On 26 January 2022, the EU’s General Court (GC) annulled the European Commission’s (EC) €1.06 billion fine on Intel for abusing a dominant position with its rebate schemes. The judgment demonstrates that the European Courts are prepared to look in detail at evidence and economic analysis in antitrust cases, and will annul EC decisions if they do not undertake an effects-based analysis or if their economic analysis has flaws.
|2000: Complaint to the EC by competitor Advanced Micro Devices (AMD).
May 2009: The EC fined Intel €1.06 billion (a then record) for abusing its dominant position in the market for x86 central processing units (“CPU”, often referred to as a computer’s “brain”)) between 2002 and 2007. Intel was found (i) to have granted exclusivity rebates to some computer manufacturers and one retailer, and (ii) to have made direct payments to computer manufacturers to halt or delay the launch of specific products containing competitors’ x86 CPUs (so-called “naked restrictions”).
The EC considered that these practices were by their nature illegal, and that it was not necessary for them to demonstrate that they had anti-competitive effects. However, the EC did perform an “as-efficient competitor” (AEC) analysis – the first time after its 2009 Guidance Paper – to demonstrate that the rebates and payments in question made it impossible for an as-efficient competitor to compete profitably.
2014: The GC rejected Intel’s appeal and upheld the EC’s decision in its entirety. It held that loyalty rebates can be categorized as abusive with no effects-based analysis being required. Intel appealed the judgment.
2017: In a seminal ruling, the Grand Chamber of the ECJ set aside the GC judgment because the GC had not examined the EC’s analysis of the AEC test and all of Intel’s arguments concerning that test, and referred the case back to the GC for a new examination.
The ECJ confirmed the previous (more formalistic) case law (Hoffmann-La Roche) on exclusivity rebates but stated that this case law needed further “clarification” where a company submits on the basis of supporting evidence that its conduct was not “capable of restricting competition and, in particular, producing the alleged foreclosure effects” (para.138). In reality, the ECJ’s judgment recalibrated its case law and stressed the “as efficient competitor principle” and the requirement to run an effects-based analysis in abuse of dominance cases. The judgment (para. 139) set out five factors to take into account in the foreclosure analysis, including the coverage (what market share was covered by the rebates) and the rebates’ duration.
For further analysis of the ECJ judgment, please see our client alert here.
The GC judgment
In its 102-page judgment, the GC applied a vigorous assessment of the EC’s analysis and held that it was incomplete and did not establish to the requisite legal standard that the rebates at issue were capable of having, or likely to have, anti-competitive effects. This is not something customary. There has only been one full annulment of an EC decision based on Article 102 since 1979, which was the Servier case – but in Servier, the Court upheld part of the Article 101 side of the decision. The only part of the EC Intel decision that survived is certain findings of the EC on the existence of “naked restrictions”. This part had not been appealed by Intel in 2014, hence it was outside the scope of the GC’s new judgment. The (now annulled) EC findings on the rebates were the most important part of the decision. And, obviously, the EC must now repay a hefty amount, corresponding to the amount of the fine plus interest corresponding to the rate applied by the European Central Bank to its principal refinancing operations, increased by three and a half percentage points.
The test for the assessment of rebates
The GC first set out the method defined by the ECJ in its 2017 judgment for assessing whether a system of rebates has the capacity to restrict competition. It noted (paras 123-126) that “three instructive inferences” can be drawn from the Hoffmann-La Roche case law.
“Although a system of rebates set up by an undertaking in a dominant position on the market may be characterised as a restriction of competition, since, given its nature, it may be assumed to have restrictive effects on competition, the fact remains that what is involved is, in that regard, a mere presumption and not a per se infringement of Article 102 TFEU, which would relieve the Commission in all cases of the obligation to conduct an effects analysis”. It is noteworthy here that the finding that the rebates in question were exclusivity rebates had not been called into question by the ECJ (see paras 97-98 of the judgment).
Where a company in a dominant position submits, during the administrative procedure, on the basis of supporting evidence, that its conduct was not capable of restricting competition and, in particular, was not capable of producing the foreclosure effects alleged against it, the EC must analyse the foreclosure capacity of the scheme of rebates. In the context of that analysis, “having regard to the wording of paragraph 139 of the judgment on the appeal, the Commission is, as a minimum, required to examine those five criteria for the purposes of assessing the foreclosure capability of a system of rebates, such as that at issue in the present case”. These are:
- the extent of the company’s dominant position on the relevant market;
- the share of the market covered by the contested practice;
- the conditions and arrangements for granting the rebates in question;
- their duration and their amount; and
- the possible existence of a strategy intended to exclude competitors that are at least as efficient as the dominant company from the market.
