Intel: General Court rewinds the clock on rebates

This article was produced by Nabarro LLP, which joined CMS on 1 May 2017.

Summary and implications

The European General Court (the Court) handed down its judgment in the Intel rebates case today. The Court dismissed in its entirety Intel's appeal against a finding from the European Commission (the Commission) in 2009 that it had abused a dominant position in the microchips market through a mixture of rebate schemes, exclusivity and other arrangements principally with its computer hardware customers. The Court also upheld the €1.06bn fine imposed – the largest ever on a single entity. But the case is of most relevance to business for the highly restrictive approach the Court takes to exclusivity and loyalty rebates.

Recap on Commission's findings

The case concerned two main types of practice:

1. discounts granted by Intel to four major computer manufacturers (Dell, Lenovo, HP and NEC) and to the retailer Media-Saturn on condition that they purchased all or most of their x86 central processing unit microchips (CPUs) from Intel (described as "exclusivity" or "loyalty rebates"); and
2. payments which were conditional on restrictions on the distribution by the customer of computers containing the CPUs of Intel's rival AMD, or on postponing or cancelling the launch of these products (described by the Commission as "naked restrictions").

Intel was found dominant in the worldwide CPUs market with a market share of 70 per cent or more, high barriers to entry and only one significant competitor in the form of AMD. Both types of practice were found to constitute unlawful abuse and the fine was levied.

Points of wider relevance in the appeal judgment

Intel appealed on several grounds, many of which go to the detailed characterisation of the individual incentive arrangements on the facts of the case. The appeal court judgment runs to some 300 pages, much of which is devoted to a detailed assessment of these factual grounds of appeal. What is of more interest for the wider business community, however, is the broader approach the Court has taken to the analysis the Commission is required to conduct in respect of exclusivity and loyalty rebates. The findings of the Court on the so-called naked restrictions are perhaps less surprising given the targeted nature of these restrictions.

Exclusivity rebates

The Court distinguished three types of rebates:

1. quantity rebates i.e. rebates linked to the volume of purchases made from the dominant firm (and based on efficiencies);
2. rebates conditional on the customer purchasing "all or most" of its purchases of the relevant product from the dominant firm; and
3. "other rebate systems where the grant of a financial incentive is not directly linked to a condition of exclusive or near-exclusive supply".

The Court took the view that the first type were generally considered not to foreclose competition; the second type were inherently unlawful without the need to show even a capability of foreclosing competition; and the third type were capable of being unlawful if, on a full analysis of all the circumstances, the scheme tended to foreclose competitors or strengthen dominance or restrict the buyer's freedom to choose his sources of supply.

The Intel case was concerned with type (2) in this three-fold categorisation – exclusivity rebates – and so that is where the Court's general observations centred.

Rejection of Intel's arguments on exclusivity rebates

The Court rejected as a matter of principle a series of arguments raised by Intel as to why its rebate schemes should not be considered abusive.

1. Intel argued that given that the conduct was historic the Commission was required to show actual adverse effects on competition. The Court rejected this: exclusivity rebates were "by their very nature" capable of restricting competition.
2. Intel pointed to the growth and success which AMD had enjoyed during the period in which the Commission had held it to have been foreclosed. The Court said, following the rejection of a similar argument in British Airways 15 years ago, it might have enjoyed even more success had it not been for the rebates.
3. Intel argued that the level of the rebate must be a relevant factor, citing the example of a one dollar rebate which AMD could easily match. The Court held that level of rebate is irrelevant. The Court similarly rejected arguments based on the customer's ability to terminate of some of the agreements on a month's notice, holding that "any financial incentive to purchase exclusively constitutes additional interference with the structure of competition on a market and must therefore be regarded as abusive."
4. Arguments on the rebates having limited coverage of the market and limited coverage of the customer's total requirements (28 per cent in the case of HP) were likewise rejected on grounds that competitors must be free to compete for the whole market.
5. Finally, Intel argued that the customers were powerful computer manufacturers which created bidding contests to extract lower prices from the two major suppliers, and that as part of these contests they offered some portion of their business for a short duration in exchange for lower prices. The Court rejected this, pointedly holding that it was of "no consequence" whether the buyer took the initiative.

Comment and business implications

What is striking about this judgment is just how rigid the approach is to exclusivity. Famously, there has been a disjunct for the last 9 or 10 years between the more formalistic approach of the European courts to exclusionary abuse (based on the Commission's former approach, and the fact that most of these cases arrive at the courts through judicial review) and the Commission's direction of travel, which is to focus on the likelihood of practices foreclosing competition. Far from bridging that gap, this case widens it. The narrow approach of the courts at least allows for contextual analysis. And that point has been emphasised, for example by the Advocate General in the Tomra case. It is reinforced by a string of pricing cases recently.

The Court sweeps all that aside for exclusivity rebates. None of Intel's arguments worked because in the view of the Court there is not even any need for contextual analysis for these kinds of rebates. The Court's position is that bar exceptional cases only making out an efficiencies defence will prevent exclusivity or loyalty rebates from constituting unlawful abuse when entered into by a dominant firm.

The dichotomy between the regulators' modernised approach to these cases and that of the courts is well illustrated here: the Commission applied in its decision a detailed economic analysis based on the "as efficient competitor" test, yet maintained in the appeal, successfully, that the whole exercise was unnecessary.

The rejection of the customer-led auctions also sits in sharp contrast with the limited carve-out for such auctions in the Coca-Cola settlement in 2005. The clear message is that firms will do better by settling these cases.

The judgment will pay careful reflection, and it may well be the subject of a further appeal by Intel to the Court of Justice. In the meantime, businesses will need to tread even more carefully with their incentive schemes in markets in which they have dominance risk.