indicate, "additivers may be able to use relatively low quality base oils" and that additive suppliers sometimes buy these volumes additional to their Exxon, Shell or BP/Mobil term contract on the spot market. As additive suppliers do not impose stringent accreditation or quality requirements, less powerful competitors ambitiously compete on price (even if this leads to additional transportation costs) to capture this demand.

(364) Pan-regional demand (Texaco, Fina, Castrol, Fiat lubrificanti, Fuchs, etc. ) is increasingly captured by Exxon, BP/Mobil and to a minor extent Shell as pan-European contracts are concluded. Occasionally, some supply is lost to a national base oils competitor who is close to the blending plant (Agip in Italy, Repsol in Spain) but more often volume is gained from these competitors. Quality, certainty of supply and no accreditation restraints are crucial for these blenders, [...]*(46).

(365) In any case, it can be concluded, [...]*, that Exxon functions as a swing producer as it supplies products to both national base oils producers (Elf, Total, Agip, KPI, MOH) and BP/Mobil and Shell whenever these producers are confronted with product shortage, refinery maintenance or need for specific grades. When the market becomes long or when production capacity is back at normal level, these clients fall back on their internal production.

As a consequence smaller base oil refiners focus on local sales and on exports

(366) Depending on their relative strength (local champions), regional integrated oil companies with blending activities might be able to win contracts for blenders for certain regional markets (for instance, an OEM first-fill for cars produced in the home country of the base oils producer but not in other countries), are present in industrial lubricants for which requirements are based on general performance rather than formulation and have developed strong finished lubricant sales positions.

(367) Small independent lubricant blenders also tend to acquire positions in niche markets. They will essentially be active in the industrial specialities field where the competitive factors are not strongly related to (integrated) strong positions in base oil productions and where geographic spread and economies of scale are less determinant. Local small blenders who have specialised in industrial lubricants will be very price-sensitive and will change base oil suppliers to increase their leverage. Customer technical assistance, flexibility and know-how in specialised applications are the primary competitive assets and smaller, local refiners often excel in this. These local blenders will not have pan-regional suppliers; they will prefer to initiate joint-development programmes with local additives suppliers and guaranteed base stock supply nearby their blending operation. The importance of base stock is also less relevant in these speciality lubricants. For example, a speciality metalworking fluid will be sold at three times the price of an industrial hydraulic oil. The base oils of which this metalworking fluid is composed will account for 5 % of the total production cost compared to 35 % for the additives.

(368) However, even for industrial lubricants and marine lubricants the base oil/additive combination is equally important despite the fact that performance accreditation can be kept "in-house". OEMs are becoming increasingly aware of the potential effects of different base oils. A base oil switch would require validation tests of a number of key products and critical performance parameters.

(369) In addition to base oil interchange and finished lubricant accreditation, regional location and OEM approvals are crucial as well. This is illustrated by a base oils refiner that profits from multi-plant base oil interchangeability, who has indicated that base oil interchangeability does not automatically lead to guaranteed sales. This company has confirmed that it has very limited sales for automotive applications and has only a few base oil/additives accreditations. According to the company, they lack regional presence and economies of scale to compete with Exxon, BP/Mobil and Shell on a pan-European level. They point out that they are increasingly losing national automotive sales to the benefit of Exxon and BP/Mobil due to increasing OEM requirements and an evolution to pan-Community demand. For 1999, only supply contracts for industrial applications have been renewed. They also indicate that, even if they had full accreditation with all major additives suppliers, they would not be able to win pan-EU supply contracts because of their local presence (cost of transport within the EEA is 10 to 12 % including handling). Only 9 % of their production is destined for the pan-European market and is exclusively used for industrial lubricants.

(370) The market dynamics described above have left the smaller base oil companies little choice but to increase their sales on the export markets. From the market investigation conducted by the Commission, it appears that between 20 and 40 % of their current production is sold on the export spot market. As indicated before, a large part of these volumes consist of heavy grades for which there is insufficient demand in the EEA. Most of the time these sales are done on a spot basis and at prices which are materially lower than on the EEA market (at current prices between 20 % and 30 % lower).

[...]*

(371) [...]*(47)(48)(49).

Competitive constraints exercised on Exxon, BP/Mobil and Shell by the smaller players: limited possibility to increase their sales in the EEA market

(372) As explained above, the ability of national players to compete with pan-regional base oil suppliers is limited by the greater transportation costs they have to bear and the smaller extent of their base oil accreditation. This restricts these smaller players to supply mainly local blenders which are mostly oriented towards industrial lubricants or to export their production. The parties have argued that the EEA base oil market is affected by strong overcapacity which limits the ability of any player to manipulate prices or production. In the course of the Commission's market investigation, competitors have put in doubt the relevance of the alleged structural overcapacity as a check on Exxon/Mobil's possible behaviour. In effect, there is no overcapacity on the market but rather surplus production that cannot be sold on the EEA market. This is due to the following factors.

Actual overcapacity is misstated by the parties

(373) The parties have defined overcapacity as the difference between nameplate capacity and actual production. Nameplate capacity appears however to be an unreliable indicator of the production capacity of a base oil refinery. Instead of speaking of overcapacity, it seems more relevant to focus on surplus production (that is, the difference between current EEA production and current EEA consumption).

(374) First of all, the parties have updated their data on nameplate capacity, admitting that the nameplate capacities mentioned in the form CO are significantly higher than actual production capacity. On the basis of production capacity that has been shut down alone, the parties have revised their initial "more than 30 % overcapacity" to less than 24 % although different appreciations of overcapacity circulate(50). These figures are still questionable. For instance, the parties had revised downward many of the publicly available nameplate capacity estimates of their own plants but not those of competitors. In addition, these figures are based on capacity without differentiating the output. Competitors like NYNAS only produce naphthenic base oil (group V) for specific industrial applications which cannot compete with paraffinic base stock used for automotive or general industrial lubricants (excluding specific applications such as printing inks or rubber) due to the price differential and technical characteristics. The same goes for those refineries which produce group III or IV base oils.

(375) There are many reasons why nameplate capacity is an overestimate of the real capacity of base oil refineries. This is for instance due to bottlenecks in the production process and refinery shutdowns for maintenance that lower the effective production capacity. An important base oil competitor explains that because a number of Community base oils plants are old, the actual average production capacity tends to be below nameplate capacity. These refineries no longer benefit from de-bottlenecking [...]*(51).

(376) The parties have abundantly cited a study of the European refinery industry made by Roland Berger. One finding of this study was that refineries, because of their very high fixed costs, tend naturally to produce at any time at their maximum capacity available. Indeed, competitors to Exxon and BP/Mobil have indicated to the Commission that their plants run at maximum capacity (except for base oil refiners producing other than group I base oils) and that it would not be economically viable to reduce the utilisation of their refineries. Cepsa indicates that even with significant volumes exported outside EEA (at lower prices than inside EEA) the plant runs at 100 %. According to Cepsa, reduced utilisation of the production capacity is economically not viable.

(377) The overall surplus production ratio calculated as the proportion of exports among merchant sales is of 26 %. However, around [...]* of this surplus production is in the hands of the merging parties. The surplus production held by third parties amounts to approximately [...]* % of the merging parties EEA merchant sales. As explained below, it is nevertheless unlikely that any of this excess production could meet the EEA demand.

Base oil production destined for the EEA market could not be easily and readily increased through re-tuning of refineries

(378) The parties have argued that base oil production in the EEA could be at no cost increased by 5 to 10 % through better crude intake and tuning of the refinery. However the market investigation conducted by the Commission has revealed that most of the EEA refineries are already running on the most appropriate crudes to produce base oils. Similarly, it would be irrational to expect refiners not to have optimised their production already through the proper tuning of their refineries. Some of the most important competitors to Exxon and BP/Mobil have indicated that they are tight on light grades.

(379) Finally, it could be argued that some refiners would rather further refine base oils in order to produce fuels and could at any moment stop that process and therefore increase their production of base oils. The parties have however explained that further refining base oils into fuels has never made economic sense for the past 10 years.

(380) A refiner can, however, decide to attribute base oils production capacity to fuels production; a smaller competitor producing group III base oils has indicated that they are considering switching to fuels production if they are not in a position (due to low demand) to raise output of base oils. [...]*(52).

(381) Crude/feedstock selection influences the volume of the effective output. For instance, it appears that smaller refineries would adjust to declining prices for export by using crude slates that allow them to steer the feedstock into increased fuel production and therefore lower base oil production. Kuwait Petroleum International (KPI), an important base oil competitor puts it as follows: "A refiner will maximise local opportunities. After that he will either not produce base oils and use the feed molecules for the production of fuels, thus not utilising the base oil refining capacity. Either he can decide to produce the base oil and sell it into export (non-EU) market" or another competitor: "We will direct feedstocks to increased fuels production if capacity utilisation rate is low".

Base oil exports from the EEA could not be easily redirected to the EEA market

(382) Base oils exported from the EEA could normally not be consumed locally because they do not meet the local requirements. Although the standard qualities (for instance a 150 solvent neutral) offered and sold by all European refiners are the same in export markets as in the EEA, demand is different in the export markets when compared to EEA demand.

(383) High grades have historically been highly in demand by the export regions for engine oils in warmer climates, for higher industrial lubricants consumption (heavy industries in emerging markets) and for higher process oils consumption (tyre production). Until Asian capacity competed with these exports, heavy grades were sold at a premium. Light grades consumption in the EEA has increased as engine lubricant requirements have gone up while there is more volume on the market for heavy grades and less demand (Asian crisis). This difference in qualities is illustrated by the material differences in prices reported on the export market compared to the EEA market, as is explained in the section below.

>TABLE>

(384) The table above shows the evolution of prices for two typical base oils within the EEA as compared to the export for the same quality. It is striking that the price differential rises from 11 to 12 % of the EEA price in August 1998 to 26 to 29 % in May 1999. If exports could be redirected in any way to meet the EEA demand, there is little doubt that the EEA refiners would have done so long ago because of the longstanding price differential and even more likely in the past year because of the material increase in the price differential.

(385) Export spot prices are significantly below EEA prices for a given base oil (for example SN 150) and this can, apart from the effect of term contracts and potential quality differences, only be explained by supply and demand effects. As the parties have indicated, a refiner will choose to dump a certain volume on the spot market as tankage approaches maximum levels and will be ready to accept market price levels covering the cost of producing the base oil versus having unused capacity. In a long market, this value could be lower than total operating costs but still could be economic for the producer. It is however important to stress that Exxon, Shell and to a minor extent BP/Mobil, have very few spot-market sales as compared to national base oils competitors. As indicated before, Exxon, Shell and BP/Mobil sales are mostly term, competitors are spot-oriented. If flexibility existed in the industry to redirect volumes from export to EEA consumption, then it is clear that, at the actual price premium of 28 %, this flexibility does not exist for producers who are mainly spot-sales oriented. As indicated before, single base oils plant refiners with limited accreditation are focused on local blenders or sell on the spot market (both within and outside the EEA). In any case, the volume of light grades exported outside the EEA is limited, as these volumes are sold on the EEA market because of higher prices and lower demand outside the EEA. The viscosity breakdown of base oil volumes exported in 1998 by the nine most important competitors to Exxon-Mobil is as follows: light neutrals ([...]* %), medium neutrals ([...]* %), heavy neutral: [...]* %. Some competitors have even described the light base oils market as becoming increasingly tight.

(386) In an internal Mobil memo(53), the effect of rationalisation on light grades is described as follows: [...]*.

(387) Based on figures provided by the most important competitors (Agip, Cepsa, DEA, Elf, KPI, Repsol, Shell, TotalFina, Petrogal) to the parties, the viscosity breakdown of base oil volumes produced in 1998 is as follows: light neutrals (45 %), medium neutrals (8,9 %), heavy neutrals (46,1 %).

Despite the higher prices available in the EEA, base oil producers do not have the possibility of increasing the output of light grades for which there is twice as much demand in the EEA as for heavy grades. EEA base oil producers cannot switch to increased light grades production without heavy investment as competitors have indicated that they have optimised LN production

(388) Competitors have indicated that the flexibility of base oils production is relatively limited. A given crude contains only a fixed potential of a certain viscosity. Selecting crudes other than the crudes currently used by the major competitors to Exxon and BP/Mobil will only have very limited (3 to 5 %) effect on light grades output. Switching crude oils also presupposes that the refiner has sufficient lubricant vacuum capacity. If he has not, an investment of around USD 50 million is necessary (figures based on competitors' estimations).

(389) Increased light grades output can be obtained by converting heavy feed molecules into light which would involve the construction of a hydrocracker (representing an investment of around USD 100 million).

(390) The parties have indicated the possibility of running more suitable lubricant feedstocks. Competitors have indicated that the economics for doing so are strongly negative as lubricant qualified feedstock comes at a premium of USD 10/tonne versus the normal feedstock (as used by the fuels refinery). Supposing that the yield of the light base oil is around 20 % of the feedstock, then this would result in a debit of close to USD 50/tonne on the production cost of the incremental light base oil, reducing gross margins by about [...]* %(54). To this have to be added transport costs of lubricant feedstock and the loss of revenues for HN sales. Several competitors have indicated that they are not able to find lubricant feedstocks without facing unviable transport costs.

(391) Volatility is the second element. A refiner has two choices: improving Noack volatility (as required by current generation lubricants) at the price of at least 10 % decreased productivity when refining more heavily or export a product with less tight volatility specifications on the less quality-driven export market. An important base oil competitor puts it as follows: "Volatility restraints on 150 SN tend to reduce the volumetric production capability for these grades as opposed to overcapacity".

(392) It is possible that some refiners choose to produce light neutrals of considerably better quality than Exxon and BP/Mobil and thus have reduced light grades production. On the other hand, the parties have stressed in their form CO that Cepsa, Repsol and Petrogal guarantee/market low-volatility light base oils and that this constitutes an individual strength of these refiners. The Commission concludes that these producers would not choose to produce such low-volatility light grades if there were no underlying market demand for this and that volatility is an important element in assessing the possibility to increase light grades production.

(393) The above shows that the surplus production on the base oil market exists mainly for the heavier base oils which do not compete on the European market for automotive applications. Refineries run at full capacity or have at least production of the required quality up to the capacity of their plants. Further conversion to group III base oils might lead to inefficient base oil plants shutting down rather than creating excess capacity and additional production in former import countries outside Europe is compensated by reduction of capacity in Europe. As currently demand for group III base oil is still marginal, the current group III production capacity will not be increased as is illustrated by the grassroot plant of Neste which is only running at a fraction of its capacity. Exxon, BP/Mobil and Shell have significant capacity for the merchant market and if regional players have excess capacity than the accreditation rules and limited geographical spread restricts their ability to compete on a European level.

(394) Exports have declined due to the Asian crisis and the emergence of new capacities in former export destinations. However, at the same time as Shell and BP/Mobil have installed high-capacity plants in Asia (Singapore), three EEA base oil plants have been shut down, reducing capacity in the EEA by more than 7 %. The overcapacity created by reduced exports outside the EEA has therefore been balanced by the closure of EEA production capacity.

Market prices and margins and the effect of overcapacity

(395) On the basis of information provided by the parties, EEA base oil prices have been more or less stable over the past 10 years. The evolution of gross margins as a percentage of selling price shows that there has been no downward trend over the past 10 years. The gross margin as calculated by Exxon for SN 150 and SN 600 remained in the range [...]*. When comparing gross margins and costs for the period [...]*, it appears that there is a high correlation between their amounts and also between their absolute variations. In other words, prices have been marginally influenced by variations in costs over that period. This is not a typical characteristic of a market allegedly affected by important overcapacity. Indeed, one would expect that in an oversupplied market, a drop in costs would be immediately followed by a drop in prices. In other words, although there is overcapacity, it affects Exxon to a much lesser extent than its competitors who are positioned in the third and fourth Solomon Cash cost quartiles.

(396) [...]*(55).

(397) [...]*(56).

(398) [...]*(57).

Unique role of Exxon

(399) The Exxon Augusta plant (located in Sicily) is running essentially to supply term customers in the export markets. [...]*.

(400) [...]*(58).

Effects of the merger on competition in the base oil market

Post-merger market structure

(401) Post merger, Exxon and BP/Mobil would have a combined market share in the range between [40 to 50]* % and [40 to 50]* % depending on whether it is calculated on the basis of the nameplate capacity or the ability of the parties to sell on the merchant market (see table above). Their closest competitor would be Shell with a market share in the region of 18 % (on the basis of nameplate capacity) and 10 to 14 % on the basis of sales to third parties. Total and Agip would have market shares below 8 %. All the remaining competitors would have market shares below 5 %.

(402) The parties have indicated that a number of national integrated base oil producers have significant capacities available for the merchant market. They would therefore, on the basis of their strong competitive position, be able to counteract readily and immediately any attempt of the parties, post merger, to increase prices. They add that the merger would not lead to the creation or strengthening of a dominant position since the market is characterised by significant structural overcapacity of [...]* % of nameplate capacity. The parties also indicate that the use of the more advanced group III capacity will even increase availability of group I capacity. Equally, the parties argue that EEA exports would be redirected to the EEA to profit from the price increase.

(403) As explained in the above sections, none of these allegations have been confirmed by the market investigation conducted by the Commission.

The merging parties could profitably raise their prices or exert price pressure.

(404) The only potential constraint on Exxon/Mobil's behaviour could come from Shell. According to information received from the market, Shell's base oil strategy is orientated on captive use and its capacity for the merchant market would be insufficient to counter the combined Exxon/Mobil market power. In addition, Shell and Exxon have joint interests and control in a number of ventures, notably the additive manufacturer, Infineum. Shell has indicated that a 5 % price increase would not be sufficient incentive to consider raising base oils output. Shell's base oils policy is to supply their internal needs first. Profitability of the remaining production output is evaluated from the combined fuels-lubricants refinery activities point of view and is heavily supported by specialities production (white oils, greases and waxes) and process oils. Shell would therefore follow the price leader in base oils production.

(405) Exxon could discipline its competitors if they were to oppose its price movements. [...]*. The prospect of seeing the market price fall because of such a move from Exxon could dissuade any of its competitors from undercutting it aggressively.

(406) Base oil competitors have indicated that in case of price increases, they will (at current prices) follow the price leader. Although the parties have emphasised the issue of EEA overcapacity, this has not been confirmed by third-party base oil competitors. The most important competitors (with multiple finished lubricant accreditations, such as Shell, Total, DEA, KPI) have all indicated that they do not have idle capacity nor export volumes which could compete economically on the EEA market. This is further illustrated by Total's acquisition of Fina. [...]*. Total has indicated to the Commission that, although it has more capacity than needed for its internal EEA market sales, it does not have sufficient light base oil capacity and that it has no possibility to increase its light grades production capacity without major investments. It is therefore the Commission's understanding that Total will not be able to supply base oils for Fina's blending activities.

Limited countervailing power of customers

Limited alternatives

(407) The parties have argued that they could not exercise any market power because they would be facing a concentrated and sophisticated demand in the person of the lubricant blenders. This argument cannot however be accepted, as the bigger lubricant blenders would have little choice but to go to either Exxon-P/Mobil or Shell. Whereas before the merger, lubricant blenders could play one of the three against the others, this possibility will be suppressed after the merger has taken place, with Exxon-BP/Mobil dominating the market place. Major lubricant blenders have clearly indicated that quality and certainty of supply prevail over price and that they tend to limit the number of suppliers to their pan-regional activities to two or, at most, three. After the merger, they will be left with little choice as Shell has only limited capacity for the merchant market that is not already tied up in term contracts.

(408) Competing blenders have stated that if Exxon-BP/Mobil were to raise prices with five percentage points, they would have difficulties switching base oil suppliers for the reasons set out above. The alternative to switching both base oil and additives supply would be considered, as one could shop around the additives companies to take existing fully accredited combinations. This could not, however, be done quickly, especially given the lead-time needed to clear existing additives, to re-stock and to obtain OEM approvals. Base oil competitors have given estimates between 12 and 18 months at a cost of between EUR 300000 and 700000. In practice, the blender would accept the price increase while starting a complete new formulation from scratch if the expected sales volume could compensate the R & D budget necessary. Lubricant blenders would be squeezed between the higher prices for base oils and the lubricants customers given the fact that the lubricant market appears to be competitive.

ATIEL proposed reforms are unlikely to have an impact on the market

(409) The parties have further argued that the ATIEL rules relating to interchangeability of base oils (and thus the ease in which base oils can be switched in lubricant formulations) will be changed in 1999. The changes proposed to ATIEL would enable smaller base oils producers to form base oils families which are interchangeable. It is true that ATIEL, a trade organisation whose membership comprises a large number of lubricant companies, has ordered a revision of its base oils interchange guidelines. This revision was requested some years ago by the national refining companies and lubricant blenders. It seems to have been motivated by indicating that multiple refining companies are at a considerable advantage in terms of pan-European supply by comparison with the single refinery companies. Since the revision has not happened as yet, it is the Commission's point of view, that assessment of the proposed merger is to be based on the existing requirements of the ATIEL base oils interchange guidelines, rather than on a potential revision in the future.

(410) Indeed, the occurrence of these changes is still very unlikely. ATIEL represents the lubricants sector and has representatives from the independent lubricant blenders and the national integrated oil companies. Although the pan-regional integrated oil companies (Exxon, BP/Mobil and Shell) are in the minority, the change of rules will have to be adopted by unanimity. The fact that a request for changing the ATIEL rules was introduced three years ago clearly indicates that the process of reaching an agreement has been tedious and long and is not accomplished as yet. Even if the proposal were to be unanimously adopted, it would still have to be adopted by the ATC and ACEA. No significant problems are to be expected from the ATC (additive suppliers), but ACEA (representing the OEM's) on the other hand will require proof that a change of ATIEL rules will not negatively affect the quality of the finished lubricant (and hence product liability) which may hinder a swift implementation (or even acceptance) of the proposed changes.

(411) Even if the revision were adopted fully in line with the proposal (towards a system whereby two or more companies could carry out testing together in order to claim interchangeability), the Commission does not believe that the potential relaxation of base oil interchange could be a material factor in assessing the operation of the market. The major European motor manufacturers have agreed to adopt, via their representation in ACEA, certain minimum technical specification requirements for gasoline and diesel engines of which the matter of base oil interchangeability is covered by the ATIEL code. In Europe, the usage of API and ACEA specifications by OEMs is broadly 50/50 as Japanese and American manufacturers will tend to use API specifications, while European manufacturers rely on ACEA. Non-European manufactured cars will therefore rely on API specified lubricants and automotive lubricant blenders will therefore meet both accreditations in order to sell sufficient volume in the market. In fact, virtually all automotive lubricants on the market give an API specification as consumers are believed to be more familiar with API specification. Additive suppliers have indicated that lubricants with ACEA accreditation are generally not marketed in isolation and would carry other approvals. Consequently, automotive lubricants have to meet both accreditations in order to make sufficient volume in the market. The proposed changes would therefore only affect the EEA first-fill market where Exxon (and to a lesser extent BP/Mobil) are very strongly represented. National base oil producers/lubricant blenders expect that less rigid ATIEL rules will eventually lead the way to a common ACEA-API standard although nothing has been accomplished on this level (or is even foreseen) as yet. As API changes specifications every two years, the actual burden of accreditation and absence of base oils interchangeability will remain unchanged.

(412) In addition, ACEA imposes only minimum technical specification requirements. Individual OEMs impose more stringent requirements. Ford for instance has its own product specifications to which the suppliers have to comply not only EU-wide but worldwide. These specifications are based on API. Volkswagen, to mention another example, only uses API standards.

Exercise of market power

(413) It can be concluded from the above developments that Exxon and BP/Mobil will hold a dominant position in the EEA market for base oils. Market investigation has indicated that the concentration will not create or strengthen a dominant position in additives or finished lubricants. However, the exercise by a combined Exxon-BP/Mobil of their market power on the base oils market and the already strong position of Infineum in additives may lead to a stronger position in finished lubricants. For future-generation lubricants development, such control over the lubricants supply chain may even lead to a situation where competition could be significantly impeded.

(414) After the merger, the combined Exxon-BP/Mobil could use its base oils dominance through various strategies. For instance, it may have the incentive to use its dominance at the base oil level in order to raise the costs of independent competitors at the lubricant markets level. To do this, it could exercise pressure upon smaller players through increased prices of base oil, through standardisation of products with an increasing number of new formulations accredited against its base stock/Infineum additive packages or even through sudden price increases, combined selling of base stock additives and restricted sales of raw materials. Shell would also benefit from such a strategy, as it would strengthen them in consolidating their position at the lubricant level. Small base oil producers would not be able to counteract since their base oils are not used as references by the additive suppliers. They are obliged, even if they have an international dimension, to supply base oils from different sources (own-production in some countries, but competitors' base oils in other countries) which forces them to spend more time and money on approvals (both API/ACEA and OEMs) and weakens their competitive position.

(415) Following the merger, a combined Exxon-BP/Mobil could have an increased incentive to challenge a changing base oil market by gaining market share from national integrated oil companies through, for instance, low prices. Only Shell would seem to adapt to low prices and the other players facing little EEA demand may have to divert their sales to the export market. This would undermine the competitive position of national players and increase the pressure on the industry to close the obsolete base oil refineries, as it would no longer be economically viable for some plants to continue operation. Capacity shut down would eliminate the current small effective overcapacity. If such strategy were successful, Exxon-BP/Mobil could afterwards again raise its prices.

(416) Through a combination of its strength in the base oil market ([40 to 50]* %) and in the lubricants markets ([20 to 30]* %), Exxon-BP/Mobil might be able to extend its dominant position on the base oil market to additives and lubricants activities. The combined entity would be in a position to capture future developments of new lubricants and establish close links with OEMs and additive manufacturers which in turn would raise barriers to entry for their competitors.

(417) The market for additive packages is determined by the market for finished products containing such additives: market power is exercised through concentrating the lubricants supply chain. As additives packages are increasingly tied to specific base stock in order to provide higher performance requested by OEMs, surplus capacity in base stock is no longer a guarantee of a competitive market. Independent blenders would increasingly take Exxon-BP/Mobil and possibly Shell base stock since these are readily tested and accredited by additive suppliers. Developing blends of additives with other base stock would come at a premium price, as additive producers are no longer assured of sufficient sales volume on these base stocks.

(418) Development, testing and accreditation for base stock other than Exxon-BP/Mobil would increasingly have to be financed by the lubricant blenders. These steps in the lubricant development otherwise (partly) financed by the additive supplier would impose significant costs on these blenders. Small independent blenders would be hit first and would disappear from the market. National integrated blenders would restrict their activities to their home countries. Highly profitable blenders, such as Castrol, Fuchs or Texaco, would also be affected by increased costs and would reach the point where they could no longer compete with Exxon-Mobil and Shell on price. Their research and development and marketing budgets would decrease, further confirming Exxon-BP/Mobil's market leadership.

Conclusion

(419) On the basis of the above, it can be concluded that the operation leads to the creation of a dominant position in the EEA base oils market.

Additives

(420) The additives market is very concentrated: the five leading suppliers hold 86 % of the world market (and more than 90 % of the engine oil additives). Lubrizol and Ethyl (respectively numbers 1 and 3 in the market) restrict their activities to additives. Texaco (number 5) has its own additives company but has no base oil production and the number 4, Chevron (Oronite), only markets lubricants in Europe. Infineum takes the second place and was formed in the beginning of 1998 by combining Shell's and Exxon's additive activities. All five producers supply both DI packages and VI improvers. Although Infineum will not be significantly reinforced by Mobil's limited ([...]* %) share of the additives market, it is to be noted that, after the merger, Infineum will have direct links with all pan-regional base oil producers who have excess capacity for the accredited automotive merchant market. In view of the limited overlap, the concentration will not create or strengthen a dominant position in the market for lubricant additives.

Lubricants

(421) In the current competitive landscape, there are several major players competing on a pan-European basis. These market participants are either vertically integrated companies such as the notifying parties or independent blenders such as Castrol or Fuchs. Both Exxon and Mobil have significant lubricant activities and face competition from (national) integrated oil companies, such as Agip, TotalFina, Elf, Repsol, Statoil, Nynas, Texaco, Shell (which is to be considered as the world's largest producer of automotive and industrial lubricants) and independent lubricant companies such as Castrol, Fuchs/DEA.