As far as the AEC test is concerned, “where the Commission has carried out an AEC test, that test is one of the factors which must be taken into account by the Commission to assess whether the rebate scheme is capable of restricting competition”.
The GC found (paras 144 and 145) that the EC had made an error or law in its infringement decision:
“… the Commission inferred from the Hoffmann-La Roche case-law, first, that the rebates at issue were by their nature anticompetitive, with the result that there was no need to demonstrate foreclosure capability in order to establish an infringement of Article 102 TFEU. Second, although the contested decision contains an additional analysis of the foreclosure capability of those rebates, the Commission took the view that, in accordance with that case-law, it was not required to take that analysis into account in order to conclude that those rebates were abusive. Third and lastly, still on the basis of that same case-law, the Commission held, inter alia, that a number of factors were irrelevant for determining whether there was abuse.
… that position is not consistent with the Hoffmann-La Roche case-law, as clarified by the Court of Justice in paragraphs 137 to 139 of the judgment on the appeal. It must therefore be found that [Intel is] correct in maintaining that the Commission vitiated the contested decision by an error of law in taking as a starting point the premiss that, in essence, the Hoffmann-La Roche case-law allowed it simply to find that the rebates at issue infringed Article 102 TFEU on the ground that they were by their very nature abusive, without necessarily having to take account of the capability of those rebates to restrict competition in order to reach the conclusion that they constituted an abuse.”
The burden of proof and the standard of proof required
The GC placed particular emphasis on the standard of proof that the EC must satisfy in order to establish the existence of an abuse of dominance. The starting premise must be the presumption of innocence, which applies in the competition law area. Thus, the standard of proof applies in the following way:
- Where the EC maintains that the established facts can be explained only by anticompetitive behaviour, if the companies concerned put forward a separate plausible explanation of the facts, the alleged infringement at issue has not been sufficiently demonstrated.
- Where the EC relies on evidence which is, in principle, capable of demonstrating the existence of an infringement, it is for the companies concerned to demonstrate that the probative value of that evidence is insufficient.
In case of doubt, the benefit should go to the defendant.
Errors in the EC’s AEC analysis
The GC then performed a painstaking assessment of the EC’s economic analysis and of the AEC test applied to the rebates that Intel granted to OEMs and one retailer. It found that some of the evidence put forward by Intel gave rise to doubts as to the correctness of the EC’s analysis and as to whether the rebates were capable of having a foreclosure effect throughout the whole of the relevant period.
The GC looked in detail at the application of the AEC test by the EC, and found a number of errors, including:
- errors in the calculation of the “contestable share”;
- errors affecting the value of conditional rebates;
- insufficiently substantiated extrapolations of the results for one single quarter-year period to the entire infringement period; and
- errors in the quantified assessment of the non-cash advantages at issue.
The GC refused on several occasions to take into account additional economic analysis (related to the application of the AEC test) that the EC provided to it during the court procedure. This was because the GC considered that taking into account of this evidence would lead to the GC substituting its own reasoning for that of the EC. The European Courts, in reviewing the legality of acts, cannot substitute their own reasoning for that of the author of the act in question.
Coverage and duration
The GC also reviewed whether the EC’s infringement decision took proper account of two out of the five criteria established by the ECJ in its 2017 Intel judgment for determining the capacity of pricing practices to have a foreclosure effect, in particular, the market coverage and the duration of the rebates at stake. It concluded that the EC’s analysis was incomplete.
The GC found that the EC had failed to determine the share of the market covered by the practice at issue. Indeed, the decision included no specific figure of market coverage, although the Commission was bound to analyse this. It also found that, although the EC had examined some factors relating to duration at its AEC analysis, there was never a full-fledged analysis of the duration of the rebates as evidence making it possible to determine the capacity of Intel’s pricing practices to have a foreclosure effect.
Lessons for the future
This ruling provides a welcome application of the principles established in the 2017 ECJ Intel judgment, namely that the EC must underpin its infringement decisions with rigorous effects-based assessment, and not rely on a formalistic approach based on “presumptions” of competitive harm.
Exclusivity rebates cannot be presumed to be an abuse of dominance once evidence is put to the EC of the absence of anti-competitive foreclosure. Companies can rely on economic analysis to assess the antitrust compliance of their rebate schemes, while the EC must carefully take such evidence into account. In short, formalism is out. Rigorous and careful analysis of anti-competitive foreclosure effects are what is needed now.
 Case T-286/09 RENV, Intel v Commission