Ever since the adoption in February 2009 of the European Commission’s Guidance Paper on its enforcement priorities in applying Article 102 (then 82) and in May 2009 of its decision in the Intel case, the correct analytical framework for rebates granted by dominant undertakings has been the subject of significant debate. The Commission sought to lay down a coherent framework in its August 2024 draft Guidelines on exclusionary practices, making a distinction between rebates conditional on the customer or supplier purchasing or supplying exclusively (or quasi-exclusively) from or to the dominant undertaking (“exclusivity rebates”) and conditional rebates not subject to exclusivity (“conditional rebates”).
Last week, the EU Court of Justice delivered its second judgment in the Intel case (“Intel II”). This completes the legal analysis of Intel’s rebates, which was the key point of contention in the case for the past 24(!) years. Are we closer to a coherent analytical framework now? Below, we compare the judgment to the draft Guidelines.
Recap – The two paragraphs that shook the competition world
Intel was the dominant supplier of microchips for personal computers in the 2000s. Its main (smaller) rival was AMD. In 2000, AMD first filed a complaint with the Commission about Intel’s business practices including rebates offered to Intel customers such as Dell, HP and Lenovo, alleging that these practices foreclosed AMD from the market.
In the early to mid-2000s, a big debate was ongoing in the EU around whether there should be a “more economic approach” to the analysis of abuses of a dominant position, which ultimately led to the Commission’s 2008 Guidance Paper (see our article on the topic here). The Commission decided to apply such a more economic approach in the Intel case, assessing Intel’s rebates not only in terms of their conditions (a formalistic approach), but also carrying out an “As-Efficient Competitor Test” (AEC Test) (an effects-based approach). It should be noted that the Guidance Paper was not applicable to the Intel case (as it was published after Intel had been given the opportunity to respond to the Commission’s charges), but the Commission stated that its infringement decision was “in line with the orientations set out in the guidance paper”. However, the Commission also stated that the AEC Test was “not indispensable” for finding an infringement. The Commission found that Intel’s rebates infringed Article 102 TFEU and imposed a fine of more than EUR 1 billion on Intel.
The Intel decision was subsequently upheld on appeal by the General Court. In its defence, the Commission strongly maintained that it was legally entitled to find an infringement on the sole basis of the existence of the rebates, without having to consider the factors that may demonstrate their capability of restricting competition. In its judgment, the General Court in essence held, first, that a finding that Intel’s rebates were illegal did not necessitate an examination of the circumstances of the case, so the Commission was not required to demonstrate the foreclosure capability of the rebates on a case-by-case basis. Second (and in the alternative), even if an assessment of the circumstances were required, it was not necessary to demonstrate anti-competitive effects by means of an AEC test. The General Court therefore did not need to vet the AEC Test carried out by the Commission in the decision. The decision was upheld.
Intel then appealed the General Court’s judgment to the Court of Justice, which delivered a landmark judgment in 2017 (“Intel I”). In two paragraphs that shook the competition world (138 and 139 of Intel I), the Court “clarified” its earlier and more formalistic case-law, in particular its judgment in Hoffmann-La Roche. The clarification related to what the Commission had to do if the undertaking submitted, during the administrative procedure, on the basis of supporting evidence, that its conduct was not capable of restricting competition and of producing the alleged foreclosure effects (para 138). In that case, the Commission is not only required to analyse, first, the extent of the undertaking’s dominant position on the relevant market and, secondly, the share of the market covered by the challenged practice, as well as the conditions and arrangements for granting the rebates in question, their duration and their amount; it is also required to assess the possible existence of a strategy aiming to exclude competitors that are at least as efficient as the dominant undertaking from the market (para 139).
Since the AEC Test had played an important role in the Commission’s assessment of whether the rebates were capable of foreclosing competitors as efficient as Intel, it was not open to the General Court not to assess it in the appeal. Therefore, the General Court’s judgment was set aside, and the case was referred back to it.