(422) Exxon's share in the 1996 EEA finished automotive lubricants market was 10 %, BP/Mobil's (on the basis of aggregation) was 13 %. For the same period, Enerfinance estimates that Shell had 13 %, Castrol 7 %, Elf 6 % and TotalFina 6 %. Exxon's share in the 1996 EEA finished industrial lubricants market was 8 %, BP/Mobil's (on the basis of aggregation) was 17 %. For the same period, Enerfinance estimates that Shell had 11 %, Castrol 6 %, AGIP 6 %, KPC 4 %, Elf 3 % and TotalFina 5 %.

(423) As to marine lubricants, Shell is market leader with 21 %. Exxon (12 %) and Mobil (18 %) have a strong presence. BP's market share is around 14 %. Castrol, Elf and Texaco have each around 10 %.

(424) Although the parties will, by combining Exxon's efficiency and Mobil's brand strength, become by far the largest lubricant blender and marketer, the concentration will not lead to the creation of a dominant position except through exploitation of their dominance in the base oil market. The lubricants industry is likely to remain competitive with competitors such as Shell, Castrol, TotalFina or Elf. Moreover, the influence and buying power of customers such as distributors or OEMs is likely to increase even further in the future.

E. REFINING AND MARKETING OF FUELS (DOWNSTREAM OIL)

GENERAL OVERVIEW

(425) Downstream activities include crude oil refining and the marketing and distribution of refined products to end-users.

The supply chain

Refining

(426) Crude oil refining generally consists of the following processes: (i) fractionation of crude oil into its component hydrocarbons, (ii) conversion, which involves changing the chemical nature of the products, (iii) treatment, which consists of reducing unwanted elements or types of specific molecules to acceptable levels and (iv) blending, which is the combination of various refined products so as to meet market specifications.

(427) Refined products are often classified as light distillates (LPG, motor gasoline, naphta), middle distillates (diesel fuel, gas oil, kerosene) and heavy products (fuel oil and bitumen).

Refined products sales and marketing

(428) Refined products are sold through either the retail distribution channel or the non-retail distribution channel. Retail sales involve sales to motorists through service stations. Products sold through retail distribution are motor gasoline and diesel. In some countries retail channels also distribute automotive LPG.

(429) Non-retail sales consist of wholesale sales to three categories of customers: independent retailers (unbranded service station operators such as hypermarkets), other independent resellers, industrial and commercial consumers (hospitals, car rental fleets, factories). The supply is done by refiners or other marketers having at their disposal massive volumes of refined products due to the fact that they own nearby either a production site (refinery) or a large storage facility (terminal).

(430) Ex refinery sales consist of sales made in large lots on a spot basis by refiners to other oil companies, traders, resellers, large industrial customers. These transactions typically involve una tantum (single deliveries) purchases of cargo lots (1 to 3 Kt barges or up to 20 Kt tanker lots), and are priced on the basis of market quotations. The delivery location is ex refinery. Ex refinery sales supply what the industry refers to as the "cargo market". Since these transactions typically involve big quantities of refined products which can travel long-distances given the relatively low transport costs, cargo market prices around the world track each other closely. These prices are published on a daily basis by organisations such as Platt's and Petroleum Argus, making it possible for buyers to negotiate using up-to-date pricing data.

Transportation of refined products

(431) Refined products that are sold ex refinery into cargo markets are collected by tanker or barge. Transportation and physical distribution of refined products consist of three different steps: (i) transportation of refined products from oil refineries to distribution terminals by means of waterway, pipeline, and rail (primary distribution), (ii) storage of the bulk fuels products at distribution terminals, and subsequent loading onto delivery vehicles and (iii) transportation and delivery to customers primarily by truck, but also by rail and barge (secondary distribution). There are a number of independent participants at each of these levels and users of these services have a wide variety of competitive options available.

Logistics

(432) Distribution terminals are an essential component for distribution of fuel products to end customers. Oil companies generally hold storage depots disseminated on the territory and pipelines to secure timely deliveries to customers, including service stations. However, storage facilities are also owned by independent terminal operators who lease storage capacity to all interested parties under short and long-term arrangements. There are generally two types of storage depots, large terminals with capacity above 50000 m3, which are supplied by massive transportation means (pipelines, ships, barges, rail tanks), smaller peripheral depots which ensure the distribution of smaller quantities of fuels to the final destination by secondary means of transport (trucks and barges).

The competitive environment

(433) The oil downstream sector is characterised by the presence of many vertically integrated oil companies active in the refining, wholesaling and retailing of refined products. However, each segment of the downstream fuels sector also includes several independent non-integrated companies.

(434) With regard, more specifically, to motor fuel retailing, the competitive environment is characterised by the presence of essentially three categories of operators:

(i) the traditional vertically integrated oil companies, which are active at all levels of the chain, notably refining, wholesaling and retailing. Within this category, it is possible to distinguish further between the majors, that is to say large vertically integrated multinational corporations which all operate in most EEA countries (Exxon, Shell, BP/Mobil), the mini-majors, namely smaller multinational corporations with their activities focused in some EEA countries (Total, Texaco) and the national champions, i.e. vertically integrated national-based companies very well established in their countries of origin (Elf, Agip);

(ii) the independent retailers, meaning those operators that are active only at the retail level and procure motor fuels from refiners and other traders. White pump stores are an example of independent retailers. They are generally lower volume (in comparison with the industry average), dealer-owned and operated facilities formerly owned by major oil companies. Single or multi-outlet operators acquire these facilities and, with minor investment, keep them operating as white brands;

(iii) the hypermarkets and supermarkets, which are non-traditional retail motor fuel suppliers. While sharing similarities with traditional independent retailers, they differ from such competitors in a number of respects. Typically, hypermarkets are large mass-merchandising retailers of groceries and other consumer goods, with a motor-fuels retailing facility located in their parking lots. Unlike most independent retailers, hypermarkets do not generally regard the sales of motor-fuels as a part of their core business, but as an additional service they offer to their customers or as an additional way to draw them to their stores. Because of the number of their customers, they usually achieve considerable throughputs and they generally compete on price. Finally, unlike a number of usual independent retailers, they generally operate in densely populated rather than rural areas.

(435) Retail outlets may be owned or leased by the supplying company (COSS) or by independent dealers (DOSS). The COSS channel can be further divided into two kinds of station operators:

(i) company-operated/commission-agent (COCO) stations operated by company employees compensated through salary and wages, or by agents compensated by commission. In both cases the supplying company retains ownership of the fuels stored on the station premises and establishes pump prices;

(ii) independent licensees/tenants, meaning sites operated by independent licensees who purchase branded motor fuels (CODO).

The DOSS channel can also be divided in two categories:

(i) dealer-operated, that is to say independently owned service stations where the supplying company sells fuels to the dealer;

(ii) agency-operated, that is, independently owned service stations where the owner sells the supplying company's fuel.

At the CODO and DODO sites the dealer can, in principle, establish the pump prices as he owns the fuel sold at the site. However, as a matter of fact the supplying company generally retains strong control over the pricing policy of these stations essentially through some mechanisms of financial compensation designed to cover the dealer's losses resulting from a pump price reduction that the company wishes to implement. Also, the contractual relations between the independent dealer and the supplying company are such that the dealer generally accepts to implement the pump price recommended by the company.

Product market

Retail sales of motor fuels

(436) Retail motor fuels sales include sales made to motorists from branded and unbranded service stations. The products sold are predominantly gasoline and diesel. From the demand side there is no substitutability between these products, as motorists must use the type of fuel suitable for their vehicle. However, from the supply side there is substitutability as refineries can be run to produce different outputs of different types of fuels. There is considerable flexibility based on the input quality of crude oil and the way in which the refinery is configured. At the distribution level, both products are always available at the same point of sales. Also, market shares for each type of fuel will roughly coincide with the aggregate market share. As a consequence, for the purposes of this case, in the retailing channel the relevant product market is constituted by the retail sales of motor fuels with no need for a further distinction.

Motor fuel retailing in toll motorways

(437) In some countries, it is possible to consider fuel retailing on motorways as a separate product market. Such a distinction is based on the different competitive conditions characterising a motorway. First, the demand is captive, in that motorists travelling on motorways are generally not familiar with the region in which they travel. As a consequence, the possibility to leave the motorway to buy fuel in a service station located off the motorway is very limited. Second, motorists choose to travel on motorways for reasons of speed, and thus they are not inclined to waste time in refuelling. The payment of a toll is another factor which further discourages motorists to exit motorways and buy fuel on off-motorway stations. In general terms, motorists pay a toll in exchange for speed and will be even less inclined to leave the motorway for refuelling purposes. A strong indication that motorways can constitute distinct markets lies in the fact that pump prices on motorways are generally higher than prices on ordinary routes.

Non-retail sales

(438) The non-retail sale of fuels cannot be similarly aggregated. The different fuels are supplied for different uses to different types of customer. The distribution channels can also differ significantly, the provision of fuel oil to a power generation plant differing from the supply of LPG for domestic heating purposes. The investment required for these distribution channels will be different from that of retail fuels. For the two fuels which are sold in significant quantities in both retail and non-retail markets, namely diesel and LPG, the use of the fuel depends on the distribution channel. Retail diesel is purchased by users of diesel powered cars and other vehicles. Non-retail diesel is typically supplied to large scale transport companies with a substantial fleet of vehicles and their own storage facilities. Similarly, LPG will be supplied to vehicles through retail sites and to households for heating purposes though non-retail channels.

(439) In the light of the above, in the non-retail channel each refined product constitutes a relevant product market.

Geographic market

Motor fuel retailing

(440) From a geographic standpoint, the market for fuel retailing is characterised by a local element in so far as the demand is constituted by motorists who are normally supplied by the service stations near to their centre of activity. As a consequence, the substitutability between service stations is geographically limited.

(441) On the other hand, there is normally a certain overlap between service stations' catchment areas, which will not only determine the competitive interactions between geographically neighbouring service stations, but will also, to some extent, have a "knock-on" or "chain-reaction" effect on more distant service stations. As a consequence, the geographic area to be retained for the competitive assessment can include several small overlapping areas. In addition, from the supply side, many important parameters of competition, such as range of products, sources of such products, quality, service level (opening hours, etc.), advertising, promotion and prices (on campaign products for example) are not decided on the local level, but on a regional or national level. Moreover, in fuel retailing taxation so heavily affects the final pump prices (in the United Kingdom taxes account for more than 80 % of the pump price) that, as matter of fact, variations in prices across most national markets tend to be relatively small. Also the various suppliers' market positions tend to be broadly similar in the different parts of national markets in which they operate. In the light of the above, for the purpose of this case the market for fuel retailing will be regarded as being national.

Motor fuel retailing on toll motorways

(442) The considerations in recitals 440 and 441 apply also to motor fuel retailing on toll motorways.

Non-retail sales of refined products (fuels)

(443) There are a number of indications that the geographic market for non-retail sales is local in scope. In this market the supply is constituted by those refiners who can sell massive quantities of refined products from their refineries or their terminals. The demand is constituted by resellers and end-users who seek a nearby source of supply, given the extra costs they have to bear when transporting these products from the point of supply to the place of final destination (a service station, a household, a factory). Accordingly, the degree of substitutability between sources of supply is geographically limited. In general terms, each point of supply can source a certain geographic area (catchment area or hinterland), whose radius is function of the transport costs that customers have to bear when transporting the products to the final destination. These costs can vary according to the means of transport. However, since customers generally use road transport, this radius is usually in the range of 100 to 150 km.

(444) On the other hand, there may be a certain overlap between the catchment areas of the various sources of supply, which will not only determine the competitive interactions between geographically proximate businesses, but will also, to some extent, have a "knock-on" or "chain-reaction" effects on more distant points of supply. As a consequence, the geographic area to be retained for the competitive assessment can include several small overlapping areas.

(445) In any event, this question can be left open as the operation does not raise any special competitive concerns irrespective of the geographic market definition being chosen in this case.

Competitive assessment

The assessment of the BP/Mobil joint venture

(446) In order to assess the competitive impact of the transaction, the position of the merged entity with respect to the structural links existing between BP and Mobil should first be appreciated. In the downstream market of oil-related products Mobil has created with BP a joint venture, combining all the European activities in this sector. In particular, the joint venture includes the refining, the supply, sale and marketing of refined fuels and lubricants. Within this joint venture BP operates the overall fuels business, while Mobil operates the lubricants business. This joint venture was notified and cleared under the Merger Regulation.

(447) In the present notification, the parties contend that while Mobil was found to have joint control of BP/Mobil within the meaning of the Merger Regulation, Mobil has no involvement in the day-to-day governance of the BP/Mobil fuel business. As a consequence, the parties claim that the combined entity Exxon/Mobil and BP/Mobil will both have every incentive to continue to compete vigorously with each other in this sector. Accordingly, it would be inappropriate to aggregate BP/Mobil's shares with those of Exxon for the purpose of assessing the competitive implications of the transaction.

(448) In essence, the parties' arguments are tantamount to saying that notwithstanding the fact that Mobil has joint control over BP/Mobil joint venture, the combined entity Exxon/Mobil and BP/Mobil joint venture should be treated as two independent undertakings competing against each other, at least with respect to the downstream fuel business. In other words, according to the parties the existence of a control relation does not influence the incentives to compete of the undertakings concerned by this link.

(449) This argument is to be rejected on the basis of a number of considerations. Pursuant to the Merger Regulation an acquisition of joint control or a creation of a full-function joint venture constitute concentrations and as such they are subject to the dominance test laid down in Article 2 of the Merger Regulation. In order to ascertain the existence of a dominant position, the relations between the JV and the parents are considered on the basis of the generally correct assumption that they achieve some form of integration and that the parent company is in a position to control the commercial policy of its JV, so that from a competition point of view they are to be viewed as being not in competition with each other.

(450) With regard to this case, it should be recalled that BP/Mobil JV was notified to the Commission in accordance with the Merger Regulation and was treated as a concentrative joint venture. Both BP and Mobil were regarded as holding joint control of the overall joint venture, irrespective of the fact that each partner had a special operational responsibility over a specific business (BP over fuels and Mobil over lubricants). This conclusion was notably based on the fact that the two parents jointly exerted decisive influence on the joint venture's policy through a supervisory committee whose tasks included decisions in the following areas: business plans, major acquisitions, closures, disposals, investment and certain other strategic decisions. Mobil was one of the notifying parties of that transaction and at that time never contested those findings.

(451) The argument put forward by the parties that Mobil is not in a position to interfere in the day-to-day management of the fuel business run by BP is not convincing: on the evidence collected in the investigation it appears that under the current state of affairs Mobil (and after the merger the combined entity Exxon/Mobil) indeed exerts a significant and profound influence on all of the relevant competitive decisions involving BP/Mobil fuel business, in particular approving, within the supervisory committee, major decisions such as the approval of the strategic business plan, the setting of financial control parameters, the approval of major acquisitions, closures, disposals and investments, and other strategic decisions. Moreover, the contractual obligation for BP to obtain Mobil's approval for such decisions involves the disclosure to the latter of a far larger amount of highly competitively sensitive information such as, inter alia, refinery disposals, refinery joint ventures, fuels performance (including forecasts, refining margins, retail margins, improvement plans, retail volumes), fuels business strategy (including competitive pressures, the competitive position, growth, productivity levels), and the fuels operating plan (including the operating environment, business development, earnings, expenditures, cash forecasts, retail divestments and longer term strategic goals). In total, the powers conferred to Mobil within the JV with BP are such to allow Mobil, and thus the merged entity, to exert a decisive influence over the JV's competitive conduct, including the fuel business operated by BP.

(452) However, the parties contend that, in the light of the exceptional circumstances existing in this case, BP/Mobil and Exxon/Mobil should be regarded as two independent undertakings capable of competing against each other. This contention is notably based on the fact that (i) BP has the operational responsibility of the fuel business and (ii) that it is the "leading" parent of the joint venture which thus can oppose and resist Exxon/Mobil interference in the competitive decisions inside the JV. Even if this assumption is accepted, account should be taken of the relevant interest stake of one undertaking in the other. In this respect, it should be noted that a link of this kind, regardless of its formal qualification, greatly reduces the incentives to compete between the undertakings concerned. It is indeed a well established principle under mainstream antitrust economics that, generally, the existence of links between two competing undertakings in the form of a significant interest stake of one in the other may change their incentives to compete. First, a link of this nature creates a strong financial interest of one firm in its competitor's welfare. This automatically can alter the dynamics of the competitive game as one firm is less interested in competing against the other than in finding a common commercial strategy profitable for both. In addition, such a link can secure access to commercially sensitive information. This in turn renders the competitive conduct of each undertaking vis-à-vis the other more transparent and thus susceptible to be easily anticipated and monitored. Also, and perhaps more importantly, a link of this nature may put one undertaking in a position that enables it to influence the strategic choices of its competitor towards decisions in line with the common interest. Finally, a link of this kind has a disciplinary effect as it can expose one firm to possible retaliations of the other in case of disagreement. All these factors may push the undertakings concerned towards a convergence of their commercial policies. It should be noted that the conduct described above is for each of the undertakings concerned absolutely rational as they are based on a profit-maximising perspective.

(453) The present case should therefore be assessed against this background. Given the relevance of the link in terms of influence exerted by Mobil (and post merger by Exxon/Mobil) over the JV's competitive conduct, as well as the presence, in the case at stake, of a number of structural features typical of a market conducive to oligopoly, the relation between Exxon/Mobil and BP/Mobil can be, ultimately, assessed in the framework of the collective dominance test. In short, the competitive assessment of the case would remain unchanged.

(454) In their response to the Commission's decision to open a second phase investigation, as well as in replying to the statement of objections, the parties raise two additional objections against the reasoning described above.

(455) First they claim that Exxon/Mobil cannot simply alter its competitive behaviour solely to the benefit of BP/Mobil because other competitors would also take advantage of any weakening of Exxon/Mobil's competitive vigour. In particular, they contend that Exxon/Mobil direct interests through the Exxon network are greater than its indirect interests through its 30 % stake in BP/Mobil. As a consequence, Exxon/Mobil would have an interest to maximise its profit rather than favouring BP/Mobil joint venture profit. A fortiori, the argument goes, BP would have no incentive to modify its competitive strategy in favour of Exxon/Mobil, as it has no interest in the latter. In this respect, it should be noted that the alignment of Exxon/Mobil and BP/Mobil's competitive strategies which could result from the merger would be aimed at maximising the profit of both entities on the market. This convergence of competitive strategies is rational behaviour taken by each firm in its own interest. As to the possibilities of the other competitors to take advantage of the "weakening" of the competitive vigour between Exxon/Mobil and BP/Mobil, the Commission considers that given the oligopolistic features of the markets in question (see below), the alignment of the competitive strategies of the two entities can cause a reduction of competition detrimental to consumers, without triggering any competitive reaction from the other players.

(456) Second, the parties also contend that the Commission approach is inconsistent with recent Community case-law. In particular, the parties refer to the Gencor/Lonhro judgment of 25 March 1999(59), where the Court of First Instance found that a company in which a merging party held a 27 % interest nevertheless was a significant competitor of its minority owner. According to the parties, the pre-merger situation in Gencor is structurally very similar to the post-merger scenario involving Exxon, BP and BP/Mobil in the fuels area. As a result of the merger the merged entity Exxon/Mobil will have its own fuel business while retaining a minority interest in a competing business, BP/Mobil.

(457) In response to the parties' line of argument, it should be noted that there are two specific points which clearly differentiate the present case from the factual background of the Gencor judgment. First, on the basis of a full assessment of the two factual backgrounds, it appears that the pre-merger situation in Gencor can in no way be compared to the post-merger situation in Exxon/Mobil. As it is clearly stated in the judgment, not only did Gencor have no influence over the day-to-day management of Lonrho, but neither on the marketing and the sale of Lonrho's platinum production. In short, the Gencor group exercised no influence over the LPD's competitive strategies (see paragraph 176 of the judgment). By contrast, as already stated above, within the framework of the BP/Mobil joint venture arrangements, Mobil retains a significant and profound influence over all of the relevant competitive decisions involving BP/Mobil fuel business. It suffices to recall in this respect that any major decision relating to the pricing policy of the joint venture would be subject to Mobil approval.

(458) Second, and more importantly, the comparison made by the parties disregards the difference between a pre-merger situation and a post-merger one. The question in Gencor was whether the competitive situation would change following the merger between Gencor and Lonrho. In that respect, the Commission (and the CFI) could found its findings on the basis of an ex post review of the competitive conduct of the two firms, and concluded that despite the minority stake of Gencor over Lonrho's platinum business, the two companies had been competing against each other in the pre-merger phase. In this respect, both the Commission and the Court clearly emphasise the fact that "LDP and Implats, by retaining separate marketing departments, competed with each other before the concentration and (...) that during the past decade LDP had, with Russia, been the main element of competition in the market" (see recitals 174 to 176 of the decision and recital 177 of the judgment). By contrast, the post-merger relation between Exxon/Mobil and the joint venture BP/Mobil is a matter of the future, to be assessed in the context of an ex ante analysis. As a consequence, in accordance with the approach which must be followed in the context of the merger control review, the Commission is to assume the worst possible scenario for antitrust purposes, that is to say the alignment of the competitive strategies of Exxon/Mobil and BP/Mobil as a result of the joint control of the former over the latter.

Mobil's interest in Aral

(459) In assessing the competitive impact of the transaction in motor fuel retailing, account should also be taken of the significant links existing between Mobil and Aral. Aral is the leading retail motor gasoline operator in Germany, and it is also active in Austria and Luxembourg. Mobil has a 28 % stake in Aral, the other shareholders being Veba (56 %), Wintershall (15 %) and a number of benzene producers with 1 % between them. Each of the three main shareholders holds [...]* % of voting rights. According to the parties, Mobil has no joint control over Aral in the meaning of the Merger Regulation [...]*.

(460) However, it can be argued that Mobil, together with Veba and Wintershall have de facto joint control over Aral based on the following considerations: [...]*.

(461) In any event, this question can be left open since, irrespective of the question of control, it is indisputable that Mobil has a very significant interest in Aral, which creates a strong financial interest of the former in the welfare of the latter. Moreover, through this stake, Mobil can exercise a relevant influence over Aral commercial policy. It has the same voting rights as the other main shareholders, and it appoints the same number of members of the supervisory committee, which in turn elects the management board. Finally, Mobil has access to a large amount of strategic information which is exchanged within Aral's decision-making bodies. It is thus clear that the existence of this link is likely to reduce Mobil's incentives to compete against Aral and vice versa. With such an important interest of the former in the capital of the latter, competition between the two operators seems to be seriously weakened.

Motor fuel retailing

(462) Taking the markets at national level, the table below shows the market shares of the parties and their competitors in those countries where the joint market share resulting from the merger is above 15 %.

(463) Motor fuel retailing market positions

>TABLE>

Parties' estimates.

(464) In percentage terms the position on individual national markets varies somewhat from country to country. In general terms, it can be noted that the merged entity would become market leader in most of these countries, with combined market shares of Exxon, BP/Mobil and Aral varying from a minimum of [20 to 30]* % (French motorways) to a peak of nearly [30 to 40]* % (in both Germany and the United Kingdom). However, the presence in all of these countries of some strong competitors (large international majors or national champions) with market shares in the range of some 15 to 20 % or more, would exclude the possibility of the combined entity alone being able to exercise single dominance.

(465) Rather, the question of possible creation or strengthening of joint dominance needs to be assessed, given in particular the structural features of the markets concerned. In this respect, markets of motor fuel retailing generally display all the characteristics normally considered necessary to allow collective dominance to arise. These characteristics are the following:

Supply concentration

(466) The concentration of the market from the supply side is the first precondition for an oligopoly to exist. An oligopoly is a market characterised by the presence of few large suppliers whose competitive actions are interdependent. Markets of fuel retailing, taken at national level, are generally concentrated, with the presence of few suppliers (between three and five) accounting for approximately two thirds of each national market. The parties contend that the markets considered by the Commission are not sufficiently concentrated. In particular, in its past practice the Commission has generally viewed a much higher degree of concentration as being likely to give rise to oligopolistic concerns. This issue will be further pursued in the sections devoted to each national market. In general terms, it should be noted that it is undisputed that the higher the number of firms participating to an oligopoly, the higher is the risk of defection in the oligopoly. Nonetheless, standard antitrust economics suggest that an oligopoly can function also in less concentrated markets. This is the case when the transparency of the market in terms of price is such that a reciprocal monitoring is easily possible even between numerous players on the market.

Product homogeneity

(467) A high degree of homogeneity simplifies tacit coordination of pricing policies. In the absence of other characteristics that are likely to differentiate one supplier from another, competition in a homogeneous product market will therefore tend to focus on prices. In the present case motor fuel is undoubtedly a commodity, with very little possibility of differentiation or customisation of the product. It is probably one of the few commodities where customers (motorists) are unable to point to any difference between two different brands of motor fuel, in terms of characteristics and performance. Motor fuel is such a fungible item of trade that refiners systematically resort to swap practices to source their service stations.

(468) In their response to the Commission decision to open the second phase, the parties contend that the retail motor fuel business is characterised by intense competition in terms of total offering to the consumer. In particular, besides price, the consumer would consider other factors such as convenience, the range of additional offerings in terms of products in the shop, the availability of other facilities (car wash), the offerings of features with other perceived value such as promotions and loyalty schemes. As a consequence, since there are wide differences between competitors' total offerings, the argument goes, there would not be much homogeneity.

(469) These considerations do not appear to be supported by the evidence collected in the investigation. As to brand recognition and product differentiation, the investigation has confirmed that these factors do not play significant roles in determining customers' preferences. The large majority of the retail operators interviewed by the Commission has indicated that brand recognition plays a minor role in motor fuel retailing. Admittedly, motor fuel retailers have tried to establish a brand recognition, in particular through marketing campaigns and loyalty schemes. However, as a general comment, it can be noted that the policy of the oil majors aimed at creating a brand image in motor fuel retailing has not proved very successful. As an exception, only in the Netherlands, and to a lesser extent in Germany, have motor fuel retailers managed to create a certain brand recognition through loyalty schemes. The matter will be further pursued in the sections devoted to the Netherlands and Germany.

(470) As to the "additional offerings" indicated by the parties, no evidence has been provided which shows that these factors have in any way played a role in conditioning motorists' preferences. Nor has the industry in its responses ever supported such a view.

(471) In fact, from the evidence collected by the Commission it appears that motorists' choice when buying fuel is driven by two essential parameters, convenient location and price. Moreover, a survey provided by the parties themselves clearly shows that most consumers in the developed countries see no meaningful difference in fuel quality among motor fuel brands. Instead, an overwhelming majority of them choose a motor fuel outlet based on locational convenience and on price.

(472) According to this survey, convenient location is the prevailing reason for buying fuel. However, contrary to what the parties claim, convenience is exclusively relevant for the purpose of defining the geographic scope of the market. In this respect, as was acknowledged in the section devoted to the market definition, the motorist's need to buy motor fuel without making a long detour determines the competitive dynamics in motor fuel retailing local. By contrast, this factor can in no way be regarded as a factor of differentiation of the nature of the products at stake.

Low technological innovation

(473) The lack of technological innovation is another feature typical of a market conducive to oligopoly and is directly linked to the homogeneity issue. Without innovation, competitors have no other choice than competing on price. Also, low technological innovation ensures that the nature of competition (and the homogeneity of products) will not substantially change in the near future. Motor fuel retailing is characterised by low technological innovation. Technological development is achieved through gradual and relatively slow incremental changes to processes and products rather than revolutionary fast moving changes.

Transparency of the market

(474) Market transparency is also one of the basic conditions characterising markets conducive to oligopoly. First, such a transparency allows the players on the market to converge towards a given price with no need of explicit coordination. Accordingly, price transparency can be expected to lead to price parallelism between the undertakings present on the market. Second, these undertakings have the possibility to mutually monitor each other's behaviour in order, inter alia, to ascertain that none of them is cheating. The market in question is characterised by full transparency, in so far as it is a retail market where pump prices are publicly known. In fact, price information is readily accessible not only at the marketing level, but at all levels of trade. For example, trading prices of fuel ex-refinery sales are published on a daily basis by respected organisations such as Platt's and Petroleum Argus. Price transparency also exists at the level of crude oil. In sum, price information is available throughout the industry, up to the marketing level. It is also worth noting that the transparency of the market is an essential condition for an oligopoly to function effectively. In motor fuel retailing the transparency is such that it makes any form of cheating by the players on the market impossible. Finally, motor fuel retailing is also a transparent market in terms of quantities sold. In this respect, it suffices to say that most operators in the industry appear to be well informed of their competitors' sales due to intense exchange of information at the level of industry associations. Also, in this sector there are a number of sources of published statistical information.

Moderate growth of the market

(475) In markets where demand is stagnant or only grows moderately there will be no impetus from market growth to foster competition in the market for two reasons. Firstly, there is no incentive among incumbent firms to compete for an increase in demand. Any gain in market share can only be achieved at the expense of competitors, possibly ending in a damaging price war with no winners. Secondly, a modest growth rate of demand does not attract outsiders to enter the market. In the case at stake, as acknowledged by the parties, motor fuel demand in Europe is relatively stagnant. More specifically, motor gasoline has been flat while diesel has captured most of the growth. Motor fuel demand in the EEA has had an average yearly growth of 0,4 % over the past five years.

Symmetry of costs

(476) Oligopolistic dominance is more likely to be sustainable if the members of the oligopoly have similar cost structures. First, similar cost structures will normally create similar incentives in a given market situation. Second, similar cost structures increase the risk of retaliation, in the sense that a supplier will know that its competitors can meet its competitive actions on an equal basis. In the case at stake, an analysis of the cost structures of fuel retailers will be conducted on a country-by-country basis taking as a reference motor fuel retailing as a stand-alone business.

Symmetry of market shares

(477) Symmetry of market shares between competitors is viewed as another factor which generally facilitates anticompetitive parallel behaviour. This is mainly because such similarity can reflect similarity of costs between undertakings. Since costs strongly influence the price policy of an undertaking, the above described symmetries render price coordination easier. In some of the national markets reviewed by the Commission below, market shares are unevenly distributed among the various players. However this factor should not be regarded as precluding in itself the possibility of oligopolistic dominance. This is because, although symmetry of market shares provides additional incentives for parallel behaviour, it cannot be considered to be a prerequisite for such parallel behaviour to exist. For instance, even in the presence of market share asymmetries, there can be symmetries of costs. This is the case in motor fuel retailing where the most important economies of scale are reached at the level of the station site. And similarly, the oligopoly members may feel that they have a fair share of the market and that they perceive each other as being on an "equal footing" and having comparable strength. They would have therefore little to win and much to lose from aggressive competitive actions. This is precisely the case of motor fuel retailing, where virtually all the alleged oligopoly members are vertically integrated operators, active across the world and with financial capacities going well beyond the position held at the level of retail in one specific national market. In sum, for these players the symmetry of national market shares probably does not play an important role. In this respect, it is worth noting that all the vertically integrated retailers have created a compensation system designed to support the financial losses that their retailing networks may suffer when exposed to intense price competition. In the light of the above, it is submitted that the asymmetry of market shares prevailing in a number of countries considered by the Commission for the purpose of the oligopoly assessment is of no relevance in the present case.

Vertical integration

(478) Similar degrees of vertical integration among the members of an oligopoly generally increase their willingness to engage in parallel behaviour and the prospects of stability of such behaviour. This results first from the fact that integration may confer cost advantages and cost similarities on the integrated players as opposed to the non-integrated ones. Second, and more importantly, strategic interests among equally integrated firms tend to coincide at every commercial level. Indeed, changes in the market conditions will affect similarly the integrated firms, pushing them to react similarly. As a consequence of that, higher stability in terms of parallel behaviour is expected in such markets. In the case in question, the oil industry is a business characterised by the presence of large vertically integrated firms. As to the consequences of this integration in the business of fuel retailing, it involves some advantages for the integrated firms such as, inter alia, security of supply and, in the long term, a better ability to face changing market conditions. Also, it is a fact that vertical integration results in a commonality of interests between the integrated oil firms covering all the sectors of the industry. Thus, this integration increases dramatically the prospect of stability of parallel behaviour by making the oil majors' interests coincide.

Demand rigidity

(479) In general terms, if demand is price-inelastic, there is a higher incentive to parallel behaviour. As total demand volumes will not decrease significantly in the case of price increases, there is therefore a strong incentive for the market players to engage in supracompetitive pricing. In motor fuel retailing the demand is rigid given that there are no real substitutes to this product. Several surveys show that in the event of a price increase, motor fuel average consumption in the long term remains stable. This is also due to the fact that there is an element of captivity in the demand.

Structural links

(480) The likelihood of oligopolistic behaviour generally increases if there are structural links between the members of the oligopoly. This is mainly because structural links make market players care about the welfare of their partners, and make them aware of each other's strategy, cost structure and plans. Such links may reduce the competitive zeal between the members of the oligopoly, or they may represent potential means of retaliation and, depending on the circumstances, they could also result in a certain common commercial interest in the market in question. The downstream oil business is characterised by the existence of several links between the major players, either in the form of corporate links (such as the BP/Mobil joint venture and Mobil's participation in Aral-Germany) or in the form of unincorporated or other commercial relationships (such as swap agreements in fuels).

Entry barriers - Absence of potential competition

(481) An oligopoly can function more effectively in the presence of high barriers to entry. These barriers may render a market difficult to contest and make it immune from competitive pressure coming from potential competitors. The motor fuel retailing business, and in more general terms the whole oil industry, are characterised by significant barriers to entry. Vertical integration has proven to be a not insurmountable barrier over the last years as is clear from the successful presence on the markets of some non-integrated retailers. However, other significant barriers render any newcomer's entry on the retail motor fuel market unlikely.

(482) First, in order to compete effectively in the market it is essential to obtain a critical mass of stations with a high average throughput. This in turn requires high levels of investment, due in part to the cost of complying with environmental regulations (the parties estimate the cost of opening a new station at EUR [...]*) as well as availability of adequate retail sites. On this latter point, in many countries licences for building new sites are difficult to obtain owing to the saturation of the market. This is certainly true for all the countries affected by the present transaction. While it can be argued that new entry could occur in the future by the reopening of the sites which have been closed in recent years, these tend to be smaller and unattractive. Also, environmental regulations make reopening an old site almost as difficult and expensive as building a new one.

(483) As a matter of fact, the only newcomers successful in establishing a significant presence in some Member States over the last years have been hypermarkets, by virtue of a mix of special conditions: (i) availability of competitive supply; (ii) ample distribution and terminal facilities; (iii) central purchasing; (iv) access to convenient commercial property; (v) relative ease of permitting; (vi) low cost structure relative to traditional competitors, and to a lesser extent some "low cost" retailers capable of establishing an adequate network in terms of number of sites and average throughput (Save in the United Kingdom, or Jet Conoco in Germany).

(484) On the other hand, many Member States have recorded an impressive consolidation due to a general decrease in margins, as a result of which a large number of small independent retailers have been forced out of the market. This is a characteristic common to all of the markets considered below. In this context, contrary to what the parties claim, the possibility that in the future "small" firms can enter the retail motor fuel market on a small scale seems remote.

(485) In sum, the only real source of potential competition in motor fuel retailing is constituted by hypermarkets which in some countries have been extremely successful in increasing their sales (in France, in less than 20 years hypermarkets have acquired more than 50 % of the market). However, the fact these operators have little or no presence in countries such as the Netherlands, Luxembourg, Austria, may indicate the existence of some structural impediments likely to prevent, even in the future, any development on that front. In other countries where these operators are already established (the United Kingdom and Germany), their real competitive potential should be measured by taking into due account all the numerous factors liable to hinder their further growth. This point will be discussed in the sections devoted to the country-by-country assessment.

Absence of buying power

(486) The ability of members of the oligopoly to rise prices can be constrained by countervailing market power of customers. Powerful and concentrated customers may either prevent the members of the oligopoly from engaging in supracompetitive pricing or provide a sufficient incentive to deviate from collusion and compete. In motor fuel retailing, purchasing power is completely absent as the demand is constituted by motorists.

Multi-market contacts

(487) Multi-market contacts can help to sustain oligopolistic dominance, because they may increase the scope for retaliation which could also take place in other markets. The oil industry is characterised by vertically integrated companies which are active at all levels of trade, from the extraction of crude oil to the marketing of fuel.

(488) Against this background, which is common to all of the abovementioned national markets, it should then be assessed whether the specific market conditions of each national market are such that an oligopoly can function effectively as a result of the merger in question. Below, the situation prevailing in each country will be discussed at length, with particular emphasis on the degree of concentration of each market, the competitive pressure exerted by operators other than the alleged members of the oligopoly, the causal link existing between the merger in question and the finding of a collective dominant position.

(489) The table below shows a comparison of motor fuel prices in the countries affected by the transaction.

Product: Retail motor gasoline - pump price

>TABLE>

Product: Retail diesel - pump price

>TABLE>

AUSTRIA

GENERAL OVERVIEW

Refining

(490) In Austria, there is only one refinery, owned and operated by OMV. Traditionally more than 70 % of petroleum products are sourced from inland refineries directly. Imports into Austria do not seem to be made by any independent reseller, but by integrated companies. Sufficient capacity is available in Austria. The crude oil is purchased by OMV and RAG and transported through an oil pipeline, from the Adriatic Sea to Vienna, where it is refined in OMV's Schwechat refinery. It follows that OMV is playing an important role in Austria: it owns the only refinery in Austria, it has the majority shares of the Adria-Vienna pipeline and of the storage Erdöl-Lagergesellschaft, it has shares of the "Autobahn-Betriebegesellschaft" and owns a proprietary pipeline between Vienna and St Valentin.

(491) OMV has contractual production sharing for its refinery with Shell, Mobil, Esso and AGIP. Combined, these companies utilise 15 % of this refinery. OMV also concludes swaps with Conoco, Agip and Aral. Shell, Mobil, Esso and Agip are also shareholders of AWP, Erdöl-Lagergesellschaft (storage) and "Autobahn-Betriebsgesellschaft". The "Paritätische Ausschuß für Kartellangelegenheiten" considered this cooperation of the majors could be close to a cartel. Majors supply each other's stations with fuel in order to save transport costs.

Infrastructure

(492) Exxon reports that several shipping companies recently offered rates for barge supply from Rotterdam to the Vienna area of [...]* DM/T for distillate and [...]* for mogas, resulting in an average volume-weighted rate for motor fuels of around [...]* DM/T. The freight rate from Rotterdam to a typical destination in Germany (the Rhine/Main area) is about [...]* DM/T. Accordingly, the average supply cost for Austria is about [...]* above Germany. This higher rate is a result of the longer shipping distance; canal tolls; the exceptionally high number of locks one must pass through in the Rhine/Danube canal, which slow the voyage; the lower water levels compared to the Rhine and a high risk of delays due to low water on the Danube (resulting from the fact that water levels on the Danube are not controlled by damming).

(493) Besides storage controlled by the majors, independent storage possibilities exist such as Tamoil in Linz, Roth Öl in Styria and smaller terminals that cover regional demand. Although no figures on capacity were provided by the parties on these storage alternatives, it has been indicated that capacities are not significant. In any case, competitors have indicated that there is no risk of foreclosure on the storage capacity as the majors are opting for renting storage capacity rather than controlling it.

Wholesaling

(494) In 1997, 27 % of the consumption on fuel and 35 % of diesel were imported. Majors, traders and "outsiders" import oil products by barge on the Danube, by train or by road. Outsiders or "B-brands" (which means smaller retail networks than those operated by OMV, Aral or BP/Mobil such as Avia, Jet, Turmöl, Rühl, IQ, AWI, Roth, Hölzl, Champion, A1, and Eurol) import fuel or buy fuel from majors and wholesalers, and apparently can choose the cheapest offer and set their prices independently. Exxon sources [...]* % of its needs from third parties purchases. The remainder comes from transfers from affiliates or from term processing arrangement at OMV Schwechat refinery. BP/Mobil purchases [...]* % of the volumes it sells from third parties. The remaining is in country refining production.

Retailing

(495) In Austria, the fuel retailers consist of branded stations, independent stations (or free stations) and white pumps. Supermarkets do not play an important role for fuel and run only very few stations.

(496) The main branded networks are those of OMV, Aral, BP/Mobil, Shell and Esso. Other branded stations are B-brands and white pumps.

(497) 64 % of the stations are branded. These stations are run by agents through a system whereby agents are bound to fix their prices according to the majors' instructions. It is important to note that majors (OMV, Aral, BP/Mobil, Shell and Esso) are shareholders of at least [...]* % of the stations in Austria.

(498) The free stations (or B-brands) are Avia, Jet, Turmöl, Rühl, IQ, AWI, Roth, Hölzl, Champion, A1, Eurol, etc. Other B-brands are partly in the hands of majors: Genol (29 % OMV) and Avanti (50 % Shell) have a net all over Austria. In Austria, white stations are not owned by independent companies, but by the majors, who prefer not to brand them (OMV for instance has a white pump network: Stroh-Tankstellen).

(499) [...]*.

(500) The service station industry is heavily regulated in Austria. Before 16 September 1981, fuel prices were regulated by the government. Prices changed only on request of the mineral oil companies if their costs increased. In 1981, the price for all fuels except diesel were liberalised. The players on the market have been OMV, Agip Austria AG, BP Austria AG, Mobil Oil Austria AG, Shell Austria AG and Aral Austria GmbH. On a local level, the majors have to compete with the B-brands (such as Avanti, Jet, Turmöl) and independents. In their surroundings majors offered their stations lower prices, in order to compete in these areas ("Price concentration"). In 1990, a new agreement (the "Branchenübereinkommen") entered into force, in order to make pricing more objective. A market information system was installed to follow the international and national fuel prices on a weekly basis. This operation did not really succeed so the agreement has been abandoned, as well as the regulation of diesel price.

(501) Austria also imposes high environmental requirements but these are no different from those required in Germany. Small stations cannot afford to meet these high standards (Exxon indicates this to be the main reason why the market will be further rationalised). Financial support is provided for closing down stations. All mentioned costs are similar in Germany.

(502) Austria has a very dense service station network (denser than in Germany). As a consequence, the average throughput per station is lower: 1422 m3 in Austria compared to 3170 m3 in Germany. The number of service stations has decreased by 23 % from 1995 to 1997. Since 1995 there has not been a new player entering the market except Jet-Conoco who have expanded their network from 29 to 55 stations.

THE PRE-MERGER ENVIRONMENT - CURRENT STATUS OF COMPETITION

Market structure

(503) The following table shows the evolution of market shares in Austria over the last four years. It appears from this table that the market was until now dominated by three companies, OMV, BP/Mobil and Shell holding together [50 to 60]* % of the market. The two following market players are Esso and Aral, each with more than [0 to 10]* % market share.

>TABLE>

(504) As explained above the remaining of the market players is composed of other smaller branded networks, supermarkets and white pumps. A number of the smaller branded networks, accounting for around [30 to 40]* % of the B-brands, are subject to financial interests from one of the top three players in the market. According to the notifying parties, market shares of supermarkets have remained flat at 2 % over the last five years and are not expected to vary significantly in the future. Similarly, white pumps have declined from 8,4 % to 8,0 % in the same time period. The parties do not expect that the competitive configuration in Austria will change significantly over the short term. This would be due to the current restrictive conditions for obtaining permits.

Competitive dynamics - costs and prices

Costs

(505) It is very unlikely that any of the other branded retailers could exercise any competitive pressure on the top three players in the market. Indeed, in the course of the market investigation, competitors have indicated to the Commission that if majors raised their prices, the independents will follow. Independents want to maintain a good relationship with the majors, as they would not survive a price war or too strong competition. This is due first to the financial interests held by OMV, BP/Mobil or Shell in some of these players. Secondly, these players' cost bases are higher than those of the A-brands. Indeed, their average throughput in 1998 was of about 824 cubic metres to be compared to an industry average of 1422 and an average of 2000 for the top three. At these low levels of throughputs, fixed costs are spread on a small basis and therefore represent a substantial share of total costs. Therefore, these differences in levels of throughputs result in significant cost differences. Thirdly, the smaller competitors depend to a significant extent on supplies at the wholesale level from the top three players which share with Exxon the production of the OMV refinery.

(506) In view of the above, it would appear that the competitive conditions of the Austrian market are dictated by OMV, BP/Mobil, Shell, Esso and Aral. Given such a small number of actual competitors, and given that the general conditions of the motor fuels retail sector are traditionally conducive to oligopolistic behaviour, there would be strong indications that the Austrian market could function under or very close to oligopolistic conditions.

(507) In Austria, account should also be taken of the fact that BP/Mobil and Aral have structural links (see general section above), and that all of these companies have specific arrangements with the OMV's Schwechat refinery.

Prices

(508) An examination of the level and variations of the motor fuels retail prices strongly suggests that the market is in deficit of competition.

(509) Indeed, prices (excluding tax) in Austria are amongst the highest in Europe. Net prices for gasoline and diesel were respectively 33 % and 29 % higher than in Germany in 1998. In addition, motor gas price shows limited correlation to crude oil price or Platt's quotations for diesel or motor gasoline. The Puwein study cited by the parties explains that the monthly variations in Germany have been 50 % stronger than in Austria and that a 10 % change of the crude oil price causes a price variation of 3 % in Austria and 5 % in Germany (excluding taxes).

(510) The parties have justified the higher prices in Austria by the combination of additional logistics costs, lower throughput and restrictions relating to the service stations shops. Logistics costs would break down in additional transportation costs (when compared to Germany) of USD [...]* and USD [...]* attributable to additional distribution costs within Austria resulting from the local terrain and retail network structure. Lower throughput would result in additional unit costs of USD [...]* which would need to be recovered through higher margins. Finally, the restrictions on shops would include a limitation on their size (80 m2 versus an Exxon standard of [...]* m2), the need for a special licence to sell tobacco, etc. This would lead to a loss of contribution to the operator's income of USD [...]*. The parties add these three items to USD [...]* and conclude that this explains 90 % of the difference between Austrian and German motor fuels pump prices.

(511) The assumption on the loss of income because of shop restrictions seems questionable. First, the parties refer to a loss of income and not to a loss of margin, which should then be of a lower value than the total income lost. Secondly, when asked about an evaluation of the additional margins generated by convenience stores in Germany, the parties referred to an item labelled "gross distribution margin - other" which featured an amount of USD [...]*/tonne for Germany ([...]*). However when looking at the same line for Austria, this item shows a higher value of USD [...]*/tonne.

(512) In any event, assuming that this USD 85/tonne relies on robust assumptions, the [...]* % non-explained price difference with Germany would still amount to around [...]*cents/litre. Given that the parties have themselves estimated that a margin over cash costs of around [...]* cents/litre would provide an acceptable return on investment, this would represent a [...]* % increase on the return [...]*.

Entry barriers - expansion - potential competition

(513) Exxon has considered both orally in meetings with the Commission's services and in its internal documents that it does not expect any competitive threat coming from supermarkets in Austria. This was confirmed by the market investigation.

THE POST-MERGER ENVIRONMENT - IMPACT OF THE OPERATION ON COMPETITION

(514) The impact of the operation is to be assessed against the above described background, having also regard to the features identified in the section devoted to the structural characteristics of retail motor fuel markets in general.

(515) The merger between Exxon and Mobil will result in the establishment of strong links between BP/Mobil, Aral and Exxon's respective activities in motor fuels retailing in Austria. This will lead to a material diminution of the incentives for Aral and Esso to challenge the leading position held by OMV, BP/Mobil and Shell. Indeed, Aral and Esso seemed to be the only networks with the capacity to exert some competitive constraints on the three top companies. Following the merger, the oligopoly would involve combined market shares of [70 to 80]* % (including the ownership or interests in non-A-brands stations) with very small competitors left in the margins with a cost basis which does not allow them to compete effectively. Moreover these small competitors would be to some extent subject to the financial interests of one of the top three players and also dependent on supplies from the same companies. The merger would therefore result in a creation or reinforcement of the collective dominance held by OMV, the trio Exxon - BP/Mobil - Aral, and Shell in the Austrian retail motor fuels market.

FRANCE - MOTORWAYS

GENERAL OVERVIEW

(516) As will be shown below, the competitive environment in the market for motor fuel retailing on French toll motorways is characterised by the presence of very few players. The supply is very concentrated and all the players are large vertically integrated oil companies. Given that no foreclosure aspects seem to arise in relation to refining and wholesaling in this market, these points are not further considered for the purpose of this assessment.

Toll motorway station as a distinct product market

(517) The operation will have a very significant impact on motorways in France. As noted above, toll motorways can constitute separate product markets by virtue of the specific features characterising them. In addition, France is one of the Member States where most motorways are toll roads, with motorists paying toll based on the distance travelled on the motorway. This factor further discourages motorists to exit motorways to buy fuel on off-motorway stations. Also, from the supply side, the insulation of the motorway channel is to do with the existence of high entry barriers (see recitals 554 to 561 below).

(518) The chart below shows the differentials between Exxon motorway and Exxon average network pump prices in the Member States mostly affected by the transaction(60).

>TABLE>

(519) As may be seen from the chart above, France is the only country among those considered for the purpose of the present case where price differences between after-tax prices on motorways and after-tax prices off-motorways are significant (in the range of [...]* %). This is presumably due to the fact that in other countries motorways are subject to a more intense competitive constraint coming from off-motorways stations.

(520) The parties consider that motor fuel retailing on French motorways does not constitute a separate market from motor fuel retailing off-motorways. This point has been also made by other oil companies replying to the statement of objections issued by the Commission in the course of the proceeding. All of them refer to the fact that owing to the large fuel capacity of modern cars, motorists can drive long-distances and buy fuel off-motorways. In addition, in France motorists are regularly informed through the publication of special brochures on price differentials between motorway and off-motorway stations. Also, according to both the parties and the other intervening oil companies, over recent years average consumption of motor fuel has decreased on French motorways, in particular if one takes into account the average traffic increase.

(521) More specifically, it is argued that fuel demand on French motorways has significantly decreased precisely because of the competitive constraint exerted by off-motorway stations. To support their positions, the parties refer to a survey carried out by the Union Française de l'Industrie du Pétrole in 1992, whereby an increase in sales volumes of motor fuels on concession motorways of approximately 410 million litres was projected in the following five years, taking also into account an increase in traffic flow and motorway infrastructure. Ex post, while the increase in traffic flow and motorway infrastructure did occur, motor fuel sales declined by almost 270 million litres during that period. On this basis, the parties draw the conclusion that there has been a migration of customers from motorway service stations to off-motorway service stations.

(522) The evidence provided by the parties is at least incomplete. The fact that there has been a decline in fuel sales on motorways of 270 million litres in five years is not conclusive of anything if taken alone. In fact, in order to show that motor fuel sales on motorways have declined in favour of sales off-motorways, the parties should have compared in a meaningful way consumption patterns in both networks. Instead, the parties have never made such a kind of assessment. More importantly, they simply disregard the fact that even in off-motorway stations, logically enough, there may have been an increase of traffic flow and road infrastructure.

(523) If anything, a meaningful comparison between motor fuel demand at motorway stations and demand at off-motorway stations shows that consumption patterns in these two networks are very similar. The table below shows the total volume sales of motor fuel on French motorways over the last five years as well as the total volume sales of motor fuel in the French market as a whole. No account has been taken of the increase of traffic flow and road infrastructure.

>TABLE>

(524) As may be seen from the table above, the trend of motorway sales over the last five years is quite similar to the overall trend. There has been a slight demand increase which is essentially due to a growth in the demand for diesel, while consumption of gasoline has slightly decreased. This trend is common to both the motorway network and the off-motorway network. On the basis of this evidence, it appears that consumption patterns on motorways are not dissimilar from those characterising off-motorway stations.

(525) A qualitative analysis of motorway stations' customers is also instructive. The table below shows an estimation of Exxon total volume sales of retail motor fuels in the years 1996 to 1998 broken down on a monthly basis.

(526) Percentage of sales by month for Exxon motorways:

>TABLE>

Figures are expressed in percentage terms.

(527) From the table above it is clear that the majority of sales of the year on French motorways occur in those months covering periods of public holidays. In particular, with respect to motor gasoline consumption, two months of the year, namely July and August, count for more than [...]* % of the overall sales; if April and May are added, four months account for around [...]* % of the total sales over the year. As to diesel, this trend is less pronounced but still present. In general terms, all the months covering periods of public holidays generate higher sales. In other words, the data above show a great seasonality of demand, which in turn implies a higher degree of captivity. Seasonal clients are essentially tourists, who are most likely to refuel on motorways. These findings again militate in favour of motorways being treated as a distinct market.

(528) Finally, the evidence that motor fuel retailing on French toll motorways constitutes a separate market lies in the fact that prices are in the range of 5 to 10 % higher than prices off-motorways. On this point, the parties claim that motor fuel retailers bear higher costs when running motorway stations. However, this fact does not explain at all why, as a matter of fact, motorists are willing to pay a premium for fuel bought on motorways. In sum, the "insulation" of the motorway channel in France from the rest of the market is precisely reflected into prices which are from 5 to 10 % higher than prices on off-motorways stations.

THE PRE-MERGER ENVIRONMENT - CURRENT STATUS OF COMPETITION

Market structure

(529) The pre-merger situation is as follows. Unlike off-motorway retailing, the market for motor fuel retailing on toll motorways in France is already very concentrated. It is characterised by the presence of only six players, all vertically integrated oil companies, of which one, Agip, has an insignificant position. The five remaining players can be divided into two categories: the national champions, namely Total and Elf; the majors, that is to say Shell, Exxon, and BP/Mobil. The two national champions are the market leaders. This is to do with their historically strong presence on national territory which on the toll motorways market has remained unchallenged. In this respect, both players have their own refineries situated in France, which secure timely and cheap supplies of motor fuel; they both own numerous storage facilities well spread over French territory; they both have a very high number of station sites which secure a capillary coverage of the motorway network. Of the two players, Total has a leading position on the toll motorways market. Following the merger with Fina, it has further reinforced its leadership, reaching a share of roughly [30 to 40]* %. Total is also, as will be shown below, the most cost efficient retailer when looking at the average throughput per site. As to the three majors, they are all mid-size players. Shell and Exxon have a comparable market position ([10 to 20]* % and [10 to 20]* % respectively), both in terms of volume sales and number of sites, while BP/Mobil follows not very distantly. The pre-merger CR4 is close to [80 to 90]* %.

(530) There are virtually no independent retailers present on the market. This is due to a number of reasons, notably the difficulty of obtaining licences on motorways, the scarcity of sites available, the high investment costs needed (these aspects will be further discussed at recitals 554 to 561 below).

(531) The table below shows an estimate of the market shares (in volume) of the parties and their competitors over the last five years.

>TABLE>

(532) The market has remained quite stable in terms of trends in market shares over the time. There are market shares fluctuations involving some players of something in the range of 10 % over a period of five years. However, the leading players maintain virtually unchanged their market position over time, with Total in the lead, followed by Elf, Shell, Exxon and BP/Mobil.

(533) It is also worth measuring the market position of the various players by reference to the number of station sites.

>TABLE>

(534) Total and Elf have a comparable number of sites (126 and 124 respectively), followed by Shell, Exxon and BP/Mobil. In terms of evolution of shares over the time, the stability is less accentuated than in market shares based on volume, but still very high. In particular, over a period of five years, Total has increased its number of sites by around 20 % (without counting the acquisition of Fina), taking the lead over Elf. As to the other leading players, with the exception of BP which in 1997 combined with Mobil their respective networks, they have maintained their position virtually unchanged.

(535) Overall, from the above considerations it appears that over recent years the market for motor fuel retailing on toll motorways in France has remained quite static and no changes have been recorded. Demand has remained quite flat over five years. The competitors have more or less maintained their position.

Competitive dynamics - costs and prices

Costs

(536) A major factor in the profitability of a retail network is the throughput. The higher the throughput, the lower are sites' average operating costs per litre. In order to make a cost comparison of the above players, the chart below shows the average throughput per site of the various players in the last five years.

>TABLE>

Source: UFIP

(537) From the table above it appears that Total has the highest throughput with 7970 m3 per site, followed by Exxon with 6482 m3. However, all the leading players have a throughput well above 5000 m3. In general terms, it can be said that the average throughput is quite high when compared to the average of the off-motorway stations. However, this is due to the fact that motorway stations are generally bigger in size than an ordinary off-motorway service station.

(538) An average throughput of between 3000 and 3500 m3 per site is viewed by the industry as allowing an ordinary service station to reach a minimum efficient scale (m.e.s.). However, given that motorway stations have higher costs to bear than ordinary stations, a higher average throughput can be taken to secure a minimum efficient scale. A throughput of 6000 m3 (twice the average m.e.s. throughput) can be reasonably taken as a breakeven benchmark. Using this approach, it appears that no noticeable differences exist between the various players in terms of cost efficiency. Three players (Total(61), Exxon and Elf) largely hit the target, while two others (Shell and BP/Mobil) are very close.