In the judgment on the referred case, the General Court for the first time examined Intel’s grounds of appeal relating to the AEC Test in detail. In that regard, it held that the Commission’s methodology in its AEC Test essentially boils down to comparing a customer’s “contestable share” of demand to the “required share” that an as-efficient competitor needs to enter the market without incurring losses. In this respect, the contestable share is the share of their demand that customers of a dominant undertaking are willing and able to switch to an alternative supplier (knowing that, where the dominant undertaking is an “unavoidable trading partner”, the customer will always want to buy a certain part of its demand (the uncontestable share) from the dominant undertaking). The General Court identified several flaws in the Commission’s AEC Test analysis, leading it to annul all aspects of the Commission’s decision insofar as they relate to Intel’s rebates.
Battle lines are drawn
Following these developments, there was plenty of debate over the extent to which there remained scope for a per se approach to rebates conditional on exclusivity. More generally, the question was raised whether the “more economic approach” had gone too far. This culminated in the Commission publishing its draft Guidelines in August 2024, to replace the Guidance paper issued in February 2009.
Unsurprisingly, the draft Guidelines cover rebates. Indeed, the Commission interpreted the position outlined in Intel I as a “presumption” against exclusivity rebates. For any conduct against which there is such a presumption, the draft Guidelines specify that “a dominant undertaking can seek to rebut the probative value of the presumption in the specific circumstances at hand by submitting, on the basis of supporting evidence, that the conduct is not capable of having exclusionary effects”. It would then be for the Commission either to show that the arguments and supporting evidence submitted by the dominant undertaking are insufficient to call into question the presumption or demonstrate the ability of the conduct to have exclusionary effects. When doing the latter, it added specifically on exclusivity clauses and exclusive dealing that this would “typically” involve factors such as the extent of the firm’s dominant position, the market coverage of the rebates, and their duration, as well as the possible existence of a strategy aimed at excluding actual or potential competitors of the dominant firm.
This, of course, was a clear reference to paragraph 139 of Intel I. However, the Commission added that such an exclusionary strategy is not legally required to establish the conduct’s capability to produce exclusionary effects. Therefore, the Commission’s position on the test for exclusivity rebates was clear: presumed unlawful. If the undertaking seeks to rebut the presumption, the Commission will assess the rebuttal and may decide that the undertaking has not done enough, in which case it will still not need to look at the various factors that may be relevant to whether a rebate is capable of restricting competition. Only if the undertaking has persuaded the Commission that it has called into question the presumption, will the Commission seek to demonstrate the capability of the conduct to have exclusionary effects, but it will not be legally required to prove the existence of a strategy to exclude as-efficient rivals.
Intel II
Of course, in the background the Court of Justice was making up its own mind on the Commission’s appeal that led to the Court’s judgment in Intel II. It is a judgment of just three Court of Justice judges, as opposed to Intel I which was handed down by the Grand Chamber. However, it provides vital clarifications on rebates and, perhaps, on exclusionary conduct more generally. We set out the key takeaways below, focussing on the impact of the Court’s findings on the draft Guidelines.
So long, Hoffmann-La Roche?
Ultimately, the Intel case was to a large extent about whether the Court’s findings in its Hoffmann-La Roche judgment remained valid. Instead of overturning that judgment, the Court “clarified” it in Intel I, by essentially saying that the formalistic approach in Hoffmann-La Roche remained good law, except if the dominant undertaking submitted, during the administrative phase, that its conduct was not capable of restricting competition, and in particular of producing the alleged foreclosure effects. If that was so, the Commission had to do more, namely analyse the extent of the undertaking’s dominant position on the relevant market, the share of the market covered by the conduct, the conditions and arrangements for granting the rebates in question, their duration and their amount, and the possible existence of a strategy aiming to exclude competitors that are at least as efficient as the dominant undertaking from the market.
Although the Court did not overturn Hoffmann-La Roche, it clearly defanged it. In particular, it did not appear to attach any material requirements to what the dominant undertaking needs to do to trigger the Commission’s additional burden of proving the additional factors listed above, including proving the existence of a strategy aimed at excluding as-efficient competitors. The only requirements are making the submission during the administrative phase and submitting “supporting evidence”.