Prices

(539) In order to fully appreciate the competitive environment in motor fuel retailing on French toll motorways, it is worth taking as a reference the French market for motor fuel retailing as a whole, which has been cited by the notifying parties as one of the most competitive in Europe.

(540) As shown in recitals 518 to 520 above, it appears that France has by far the highest price differential. The comparison with figures of different countries can be sometimes misleading due to the different competitive conditions existing in each national market. In some countries such as Austria and Netherlands, for instance, overall pre-tax off-motorway prices are extremely high and this can in part explain a lower differential with motorway prices. Also, with the exception of France, in all those countries motorways are free of charge. This militates in favour of a more intense competitive pressure exerted by off-motorways stations on motorway stations. However, the situation in France can be compared to the United Kingdom since pre-tax pump prices in the latter country are close to pre-tax prices in France. Nonetheless, the United Kingdom has a differential which is more than five times smaller than France. Also the situation in Germany can be instructively compared to France since price differences are not very pronounced.

(541) The chart below shows a comparison between Exxon motorway prices and Exxon network prices in France, the United Kingdom and Germany.

>TABLE>

(542) It can be noted that the price differentials between these countries are very significant and can hardly be explained by higher costs in France as opposed to the United Kingdom and Germany. Even considering that some of the costs listed by the parties for the French motorways are absent in the United Kingdom and Germany (such as concession rents), the other costs appear to be common extra costs that a retailer has to bear when running a service station on a motorway irrespective of the country.

(543) A comparison of gross margins and costs borne by Exxon on motorways in France, the United Kingdom and Germany can be illustrative in this respect. In particular, it is instructive to compare what the parties name in their submission "other wholesale costs", meaning those costs which cover all direct (such as land rentals/taxes, advertising, retail personnel) and indirect (such as corporate overhead, systems) costs specifically associated with running and managing the retail business in that country. Transportation and delivery costs are not included.

>TABLE>

Figures are expressed in USD for metric tonnes.

(544) From the table above it appears that those costs that the parties consider as being the cause of the price surcharge on French motorways are lower in France than in the United Kingdom and Germany. By contrast, gross margins and prices in France are substantially higher than in the United Kingdom and Germany.

(545) Nor can it be claimed that the comparison of the above data is misleading due to the different structure of the Exxon retail network in each country. In this respect, it suffices to say that the Exxon motorway network in France is exclusively composed of company owned stations, that is to say the stations where Exxon bears the highest possible costs.

(546) In their response to the statement of objections, the parties note that the costs data cited above as representative of motorway stations include in fact motorway and off-motorway stations. In this regard, it should be noted that the information upon which the Commission has relied was submitted by the parties in response to a request of information unequivocally requiring the cost data relating to motorway stations. The Commission has thus assumed that the parties regard these data as somehow representative of the costs of their motorway stations as well.

(547) In any event, even basing the assessment on those data that the parties do not contest, it appears that the margins generated on French motorway stations are inexplicably high. In this respect, as has been noted above, pre-tax pump prices on French motorways are on average between [...]* and [...]* % higher than overall French pre-tax pump prices. The parties contend that these differences derive from the higher costs borne by a retailer when running a motorway station. In particular, they claim:

(i) that all French "concession" motorway service stations (which constitute [...]* % of Exxon stations) must pay volume-related rents of approximately [...]* centimes/litre to the motorway concessionaire;

(ii) that motorway service stations incur higher investment costs to provide additional facilities required by the concessionaire: these, typically, include additional fuel storage tankage, additional washing, parking facilities etc., and the resulting extra cost amounts to [...]* centimes/litre;

(iii) that motorway stations are required to be open every day for 24 hours. The additional costs are about [...]* centimes/litre;

(iv) that the use of fleet cards is higher on motorways, resulting in additional costs of about [...]* centimes/litre.

All these costs allegedly account for a large part of these differentials - [...]* % for motor gasoline and [...]* % for diesel.

(548) However, even accepting the generous estimates made by the parties, there would still remain an unexplained significant differential of something in the range of 20 to 30 %. In short, on the basis of the calculation provided by the parties, there would still be a margin of up to 10 centimes/litre which is not explained by any extra costs.

(549) In the light of the above it appears that overall margins and ultimately, profits made by the motor fuel retailers on French motorways are higher than those generated on off-motorway stations in France. These returns are also higher than those generated in the motorway stations of the other Member States considered above.

Price trends

(550) A comparison of the price trends over the last four years between motorway and off-motorway pump prices in France is also quite illustrative.

>TABLE>

(551) From the table above it appears that in absolute terms pre-tax prices have gone down on both motorways and off-motorway stations owing to a crude oil price decrease. However, in relative terms it appears that pre-tax pump prices have followed different trends, with off-motorway prices decreasing much more dramatically. In this respect, it is striking that the differentials between motorway and off-motorway prices have increased by something in the range of 30 to 40 %.

(552) In essence, the table above shows that, first, the market for motor fuel retailing on toll motorways in France is unrelated to price trends prevailing in the neighbouring market of off-motorway stations; secondly, the players on the toll motorway market seem to have reached a supracompetitive equilibrium which is not disturbed by the intense competitive game involving the various players on the off-motorway market.

(553) Finally, with respect to the degree of price competition existing between the various retailers on this market, it is worth noting that price differences are very small. In general terms, pump prices track each other closely and variations between different retailers' prices are insignificant. This lack of price competition is again the indication that each retailer is satisfied with his returns and that the market has reached a supracompetitive equilibrium.

Entry barriers - expansion - potential competition

Entry barriers

(554) It is acknowledged by the industry that a number of factors make entry on the toll motorway motor fuel retailing market more difficult than in the retail service station market as a whole. This is the case with France as well.

(555) First, building a new site on a motorway requires some investment. The parties estimate that the cost of building a new site on a motorway can be from five to 10 times higher than the costs relating to an ordinary site. According to the parties a double-sided site can cost up to USD [...]* and a single site can cost USD [...]* while an off-motorway site generally costs around USD [...]*. The extra costs result essentially from the requirement to provide additional facilities such as fuels storage tankage, additional washing, parking facilities etc.

(556) Secondly, the process for obtaining approval for a new site can take a very long time. Overall, a period from one to two years is required in France to complete the construction of a site on a motorway, counting also the time for obtaining the requisite permission and finalising the tendering procedures.

(557) Thirdly, the construction of new service stations is heavily regulated and carefully controlled by the highway authorities in France. Inter alia, the spacing between service stations on motorways is also regulated. This clearly restricts the opportunities for developing new sites.

(558) Fourthly, the market is already quite saturated in terms of existing sites and this further reduces the possibility of creating new outlets.

(559) As a matter of fact, the scarcity of new sites together with the other factors cited above has traditionally prevented the entry by newcomers on the toll motorways market in France. It is quite instructive that over the last 10 years, only three motorway service stations have been awarded by tender to companies which, prior to 1989, were not represented on the French motorways. One station has been awarded to Leclerc (a French hypermarket company) and two stations have been awarded to Dyneff (a large independent reseller).

(560) The parties claim that there is an element of contestability in the market due to the fact that concessions for exploiting sites are of limited duration and are periodically re-assigned by the highway authorities. Motorways in France are owned by either the central government, by local governments, or by wholly or partly privately owned companies. The latter are awarded the rights to build and operate motorways under "concessions" from the national/local governments. They in turn award concessions for retail service stations to motor fuels retailers. In the case of the "concession" motorways, their duration tends to be between 15 and 30 years, although those awarded directly by the local or national governments are generally for 30 years.

(561) Against this background, it should be noted that this system, while theoretically designed to create some form of competition for the reallocation of an existing site, has never functioned properly. As a matter of fact, the parties admit that they have no knowledge of a new bidder acquiring existing motorway service stations at the end of an initial concession period. This is because the industry practice is for the existing supplier to obtain an extension of the initial concession, usually in conjunction with an agreed investment in or modernisation of the service station, before the initial concession period expires.

Potential competition

(562) It should also be noted that no real threat of potential competition exists on the toll motorways market in France. In this respect, the major source of competition in off-motorways motor fuel retailing in France has come from supermarkets and hypermarkets (they account for more than 50 % of the total sales in France). These players have in particular developed fuel retailing as an ancillary activity complementary to their core business. The expansion of their sales is essentially based on the possibility to keep prices low by achieving cost efficiencies. This in turn is made possible thanks to the logistic synergies existing between their core activities and the activity of motor fuel retailing. In particular, the hypermarket format places a conveniently located motor fuel forecourt on the car park to take advantage of the large number of customers coming to shop at the superstore. This provides the operator with high throughputs at low operating costs. However, all these factors cannot be easily recreated on motorways. As a matter of fact, to date, with one small exception (Leclerc has one station with lower than average prices), none of the supermarket chains is present on French motorways. Most of the supermarket and hypermarket chains active in motor fuel retailing have indicated in their responses to the Commission enquiry that they do not intend to enter the toll motorways market precisely for the reasons stated above. Also, they point to the fact it is all the more difficult to enter the toll motorways market without a sufficiently large number of stations.

THE POST-MERGER ENVIRONMENT: THE IMPACT OF THE OPERATION ON COMPETITION

(563) As indicated above, the pre-merger situation shows already a deficit of competition which can be explained by a number of factors. The market is very concentrated with five operators (TotalFina, Elf, Shell, Esso and BP/Mobil) accounting for almost 90 % of the sales. There is no presence of any aggressive competitor as in the off-motorway market. The market is well isolated from any competitive threat by high entry barriers. Also, independent retailers and in particular hypermarkets which theoretically would be ideal candidates to enter this market, have shown no interest or have not been in a position to do so.

(564) The impact of the present operation is to be assessed against the above described background, having also regard to the features identified in the section devoted to the structural characteristic of retail motor fuel markets in general. First, the operation causes a further increase in market concentration. In essence, the operation results in the elimination of one mid-size competitor from the market.

(565) More importantly, the operation will change the structure of the market as it will result in the constitution of a new force constituted by the couple Exxon/Mobil BP/Mobil, holding together a market position roughly comparable to the competitor number two. In sum, the operation has a significant adverse effect by rebalancing the internal oligopoly equilibrium. Instead of five players, three of which hold relatively low market shares, the new landscape will be characterised by the presence of two big players, namely Total/Fina with [30 to 40]* % and Elf with [20 to 30]*, plus a comparable third force constituted by Exxon and BP/Mobil, holding together [20 to 30]* % and a smaller player, Shell, at [10 to 20]* %.

(566) The competitive relationship between the members of the oligopoly will be changed in that the scope for retaliation will be increased. With higher and more balanced market shares of the participants, the consequences of competitive actions in terms of retaliation are potentially more widespread and serious than with the current supply structure.

(567) In the pre-merger scenario, the three mid-size players could still represent disturbing factors in the oligopoly equilibrium. In particular, they could legitimately aspire to expand their market shares in the light of their good cost-efficiency scores. Also, these three players represent the only multinational super majors, thus being in a position to constitute a constant competitive threat to any national leader. After the merger, the market structure will be such that no competitive move will ever be conceivable without triggering a strong and widespread reaction. Due to the low growth of fuel demand, any attempt from one of the various players to increase its sales will inevitably have to occur to the detriment of the other players. In addition, given the homogeneity of the product, the only possible competitive strategy aimed at gaining market shares consists of reducing prices. However, any price cut will inevitably trigger a retaliation from the other players which will feel directly targeted. Given in addition that three players out of five have a very capillary presence on the territory, and that they are equally cost-efficient, competition on price will be avoided at all costs. Shell, which is the only "small" member of the oligopoly left, and could have different incentives, would however be running very serious risks if it were to compete vigorously, since it has a smaller market coverage than the bigger retailers, and therefore it is easier to target. In essence, the merger will render the interaction between the few dominant players easier to implement and monitor, while further exacerbating the poor economic performance of the market.

(568) In the light of the above, it appears that the operation will result in the strengthening of a collective dominant position held by TotalFina, Elf, Shell, Esso and BP/Mobil as a result of which effective competition would be significantly impeded in the common market.

GERMANY

GENERAL OVERVIEW

Refining

(569) There are currently 17 fuel refineries in Germany, on the coast or inland. The main refiners are Exxon, Shell, BP/Mobil and DEA who own approximately 42 % of total capacity. These three companies are vertically integrated in the sense that they are also active at the stage of motor fuels wholesaling and retailing. Most other refiners (such as Conoco, Elf, Total, Agip, OMV) also are vertically integrated oil companies. Non integrated refiners such as Ruhr Oel, Wintershall, Beta, Holborn account for 43 % of the total German production. These refineries have a total capacity of 2186 KBD. Consumption in the whole of Germany (diesel/Mogas/Mofuel) reached a total of 622 million bbl in 1996. According to Eurostat, Germany has net imports of around 21 % of its production capacity, although this figure does not include the recent built Leuna refinery. It is important to note that the inland refineries are unable to sell their excess production on the cargo market and have therefore to market it in Germany.

Infrastructure

(570) Refineries and storage capacity are strategically located close to the major consumption areas and regional storage capacity provides product supply in the more remote parts of Germany. The river Rhine connects the Western located refineries (Gelsenkirchen VEBA, Wesseling DEA, Godorf Shell, Karlsruhe Esso/DEA/VEBA/Conoco) and indirectly via the Mittellandkanal the refinery of Wintershall. The southern refineries (Ingolstadt Esso, Vohburg BP/Mobil/AGIP/RuhrOel) are accessible via the Danube river. Three refineries are connected to the port of Hamburg (Heide DEA, Hamburg Holborn and Hamburg Shell). Inland refineries which are not connected to a port or river mainly serve the region in which they are situated or important storage concentration as it is the case for Leuna (Elf) for the Dresden area, Burghausen (OMV) for Munich and Austria (via pipe connected to the Danube) and Schwedt (DEA/Ruhroel/AGIP/Total/Elf) for the Berlin area. The regions with high consumption levels (Hamburg, Hanover, Cologne, Frankfurt, Stuttgart, Munich, Berlin, Dresden and so forth) are supplied by ex-refinery sales from nearby refineries.

(571) Pipelines (both product pipelines and the NATO jet pipeline) cover the western and southern parts of Germany while the eastern part of Germany is served by important storage capacity connected by both pipelines and the river Elbe. Pipelines connect all major storage locations with the refineries, although a distinction should be made between proprietary product pipelines and the NATO pipeline system which is running at full capacity and whose primary function is to transport jet fuel to military airports.

(572) Most of Germany can be supplied from the cargo market by barge traffic on the Rhine, Danube and Elbe (which connects Hamburg with eastern Germany). Due to the varying water levels, barges are restricted to a maximum of 1000 tonnes. Barge transportation on the Rhine can be affected by both high and low water levels. While rising water levels have virtually no impact on freight rates as long as traffic has not been discontinued (which is very rare), low water levels do have an impact on freight rates, because loads must be reduced in order to ensure clearance through shallow water. During these periods, freight rates go up, although the impact on rates also depends on the actual demand situation upstream. In periods of low demand (particularly summer, when demand for gas oil is low), low water levels may have little effect on freight rates. In periods of high demand, low water levels may substantially affect freight rates. The duration of low water levels on the Rhine typically ranges from one week per year to one month per year.

Wholesaling

(573) From the market investigation it appears that access to the wholesale market is relatively easy for retailers. Typically sales prices charged to independent resellers are based on the OMR (the Oil Market Report is a local equivalent to Platt's) quotation with an agreed differential to cover factors such as location differentials and any service provided. OMR quotations are the price that a reseller would have to pay if he wanted to buy motor fuels from storage facilities close to the Rhine. They parallel Platt's quotations but are higher by approximately 10 % to 15 %. According to the parties, barge freight from the ARA (Antwerp-Rotterdam-Amsterdam) area to the Karlsruhe area should account for approximately half the difference between the Platt's and OMR quotes. The other half of the difference represents all other costs that a company transporting product into Germany and selling it to customers there would incur, including handling costs (such as costs of loading and offloading barges), terminaling/storage costs, local port charges (if any), insurance, costs for order and sales staff, etc., and any margin for the company selling the product.

(574) The parties have estimated that 55 % of storage capacity in Germany is in the hands of independent companies. This has been confirmed to the Commission by independent storage suppliers such as Petroplus, Oiltanking and Van Ommeren and by direct customers such as Mabanaft (trader), AVIA and Baywa (wholesalers). These independent companies offer their services to all kinds of clients, even to majors.

(575) In Germany, traditionally more than 70 % of petroleum products are sourced from inland refineries direct. Imports are generally directed to independent resellers, but they can also stem from intra-group trading. Currently, most competitors to the parties only buy limited volumes in order to safeguard independence from the majors controlling the refineries. Non-integrated competitors to the parties have indicated that in the past years, more than 90 % of the purchased product was supplied by in-land refineries or purchased from majors who obtain the product via exchanges or affiliates purchase. The non-integrated competitors purchase from both inland refineries and the cargo market (ARA). Sourcing depends on the conditions offered; as the price differential between cargo and inland refineries changes all the time, sometimes prices on the cargo market are lower than inland. For instance, it has been explained to the Commission that at the moment the inland refineries offer a more attractive price than the cargo plus transport and storage. It has appeared that unless cargo prices are significantly lower, it is more interesting to buy ex-inland refinery.

(576) As to the vertically integrated retailers, they obtain motor fuels from three sources: own production, exchange agreements with other refiners or through purchases from third parties [...]*. BP/Mobil purchases [...]* % of its needs from third parties. Exxon has exchange agreements for motor fuels with [...]*. This is partly due to the fact that Exxon mainly operates refineries in southern Germany.

(577) A number of actors, such as Avia, Mabanaft and Petroplus (wholesalers and traders), are present at the wholesaling level, which ensures that non-integrated retailers can obtain supplies of motor fuels. It is to be noted however that the aforementioned wholesalers have used supplies via traders only in a limited way as most wholesalers have also (limited) trader activities or have sufficient combined buying power to order vessels/barges on the cargo market.

Retailing

(578) Four categories of retailers can be distinguished in Germany: A-brands, B-brands, independents and supermarkets. A-brands include vertically integrated oil companies such as Exxon, BP/Mobil, DEA and Shell but also Aral, where, as explained above, Mobil has a 28 % stake (with the other main shareholders being Veba (56 %) and Wintershall (15 %)), and which benefits from captive supplies from its shareholders Mobil (through BP), Veba and Wintershall. B-brands include smaller integrated companies such as Elf, TotalFina or Jet/Conoco.

(579) From 1994 to 1997, the number of service stations has fallen by 7 %. During the same period the volume of motor fuels sold in service stations remained stable at 50 million cubic metres.

(580) About 60 % of the service stations are branded. Sometimes the oil company owns the property and the building/investments on it. Sometimes the oil company owns only the building and investments whilst the property belongs to the retailer. The majors operating in the German market have initiated an "agency contract system". This is specific to Germany, where personnel costs are said not to allow COCO structures. The agency system ("Agentursystem") can either include the whole business of the station or can also exclude certain parts such as the shop or car wash.

(581) A distinction can be made between dealer-owned stations and company-owned stations. There are only a few "test stations" of company operated stations. As far as dealer-owned stations are concerned, oil companies still have a very strong position. Agents depend on the oil company, which sets the price. For instance, the dealer-operated service stations bearing the Esso brand in Germany are all under agency contract. Under such contracts, Exxon owns the fuels stored on the station premises and sets the retail price. The agent system allows better and more secure margins [...]*(62).

(582) The dealer is remunerated through a commission based on sales volumes and unrelated to pump price. The agent receives only a margin on every sale, of typically about 2,5 to 3 pfennig per litre for dealer operated and 5 pfennig per litre for dealer-owned stations. The commission is higher for dealer-owned stations because the dealer has then to bear more costs than in a company-owned station.

(583) As far as company-owned dealer-operated (agents) stations are concerned, the contracts have a long duration (normally about 10 years, but can be extended twice for a period of five years), which makes it difficult for competitors to enter. The length of Exxon's dealer supply contract varies between [...]*. It amounts to [...]* in average.

(584) Convenience stores are a major element for the profitability of service stations in Germany, as they subsidise the fuel price. They are run by subsidiaries of the oil companies or by companies like "Lekkerland". The oil company receives a rent in return, although for some companies the "agency system" also applies to the shop.

(585) Owing to the severely regulated opening hours of shops and supermarkets, the convenience store has become an important retail facility in Germany (4 % of food and beverages sales in Germany). The convenience store attracts customers and aims at enhancing the service and quality image of the fuel retailer. The rent received from the shop business generates important profits offsetting lower demand for fuel ([...]*). The importance of a convenience shop is expected to increase further. It is considered by the majors to be a powerful differentiation from white pumps and B-brands.

(586) [...]*. Majors therefore invest heavily in service offering, based on the significant 70 % brand loyalty of the German customer, as a strategy to maintain profitability and market shares differentiating A-brands from "price-cutters". [...]*.

Western Germany and eastern Germany

(587) Given the differences in terms of supplies, welfare and even market players between the eastern and the western parts of Germany, it could be argued that both form two separate relevant geographic markets. Indeed, there are some differences in prices, and market shares are not similar.

(588) However, this assertion was not confirmed by the Commission's investigation for the following reasons. First, the price difference between both regions is limited and in some cases inferior to price differences within western Germany itself. Typically prices are slightly lower in the eastern part of Germany. This is in fact due to massive entry into this market in the early 1990s, when traditional western Germany retailers tried to duplicate their market share in the new Länder. This has resulted in excess capacity and therefore pressure on prices. Secondly, the manifest market shares differences between western Germany and eastern Germany have a historical basis. After German reunification, the monopolistic position of Minol was suspended when Elf in 1992, having a market share of less than 10 % in western Germany, acquired Minol and therefore market leadership. Up to now, Elf has kept this position of market leader although its market share has been reduced to 24 %(63) by competitors. Minol had a deteriorated network, and competitors have gained market share with new service stations at strategic locations. As the deteriorated Minol network is likely to be further rationalised, it is to be expected that the eastern German market is going to evolve slowly to the market shares ventilation in the west. Thirdly, there is no difference in legislation, fiscal provisions or environmental obligations. Fourthly, there is no natural barrier between the eastern and the western parts of Germany. This means that a uniform price increase in the western part would be defeated through marginal competition at the limit between the eastern part and the western part, through a sort of ripple effect.

THE PRE-MERGER ENVIRONMENT - CURRENT STATUS OF COMPETITION

Market structure

(589) Overall, the majors (ARAL, Shell, Esso, BP, DEA or "A-brands") have more than 65 % of the market. Hyper/super markets are at around [0 to 10]* %, White pumps [0 to 10]* % and the most important B-brand JET-Conoco around [0 to 10]* %. According to the parties' 1997 figures, Aral is market leader with [20 to 30]* %, Shell has [10 to 20]* %, Exxon [10 to 20]* %, DEA [0 to 10]* %, BP/Mobil [0 to 10]* %, Elf [0 to 10]* %, TotalFina [0 to 10]* %, Agip [0 to 10]* % and KPI less than [0 to 10]* %. The price-cutters accounted for [10 to 20]* % (with Hypermarkets [0 to 10]* %, JET Conoco [0 to 10]* % and white pumps [0 to 10]* %). The smaller B-brands (such as Avia, Tamoil ...) accounted for around [0 to 10]* %.

(590) There has been little variation in market shares of A-brands or even B-brands from 1994 to 1998 as can be seen from the following table.

>TABLE>

Competitive dynamics - costs and prices

(591) According to the notifying parties, pump prices vary only marginally in Germany. The price differences are explained by higher operating costs in rural areas and transport costs to a minor extent for remote areas. The market investigation has revealed that the German market is typified by a situation whereby all A-brands charge prices at a similar level whereas the remainder of the competition prices at 1 pfennig below A-brand prices. Supermarkets offer prices at 3 pfennig below A-brands. This price structure does not seem to vary over time.

(592) Aral is the market leader. Price changes seem to be initiated by A-brands according to variable patterns and are followed by the rest of the competition. The majors have most impact on prices. All vertically integrated oil companies (Aral, Shell, Esso, BP and DEA) show similar company structures. Compared with their competitors they have very good access to the purchase and supply markets. Only the five aforementioned companies have a condensed retail network for the whole relevant geographical market. Multiple links exist between these five companies (materialised in JVs for refineries, exchanges, storage and airport fuel supply). This leads to identical or at least comparable purchase costs between A-brands.

(593) Despite their lower price, B-brands and supermarkets tend to follow price trends. It results from the Commission's market investigation that price changes are mainly initiated by the majors and it happens very rarely that one of the majors does not follow. Price wars occasionally occur for a limited period in time, for instance when a supermarket opens a new fuel retail outlet. Competitors have indicated that if the majors would raise the fuel prices about 5 to 10 %, the other companies would rather follow, in order to take advantage of higher margins. The fuel retail industry has understood that not much market share can be gained by competing with the majors in lower prices. In addition, competitors have indicated that due to the low margins and high investment costs (for meeting environmental requirements) they could not sustain a price war.

(594) According to competitors, B-brands are able to offer lower prices because they operate often more efficiently and have a different cost structure than the majors as their overhead costs are most of the time slightly lower (no marketing and advertising costs, smaller shops, less site investment, less personnel) and as they offer more limited service. B-brands sell at a price 1 pfennig less than the majors. However they are usually of very limited size when compared to the A-brands. Companies such as Elf, Agip, Fina, Total and OMV can claim, individually, at most 3 % of market share.

(595) [...]*.

(596) Whereas the majors did not react when Conoco entered the market in the late 1980s, the market investigation reveals that today when a new competitor tries to offer price differences of 1 pfennig and more, there is a localised price war in the neighbourhood. It was explained to the Commission that when two years ago, Conoco tested unmanned service stations at a price of 2 pfennig lower than the majors' price, the majors reacted, and as Conoco needs their acceptance and supplies, it stopped the price war and raised prices.

(597) [...]*. That reduces the potential volume shift from price-cutters to a maximum of two-thirds of their existing volume.

(598) Although some of the small/independent companies offer cheaper pump prices (1 %), this is not sufficient to permit the conclusion that there would be competition on prices. Independent companies are only active in certain regions. They typically have lower throughputs (1600 against 3000 in average for the industry) and therefore higher fixed costs per unit of product sold. They compensate these lower throughputs through lower service to remain competitive. Even if throughputs are increased, these independent companies are not in a position to exercise competition on prices as their variable costs are distinctly higher than those of the majors. In a graph representing overall costs as a function of throughput(64), Exxon has shown that costs remain fairly stable for traditional service stations, as German service stations can be qualified. It is the Bundeskartellamt's assessment that the independents do not and cannot exercise competitive pressure upon the five A-brands companies. This point was confirmed by the Commission's market investigation. The independent retailers seem to have too low margins to start competing on the basis of prices against the majors. Indeed, since most of the time their presence is geographically focused, they would suffer more from a price competition than the A-brands would.

(599) The majors compete with the supermarkets and offer low prices during the week, price reductions on Fridays and higher prices on Sundays, when the supermarkets are closed.

(600) In Germany, hyper- and supermarkets do not have such an important competitive impact on the market as in France, but exert in certain areas a certain price control on the majors. Supermarkets have a rather small market share, at just below [0 to 10]* %. [...]*. Hypermarkets are not a significant competitive threat to the majors for the following reasons:

- hypermarkets offering fuels are highly fragmented and have no nationwide chain,

- their price is typically 3 % to 4 % lower than that of the majors but they cannot offer the same service to customers as they have limited opening hours,

- they have no shop offering on the forecourt, and

- they accept only cash payments.

(601) Contrary to what happened in France (where their entry in the market was contemporaneous) supermarkets are not integrated into the supply chain. They have not captured significant throughputs and do not benefit from any specific cost advantage when compared to A-brands. [...]*(65). Only limited competition is therefore to be expected from the supermarkets.

(602) The majors concentrate on the big cities, whereas independents are more active in rural areas and small/medium-sized towns, where there is less competition. The throughput is higher in towns, but the costs are lower in rural areas, as there are no big investments necessary, personal contact is more important, etc. The fuel price is normally the same or higher in rural areas.

(603) In Germany the "agency system" has helped to avoid price wars, as the majors have absolute control over prices. Majors control the prices, as they are operating with agents, not with independent dealers. The agent has to accept the prices, and sometimes prices are even changed automatically by the computer system.

(604) According to third parties, fuel retailers have knowledge of pump price increases before they are put in practice allowing for sufficient time to change prices at the pump. On the basis of market observations in southern Germany(66), the conclusion was that there is no significant competition between the A-brands. Based on (public domain) spot checks in 1990, 1991, 1993, 1997 and 1998 this study has assessed that price increases were initiated only by the five majors (A-brands) and that companies take the initiative for price increases on the basis of an agreed (although variable) schedule. In a competitive market, competitors would consider waiving the opportunity to increase prices in order to improve their image in the eyes of the consumer.

Barriers to entry - expansion - potential competition

(605) In the western part of Germany, access and expansion are very limited. This is due to property restrictions, numerous planning issues, and specific legislation limiting the opportunities for building new outlets (local authorities). It should also be noted that most of Germany is already highly populated. This makes the creation of new outlets more difficult and more costly. Environmental considerations also play an important role in deterring entry and expansion. Environment legislation is stricter than in most other European countries. The stage II and liquid-tight pavement require significant investment, which is expected to trigger further network reduction. The agency system allowing oil companies to prolong the duration of the dealer operating the service station forms an important barrier for new entrants. Finally, loyalty systems and branding (as representative of the quality of the service) are considered important by the German consumer and make entry even more difficult.

(606) Market entry was easier in the eastern part of Germany, as the ailing Minol network had to be rebuilt and the typically long duration of western German retail contracts (for property and supply) was not hindering new entrants. As majors wanted to attain the same market shares in eastern Germany as those in the west, service stations were oversized (3 to 4 times bigger than the stations in western Germany) and faced low throughputs. At the moment there are too many stations in the east, with too much capacity, and as a result of this most stations face financial problems; especially the white stations. Hypermarkets have market shares in the east of 10 to 12 % but are faced with low throughputs as well and serious competition from supermarkets in former eastern bloc countries.

(607) Overall, the German market is a saturated market where throughputs are low and demand is projected to decrease. 600 service stations have been closed down since 1998 and only 150 new service stations have been opened: this trend is likely to continue in the future.

THE POST-MERGER ENVIRONMENT - IMPACT OF THE OPERATION ON COMPETITION

(608) The impact of the present operation is to be assessed against the above described background, having also regard to the features identified in the section devoted to the structural characteristic of retail motor fuel markets in general.

(609) The merger will result in three of the A-brands to have strong links between them. Overall, Aral, Esso and BP/Mobil will hold together 40 % of the retail market. The two other A-brands, Shell and DEA, will respectively have 14 % and 10 % of the market.

(610) The creation of a link between three of the major competitors in the German motor fuels retail market will lead to material incentives not to compete against each other for these companies. It is unlikely that other A-brands such as Shell or DEA would challenge absence of significant competition on the part of Aral, Esso and BP/Mobil. Indeed, as explained in recitals 591 to 604 above the German market is characterised by the absence of any real challenger to the A-brands and a quasi-immediacy of pricing instructions from the majors. In such a competitive landscape, Shell (which has traditionally focused on quality rather than price) and DEA would rather seek to benefit from higher margins at stable market shares rather than lower margins for the same market shares.

(611) For the same reasons as indicated above it is unlikely that other non-A-brand competitors would challenge such a situation. First, they hold altogether less than 35 % of the market, with no one having more than 5 % market share. They would be more harmed by a price war with the A-brands than the A-brands themselves (which could cross-subsidise thanks to their greater size). Secondly, they depend to a certain extent on the A-brand companies for obtaining their supplies of motor fuels. Thirdly, apart from Jet-Conoco, they bear a higher cost basis than most A-brands.

(612) Therefore, the operation will result in the creation of an oligopolistic dominant position by the A-brands, namely Exxon, BP/Mobil, Aral, Shell and DEA.

LUXEMBOURG

GENERAL OVERVIEW

Refining

(613) As no refineries are located in Luxembourg, the supply of motor fuel is to be secured through importation from neighbouring countries, primarily from the ARA regions via secondary means of transport, essentially rail, barges and trucks. More specifically, a marketer seeking to import products into Luxembourg has two options: arrange for capacity in one of the terminals located in Luxembourg and purchase barge lots from fuel suppliers in the ARA region, or deliver barge lots to other terminals in neighbouring countries and truck products into Luxembourg.

Wholesaling

(614) As most retailers operating in Luxembourg are vertically integrated oil companies, the wholesale activity is very limited. With respect to logistic, no risk of foreclosure exists at the level of terminals available for storing motor fuel. There are essentially four active terminals in Luxembourg, of which the one strategically situated at the riverport of Mertert (approximately 56000 m3), is owned by an independent terminaling company. Owing to the small size of the country, many depots located in neighbouring countries represent competitive alternatives and are currently used by some of the retailers operating in Luxembourg.

Retailing

(615) The market for motor fuel retailing in Luxembourg is characterised by two specific features. First, motor fuel demand is stimulated by the government policy of low petroleum duties, which results in pump prices lower than in neighbouring countries. As a result, volume sales are large, relative to the size of the country. Also, most of these sales concern "non-resident" customers from neighbouring countries. Secondly, motor fuel pump prices are capped by the government, which fixes maximum pump prices. As it will be seen below, the price regulation has favoured the alignment of retailers' pump prices on the maximum ceiling fixed by the government.

THE PRE-MERGER ENVIRONMENT - CURRENT STATUS OF COMPETITION

Luxembourg as a distinct geographic market

(616) While it can be argued that Luxembourg cannot constitute a distinct geographic market because of its geographic location, the specific market conditions existing in this country militate in favour of a separate market. In this respect, it is worth noting that due to a light tax burden, within the Community Luxembourg has by far the lowest after tax prices (USD 936 per Mt). If compared with its neighbouring countries, Luxembourg motor fuel price differences vary from a minimum of approximately 15 % with Germany, up to 40 % with France (see table above).

Market structure

(617) From the supply side the competitive environment is characterised by the presence of a number of vertically integrated oil companies operating retailing activities. Five players hold each more than 10 % of the market, and two other are above 5 %. None of them has significant advantages in terms of access to sources of supply since no refineries are located in Luxembourg. However, due to the small size of the country, and the relative proximity to some important refining regions (ARA), retailers can procure motor fuel and supply their network with no significant extra cost compared to the European average. No hypermarket and supermarkets are active in Luxembourg. Independent retailers have a very marginal position.

(618) The pre-merger situation is as follows. The table below shows an estimate of the market shares (in volume) of the parties and their competitors over the last five to six years.

>TABLE>

(619) The market appears to be moderately concentrated. Shell is the market leader with [20 to 30]* %, followed by Aral at approximately [10 to 20]* %, and then Total which, following the merger with Fina, has increased its market share up to around [10 to 20]* %. Exxon is a mid-size player with a share of [10 to 20]* %. BP has a small share of [0 to 10]* %. The pre-merger CR4 is close to 70 %.

(620) In terms of market shares trends, the market shows a certain evolution at the bottom of the ranking (probably explained by the small quantities in question) while the position of the leading players remains quite stable over the time. More specifically, in a period of five years, Total has doubled its market position (disregarding the merger with Fina) while both BP and Texaco have registered a decrease of roughly [...]* % of their sales. By contrast, the stability of the leading players on the market, namely Shell, Aral Exxon and Q8, is higher. [...]*.

(621) If, then, a shorter temporal horizon is taken as a reference (three or four years), the static nature of the market appears all the more accentuated. With the exception of Total, which has recently merged with Fina, market share variations are absolutely insignificant.

(622) It is also worth measuring the market position of the various players by reference to the number of station sites.

>TABLE>

(623) Based on 1997 figures, it appears that three operators have a comparable number of sites, namely Shell, Aral and Q8 with 52, 50 and 48 stations respectively. Total/Fina and Exxon have 34 and 30 sites respectively. BP owns only three sites. In terms of evolution over the time, again the stability is quite pronounced. Among the leading players, the most important variation concerns Q8 which registers the highest decrease, with a reduction of eight stations in five years. It is followed by Exxon which has reduced of five stations its total numbers over a period of five years.

(624) It is also striking that the overall number of sites is extremely high given the size of the country. As it will be seen below, such a concentration of stations in such a small territory seems to hint that further expansion in terms of new sites seems to be unlikely.

(625) Overall, from the above considerations it appears that the last years the market for motor fuel retailing in Luxembourg has remained quite static with seven main players (Shell, Aral, Exxon, Q8, Texaco, TotalFina and BP/Mobil) and no dramatic changes have been recorded. The demand has remained quite flat over five years. The competitors have more or less maintained their position.

Competitive dynamics - costs and prices

Costs

(626) A major factor in the profitability of a retail network is the throughput. The higher the throughput, the lower the sites' average operating costs. In order thus to make a costs comparison of the players on the market, the chart below shows the average throughput per site of the various players in the last five years.

>TABLE>

(627) Taking as a reference the year 1997, for which complete data are available, it should be noted that BP has by far the highest throughput. However this is due to the fact that it owns just three very big sites. As to the other retailers, Total has the highest throughput with 6609 m3 per station (however this throughput refers to the situation before the merger with Fina), followed by Exxon with 5057 m3 (6795 m3 in 1998), then Texaco and Shell with a comparable average throughput of 4900 m3. Q8 has the lowest average with 2542 m3 per site.

(628) However, in general terms it can be said that the average throughput of the industry is very high in Luxembourg when compared to the average of other countries.

>TABLE>

(629) An average throughput of more than 3000/3500 m3 per site is viewed by the industry as allowing a retailer to reach a minimum efficient scale so to cover his fixed costs. In this respect, an internal study conducted by the parties shows that an ordinary service station achieves efficient scale precisely if it generates a throughput of around [...]* m3 and realizes a gross margin of [...]* cents per litre. On this basis, it appears that none of the leading operators has over the others a significant advantage in terms of operating costs per site.

Prices

(630) In order to fully appreciate the competitive environment existing in motor fuel retailing in Luxembourg, it is worth comparing pre-tax price trends in the countries affected by the operation.

Product: Retail motor gasoline - pump price

>TABLE>

Product: Retail diesel - pump price

>TABLE>

(631) As may be seen from the table above, pre-tax prices in Luxembourg have gone down over the last three years (from 401 USD per MT in 1996, to 321 USD per MT in 1998), but this is essentially due to the price decline of crude oil. In general terms, pre-tax prices are in the average band of Community prices.

(632) It is also worth comparing gross margins in the countries affected by the operation. Exxon gross margins are taken as a proxy of the industry average gross margins.

>TABLE>

(633) As it can be seen from this table, while not reaching the peak of the Netherlands and Austria, gross margins in Luxembourg are nonetheless very high by comparison with the French margins which have been described by the parties as among the most constrained by competition in Europe. In percentage terms, Luxembourg margins are around [...]* than the French margins. No special extra costs related to some peculiarity of the Luxembourg territory can explain these differences. The table below shows a comparison of Exxon transport and delivery costs as well as Exxon "other wholesale costs" in the abovementioned countries.

>TABLE>

(634) From the table above it appears that transport and delivery costs are well in the European average. On the other hand, the "other wholesale costs" are substantially higher in Luxembourg than in France. However, these differences still do not explain the whole differential existing at the level of the gross margins between Luxembourg and France. More specifically, even deducting these extra costs, Luxembourg margins still remain around [...]* higher than gross margins in France. Given that Luxembourg service stations have higher average throughputs (and hence probably lower average costs) than French service stations, and in view of the product homogeneity of motor fuels, such high margins would suggest a substantial deficit of competition on the market.

Specific features of the Luxembourg market - price regulation

(635) A specific feature of the market in Luxembourg is the fact that the government sets a maximum price per grade of motor fuels. More specifically, it is the Ministry of Economic Affairs which operates a system of maximum wholesale and retail pump prices. In concept, the maximum price formula is based on four main components: (i) cargo market quotations; (ii) a base (wholesale) margin for the supplying oil company; (iii) an additional compensation for various costs (rail transport costs, compulsory storage costs, and environmental investments); and (iv) a dealer margin.

(636) Most price changes take place as a result of changes in cargo market prices, and do not result in any change in the marginal components within the formula. The official (wholesale) margin, based on the prevailing actual maximum prices and daily cargo market prices, is calculated daily by the Ministry (and the industry). These calculations include excise taxes and VAT. If there has been a rise in cargo market prices since the prevailing maximum prices were established, the actual margins available to the industry will fall below the defined base margins, whereas if cargo market prices have fallen, the converse is true and actual margins will be above the base margin levels. Foreign exchange rate fluctuations will also cause changes in actual versus base margins, since the maximum prices are determined in local currency. It is these variances that trigger changes in the maximum prices. The Ministry works on the basis that changes to the maximum prices are justified if the differential between the maximum and actual (calculated) margins for a particular product exceeds LUF 0,25 per litre for 10 consecutive days or exceeds LUF 0,4 per litre for more than two consecutive days. However, changes to the maximum prices are always subject to government approval, which means that they do not immediately follow such periods of price deviation. The industry - and the Ministry - track the amounts of under- and over-recovery (caused by cargo market and exchange rate fluctuation) versus the base margins. Under-recovery is typically compensated by a deliberate delay in the reduction of maximum prices when cargo market prices fall, and over-recovery is offset by delays in implementing maximum price increases when cargo market prices rise.

(637) Also, an adjustment in the price formula can occur at the request of the industry in order to cover increases in costs. In that case, an application for price-margin increases is to be made by the local petroleum industry association. However, price changes arising from or including an increase in margin are rare compared to those reflecting cargo market and exchange rate changes. In this respect, it is worth noting that over the past 12 months up to June 1999, there have been 12 maximum-price changes for motor gasoline and 11 maximum-price changes for diesel. The last margin changes took place in August 1998.

(638) While it is undisputed that the government intervention in fixing maximum price is a distorting factor in the proper functioning of the market, on the other hand it does not prevent retailers from competing against each other by pricing below the government price ceiling. However, as a matter of fact, most competitors apply the maximum sale price determined by the government, with very minor discounts granted in some rare cases.

(639) In essence, the retailers align their prices on the government price and hardly deviate from that price. From the investigation conducted by the Commission it appears that among all fuel retailers, few operators are currently selling motor fuel at a discount price: some white pump operators are currently pricing at between 0,1 and 0,4 F/litre below the maximum prices, and some branded stations, namely Total and Exxon, are currently pricing at 0,1 F/litre below the maxima.

(640) In short, at the moment, the competitive game seems not to be properly functioning. There is in practice very little competition between the seven main motor fuel retailers which systematically adapt their own pump prices to the government maximum price.

Entry barriers - expansion - potential competition

(641) Beside the entry barriers common to the sector, the major impediment for a newcomer to enter the Luxembourg market for motor fuel retailing is the saturation of the market in terms of number of station sites. This greatly reduces the possibility of creating new outlets.

(642) As to the possible entry on the Luxembourg market of aggressive competitors, it should be noted that there are no hypermarkets forecourts in Luxembourg. More importantly, the barriers to their possible future entry appear insurmountable. First, both the national and local governments of Luxembourg have very stringent permitting laws and regulations. Secondly, the access to commercial property is limited because of the small size of the country. Thirdly, available commercial properties are very expensive. Fourthly, Luxembourg is too small in size to sustain a significant growth of supermarkets. Finally, the demand for motor fuel in Luxembourg is constituted by the transit traffic of foreigners attracted by low prices. It is clear that this demand cannot be captured by hypermarkets selling motor fuel.

THE POST-MERGER ENVIRONMENT- THE IMPACT OF THE OPERATION ON COMPETITION

(643) The impact of the present operation needs to be assessed against the background described above, regard being had to the features identified in the section devoted to the structural characteristic of retail motor fuel markets in general.

(644) In the current situation, and despite the presence of seven main players, the performance of the market for motor fuel retailing in Luxembourg in terms of price competition is very poor. Because of the presence of a price cap fixed by the government, most motor fuel retailers align their pump prices on that ceiling and avoid competing against each other. The margins secured through the government formula are generous and allow also the most inefficient retailers to stay on the market without suffering the competition. As a matter of fact, very few players on the market deviate from the government price which functions like a cartel benchmark. Indeed, pump prices hardly vary by virtue of the initiative of one or the other player, but only further to the adjustments decided by the government.

(645) However, while it is undisputed that the competitive dynamics are distorted by the existence of the government regulation fixing the maximum possible pump price, it is a fact that competition could well function below the price cap. Instead, motor fuel retailers prefer avoiding any price competition. In sum, to date the market seems to be characterised by a substantial deficit of competition.

(646) While as a direct result of the operation the combined entity will jointly control a small-size player (Exxon [10 to 20]* % and BP/Mobil [0 to 10]* %), account should however be taken of the link existing between Mobil and Aral, which is the competitor number two in Luxembourg with [10 to 20]* % of the market.

(647) Based on this approach, the operation will then substantially increase the degree of concentration of the market, since the trio constituted by Exxon/Mobil, BP/Mobil and Aral will account for roughly [30 to 40]* % of the market.

(648) More importantly, the operation will change the structure of the oligopoly as it will result in the creation of a new aggregation of players having a considerable market power. As a consequence, the market will be characterised by the presence of a leading aggregation of players, namely Exxon-BP/Mobil-Aral, a second force, Shell, plus a number of "followers". In this respect, the operation will result in the existing oligopoly being significantly tightened so that any possibility of competition between the various players on the market will become even more remote. After the merger, the market structure will be such that no competitive move will be ever conceivable without triggering a strong and widespread reaction. For instance, if the government were to raise or even eliminate its current price ceiling, the operation would make it even more unlikely for a market participant to adopt a more competitive stance, as such action would be even more likely to result in a strong and widespread reaction. Similarly, any potential entrant would face stronger and more widespread hostility from the current market players.

(649) In the light of the above, it appears that the operation will result in the creation or the strengthening of a collective dominant position held by Shell, Aral, Exxon, Q8, Texaco, TotalFina and BP/Mobil as a result of which effective competition would be significantly impeded in a substantial part of the common market.

NETHERLANDS

GENERAL OVERVIEW

(650) The Netherlands is within the EU one of the countries mostly affected by the operation as far as motor fuels retailing is concerned. Two categories of players are present, the integrated oil majors (Shell, Exxon, BP/Mobil, Texaco, TotalFina and Q8) and non-integrated independent, branded or unbranded retailers. The retail market is characterised by a high degree of concentration, with the oil majors accounting for more than [80 to 90]* % of the volumes traded, the remaining being shared among the independent operators. As opposed to other national markets, notably the United Kingdom and France, there are currently no supermarkets active in fuels retailing in the Netherlands. The wholesale supply of fuels is organised locally, with both integrated and non-integrated retailers supplying their retail networks from one of the five local refineries (Shell, Exxon, Total, BP-Texaco and Q8). The number of service stations has been decreasing over the recent years (3980 service stations in 1997, making an average throughput of 2502 m3, which is one of the lowest among the relevant markets examined). The pump prices are among the highest. The planning and permissions policy of the government is one of the most restrictive among the relevant markets considered, making prospects for expansion of the retail network look poor. Due to the topography of the country (i.e., soft absorbent soil), the environmental legislation is one of the most requiring in the EU, having as a consequence an increase in operating costs.

Refining

(651) The major oil companies (Shell, Exxon, BP/Mobil, Texaco, TotalFina and Q8) are the main sources of supply of motor fuels. They operate the five refineries in the Netherlands, as can be seen in the following table. There is a sixth refinery in Amsterdam, owned by Smith & Hollander, which, however, does not produce motor fuels, but mainly asphalt products.

Refining capacity in the Netherlands (1997)

>TABLE>

Roland Berger & Partners GmbH, 1997.

Wholesaling

(652) The Netherlands is a net exporter of refined petroleum products. Motor fuels intended to supply the retail network are locally sourced from one of the above refineries. The major oil companies are therefore the main wholesale suppliers of fuels. They trade their products either through their own marketing organisations or through independent operators, who then resell to the retail market.

Retailing

(653) At the retail level, there are two categories of operators, that is, all of the above integrated major oil companies, on the one hand, and a number of non-integrated, independent retailers, on the other hand. As mentioned in the preceding paragraph, independent retailers source practically all of their motor fuels from one or various of the above refiners.

(654) Hypermarkets are not an active force in fuel retailing. This is mainly due to the stringent environmental policy imposed by the government on service stations and to the fact that, in general, supermarkets in the Netherlands do not have large forecourts to accommodate service stations, as the case may be in France or in the United Kingdom. Reportedly, supermarkets used to be active in fuels retailing in the Netherlands, until 15 years ago. They were facing, however, increasing environmental cost and burden, and decided to leave the market (Exxon, for instance, bought 15 stations from Miro, a supermarket chain, and closed them down three years later). Under the Dutch permissions policy, once a service station has closed, its owner loses his licence. Accordingly, those who exited the market in the past years are unlikely, under the current governmental attribution system, to re-enter it.

THE PRE-MERGER ENVIRONMENT - CURRENT STATUS OF COMPETITION

Market structure

(655) The characteristics of the market are conducive to an oligopolistic structure among the integrated majors. Despite the asymmetry of market positions, the stability of market shares and the outstanding shared characteristics of the majors (vertical integration, similar refining, wholesaling and retailing cost structures, similar return on capital expenditure), reflect a market that is functioning at margin and price levels independent of the competitive dynamics of the ARA area.

(656) Shell is the leading player and national "champion". Reportedly, its predominance in the Dutch market is due to historical reasons associated with the discovery, jointly with Exxon, of the Groningen natural gas field. The two companies were encouraged by the government to reinvest in the Netherlands the revenues from the gas discovered. Shell reinvested those revenues in the downstream sector, building refining capacity and an extensive distribution network(67). Shell's market share in the retail market in 1998 was [30 to 40]* %.

(657) Besides Shell, the next larger player is BP/Mobil which accounts for [10 to 20]* % of the retail market. It is followed by Texaco, Exxon and TotalFina with market shares in the range of [10 to 20]* % and by Q8 with a smaller presence at retail level. Those companies account for some [80 to 90]* % of the market, leaving thus the remaining part to a number of independent retailers and white pumps, whose basic characteristic, differentiating them from the majors, is their absence of vertical integration, and their consequent supply dependence on the majors.

Retail market shares 1998

>TABLE>

Parties' data.

(658) The stability of market shares over the last years reflects a lack of intense competition and an absence of market entries. Overall, no noticeable fluctuations were recorded in the relative competitive positions of the various competitors. The following table shows the movement of market shares at the retail level over the last five years.

Movement of retail market shares

>TABLE>

Parties' data.

(659) The above market positions are also reflected in the distribution of outlets across Netherlands territory. The retail market is dominated by the major refiners, accounting for more than 75 % of the existing service stations. Although the number of outlets has decreased over the last years, as a result of environmental rationalisation and restrictive planning policies, the majors have been able to further develop their network to the detriment of independent retailers.

Number of service stations in the Netherlands

>TABLE>

Parties' data.

Competitive dynamics - costs and prices

Costs

(660) Furthermore, not only do the major refiners drive the largest retail volumes of fuels, but also they do so at the most cost effective way, by realising the largest volume per site ratios. The following table shows the average throughputs realised by the majors and the independents in the Netherlands.

Average throughput (m3)

>TABLE>

(661) As of 1997, the average throughput of the majors was almost twice as large as the average throughput of the independents. As a result of the concentration, the differential will increase further. This difference in throughput is explained by two main reasons: the larger size of the outlets operated under the majors' brands (and almost exclusively the large outlets located along motorways); and the prime location of the majority of their outlets. According to a report commissioned by the Netherlands Government, the major oil companies have been able, thanks to their financial surface, to build and operate larger outlets located at prime locations. Moreover, they have been able, historically, to establish an extensive network of high-throughput outlets along the Dutch motorways. Until 1994, concessions were assigned to the major oil companies in accordance to their existing market shares. As this permission policy proved to be unjustifiably restrictive, it was abandoned and replaced by a policy whereby new concessions were given to the highest bidder for an unlimited period of time.

(662) A higher throughput reduces the cost basis of the retailer. In this sense, it goes without saying that the majors may expect to have better profit margins than independent retailers. Even though their absolute gross margins may be similar, their relative costs are different, with the majors being able to spread them across a larger turnover basis. This situation favours the creation of economies of scale in the operation of the majors' retail network; their profit margins increase as their turnover rises. The Netherlands Government report and the market investigation indicated that, within the current competitive environment, whereby no new massive concessions are expected to be attributed, economies of scale may be a prerequisite for a player to remain viable in the market. The shrinking network of the independents may be, to a large degree, explained by the absence of economies of scale of smaller networks and outlets.

Prices

(663) Prices in the Netherlands are high compared to the fuel economics of the ARA range. Although ex-refinery prices are aligned to Platt's levels, wholesale and retail margins and prices appear to be disproportionately higher. This has resulted in higher pump prices in comparison to neighbouring markets within the same supply environment. For instance, the last five years, pump prices were up to 40 % higher than in France which has been described by the parties as among the most constrained by competition in Europe. The following table shows that the 1998 wholesale and pump prices are higher in the Netherlands than in other neighbouring markets aligned to the Platt's ARA quotations.

>TABLE>

(664) This difference may not be attributable to any higher excise duties and VAT applicable in the Netherlands (taxes are more or less the same or even slightly lower in the Netherlands), but mainly to the substantially higher gross distribution margins, which more than doubled the French ones over the same period. The table below indicates the variation of the gross margin (wholesale price vs retail price, tax excluded) in four neighbouring geographic markets aligned to the Platt's ARA quotations.

Average gross margin

>TABLE>

Extrapolated from parties' data Annex 7.7.

(665) Nor can these substantial differences be attributed to higher costs. The major components of distribution costs are terminaling and transportation. In the Netherlands, the sources of supply are numerous and well located; the five local refineries supply practically all of the retail market. The distances to be travelled are relatively small, making direct distribution from the source of supply to sales points a matter of several hours. Storage and other logistic costs are therefore substantially reduced, whereas transportation costs are similar to the other benchmark markets (barge: 1 % of Mogas non-retail price; rail: 2 %; pipeline: 1 % and truck: 3 %).

(666) In their response to the statement of objections, the parties contend that the higher distribution margins in the Netherlands are largely attributable to high retail margins which accrue to service stations operators rather than the supplying oil companies. In addition, in the Netherlands fuel retailers bear also additional extra costs due to the loyalty schemes and the strict environmental regulations.

(667) While it is undisputed that dealer margins in the Netherlands are among the highest in Europe, as noted above, it is a fact that also wholesale prices and margins are inexplicably high in the Netherlands, in particular given the structural characteristics of the region. The table below shows a comparison between Exxon gross margins less dealer margins in the Netherlands and France.

>TABLE>

(668) It appears that Dutch wholesale margins remain in the range of [...]* higher than French margins. These differentials can hardly be explained by the other extra costs claimed by the parties (fidelity schemes and environmental regulations), which, incidentally, should be mostly covered by the dealer owning the site.

(669) In sum, the higher price environment in the Netherlands shows a deficit of competition. The core of the Dutch retail market may appear to be unrelated to price trends prevailing in the neighbouring markets. Prices are structurally higher across the territory, except for some frontier areas in direct competition with a neighbouring market. This is the case of the area along the Netherlands-German frontier, where lower German prices have influenced the pricing of the Dutch outlets. It is quite characteristic that in order to prevent a deterioration of the margins in this area, the Netherlands government decided to establish a compensation system to sustain independent retailers.

Barriers to entry - expansion - potential competition

Entry barrier

(670) The major barriers to entry and/or expansion in the Dutch retail market are linked to the stringent environmental policy and the consequent restrictive concessions policy of the government. Part of the reduction in the number of service stations was due, according to the Netherlands Government Report, to incentives to close given by the government to service stations that did not comply with the requiring environmental legislation (soil contamination prevention). Operating licences are issued by city and provincial authorities. In general, such authorities do not make available new locations for service stations, considering that the market is already saturated. When they do so, they impose serious requirements in terms of safety and environmental protection. In doing so, they appear to be more keen on giving new permissions to the oil majors rather than to independents or to white pumps. Reportedly, an independent applicant may have to wait up to four years for a new licence, as opposed to oil majors which can obtain a licence much quicker, in particular when side-infrastructure work is part of the package bid. Moreover, given that tenders for new licences, or for the replacement of old ones, are attributed to the highest bidder, the oil majors are often in the best position to expand their network.

Potential competition

(671) The competitive potential of the independent retailers is substantially reduced. Currently, the market appears to be functioning at a supracompetitive price equilibrium, with both the integrated and the non-integrated retailers being winners. Major oil companies are active at both levels of the supply chain, that is, as fuels wholesale suppliers and retailers. Independents are only active as retailers and obtain almost all of their fuels requirements from the local majors, in most cases through exclusive yearly renewable supply agreements. The revenues of the integrated oil majors may be realised at various stages of the supply chain, whereas those of retailers only at retail level; gross wholesale margins accrue to the oil majors, in contrast to retail margins, accruing to both the majors and the independents.

(672) It is clear that the integrated oil majors have a tremendous financial advantage over the independents. They have a higher degree of flexibility in case of price competition. Reportedly, the oil majors may offer loss compensation schemes to their retail network, in case of a local price war. Their financial availability and the fact that they hold the sources of supply, has made it possible to the larger oil majors, in particular Shell, to profile themselves as price leaders. The market investigation has shown that when retail prices increase, independent retailers follow the price rise, considering it more attractive to increase their profit margin through higher prices, than to sell more volume at a lower margin. Since most outlets are already functioning at peak throughput capacity, any increase in throughput would involve an expansion of the filling stations or of the retail network. This is made difficult by (i) the high incremental cost linked to an expansion of filling stations and (ii) the fact that the possibilities for an extension of the retail network are limited, for the reasons stated in recitals 670 to 671.

(673) The likelihood of a price war initiated by independents is substantially reduced for a series of reasons. The independents might be able to increase the volume traded, however this can only be done up to a certain ceiling. If they decide to engage in a price war, they may put in jeopardy their future supply contractual arrangements with the oil majors. A price war from their side could be of limited duration, as the oil majors will be able to match the lower prices, without this affecting their financial situation. In the longer run, independents will have to turn to other sources of supply, such as the cargo market. However, the investigation has shown that, in general, independent retailers are reluctant to do so. Indeed, cargo market prices are not necessarily more competitive than prices at the local refineries. This is more so, considering that in order to achieve better prices at the cargo market, the volumes purchased need to be substantial. In the Netherlands, the average throughput of independent retailers is in general lower than in other neighbouring countries, where larger independents and supermarkets obtain supplies from the cargo market (such as in France or in Germany). Most independents require not more than 4000 litres per site per week, whereas the cargo market prices are attractive when the volumes purchased are above 10000 litres. This would compel independents to team up so as to pull joint purchases together; additionally, handling and storage costs would consequently eliminate any possible price advantages. Moreover, independent retailers have stated that the quality of the products may not be guaranteed when purchases are made from the cargo market. In general, they chose to buy products from the local refineries in order to ensure product quality and to get other advantages from being loyal customers to these refineries. As a result, independent retailers have become price followers for as long as their retail margins are not squeezed, in which case they might consider exiting the market, usually by selling their outlets to one of the majors.

(674) The low prospects of potential competition from the independents may be well illustrated in the following case study, quoted from the Netherlands Government Report. The local authorities of the city of Almere persuaded a white pump to enter the local market (by facilitating the procedure of obtaining a concession) in order to break the local oligopoly of the existing service stations. In return for the concession, the white pump committed itself to offering a retail price 4 % lower than the immediate competitors.

(675) The white pump was installed in the vicinity of three existing service stations. Another six service stations were active further away but still within the borders of Almere. All existing service stations were under the majors' brands.

(676) The first year, the white pump achieved an adequate turnover and so did the three service stations which had followed the newcomer in reducing prices. The volume of sales in Almere had not increased during that period and consequently the six service stations which did not follow the price reductions lost market share. After a year, all nine service stations offered the same price as the white pump.

(677) After the first year, the white pump increased its pressure and reduced its price by 7 %, 9 % and for a time even 12 %. At each price decrease, the competition followed swiftly. And also when the white pump increased its price, every service station went along with it. In other words, the white pump was transformed from a newcomer into a price-leader.

(678) In this case, since all service stations offered the same price, attracting extra turnover for the white pump was only possible through the discounter's goodwill which it had cultivated in this local market. Despite the fact that the white pump was appreciated for being a price fighter, this goodwill effect faded away after a certain period of time. When the white pump retailer reduced its price by 12 %, it was barely able to break even. The service stations of the major oil companies, on the other hand, were able to follow the price-leader swiftly since they received compensation from the oil companies for the reduced distribution margin. Thus the white pump was not in competition with the local service stations, but with the oil companies. At that moment, the price of fuel in Almere had returned to the level at which the white pump was introduced. At that same moment, the white pump decided to sell the operation to the highest bidding major oil company.

(679) In the light of the above, it can be concluded that the potential competition from the independent retailers cannot exert a substantial pressure on the major oil companies.

THE POST-MERGER ENVIRONMENT - IMPACT OF THE OPERATION ON COMPETITION

(680) The impact of the operation is to be assessed against that described background, regard being had to the features identified in the section devoted to the structural characteristic of retail motor fuel markets in general.

(681) As a result of the operation, the oligopolistic structure of the market will be further reinforced, in presence of three competitors, accounting jointly for more than half of the market, a cluster of two comparable mid-sized players, that is to say Texaco and TotalFina. The rest of the market will remain scattered among small fringe independent retailing companies.

(682) The notified operation will thus lead to a further concentration among the oil majors. The current oligopolistic environment will be further tightened up, thus making interaction among the players easier to implement and to monitor. In a post-merger scenario, the poor economic performance of the Dutch retail market will be further exacerbated. The independent retailers have no incentives to fight on prices, as they can, on the contrary, benefit from price increases by the majors at retail level. Overall, under such circumstances, the final consumer will be worse off. It can be concluded that the concentration will lead to the creation or reinforcement of a collective dominant position in the retail market, held by the major oil companies (Shell, BP/Mobil, Exxon, Texaco and TotalFina).

THE UNITED KINGDOM

GENERAL OVERVIEW

Refining

(683) The United Kingdom is a net exporter of refined petroleum products. Traditionally more than 85 % of petroleum products are directly sourced from inland refineries. Imports (usually shipped from continental Europe or Scandinavia) are generally directed to independent resellers, but they can also result from refiners' intra-group trading.

(684) There are currently 10 major fuels refineries in the United Kingdom. Most of these are located in the southern half of Britain, and are connected to the main pipelines covering most of England and Wales. They are also often accessible by ship. It should be noted, however, that Scotland has only one refinery (BP/Mobil's Grangemouth site), which serves most of the local distribution terminals.

(685) The main refiners in the United Kingdom are BP/Mobil, Exxon and Shell, who own approximately 57 % of total capacity. However, most other refiners (such as Elf, Total/Fina, Texaco and Conoco) also are vertically integrated oil companies with retail activities in the United Kingdom, and individually often account for more than 10 % of total domestic refining capacity.

Wholesaling

(686) Imported or nationally refined motor fuels are then either sold on the cargo market and exported, or transported to retailers, end-users or resellers. In some cases, petrol and diesel are collected directly at the refinery gate or the import storage facilities and distributed by road. However, in most cases, motor fuels are transported to distribution terminals, which are located throughout the United Kingdom and are supplied by inland pipeline network, sea or (less often) rail. Motor fuels are then collected by road tankers and distributed to nearby customers.

(687) In England and Wales, most motor fuels are transported by one of the six pipelines connecting refineries to distribution terminals. In Scotland, there is no such network, and most products are shipped from BP/Mobil's Grangemouth refinery.

(688) The non-retail sector in the United Kingdom appears to be competitive at present. This is firstly indicated by the fact that, at present, no company appears to hold more than 20 % of non-retail sales of motor fuels to third parties. This is also confirmed by the fact that most of the companies which replied to the Commission enquiry indicated that the non-retail prices of motor fuels generally reflect the international Platt's index and the transportation costs from the refinery or the import terminal concerned, and that they had no difficulty in getting supplies.

(689) The Office of Fair Trading (OFT) reached similar conclusions in its report on competition in the United Kingdom petrol sector(68). It even found that, because of these competitive conditions, "mini majors" have tended to return wholesaling losses in almost all of the last five years.

(690) One of the most important factors for such competition appears to be the chronic excess refining capacity in the United Kingdom, as total refinery output for petroleum products is almost 20 % higher than the domestic demand, which seems to have levelled off since 1990. Although this gap is expected to be somewhat bridged after the closure of Shell's Shellhaven facility (the smallest refinery in the United Kingdom), it will still represent approximately 15 % of total demand.

(691) Another explanation is that a significant number of independent companies own and/or operate distribution terminals, from which they can import products from the cargo market and supply products to non-refiner retailers (such as hypermarkets) or unbranded retailers. Their presence ensures that large end-users customers and non-refiner retailers can readily find supplies at competitive prices.

THE PRE-MERGER ENVIRONMENT - CURRENT STATUS OF COMPETITION

Market structure

(692) The United Kingdom market is characterised by the presence of a substantial number of competitors: in addition to the three main refiners (Exxon, BP/Mobil and Shell) which still collectively hold more than 50 % of the market, a number of other oil companies (Total/Fina, Elf, Texaco and Conoco) are also active in the market, as well as hypermarkets and independent retailers.

(693) The evolution of market shares (in terms of volumes) between 1994 and 1998 can be summarised as follows:

>TABLE>

(694) The retail sector may be broadly divided into four main tiers: (i) the "majors" (Exxon, BP/Mobil and Shell); (ii) the "mini majors" (Total/Fina, Texaco, Conoco/Jet and Elf); (iii) the hypermarkets (Tesco, Sainsbury's, Safeway and Asda), and (iv) the independent retailers.

(695) Exxon, BP/Mobil and Shell ("the majors") are the three main retailers in the United Kingdom, with market shares (in terms of volumes) above [10 to 20]* % each (and rising). They have an extensive geographic coverage, as indicated by their high number of sites (more than 1800 each).

(696) They achieve relatively low costs through (i) low supply costs (due to direct supplies from in-house refineries or swaps with other refiners), (ii) high average throughput per site (approximately [...]* m3 and over) and (iii) forecourt shops, which yield additional revenues and profits.

(697) Their market shares (in terms of volumes) have slightly increased since 1994, and currently reach a total of [50 to 60]* %.

(698) Total/Fina, Texaco, Conoco/Jet and Elf also are vertically integrated oil companies with refineries in the United Kingdom. They do not belong to the same category as the super majors, as they have substantially lower market shares and more limited presence (fewer than 1000 sites for all companies except Texaco).

(699) Although they also obtain supplies directly from in-house refineries or swaps with other refiners, they would appear to face higher supply costs than the super majors, as they have returned wholesaling losses in almost all of the last five years. Finally, most of these companies achieve substantially lower throughputs (approximately 2000 m3 and less) than the super majors and would therefore bear higher retail costs.

(700) Finally, unlike the super majors, their collective market shares appear to have declined by more than 23 % since 1994, and currently amount to a total of approximately 20 %.

(701) The hypermarkets (Tesco, Sainsbury's, Safeway and Asda) sell motor fuels from filling stations established on the car parks of their premises. They are mainly located in urban or densely populated areas, and therefore have a limited geographic coverage, as is further confirmed by the relatively low number of filling stations each hypermarket brand operates (below 200 sites).

(702) Hypermarkets get most of their supplies from refiners (and, for some of them, mainly BP/Mobil). However, they usually procure large volumes of motor fuels and therefore also source from traders or importers, which can represent up to approximately 35 to 40 % of their supplies.

(703) Hypermarkets' filling stations generally achieve much larger throughputs (typically 10000 m3) than sites owned by other retailers. This gave them substantial cost advantages over all other retailers before the super majors reduced the number of filling stations they operated in order to reach higher throughputs.

(704) Hypermarkets have gained considerable market shares over the last 10 years. They currently amount to a total of [20 to 30]* %, and Tesco has become the fourth largest fuels retailer in the United Kingdom. According to the OFT(69), this dramatic growth was primarily underpinned by their sustained and substantial price advantage (between 2,5 and 4 pence per litre) over their competitors. It has also obviously been fuelled by the strong expansion (over 10 % per annum between 1994 and 1997) of the number of filling stations established. However, as will be further explained below, there are indications that, due to capacity constraints, barriers to expansion and rising costs, the hypermarkets' market shares should not increase substantially in the incoming years.

(705) Independent filling stations usually are "historical" sites which were set up a long time ago. They usually do not sell fuels only, but also other products (such as heating oil, groceries etc.).

(706) Although they apparently find supplies easily, independent retailers are not vertically integrated and therefore usually bear higher supply costs. The results of the Commission enquiry suggest that their retail costs are also higher than those of other retailers, mainly because of their low throughputs (typically 500 m3). A number of such sites are located in rural areas, where they can charge higher (by up to 3 to 4 pence per litre) prices in exchange for additional service or convenience. For instance, most of these sites also operate village shops, post offices etc.

(707) Their aggregate market shares are now below 6 % and keep declining.

Competitive dynamics - costs and prices

A historical perspective

(708) There are indications that, prior to 1995, the motor fuels retail sector was substantially less than competitive: [...]*.

(709) In late 1994 or early 1995, the majors apparently realised that the growth of the hypermarkets was a serious threat to their business, and could potentially lead to the same market structure as in France, where hypermarkets hold approximately 50 % of market shares and prices before tax are low. Some of them, and especially Exxon, decided to try to raise the hypermarkets' costs and lower their own costs by drawing volumes away from the hypermarkets and back to the majors. In particular, Exxon engaged in both an intense price war with the majors (known as the Pricewatch campaign) and drastic cost reduction schemes.

(710) This caused the prices to drop considerably and to reach such levels that all retailers apparently reported losses in the first half of 1996. Price rises were then attempted, which only began to deliver results if they were supported by Exxon. They have appeared to rise since that period, and even to exceed the levels historically priced by the hypermarkets for the same cargo market prices.

(711) Finally, during the same period, the majors [...]* tried to bridge their cost differentials with hypermarkets [...]*.

(712) Up to 1995, the market did not appear to be fully competitive. This is indicated by the fact that, in early 1994, pre-tax petrol retail prices were 20 % higher than in France, before dropping by more than 30 % within two years [...]*.

(713) In that context, hypermarkets were apparently not taken very seriously, and, as the OFT indicates: "up to 1995, the supermarkets were left largely unchallenged to operate at prices significantly lower than those being offered in the rest of the market. (...) Oil company attempts to respond to the supermarket threat through (price support schemes) to those sites most badly affected were unsystematic, typically at a local level, and without major publicity".

(714) [...]*

(715) In late 1994, the hypermarkets' shares of volumes reached approximately 17 %. As the OFT indicates "the continuous loss of volume to the rapidly expanding supermarket network prompted the major oil companies into a more severe reaction than had been seen hitherto".

(716) [...]*. In that context, Exxon feared that "Esso retail's market position and profitability (would) erode substantially over the next 5 to 10 years".

(717) This resulted in a strong reaction with two different arms. The first of these arms was a severe price competition, which was expected to draw volume back to Exxon and away from the hypermarkets and hence bridge some of the gap between the costs of Exxon and those of the hypermarkets. It also aimed to prevent further entry or expansion by the hypermarkets. The second of these arms was a drastic cost reduction scheme.

The Pricewatch campaign

(718) The first retailer to move was BP who, in late 1994, announced that it would be pricing nationally in response to the supermarkets. It was soon to be followed by Exxon who, in autumn 1995, tried out its "Pricewatch" campaign in Scotland and north-east England, and launched it nationally in January 1996. Shell also declared that it would match Exxon's offer in all Pricewatch areas.

(719) The Pricewatch campaign consisted in matching the hypermarkets' pricing levels and communicating on this scheme, so as to draw volume back to Exxon's sites and away from the hypermarkets' filling stations. The other majors followed the same approach, even though to a lesser extent.

(720) In the first half of 1996, the effects of such price competition appear to have been devastating on the retailers' margins. [...]*. Consequently, gross margins dropped and even became 50 % lower than in France. At such levels, no motor fuels retailing operation appeared to break even.

"Price restorations"

(721) These losses appear to have prompted attempts by either hypermarkets, BP or Shell to lead price increases. Most of these early attempts failed, [...]*. This movement was continued in 1997, when retail prices reached the levels which would be traditionally charged by hypermarkets given the cargo market prices.

(722) However, and contrary to the pre-Pricewatch situation, this movement even continued in 1998, when hypermarkets apparently accepted that retail prices should rise above the levels which would have traditionally been charged before Pricewatch.

Cost movements

(723) In 1994 to 1995, the cost differential between the hypermarkets and the majors was apparently in the order of GBP 0,02 to 0,03 per litre in the favour of hypermarkets. In this context, any price competition from the majors would probably have failed without an additional cost-oriented scheme.

(724) Consequently, and in addition to the price competition between the majors (and especially Exxon) and the hypermarkets, the 1995 to 1998 period was also characterised by substantial cost-oriented actions.

(725) Although such actions covered a number of subjects, the most visible (if not important) scheme consisted in reducing the average site throughput differential between the hypermarkets and the majors' filling stations. This appears to have been primarily achieved through two different actions:

- Price competition with the hypermarkets: it aimed to draw volumes back to the majors (and therefore increase their total volumes) and away from the hypermarkets (and therefore reduce their total volumes). As a result of Pricewatch, Exxon's volumes appear to have grown by approximately [...]* % in the first half of 1996,

- Elimination of less economic filling stations: between 1994 and 1997, Exxon lost 11 % of its branded filling stations, BP and Mobil lost 10 % of their combined sites and Shell lost 30 % of its service stations. It should also be noted that, because most of the sites lost were dealer-owned or dealer-operated, this campaign also resulted in higher control of the majors over their retail network.

(726) Consequently, and as a result of other cost-cutting actions, [...]*.

(727) Indeed, all the other competitors appear to have mainly lost ground with the majors and the hypermarkets in that period: they apparently were not able to follow the pace imposed by the hypermarkets and the oil majors, as their market shares of each of these groups of competitors have dropped by more than 23 % between 1994 and 1998.

Mini-majors and independent retailers no longer exert competitive pressure

(728) The changes in the United Kingdom motor fuels retail sector have resulted in a new market structure, where the mini majors and the independent retailers can no longer exert a real competitive pressure.

(729) This is clearly indicated by the fact that the mini majors and the independent retailers were not able to follow the pace imposed by the hypermarkets and the oil majors; this is also suggested by the fact that they now face much lower average throughputs (less than 2300) than the oil majors (more than [...]*) or the hypermarkets; and this is finally confirmed by the fact none of these individual competitors currently has a market share over 6 %, and that [...]*.

(730) This is also supported by the OFT report, which states that "independent retailers' profit margins (...) are likely, in most cases, to be lower than those of other sites in comparable areas and may be unsustainably low for many such retailers".

Hypermarkets will become price followers

(731) There are also indications that hypermarkets will no longer seek or be able to lead price competitive actions, and will become followers to the majors instead. In particular, hypermarkets indicated that, although they will continue to set their prices so as to match the lowest price displayed at any of their main competitors' sites, they will now tend to move away from a policy of actively seeking to post lower pump prices than key oil company competitor outlets.

(732) [...]*. It is also confirmed by the fact that, although gross margins approached historical levels in 1997, they are expected to increase in the next few years.

(733) A first explanation for this evolution is that hypermarkets will have very little scope for aggressive competitive action, as they no longer have real cost advantages over the majors. [...]*. In addition, the results of the investigation carried out by the Commission suggest that the costs of the hypermarkets should rise in the future, as new sites will have lower throughputs and hence will face higher costs than the existing ones.

(734) The Commission enquiry shows that another explanation is that, after Pricewatch, hypermarkets have learnt that any aggressive pricing will be followed by the majors, even if that entails severe losses in the retail sector. In such a case, there would be no incentive for price reductions, as they would not result in higher market shares but lower profits.

(735) The fact that hypermarkets probably have nothing to gain from aggressive pricing is further confirmed by the fact that they cannot significantly expand their sales volumes. The results of the Commission enquiry indicate that it would be very difficult for them to expand sales volumes through their existing outlets or to obtain planning permissions for new sites, and that, although hypermarkets plan to open additional filling stations, these stations will have lower throughputs (and hence profitability). Hypermarkets have also indicated that they did not anticipate any significant growth in their market shares.

(736) These conclusions would be supported by the OFT report, which states that "supermarket petrol sites appear to be close to the limit of their growth. (...) Further growth in volumes can only come from greater market share and even this is capacity constrained. As petrol sales reach their limit, the incentive to sell at low margins (...) will be reduced".

(737) Finally, some hypermarkets have recently entered into joint ventures with the majors for the provision and operation of hypermarket-branded shops on major-branded filling stations. Examples of such ventures are BP/Safeway and Exxon/Tesco. Such collaboration is presently limited to the major-branded sites concerned and has no impact on either the operation or the supply of the hypermarkets' own filling stations. However, it will obviously create long-term relationships between the hypermarkets and the majors, which could eventually extend to the hypermarkets' fuels retail business, and would in any case probably reduce the incentive for hypermarkets to compete with the majors.

(738) Consequently, although hypermarkets have traditionally acted as price leaders in the motor fuels retail sector, their market shares and capacities are levelling off and they appear to be becoming price followers to the majors. This is further confirmed by the results of the Commission enquiry, where hypermarkets have indicated that, in the case of movement in the wholesale prices, they would generally look at the majors' reaction before deciding on their own movement.

The majors become the clear market leaders

(739) On the contrary, the majors appear to have been strengthened rather than weakened by the fierce price competition of 1995 to 1996. This is firstly clearly indicated by the fact that their combined market shares have increased by approximately [...]* % between 1995 and 1998.

(740) Furthermore, although current gross margins are much lower than those which existed in early 1995 (before aggressive competition with the hypermarkets started), some of the majors' current profits appear to be even higher than in early 1995.

(741) In addition to increased profits, the majors have also obtained a better grasp on the market, as most of the damage of the price war era has been borne by the mini-majors and the independent retailers. Hypermarkets no longer appear to be in a position where they can substantially threaten Exxon, BP/Mobil and Shell, and a substantial number of dealer-operated filling stations (where prices are less easily controlled) has been eliminated.

(742) In the light of the above, it appears that the majors are now becoming the clear market leaders of the motor fuels retail sector in the United Kingdom, and that any significant competition could therefore only arise from rivalry between the members of this league, with the other competitors having lost any substantial capacity or incentive to fight against Exxon, BP/Mobil and Shell.

(743) The parties respond that ex-tax pump prices and gross distribution margins in the United Kingdom remain among the lowest in Europe, that this indicates intense competition, and that the broad historical trend of declining margins in response to intense competition is likely to continue. However, although ex-tax pump prices and gross distribution margins in the United Kingdom may be relatively low, so, it would appear, are costs. For instance, Exxon's costs appear to have dropped by [...]* % between 1994 and 1997. By contrast, and in contradiction with the parties' objections, [...]*. Such figures do not appear to support any of the parties' assertions.

(744) The parties also submit that the documents filed with the Commission demonstrate that any possible price increase by the majors would not be followed by hypermarkets, and that the hypermarkets' policy to be unbeatable on price in their catchment areas is not likely to change following the notified transaction. This would indicate that the majors cannot exercise any market power. However, as explained above, hypermarkets may continue posting the lowest prices in their catchment area, but also follow price rises. [...]*. It is also confirmed by the fact that hypermarkets have indicated that they would now tend to move away from a policy of actively seeking to post lower pump prices than key oil company competitor outlets, and that, in the case of a price rise, they would generally look at the majors' reaction before deciding on their own movement. It is, finally, consistent with the fact that hypermarkets find it difficult both to expand sales at existing outlets and to open competitive service stations. In that context, the fact that some hypermarkets have indicated that they would not completely follow a price rise by 5 %, is probably due (as their replies suggest) to the fact that a 5 % price rise is very unusual in the market for retail fuels.

(745) In addition, the new market structure appears to be to provide strong incentives for oligopolistic behaviour, which appear to be even more potent than they have been before the rise of the hypermarkets.

(746) The fact that the market structure is conducive to oligopolistic behaviour is clearly indicated by the general characteristics of the motor fuels sector in Europe (that is, product homogeneity, transparency, absence of technological evolution, demand rigidity, absence of countervailing buying power, high entry barriers, and so on) and those of the relationships between Exxon, BP/Mobil and Shell (vertical integration, multi-market contacts and such like).

(747) But, in addition to these general features, the structure of the United Kingdom market as it has resulted from its recent evolution even increases such incentives, to a degree arguably unseen for many years (if ever).

(748) Firstly, the United Kingdom market for retail sales of motor fuels appears to have been particularly stagnant in the last few years, as the volumes sold in the United Kingdom appear to have grown by less than 2 % from 1994 to 1998.

(749) Secondly, there is clearly a symmetry between Exxon, BP/Mobil and Shell's market shares, which, in 1998, ranged between [10 to 20]* % (for Shell) and [10 to 20]* % (for Exxon). Although it can be argued that there has been no stability in these market shares in the last four years, what is more important is that (i) all of these market shares have grown (thereby reflecting the increase in the majors' influence on the motor fuels sector), and (ii) the symmetry in the market shares of the majors has increased in the last few years, as the difference between the highest and the lowest of these market shares has been reduced by 37 % since 1994.

(750) Thirdly, there also appears to be a strong correlation between the cost structure of these companies, [...]*. The parties disagree with this point, and submit that the operating cost differences between the various participants in that market are much wider than this. [...]*. The parties also submit that, even within Exxon, there are cost differences dependent on service station format and offering. However, first, it should be noted that the national strategy of the market players (and, at least, of Exxon) is defined on the basis of the average break-even requirements, and that, therefore, it is the average cost structure of the market players which should be taken into account when discussing the symmetry of their cost structure. Secondly, the fact that the costs may vary to a large extent even within a given competitor further confirms that a difference of approximately GBP 0,01 per litre between the average costs of several competitors is indicative of a symmetry of costs between these competitors.

(751) Fourthly, Exxon, BP/Mobil and Shell appear to have symmetric incentives, as all three majors achieve a national coverage of the United Kingdom market and have the same degree of vertical integration. They also share ownership in some storage facilities and pipelines.

(752) Finally and perhaps most importantly, the majors now have a considerable market power over the United Kingdom motor fuels retail sector. This significantly enhances the benefits that the majors can derive from any coordinated behaviour and therefore also substantially increases any incentive for them to act accordingly.

(753) Consequently, although it can be concluded, as the parties submit and the OFT reported in 1998, that the motor fuels retail sector in the United Kingdom is still competitive, it appears that the majors have seized control of this market again, that any significant competition could therefore only arise from rivalry between the members of this league, and that such a rivalry has become more unlikely due to the above described developments.

(754) The parties respond that competition in the United Kingdom retail motor fuels industry is intense, on the grounds that the United Kingdom competition authorities' reviews of that market concluded that it was highly competitive. However, as explained above, the market appears to have evolved significantly in the last few years, with the fierce price competition of the mid-1990s giving way to a control by the majors. And, although the latest report by the OFT indicates that the market is operating competitively, it also concludes that "the market will require monitoring to ensure that competition between the major players does not become muted should it become more concentrated".

Entry barriers - expansion - potential competition

(755) Exxon, BP/Mobil and Shell would not appear to be challenged by the prospect of possible market growth or entry, as the volumes sold in the United Kingdom appear to have been relatively stagnant since 1994 (they have grown by less than 2 % in three years) and as market entry appears to be very unlikely.

(756) In particular, the OFT report states that "entry into retailing is currently unattractive given the low margins available on fuels sales and the costs of complying with environmental legislation. In order to earn sufficient returns, entry has to be on a fairly large scale and it can be difficult to obtain planning permission for greenfield sites. It might be argued that new entry could occur in the future by the reopening of the large numbers of sites which have been closed in recent years. However, these will tend to be the smaller sites and, because of environmental regulations, sites have to be completely closed and tanks filled with concrete. Such sites will often be put to an alternative use. This makes reopening an old site almost as difficult as building a new one".

(757) And, although the parties submit that new entry into retail fuels marketing in common, they can only cite the examples of the hypermarkets (which are not really new entrants) and of the "Save" network (which has a very limited market share and lost 15 % of its filling stations in 1997).

THE POST-MERGER ENVIRONMENT - THE IMPACT OF THE OPERATION ON COMPETITION

(758) The impact of the present operation is to be assessed against the above described background, having also regard to the features identified in the section devoted to the structural characteristic of retail motor fuel markets in general.

(759) As explained above, the United Kingdom market for the retail sales of motor fuels is now clearly led by the majors, with the other competitors having lost any capacity or incentive to engage in substantial competition against Exxon, BP/Mobil and Shell. The structure of the market is significantly conducive to oligopolistic behaviour between the three majors, who have similar features and comparable market shares.

(760) Third parties have indicated that there has been competitive pricing between Exxon and BP, and that it is this dynamic which may be assumed to influence United Kingdom market pricing in the future. [...]*. Given the majors' new and strong incentives for oligopolistic behaviour, it is unlikely that this rivalry could continue to the same extent in the future.

(761) As explained above, the notified concentration will create a structural link between Exxon and BP/Mobil, which will result in Exxon being able to exert a significant influence in the BP/Mobil venture and, more generally, will create new incentives for a cooperation between these two companies. Given the additional presence of strong market-related incentives for such a cooperation, it is very likely that, following the merger, any real competition Exxon and BP/Mobil will be eliminated.

(762) There is also strong evidence that Shell would not seek to challenge the competitive position of Exxon and BP/Mobil. It is very likely that Shell will instead join Exxon and BP/Mobil to form an oligopoly.

(763) It would appear that, even prior to the merger, Shell has not traditionally been a very aggressive price competitor. [...]*. Moreover, Shell appears to have been one of the main participants in the price restoration movement, [...]*.

(764) Instead, third parties have indicated that Shell's policy seems to be to seek a premium quality brand position which applies not just to fuels. This would appear to have led Shell to price its products above the market.

(765) There is also no indication that Shell would seek or be able to change this strategy following the merger. [...]*. Secondly, unlike Exxon and BP/Mobil, Shell is not engaged in joint ventures with hypermarkets and therefore cannot propose similar offerings. Lastly, Shell's current strategy appears to be relatively successful, as its market share of motor fuels retail volumes has increased by 13 % between 1994 and 1998 (although, as indicated above, such growth should be limited in the future).

(766) Furthermore, Shell is already engaged in a number of partnerships with BP/Mobil and Exxon: Shell jointly owns a number of distribution terminals in the United Kingdom with either BP/Mobil or Exxon, and Shell and BP/Mobil both participate in the ownership of two of the main pipeline networks covering England and Wales. Shell's links with BP are also expected to increase after the closure of Shell's Shellhaven refinery, when Shell will probably source a substantial part of its refined fuels from BP/Mobil's Coryton refinery.

(767) Finally, Shell would be all the less likely to move to a more competitive posture as the incentives for oligopolistic described above are now stronger than they probably have ever been for years.

(768) In the light of the above, it can be concluded that, after the operation, the majors will be engaged in an oligopoly on the United Kingdom motor fuels sector.

Unique competitive position in the wholesale sector

(769) Exxon, BP/Mobil and Shell are active in non-retail sales of motor gasoline and diesel to third parties. According to the figures in the possession of the Commission, their current combined shares in that sector would be in the order of [30 to 40]* %.

(770) Furthermore, these market figures underestimate the majors' actual market power. In that respect, it should be noted that, even after the closure of the Shellhaven refinery, the majors will collectively represent approximately [50 to 60]* % of the total United Kingdom refining capacity; they will also own together approximately [40 to 50]* % of the distribution terminals in the United Kingdom, and they will participate (together with non-majors) in the joint ownership of an additional [0 to 10]* % of distribution terminals; and finally, Exxon's Fawley and BP/Mobil's Coryton sites will be the only two refineries in the south and south-east of England.

(771) Exxon, BP/Mobil and Shell will therefore have a unique competitive position. Firstly, as refiners, these companies will have the opportunity to engage in swaps arrangements with other refiners, and therefore to obtain nationwide supplies at better conditions than those for non-refiners. And secondly, they will have significant competitive advantages over other refiners, because they will reach unequalled volumes of sales and will gain unequalled geographic coverage of the United Kingdom territory. For instance, the majors' refining capacity will be more than five times as high as that of their nearest competitors (Texaco and Conoco), and the number of distribution terminals which they own or share will be considerably higher than that of their next competitor.

(772) The majors will depend on swaps with other refiners to a much lesser extent than other refiners will depend on swaps with them. Therefore, a possible consequence of their cooperation is that they would have the incentive and the possibility to impose harder conditions on the swap arrangements they have with other refiners.

(773) Another possible consequence of the post-merger situation is that the majors will be in a better position to deter local competitive actions (such as entry) by other refiners or other wholesalers, as, given their more extensive geographic coverage, they could offset the costs resulting from local competitive actions over larger volumes and territories.

(774) A third possible consequence of the post-merger situation is that Exxon will influence BP/Mobil (or even Shell), currently a major supplier to the hypermarkets, to offer higher prices or restricted quantities to these customers. Some hypermarkets have expressed concerns that such behaviour could raise their costs and therefore undermine their competitiveness in the downstream motor fuels retail sector.

(775) Given the refining overcapacity and the substantial amount of wholesalers in the United Kingdom, the operation will probably not create a dominant position in the non-retail sector in this country. However, the majors will have a unique competitive position in that sector which, as explained above, will increase their competitors' supply costs.

(776) The parties submit that the fact that companies other than the majors will control approximately 46 % of the United Kingdom refining capacity and 58 % of United Kingdom distribution terminals demonstrates call into question any conclusion that the majors will be able to exercise market power or be able to raise other competitors' supply costs following the transaction. As explained above, it is in the light of these figures and of the Commission enquiries that it may be concluded that the wholesale sector will remain competitive. However, the parties' arguments apparently fail to address the facts that the other competitors on the wholesale sector are much smaller than any of the majors, that they usually have limited geographic coverage, and that they either do not have access to swaps or depend on swaps from the parties to a much larger extent than the majors depend on swaps with them. Consequently, they do not call into question the fact that the majors will be able to address their rivals' supply costs or successfully engage in local competitive actions.

(777) According to the parties, evidence also indicates that the integrated firms treat refining and marketing as separate businesses. The parties therefore consider that vertical integration does not give the majors any advantage versus non-integrated players, and that this is indicated by the growth in the presence of non-integrated firms, especially hypermarkets or chains such as British Fuels. However, the fact that refining and marketing are treated as separate businesses does not preclude the fact that vertical integration may and already does give competitive advantages to integrated firms, as refiners have access to safe supplies and to swaps, and, as far as the majors are concerned, who also benefit from extensive geographic coverage and distribution infrastructure (such as pipelines and distribution terminals). Furthermore, it should be noted that, apart from hypermarkets (whose size, presence and brand-recognition makes them somewhat exceptional competitors), non-integrated firms only account for a very small (less than 6 %) and declining fraction of retail sales.

Unchallenged market leadership in retailing

(778) The majors are the three main retailers in the United Kingdom, and even more significantly, the two main majors. Their combined market shares are approximately [50 to 60]* % - approximately [...]* times as high as those of their nearest competitor (Tesco, which is engaged in a joint venture with Exxon).

(779) As was explained in recitals 728 to 738 above, the recent evolution of the motor fuels retail sector has been characterised by the decline of the traditional competitors of the majors, i.e. the mini majors and the independent retailers. It has also resulted in the hypermarkets moving from price leadership and aggressive competitive posture to a position of price followers with limited expectations of market share growth.

(780) More specifically, the current market dynamics shows that the majors keep gaining market to the expense of the mini majors and the independent retailers. It can therefore be expected that, because of this evolution and of the capacity constraints of the hypermarkets, the majors' combined market would reach even higher levels in the next few years and that these companies would definitely secure their control over the retail market.

(781) These conclusions will even be reinforced after the operation since, as was explained in recitals 769 to 777 above, the majors will now have the chance to raise the other retailers' supply costs and therefore weaken these competitors' positions in the retail sector.

(782) Finally, given the usually limited geographic coverage of the hypermarkets and the independent retailers, and, by contrast, the majors' extensive national presence, Exxon, BP/Mobil and Shell could collectively engage in selective competitive actions in some areas, and significantly damage competitors while offsetting the costs associated with these actions through limited national price increases. The parties disagree with the fact that hypermarkets have limited geographic coverage, on the grounds that [...]* % of Exxon service stations and [...]* % of BP/Mobil service stations are within five kilometres of a hypermarket. However, it should be noted that only a fraction of hypermarkets are equipped with service stations: according to some hypermarkets, only 35 % to 50 % of their outlets are equipped with an adjacent service station. It follows that, even though a parties' site may be close to a hypermarket, it may not face any competition from this hypermarket. Furthermore, the figures provided by the parties refer to hypermarkets as a whole. Their figures may therefore be misleading, insofar as, although hypermarkets as a whole may reach national coverage, each hypermarket chain individually has a much more limited presence.

(783) Consequently, after the operation, the majors' market positions could not be successfully challenged by non-major competitors.

(784) In the light of the above, the notified concentration will create a oligopolistic dominant position by Exxon, BP/Mobil and Shell on the market for the retail sales of motor fuels in the United Kingdom, as a result of which effective competition would be impeded in that Member State which is a significant part of the EEA.

F. AVIATION LUBRICANTS

(785) Aviation lubricants (also referred to as ester-based turbine lubricants or aviation synthetic turbine oils) are used in the lubrication of either turbofan engines (commercial and military aircraft) or aero-derivative applications (those which were developed on the basis of aircraft engine technology, such as some marine turbine engines and some turbo-lubricated power generators and other industrial accessories). As it will be seen in the paragraphs on product market definition, commercial aircraft lubricants, military aircraft lubricants and lubricants for aero-derivative applications constitute three distinct product markets.

(786) Their production process follows more or less the typical stages of regular lubricants manufacture, that is, the blending of synthetic base stocks with additives, although ingredients are different than those used in the manufacture of regular lubricants.

(787) As it is the case for automotive lubricants, the manufacturing process and the finished lubricant requires, prior to commercialisation and use in airborne equipment, approvals from relevant organisations, in that instance of the relevant military authorities and OEMs. However, in contrast with automotive lubricants and due to the higher safety standards that aviation lubricants have to comply with, the approval process for aviation lubricants is more stringent and lengthy.

(788) There are three main categories of buyers of aviation lubricants, that is, commercial airlines, the militaries, and various industrial customers.

MARKET DEFINITION

Relevant product market

Aviation and regular lubricants are not part of the same market

(789) The parties advance the argument that aviation lubricants are part of the same market as regular (automotive and industrial) lubricants, examined above. Although the two categories of lubricants are not substitutable from the demand side, the parties claim that they are substitutable from the supply side and accordingly, regular lubricants manufacturers which are not active in aviation lubricants can, readily and in good time, switch their production process to that of aviation lubricants. Three points support their view: first, the raw materials required for the blending of aviation lubricants (i.e., polyol ester base stock and additives) are largely available; secondly, the blending equipment is the same for both categories of lubricants; thirdly, there is an overall blending overcapacity, which makes the reassignment of blending facilities economically possible. Therefore, supply-side substitution with other automotive lubricants would be possible by cleaning the blending equipment used previously for automotive lubricants and converting it, within a relatively short period of time, into equipment dedicated to the manufacture of aviation lubricants.

(790) The market investigation concluded that regular (automotive or industrial) lubricants and aviation lubricants constitute two separate product markets. From the demand side, there is clearly no substitution between the two; the higher safety requirements of aviation lubricants and the different operational environments of automobile and aircraft engines eliminate any possible demand-side substitution with regular automotive lubricants. The parties do not deny this statement. From the supply side, the arguments of the parties are not conducive to the existence of one single market, as they only rely on purely logistic aspects of supply substitution. Even assuming that the substitution of blending equipment could be possible technically (although other manufacturers questioned have expressed reservations as to such a possibility) and that a blender of regular lubricants could start blending aviation lubricants, full market entry cannot, however, be achieved easily and readily, as it is characterised by significant barriers which will be discussed in the following paragraphs. Additionally, not only are not regular lubricants substitutable to aviation lubricants, but as it will result from the following paragraphs, there exist even narrower product markets for aviation lubricants used in commercial aircraft engines, military aircraft engines, and aero-derivative applications.

Commercial, military and aero-derivative applications

(791) The existence of three narrower markets for aviation lubricants is evidenced by the lack of demand and supply-side substitution among the three categories mentioned above. The lack of demand substitution among the various categories of aviation lubricants is a result of a continuum of quality characteristics which limit the intended use of the three respective products. Aviation lubricants used in commercial aircraft engines have to fulfil higher quality and safety requirements than those used in military aircraft engines (military engines are overhauled very frequently and time between overhauls is between 200 to 1000 hours compared to 20000 hours for commercial engines). The lack of supply-side substitution is a result of the different specification requirements and amount of time necessary for market entry in each of the narrower aviation lubricants markets. In order to achieve full market entry in the market for commercial aircraft lubricants, the potential supplier will need to observe some typical steps required prior to the commercialisation of the product. The typical procedure, outlined below, could take up to six years for lubricants used in military aircraft engines and up to 12 years for lubricants used in commercial aircraft engines.

(792) In order to have a marketable product to satisfy the needs of the majority of airlines operating mixed fleets of aircraft, a supplier has to make extensive research in view of the product formulation. The cumulative time required for this first step may vary from three months for incumbent manufacturers to four years for new entrants. At this stage, a supplier may not chose to engage in any military and/or commercial accreditation process, but to manufacture and commercialise products intended for industrial applications (meaning turbine engines used for oil drilling, gas pumping, electrical generation or marine turbine engines). Subsequently, for a supplier to start selling products intended for military or commercial aircraft, the respective military and OEM approvals are necessary. First, a supplier needs to obtain the approvals of governmental military agencies (e.g., US Navy, UK MoD, French DGA and so on). This process entails laboratory testing (glassware and rig) and engine bench testing. It may take up to two years and is usually a prerequisite for the application for OEM testing, or otherwise required prior to granting full approval from OEM. Again, a supplier may not choose to proceed with OEM approval as this is only required for products intended for commercial airlines' use. In such a case, a supplier may focus on lubricants intended for military customers. According to the parties, the military market is bid driven, not brand loyal and more selective.

(793) The following step for full market entry in the commercial aircraft lubricants market is OEM approval. It consists of full accreditation in the majority of aircraft engines and accessories in use by commercial airlines. This usually comprises seven major engine manufacturers (such as General Electric, Rolls Royce, Pratt & Whitney, Allied Signal, Snecma, International Aero Engines, CFM International); 20 major engine models; four major accessory manufacturers; and four major accessories (APU, CSD, IDG and ATS)(70). The process comprises laboratory testing (glassware and rig), engine bench testing by model and a lengthy in-flight service evaluation by model. According to parties' estimates, the cumulative time required from the start of the formulation to full approval and commercialisation of commercial aircraft lubricants can take up to 12 years. However, the parties claim that there is downward trend in the time needed for obtaining commercial OEM approvals. Some engine manufacturers would not require flight evaluation for some of their engines. For instance, Pratt & Whitney have revised their program and require evaluation only in one engine thus reducing the time and cost required to go from military to commercial approval to [...]* years and USD [...]*. The parties explain that the average time period required for commercial OEM approval would now range from [...]* to [...]* years and cost from USD [...]* to [...]*. In any case, this still amounts to a considerable barrier to entry.

(794) From a technical viewpoint, commercial aircraft lubricants could be used in military applications, however the reverse is not true. This is due to the additional testing required for civil aviation approval. Moreover, lubricants used in aero-derivative applications are required to pass very limited approval processes. From a technical viewpoint, both commercial and military lubricants could be used in those applications, however the reverse would be technically unfeasible. An illustration of the above one-way substitution is that most airlines use the same product for all their aero-applications (both flying and ground equipment) and for this reason they only buy the higher quality lubricants for commercial aircraft engines which have obtained OEM approvals(71). Therefore, both supply and demand substitutability can be said to operate in one direction. This is furthermore illustrated by the reduced number of suppliers to commercial airlines (Exxon, Mobil and to a lesser degree Shell and Castrol) and the larger supplier base to the militaries (the above majors plus a number of smaller specialised manufacturers, such as NYCO, Henkel, Hatco, etc.).

(795) On the basis of the above evidence, it is fair to conclude that the aviation lubricants sector is constituted by three distinct product markets, divided according to their intended use. The operation will materially affect the market for commercial aircraft lubricants.

Relevant geographic market

(796) The parties suggest that the geographic market is worldwide, stating that most manufacturers supply their worldwide network from one or a few plants; that aviation lubricants travel easily and at low cost; that prices are similar in the different regions of the world; and that there are no trade barriers. The market investigation has confirmed the above definition for all types of aviation lubricants. Therefore, the geographic market is considered to be the world.

COMPETITIVE ASSESSMENT

(797) Exxon and Mobil account for [60 to 70]* % of the worldwide production and sales of ester-based turbine lubricants (Exxon: [30 to 40]* %; Mobil: [20 to 30]* %). The other main suppliers are Shell ([0 to 10]* %), Castrol ([0 to 10]* %), NYCO ([0 to 10]* %) and others ([20 to 30]* %).

(798) In the market for commercial aircraft engines lubricants, the combined market share of the merging parties is significantly higher. According to parties' estimates, their combined market share at the end of 1997 was [80 to 90]* % (Exxon: [40 to 50]* %; Mobil: [40 to 50]* %, Shell: [0 to 10]* %; others: [0 to 10]* %). Airlines and competitors questioned during the market investigation, indicated that the combined entity would wind up with more than [0 to 90]* % of the market and that, at present, Shell and Castrol would be the only available alternative suppliers, with [0 to 10]* % and [0 to 10]* %, respectively (it is worth noting that Shell has no customers in any US airline and sells marginally to EU commercial carriers).

(799) Exxon and Mobil appear to have nearly insurmountable market leadership which according to some competitors is due to historical market conditions. The aviation lubricants industry has suffered from poor profitability and a high cost base as a result of the continuous need for research and development triggered by technological advances in aircraft manufacturing. As a consequence, many suppliers withdrew from the market. All the questioned parties have indicated the high level of commitment that the commercial aircraft engines market commands. The merging companies have been committed there from the beginning of the development of the jet turbine technology and have evolved into the supplier of choice of all the main airlines. According to the parties, their larger share in that market is due to their decades-long dedication in it, whereas other aviation lubricants suppliers have focused on the other two markets (Castrol in the industrial and the military, Shell in the military and other aviation products, Hatco in ester-based stocks sales, rebranding and military sales). The parties claim that they do not hold any insurmountable technological leadership because the technology required to compete would be mature. This is probably true but does not change the fact that there are high barriers to entry in the market due to approbation procedures which make unlikely entries even if prices were to be raised.

(800) In fact, the remaining manufacturers of aviation lubricants tend to be specialised lubricants and chemicals manufacturers that concentrate mainly on aero-derivative and some of them on military applications. Hatco (US), is a major supplier of ester base stock and is mainly a supplier to military and aero-derivative applications which has certain commercial co-approvals with Shell; NYCO (FR), is an independent supplier with a range mainly of military aviation products and certain engine manufacturers' approvals; Anderol (US), is mainly a supplier for aero-derivative applications, with some reduced activity in military aviation applications; Henkel (US), is a supplier of ester and has a specialised activity directed towards aero-derivative and military applications and no approvals in commercial aviation lubricants. Other manufacturers of ester-based turbine lubricants and the product markets that they serve include the following: Arpol Petroleum (US-rebranded-military); BP Marine Ltd (British-rebranded-industrial); Caltex (US-rebranded-industrial); Delta Petroleum (US-rebranded-military, industrial); Hexagon (US-rebranded-military, industrial); Petron International (US-rebranded-military); Total (French-rebranded-military); Velsicol (US-rebranded-military, industrial).

(801) It is made clear that Exxon and Mobil are the leading suppliers of own aviation lubricant products destined to commercial aircraft engines, and have the largest manufacturing capacity for aviation lubricants. Actual competition in the market for commercial aircraft lubricants is reduced to the two merging suppliers (their respective price increases over the last five years were of [...]* % and [...]* %). Any remaining bargaining advantage for customers will therefore be removed through the elimination of competition between those two.

(802) The parties claim that, were the combined entity to raise the prices above competitive levels, established suppliers would expand their sales and potential competitors would enter the market. The investigation has shown that the demand for commercial aircraft engines lubricants is inelastic as a consequence of the low cost of aviation lubricants in the total cost of aviation products of commercial airlines and the accrued priority of such customers to ensure the highest level of flight safety, which increases switching costs and prevents them from shopping around. For instance, the parties have explained that customers have an incentive to continue purchasing from their current supplier as they would incur costs to change jet oil brands, including notifying and filing forms with their local government's aviation authority (including the Federal Aviation Authority if they fly in the US), changing all copies of their operating and maintenance manuals and issuing notices to all airports into which they fly. Moreover, according to the parties, airlines make lubricants purchasing decisions based on technical and historical performance evaluations which are not, therefore, strongly related to price.

(803) With a combined market share of [0 to 90]* %, the merged entity would be indeed in a position to behave independently from its competitors and customers. The parties argue that despite the low price elasticity of demand, new market entry will be triggered more likely by existing competitors such as Castrol, Hatco and Shell who could expand their production of aero-derivative and military engine lubricants to commercial engine lubricants. In addition, they claim that these new entries would be sponsored by airlines in response to higher prices. The market investigation has shown that market entry may not be necessarily encouraged under the prospected demand growth (1 to 2 % over the next five years), and may be substantially hampered by the high incremental entry cost, pertaining to the approval requirements and the concern to ensure profitability through driving sales volumes. In addition to this, the parties have stated that for a supplier of commercial aircraft engines lubricants to compete in this market in a viable and optimal way, a minimum market share of 20 % may be required. In fact, existing suppliers are reduced and share a smaller customer base, whereas there have been no new entrants in the market for lubricants for commercial aircraft engines during the past 20 years.

(804) The parties argue that their airlines customer base has significant buyer power and point out that the concentration trend in the airlines industry is increasing the countervailing power of airlines. However, besides the fact that aviation lubricants do not constitute a significant cost area for airlines, the investigation has shown that the concentration trend will lead strategic alliances between airlines to a focus on cost and seat sales and, unless their engineering departments combine, such power may never become relevant for lubricants suppliers. Moreover, the parties suggest that a supracompetitive price increase would have adverse effects on various other aviation products which the same airline customers may purchase from the same supplier. However, airlines questioned view aviation lubricants as a stand-alone product and consider that, if there are no available lubricant substitutes at sight, the existence of other aviation products cannot influence ongoing commercial relations in commercial aircraft engine lubricants.

(805) The operation will bring together, and will eliminate competition between, the main currently competing suppliers, and lead to the creation of a dominant position held by the merged entity on the market for aviation lubricants used in commercial aircraft engines.

G. AVIATION FUELS

PRODUCT MARKET

(806) Aviation (or jet) fuel is a kerosene-type fuel used for the powering of jet aircraft engines. Although similar to domestic kerosene (used for heating purposes, in particular in the United Kingdom), it meets strict performance specifications. According to the previous decisions (BP/Mobil, Shell/Gulf Oil), aviation fuels constitute a separate product market from other motion fuels (such as motor gasoline, automotive diesel and marine fuels). The notifying parties also support this view.

GEOGRAPHIC MARKET

(807) The parties state that aviation fuels is a worldwide market, or at least, wider than the EEA. To support this view, they refer to the fact that aviation fuels are sold on a bidding basis for global supply contracts with the various airlines, which require suppliers to deliver product at airports around the world. Moreover, pricing is Platt's-related thus making aviation fuels prices in the cargo markets around the world to track one another closely.

(808) The Commission does not consider that these facts are conducive to one single worldwide market. One distinction needs to be made between manufacturing of aviation fuels, which may be of worldwide relevance, and actual supply of aviation fuels to specific locations, whereby infrastructure logistics may limit the geographic scope of competition.

(809) Although ex-works and cargo market prices may be aligned, this is due to the commodity character of the product, and it reflects a general price alignment among refined petroleum products. Given the uniform pricing of raw materials (crude oil) and the, more or less, uniform cost structure of the refining processes and transportation costs around the world, it may be expected, as it is the case for other refined petroleum products, that aviation fuels indicate a uniform worldwide price. However, when it comes to the supply of aviation fuels to specific airports, the worldwide character of this market fades away. Although airlines launch global tenders, they do not necessarily select one and the same supplier for all their worldwide requirements; on the contrary, they select the most attractive bidder on an airport-specific basis depending on the relative strengths and advantages of suppliers in a given location. Moreover, prices charged to load the product into-plane, may differ from one contract to another. The competitive environments may vary from one airport, or location, to another. This is mainly due to the distribution infrastructure for the supply of aviation fuels to the various airports, which is in general dedicated to a specific airport. Therefore, suppliers need not only produce aviation fuels to be able to win contracts with airlines; they also need to have access to the specific distribution and loading infrastructure that will enable them to effectively market their product at the final stage of the supply chain, that is, into the aircraft.

(810) Competitors questioned during the market investigation also contested the worldwide scope of this market. They indicated that airport-specific markets may exist, even in those cases where several airports are located in the same area (for instance, the London airports, Heathrow, Gatwick and London City). In that sense, from the demand side, if price of jet fuel increased at one airport, an airline would not be able to switch to another airport in order to benefit from the cheaper jet fuel, mainly as a result of the constraints on slot availability. From the supply side, an oil company's ability to switch from supplying one airport rather than another, is dependent on access to the logistics infrastructure. Supply-side substitutability is therefore also limited.

(811) The parties admit that access to airport supply infrastructure is important. However, they state that there is strong competition among a variety of suppliers, with many of the same suppliers appearing at different airports, or others having formed joint ventures in the areas of airport storage and into-plane activities.

(812) Although this statement is not incorrect, the investigation has shown, however, that at least in one European airport - Gatwick in London - the existing distribution infrastructure may severely limit access of new suppliers to this airport, while it may materially affect access conditions for existing suppliers, as a result of the merger. Gatwick is supplied through a network of dedicated pipelines, one of them belonging to Exxon and the other to a consortium between Shell, Mobil, BP and Texaco. In that sense, the offer and demand of aviation fuels in Gatwick cannot be grouped with that of another airport, not even closely located airports, such as Heathrow or London City. Airlines operating from Gatwick could therefore not profit from a price variation occurring in another airport in order to switch their demand to that airport. Supplying companies would not be able to switch from supplying Gatwick rather than another airport, since this is dependent on access to the logistics infrastructure. In this sense, the demand and supply conditions for Gatwick seem to constitute a closed system, which makes substitution between airports unfeasible, since demand and supply can only be matched locally. Under these circumstances, it may be concluded that the supply of aviation fuels to Gatwick airport constitutes a distinct geographic market.

(813) Gatwick airport constitutes a substantial part of the common market. In terms of air traffic, Gatwick ranks as the fifth airport in the EEA. Moreover, in terms of consumption of aviation fuels, Gatwick represents 25 % of the total consumption of aviation fuels in the United Kingdom (Heathrow accounting for 55 %, the remaining United Kingdom airports for 20 % altogether) and 6 % of total EEA sales of aviation fuels(72).

Supply process and infrastructure

(814) According to the parties, airlines are the primary aviation fuels customers. They buy fuel into-plane, meaning that suppliers need access to logistics infrastructure to transport the product from their refinery or import terminal to the airport and into the aircraft. Such infrastructure may be owned by the major oil companies or by third parties, including the State. Most airports are supplied by pipeline, usually direct from several refineries. Rail and barge may also be used for refinery-to-airport transportation, although such alternatives may be more costly. At the airport, jet fuel must be stored and allowed to settle for a minimum of two days. From the airport storage, jet fuel is transported to the aircraft stands either through a hydrant pipeline system, or by special trucks. This last step is also called "into-plane operation". In the areas of airport storage and into-plane activities, the suppliers have formed joint ventures to reduce cost and achieve efficiencies. However, this may not be true of moving the fuels ex-refinery or ex-import terminal to the airport, where proprietary or fully utilised pipeline networks may severely constrain access and therefore successful market entry and/or expansion. This is particularly the case at Gatwick airport.

Aviation fuels supplies to Gatwick airport

(815) As a result of the merger, the combined entity will share [40 to 50]* % of jet fuels supplied to Gatwick. This market share will be combined with a high percentage ownership of the most competitively advantaged means of supply (industry-owned pipelines, such as West London pipeline, the Walton Gatwick pipeline and the existing Exxon system, directly linking its Fawley refinery to Gatwick). According to competitors, all other means of supply are either constrained by limits on capacity or severely disadvantaged from a competitive point of view. Furthermore, Exxon will control a substantial part of jet-fuel refining capacity for south-eastern England, since it owns the Fawley refinery and will acquire joint control of the Coryton refinery, presently belonging to BP/Mobil. It has to be noted that the only other available refinery in this area, Shell Haven, is due to close at the end of this year, which means that the combined entity will have the totality of refining capacity in south-eastern England (Mobil is entitled to [...]* % of the jet fuel offtake from Coryton for inland distribution - note that aviation as well as marine fuels are not part of the existing BP/Mobil joint venture). The parties combined share of jet fuel supplies to Heathrow and City Airport are respectively [30 to 40]* % and [40 to 50]* %. In addition, the parties would not own a high percentage of the relevant infrastructure that is used to supply these airports.

(816) Following the merger, the combined entity will control a substantial share of total pipeline logistics serving Gatwick and an even more considerable share of industry-owned pipelines (see attached supply charts, submitted by the parties and competitors). Exxon owns a pipeline linking its Fawley refinery to Gatwick. No third party is given access to this pipeline (apart from British Airways which has been given limited throughput rights for own-supply). The only alternative direct route to Gatwick is the UKOP pipeline, via the Walton and West London pipelines. This route is reported to lack sufficient capacity at present. The ownership in these pipelines is shown in the following table:

Industry-owned pipelines to Gatwick airport

>TABLE>

(817) Other alternative means of access are not competitive. Transport by rail would cost an additional GBP 5 per tonne. The GPSS (an open access government-owned pipeline) has no direct access to Gatwick, but needs to go through the Walton pipeline ([...]* % Mobil). This alternative is not viable, first because it would cost GBP 5 extra per tonne, and secondly because the GPSS-Walton line is full. An alternative access, via the GPSS route from the north Thames estuary via Aldermaston, is in practice excluded, as at certain key points the flow operates in the wrong direction for supply to Gatwick. Even assuming that access may be granted to the Walton pipeline, which has direct access to Gatwick, one has to pass through the West London pipeline ([...]* % Mobil), which is, reportedly, of limited capacity.

(818) Given that the cost of accessing the airports via industry-owned pipelines is approximately GBP [...]* per tonne, there appears to be a significant competitive disadvantage for suppliers which need to rely on these alternative routes. As a result, the combined entity will effectively have a logistical advantage over all other competitors for any extra supply capacity. This competitive disadvantage is even more severe for those players who do not share existing pipelines. This may act as a serious deterrent to market entry, given that it is the logistical cost that determines the margin and therefore the price that can be charged to an airline (the cost of the product being largely based on Platt's and that the other main costs are access and into-plane costs, the latter being the same for all players that have managed to get access to the airport).

(819) In their response to the statement of objections, the parties contend that besides the two main pipeline routes for the supply of jet fuel to Gatwick airport, there are at least three alternative routes, namely various parts of the government-owned GPSS pipelines. However, it should be noted that all these alternatives routes are indirectly connected to Gatwick via the Walton pipeline. In so far as they do not secure direct access to Gatwick Airport, they cannot be regarded as an alternative to the two main pipeline routes.

(820) The parties also note that while the main pipeline routes to Gatwick are all either full or close to capacity, expansion of the dedicated jet fuel pipelines is relatively easy should the owners so decide. However, it is a fact that as a result of the merger Exxon/Mobil will control solely one pipeline route while holding a significant interest in the only alternative route to Gatwick. The combined entity is therefore in a position to be able to restrict the access to this infrastructure, as well as limit its expansion. In sum, the control by the merged entity of most of the essential infrastructure, access to which is necessary to supply jet fuel to Gatwick Airport, can result in a foreclosure of the other jet fuel suppliers to the market of supply of jet fuel to Gatwick Airport.

(821) In terms then of sources of supply, it is worth noting that after the closure of Shell Haven, Exxon's Fawley and Mobil's Coryton refineries will become by far the most competitive sources of jet fuels supply to Gatwick. [...]*. Other United Kingdom refineries with inland distribution access to the south-east of England have no excess jet fuels capacity (the United Kingdom was a net importer of 1,6 million tonnes of jet fuels in 1998). Furthermore, other than Fawley and Coryton, the principal source of supply is the cargo market. The cargo market may result in a significant price disadvantage, as currently, the cif/fob cost differential is around GBP [...]* per tonne. In addition, there is currently no competitive access to Gatwick from the principal import terminals, since access will become dependent on industry-owned pipelines and/or on GPSS pipelines or rail, the latter two being non-competitive or impractical, as seen above.

(822) The parties contest the fact that they would control the most competitive sources of supply for the market in question. In this respect, they point to the presence of a number of competitors supplying Gatwick Airport from refineries located outside the south-east of England. In this respect, while it is undisputed that other operators supply jet fuel to Gatwick from production facilities not located in proximity of the airport, there still remains the fact that as a result of the merger the combined entity will be in a position to supply the jet fuel market from the best located refineries. In other words, the proximity of Exxon and Mobil refineries to Gatwick Airport constitutes a significant competitive advantage.

Creation of a dominant position in jet fuels supplies to Gatwick airport

(823) Exxon currently enjoys significant shares in the supply of jet fuels to Gatwick. As a result of the merger, this position will be further enhanced. The combined entity will be the supplier of [40 to 50]* % of jet fuels to Gatwick. The parties agree with this estimate. In addition to its own pipeline, Exxon will acquire Mobil's capacity shares in the three industry-owned pipelines ([...]* % of UKOP, [...]* % of West London and [...]* % of Walton Gatwick). Moreover, the combined entity will have the only competitive refining capacity in the region.

(824) Furthermore, according to competitors, Exxon has up to 1,1 million tonnes spare capacity on its own pipelines. It could use this extra capacity to capture any incremental growth in demand at Gatwick, which is anticipated to double over the next 10 years. Other suppliers have no spare refining or pipeline capacity to exploit any further market growth.

(825) A competitor and one major airline customer operating in the United Kingdom airports, including Gatwick, have complained about the supracompetitive position that the combined entity will acquire with respect to jet fuels supply to Gatwick. On the basis of the elements explained in the preceding paragraphs, it is the Commission's view that the operation will result in the creation of a dominant position in the market for the supply of aviation fuels to Gatwick airport, in the United Kingdom.

IV. COMMITMENTS AND ASSESSMENT

(826) On 3 September 1999, the parties offered certain commitments to remove the competition concerns which the Commission had identified in its statement of objections of July 26, 1999. On 20 September 1999, the parties submitted amended commitments taking into account certain adjustments required by the Commission in view of, in particular, the results of the market test. The commitments will be summarised and assessed in the following points, following the order of the relevant markets on which the Commission stated objections as followed above in the assessment part of this decision.

(827) The full text of the commitments is attached to this Decision and forms an integral part of it.

A. NATURAL GAS

DUTCH WHOLESALE TRANSMISSION MARKET

(828) Within [...]* from the date of the Commission's decision approving the concentration ("the effective date"), the parties will reach a binding agreement on the sale of Mobil Europe Gas Inc. ("MEGAS") to a purchaser to be approved by the Commission. MEGAS shall be sold as a viable going concern, including its contracted supply and sales portfolio, and all associated transport and other service agreements. The parties shall also nominate an independent and experienced trustee to oversee the ongoing management of MEGAS pending completion of the sale.

(829) MEGAS is the Mobil entity that sells natural gas in competition with Gasunie on the Dutch market. This company shall be sold with its contracted supply portfolio. It can be noted that Mobil's upstream production affiliate in the Netherlands has concluded depletion contracts for all of its concessions with Gasunie, MEGAS or Mobil's German affiliate. The rights and obligations of MEGAS under these contracts will be passed on to the purchaser so that the purchaser will have access to Dutch gas in the same proportion as that to which MEGAS would have had access to if it remained within the Mobil group. Therefore, the purchaser will be in the same unique position with regard to the TPA contracts with Gasunie as Mobil before (see recital 218). As the purchaser has to be a viable existing or potential competitor with the ability to maintain and develop MEGAS as an active competitive force, the Commission will take into account the possibility for the purchaser to have access to natural gas from other sources than the Netherlands and to bring this gas to the Dutch market. If these conditions (that are to be checked by the Commission) are fulfilled, the status quo ante to the transaction will be re-established so that the transaction will, subject to fulfilment of the commitment, not lead to the strengthening of Gasunie's dominant position in the Dutch wholesale transmission market.

GERMAN LONG-DISTANCE WHOLESALE TRANSMISSION MARKET(S)

Divestiture of Exxon's 25 % interest in Thyssengas

(830) Within [...]* from the effective date, the parties will reach a binding agreement on the sale of Exxon's 25 % equity interest in Thyssengas GmbH to a third party approved by the Commission. Within 10 days from the effective date, the parties will nominate an independent and experienced trustee to represent Exxon's interest and exercise Exxon's 25 % voting rights in Thyssengas.

(831) The Commission considers that, overall, the divestiture of Exxon's equity interest in Thyssengas compensates for the strengthening of the dominant positions that would otherwise have occurred as a consequence of the merger. This strengthening relates to Mobil's activity on the long-distance market with a current market share of around 1,5 % and Mobil's equity interest of 7,4 % in Ruhrgas. It is noted in this respect that Thyssengas has a market share of around 7,3 % on the long-distance market and is a member of the defined German oligopoly.

(832) Once the parties are no longer involved in Thyssengas, the Commission considers that this company has more incentives to challenge the other oligopoly members and, in this way, "replaces" the potential of Mobil. This was contested by a third party, which argued that Exxon's withdrawal from Thyssengas will not move Thyssengas out of the oligopoly group and turn it into a competitor, compensating for the elimination of Mobil. This third party explained that this was due to the remaining equity links between Thyssengas and the other oligopoly members via RWE-DEA (50 % shareholding in Thyssengas) and Shell (25 % shareholding in Thyssengas).

(833) For the following reasons, the Commission is satisfied that the proposed withdrawal from Thyssengas will restore competition up to its pre-merger level. First of all, Mobil was also linked to Ruhrgas and operated closely together with BEB for its German production. Secondly, RWE-DEA's shareholding and influence in Thyssengas are clearly more important than its 3,5 % shareholding in Ruhrgas (where it is a member of the Bergemann pool) and its obligation to supply 60 % of its German gas production to Ruhrgas (this supply volume accounts for less than 1 % of German natural gas consumption). RWE-DEA has therefore all the incentives to develop and strengthen Thyssengas. With regard to Shell, it is noted that as a consequence of Exxon merging with Mobil and Exxon divesting from Thyssengas, the interests between Exxon and Shell will, for the first time in more than 30 years, no longer be parallel on the German natural gas markets. It is difficult to assess the impact thereof on the competitive behaviour of the different ventures (competitive interaction between BEB-Thyssengas, BEB-Ruhrgas, Thyssengas-Ruhrgas and so forth), but it can be concluded that this brings back an element of instability into the operation of the oligopoly.

(834) In assessing the suitability of the purchaser of Exxon's equity interest in Thyssengas, the Commission will duly take into account the incentives for the competitive behaviour of Thyssengas attributable to the identity of the purchaser.

Voting rights in Erdgas Münster

(835) The parties shall use their reasonable best efforts to obtain the agreement of the other shareholders in EGM to reallocating a certain percentage of the voting rights currently held by Mobil to such shareholders (pro rata to their current voting rights) so that BEB and Mobil (or Exxon Mobil) combined will hold less than 50 % of the voting rights in EGM. The commitment contains a reallocation for both legal entities within EGM.

(836) As part of the market test of the proposed commitments, the other EGM shareholders have indicated to the Commission that they do not oppose the reallocation along the lines of the details of the commitment. On this basis and in view of the fact that the agreement with the shareholders will be subject to Commission approval, the Commission is confident that the parties will not obtain the control of the corporate governance structures of EGM

(837) Therefore EGM's incentives to compete will return to their pre-merger level. The Commission considers that EGM maintains the possibility to develop into a potential competitor on the long-distance wholesale transmission market in a future liberalised environment.

UNDERGROUND STORAGE FACILITIES SERVICING THE SOUTH OF GERMANY

(838) For a period of 10 years following the effective date [...]*, the parties will offer to enter into a binding agreement with third parties for the sale of all Mobil's rights to one or more of the depleted reservoirs suitable for conversion into storage facilities servicing the south of Germany until it has sold a combined estimated working gas volume, after conversion, of approximately 600 MCM.

(839) If the parties and a potential purchaser fail to agree on a fair market price within six months from that purchaser's written expression of interest in a particular reservoir, the parties will, at the potential purchaser's request, acquire a fair market valuation from three independent experts approved by the Commission. The parties agree that they will be bound by the value assessment of such experts.

(840) The commitment to sell one or more depleted reservoirs with a total estimated working gas volume of 600 MCM would represent the possibility to increase the current relevant storage volume by around [...]* %. It is also clear from the market investigation that such a volume would be sufficient to compete with Ruhrgas in the relevant market. No respondent to the Commission's market test has objected to the proposed remedies as a means to remove the competition concerns. Therefore, the Commission considers that the commitment ensures that the barriers to entry for potential competitors will not be increased as a consequence of the concentration. It can therefore be concluded that in those circumstances, Ruhrgas' dominant position in the market will not be strengthened.

B. BASE OILS

(841) Within [...]* from the effective date, the parties have committed themselves to transfer control (within the meaning of the Merger Regulation) over one or more base oil businesses together encompassing approximately [...]* barrels per day of base oil manufacturing capacity to BP Amoco and/or one or more third parties to be approved by the Commission, in one of the following ways or a combination thereof:

(a) transfer (or return) of ownership [...]* or equivalent capacity; or

(b) a long-term lease agreement or similar arrangement of [...]*, renewable at the lessee's option, pursuant to which the operational and strategic management of the base oil capacity concerned is transferred, such that the transferee can unilaterally decide, without informing the parties, on all operational and strategic matters. Such arrangement is to be subject to approval by the Commission and the lease agreement will provide that the lessor will not have any access to competitively sensitive information of the lessee. Where such information is necessary in the exercise of the normal rights of the lessor, this information will be supplied to an independent representative subject to confidentiality obligations with the exception of a limited number of items (such as spills that require action on the part of the lessor), to be identified in the lease agreement.

(842) In addition to the base oil manufacturing facilities, the base oil business(es) will include all necessary personnel, supply contracts, customer lists and contracts related solely to the divested facilities, access to technology and the provision of technical support to run a base oil business.

(843) During the transition period, the lubricant leg of the BP/Mobil joint venture (which includes the base oil activities) will be managed entirely independently from the rest of the Exxon-Mobil group with no passing of information and no reporting to persons outside of the joint venture. Certain transitional measures with regard to the BP/Mobil JV that are explained in section C below are also clearly relevant for the base oils remedy. If the base oils business were not purchased by BP Amoco, the parties have undertaken to keep it separate from their activities by appointing a hold separate trustee.

(844) The undertakings also provide that the parties should help the purchaser in order to enable him to obtain the necessary quality accreditations. In addition, the parties have committed to transfer supply contracts solely related to the divested businesses to the purchaser. These contracts amount to more than [...]* % of the sales. Finally, Mobil will continue (at the option of the purchaser) to source what it currently uses from the divested refineries for a period of [...]*. This represents around [...]* % of the production of these refineries.

(845) The resulting limitation of overlap from the concentration will depend on the outcome of the ongoing negotiations between BP Amoco and Mobil on the dissolution of the lubricant leg of their joint venture. However, the commitments ensure that, whatever the outcome of the negotiations on the dissolution of the lubricant leg of the BP/Mobil JV, Exxon Mobil's share of the merchant market would stand at no more than around [less than 40]* %. On this basis, the modified transaction would not result in the creation of a dominant position held by Exxon Mobil on the base oils market.

(846) When, instead of a sale of the manufacturing plants, control has to be changed by means of a long-term lease arrangement, the Commission also needs to approve such an agreement. In this respect, the Commission considers that the outline of such an agreement as described in the commitments is a sufficient starting basis for the Commission to assess whether the agreement would lead to a change of control and allow the lessee to operate the business as an active competitive force.

(847) Third parties explained that a lease arrangement would be workable in principle if the term of the lease was sufficiently long. They also explained that any purchaser of the divested businesses should have some lubricant-blending activities in order to be a viable competitor. The Commission believes that a 15 years period renewable at the option of the purchaser should ensure enough duration for the purchaser to operate the business as a long term undertaking. As to the importance of having a lubricant activity for the purchaser, the Commission will verify the suitability of a purchaser or lessee at the time the parties will submit a proposal to the Commission and the integration of the base oils business within its other activities shall be taken into account.

C. MOTOR FUEL RETAILING

(848) In response to the Commission's conclusion that the concentration will create or strengthen oligopolistic dominance over Austria, French toll motorways, Germany, Luxembourg, the Netherlands and the United Kingdom, the parties have undertaken to dispose of Mobil's share in Aral and to terminate Mobil's participation in the fuels part of the BP/Mobil joint venture.

ARAL

(849) Mobil's rights in Aral are held by Mobil Marketing und Raffinerie GmbH (MMRG), a wholly owned subsidiary of Mobil. [...]*.

(850) The parties commit to withdraw completely from Aral within [...]* from the effective date. [...]*

(851) [...]*

(852) This undertaking entirely removes the incentives for Exxon or BP/Mobil and Aral to coordinate their competitive behaviour and therefore eliminate the objections raised by the Commission.

BP/MOBIL JOINT VENTURE

(853) The parties undertake to withdraw entirely, within [...]* from the effective date, from the joint venture fuels activities. This should be achieved through one of the two following alternative remedies: (i) sale of Mobil's interest in the joint venture to BP Amoco (or, should BP Amoco not wish to buy this share and accept a third party, to a suitable third party), or (ii) dissolution of the BP/Mobil joint venture, with essentially all of the fuel assets being acquired by BP Amoco or a suitable third party.

(854) During the interim period, the parties have undertaken (i) to appoint independent representatives at the joint venture committee, who will vote in the best interest of the joint venture and will not exchange any fuels, base oil or lubricants information with Mobil or Exxon-Mobil and (ii) to eliminate access by personnel of the JV lubricants business (operated by Mobil) to information relating to the fuels business.

(855) The Commission considers that this offers the best protection possible during the interim period. The Commission recognises that the implementation of these transitional measures depends to some extent on the other partner in the joint venture, BP Amoco. However, as an operator of the fuels business, BP Amoco will have an incentive to make sure that no information can be transmitted to Exxon Mobil. Nevertheless, BP Amoco has objected to the appointment of independent representatives at the joint venture committee, arguing that this amounts to an assignment of Mobil's rights which is prohibited by contract. It seems, however, that the fact that the parties commit themselves vis-à-vis the Commission to appoint certain independent representatives, to be approved by the Commission, might not preclude the parties from nominating these representatives as Mobil representatives pursuant to the JV agreement. It is also not sure that BP Amoco and Mobil will find ways to agree on how to isolate the information systems relating to the fuel businesses from the lubricant businesses. However, the parties have committed to cut unilaterally such links if they could not agree with BP Amoco within a short time. The parties have also committed to appoint an independent observer, whose task will be to monitor their compliance with the offered remedies.

(856) BP Amoco has argued that the time period for Mobil's withdrawal from the BP/Mobil joint venture should not be longer than [...]*. This is based on two reasons. First, Mobil would not have to look for a buyer since only BP Amoco could purchase their interest in the BP/Mobil joint venture. Secondly, the interim period would be harmful to competition because of the possibility of Exxon-Mobil's benefiting from information relating to BP/Mobil fuel-retailing activities.

(857) However, the Commission considers that the duration of the interim period is reasonable. In determining the right time period, the Commission had to balance on the one hand the need for a quick resolution of its competition concerns together with the possible harm the continued presence of Mobil within the JV might do to one of the main competitors to Exxon against, on the other hand, the need for the parties to get a proper consideration in the face of a monopolist buyer. Indeed, BP Amoco could be inclined to dictate the terms of the sale should the interim period be too short. This interim period should not be used by the parties to unduly prolong Mobil's participation in the JV. That is why the commitments also provide for (i) an obligation on Mobil to accept an expert resolution of any issues linked to the valuation of Mobil's interest should BP Amoco so desire and (ii) an obligation on the parties to negotiate in good faith with BP Amoco. The reports of the independent observer will inform the Commission on all these issues, thereby enabling the Commission to order under certain circumstances a shorter time limit for the divestiture of Mobil's stake or the dissolution of the joint venture.

CONCLUSION

(858) The commitments concerning fuels would eliminate all of the overlap between Exxon's and Mobil's fuels business in all of the markets where the Commission concluded that an oligopolistic dominance might be created or strengthened, and possibly in the whole of the fuels sector in Europe. Consequently, such dominant positions will not be created or strengthened.

D. AVIATION LUBRICANTS

(859) Within [...]* from the effective date, the parties will divest themselves of Exxon's worldwide aviation lubricants business with commercial airlines to a purchaser approved by the Commission. The business comprises product R & D, marketing and sales, blending and packaging facilities, product quality management and technical support and the associated marketing and sales, product quality management and technical support personnel. A trustee will ensure that, during the interim period, the Exxon and Mobil businesses are kept separate.

(860) The market test conducted by the Commission confirmed that the divestiture of Exxon's aviation lubricant activities would be preferable to that of Mobil's aviation lubricants business (as initially proposed by the parties). Indeed, Mobil's activities are integrated within the Mobil group. Mobil produces its own ester base stock and also produces five proprietary additives. It did not appear to be possible to offer Mobil's blending facility as it was part of a wider complex. Respondents to the Commission's market test therefore indicated their doubts as to the capacity of a purchaser of Mobil's aviation lubricant business to operate as a viable competitive force able to constrain Exxon Mobil's actions. When informed about the Commission's doubts as to the adequacy of the divestiture of Mobil's aviation lubricant business, the parties offered to shed Exxon's business. Contrary to Mobil's aviation lubricant business, Exxon's aviation lubricants business is not closely intertwined with other Exxon businesses (supply of raw materials to a large extent from third parties, separate blending facility).

(861) On this basis, the Commission considers that no dominant position will be created on this market and that its competition concerns expressed in the statement of objections are eliminated.

E. AVIATION FUELS

(862) The parties have undertaken to sell, within [...]* from the effective date, aviation fuel pipeline capacity from the Coryton refinery to Gatwick airport equivalent to Mobil's 1998 sales volumes at Gatwick. This undertaking should remove the bottleneck control that the parties would have gained through their merger on the supply of jet fuels to Gatwick Airport. Therefore, no dominant position will be created on the market for the supply of aviation fuels to Gatwick airport,

HAS ADOPTED THIS DECISION:

Article 1

On condition that the commitments summarised in recitals 826 to 862, and laid out in detail in the Annex, are fully complied with, the concentration notified on 3 May 1999 consisting of the merger between Exxon Corporation and Mobil Corporation is declared compatible with the common market and the functioning of the EEA Agreement.

Article 2

This Decision is addressed to:

1. Exxon Corporation 5959 Las Calinas Boulevard Irving Texas USA

2. Mobil Corporation 3225 Gallows Road Fairfax Virginia USA

Done at Brussels, 29 September 1999.

For the Commission

Mario Monti

Member of the Commission

(1) OJ L 395, 30.12.1989, p. 1; corrected version published in OJ L 257, 21.9.1990, p. 13.

(2) OJ L 180, 9.7.1997, p. 1.

(3) OJ C 87, 7.4.2004.

(4) Case IV/M.727, BP/Mobil of 7 August 1996 (OJ C 381, 17.12.1996, p. 8). The activities of the joint venture do not include marine and aviation lubricants.

(5) Turnover calculated in accordance with Article 5(1) of the Merger Regulation and the Commission notice on the calculation of turnover (OJ C 66, 2.3.1998, p. 25). To the extent that figures include turnover for the period before 1 January 1999, they are calculated on the basis of average ecu exchange rates and translated into euro on a one-for-one basis.

(6) Parts of this text have been edited to ensure that confidential information is not disclosed; those parts are enclosed in square brackets and marked with an asterisk.

(7) See Commission decision in Case IV/M.1200, Arco/Union Texas (OJ C 16, 21.1.1999, p. 8) Commission decision in Case IV/M.88, Elf Enterprise (OJ C 203, 2.8.1991) and Commission decision in Case IV/M.85, Elf Occidental (OJ C 160, 20.6.1991).

(8) Prices in the crude oil trade markets may be significantly influenced by mere announcements following an OPEC meeting.

(9) OJ L 204, 21.7.1998, p. 1.

(10) The coal and steel aspects of the case have been dealt with by the Commission in Case IV/ECSC.1252, RAG/Saarbergwerke. The German decision is summarised in the BKartA's biannual working report and its press release of 4 February 1998.

(11) It is not necessary for the purpose of this case to decide whether there is a separate LCV and HCV gas market in the Netherlands, as the outcome of the case would not be materially different.

(12) [...]*

(13) [...]*

(14) [...]*

(15) [...]*

(16) [...]*

(17) [...]*

(18) [...]*

(19) Joined Cases C-68/94 and C-30/95 France and Others v Commission (1998) ECR I-1375.

(20) Judgment of 25 March 1999 in Case T-102/96 Gencor v Commission (not yet reported).

(21) DEUDAN (pipeline bringing Danish gas to Germany): [...]* %, BEB [...]* %; NETRA (pipeline bringing Norwegian gas to North Germany): BEB [...]* %, Ruhrgas [...]* %; Salzwedel-Berlin (extension of NETRA to Berlin): BEB [...]* %, Ruhrgas [...]* %, VNG [...]* %; NETG (pipeline bringing Dutch LCV gas to Germany): Ruhrgas [...]* %, Thyssengas [...]* %; METG (pipeline bringing Dutch HCV gas to Germany): Ruhrgas [...]* %, Exxon and Shell each [...]* %; SETG (extension of METG): Ruhrgas [...]* %, Exxon and Shell each [...]* % (the NETG-METG-SETG pipelines are extensions of one another); EVG (pipeline link in Central Germany): Ruhrgas [...]* %, VNG [...]* %.

(22) [...]* - Annex 49.21 (italics added). Also [...]* - Annex 49.26.

(23) Table 3.13 in "Competition and Liberalisation in European Gas Markets - A Diversity of Models", Jonathan P. Stern, Royal Institute of International Affairs, 1998.

(24) Bresnahan, T.F. and Salop, S.C., "Quantifying the competitive effects of joint ventures", International Journal of Industrial Organisation, vol. 4 (1986) pp. 155-175.

(25) API stands for American Petroleum Institute and regroups oil companies; ACEA stands for association des constructeurs européens d'automobile and ATIEL stands for association technique de l'industrie européennes des lubrifiants. Their respective roles are further explained in this section.

(26) Group II is not currently used in the EEA.

(27) As confirmed by the competitors, all DI manufacturers have generic approvals in Exxon, Mobil and Shell base oils that are available for base oil customers.

(28) Exceptions exist for group III and IV base oils.

(29) As described in documents such as "API 1509" and "Atiel code of practice".

(30) Case IV/M.1137, Exxon/Shell of 25 May 1998.

(31) [...]*

(32) [...]*

(33) [...]*

(34) [...]*

(35) [...]*

(36) [...]*

(37) [...]*

(38) [...]*

(39) [...]*

(40) BP/Mobil have spent approximately USD [...]* million in order to obtain interchangeability between its (then) [...]* European refineries.

(41) [...]*

(42) [...]*.

(43) [...]*

(44) [...]*

(45) [...]*

(46) Toll blending refers to the blending of lubricants under the brand of and according to the specifications of a lubricant blender (who may be without production capacity in that region).

(47) [...]*.

(48) [...]*.

(49) [...]*.

(50) [...]*.

(51) [...]*.

(52) [...]*.

(53) [...]*.

(54) [...]*.

(55) [...]*.

(56) [...]*.

(57) [...]*.

(58) [...]*.

(59) Case T-102/96, Gencor v Commission, not yet published.

(60) Since no industry data on average pump prices on motorways are available, Exxon pump prices on motorways are taken as a proxy of the industry average pump prices. Similarly, the parties could not isolate data concerning solely toll motorways. The Commission therefore took data covering all motorways as a proxy for toll motorways.

(61) Account should also be taken of the fact that Total has recently acquired Fina which has the poorest performance in terms of throughput per site.

(62) [...]*.

(63) On the basis of the market shares of the main eastern German competitors 1998 provided by the parties.

(64) [...]*.

(65) [...]*.

(66) Study Bayerische Landeskartellbehörde 1998.

(67) [...]*.

(68) Competition in the supply of petrol in the United Kingdom - A report by the Office of Fair Trading: May 1998.

(69) See footnote 68.

(70) Auxiliary power units, constant speed drivers, integrated drive generators and air turbine starters.

(71) Anecdotal evidence is that, in order to reduce possible safety risks, airlines use one and the same OEM-approved lubricant for stationary and airborne applications to ensure that a non-approved aero-derivative lubricant cannot be mistakenly used in an aircraft engine.

(72) In its decision of 6 July 1994 (Case IV/M.460 - Holdercim/Cedest) (OJ C 211, 2.8.1994), the Commission had stated that markets representing "less than one per cent of the Community consumption of the product" constitute an insubstantial part of the common market.

ANNEX

The full original text of the conditions and obligations referred to in Article 1 may be consulted on the following Commission website:

http://europa.eu.int/comm/ competition/index_en.html

CONTENTS

>TABLE>