Nonetheless, the Commission cites Hoffmann-La Roche 10 times in the parts of its draft Guidelines that deal with exclusive dealing and exclusivity rebates. It was therefore interesting to see whether, in Intel II, the Court continued to pay homage to its old case-law. The answer is no. The judgment does not even feature once in Intel II except in an introductory paragraph. It seems the Court is ready to move on and to view paragraphs 137-139 in Intel I as the legal test for rebates, relegating the analysis in Hoffmann-La Roche to the history books.
Is there a “presumption” against exclusivity rebates?
But the implications for the Commission’s draft Guidelines go further than changing a few footnote references to Hoffmann-La Roche. The judgment in Intel II also calls into question the approach the Commission intends to take in the draft Guidelines to the undertaking’s submission of evidence that triggers a requirement for the Commission to prove the capability of the rebates to restrict competition.
The Commission takes the view in the draft Guidelines that exclusivity rebates are presumed to have exclusionary effects. The undertaking can “seek to rebut the probative value of the presumption in the specific circumstances at hand by submitting, on the basis of supporting evidence, that the conduct is not capable of having exclusionary effects”. The Commission then sets itself the task of examining whether the presumption is rebutted. According to the draft Guidelines, infringement of Article 102 is confirmed if the Commission (i) shows that the arguments and supporting evidence submitted by the dominant undertaking are insufficient to call into question the presumption; or (ii) provides “evidentiary elements” demonstrating the capability of the conduct to have exclusionary effects (draft Guidelines, para 60(b)).
However, when reading Intel II it does not appear that this is what the Court had in mind in paragraphs 138 and 139 of Intel I. In particular, paragraph 277 of Intel II clarifies what “trigger[s]” the Commission’s “obligation” to assess the possible existence of a strategy aiming to exclude as-efficient competitors from the market. It is not the fact that Intel managed to “call into question” a presumption, but the fact that Intel submitted, during the administrative procedure, on the basis of supporting evidence, that its conduct was not capable of restricting competition. The same follows from paragraphs 330 and 331 of Intel II: if the dominant undertaking makes such a submission during the administrative procedure, on the basis of supporting evidence, then this “requires” the Commission to carry out an analysis to determine the existence of the capability of the dominant undertaking’s conduct of restricting competition. That the Commission is “required” to do so, is also clear from other paragraphs in the judgment (e.g., para 180). In other words, Intel submitted a reasoned defence, supported by evidence. That alone meant that the Commission had to engage with that defence and prove the capability of Intel’s rebates to restrict competition.
While the submission by the dominant undertaking must of course meet certain basic thresholds, it is clear from Intel II that the Court did not have in mind that the submission must be sufficient to call into question a presumption. Rather, the requirement set in paragraph 138 of Intel I and throughout several paragraphs of Intel II is a gateway question: something the dominant undertaking must do to “trigger” the need for the Commission to dig deeper. If the dominant undertaking makes a reasoned submission supported by evidence that its conduct was not capable of restricting competition, then that does not first require an interim decision by the Commission about whether the submission is sufficient to rebut the presumption and thus shift the burden to the Commission.
The concept of a “presumption” carries within it a shifting of the burden of proof. Here, the burden does not shift. The dominant undertaking must decide to contest the case during the administrative procedure, and it must do so on the basis of supporting evidence, but the burden remains on the Commission. Once the submission is made, this triggers the requirement for the Commission to dig deeper and provide additional evidence and analysis.
AEC Test, back from the dead
The AEC Test was pronounced dead by the General Court, resurrected by the Court of Justice, and condemned to insignificance in the draft Guidelines. But it is back with a vengeance in this judgment.
As already explained above, if the undertaking submits, supported by evidence, that its rebates were not capable of having the alleged foreclosure effects, the Commission must analyse the possible existence of a strategy to exclude as-efficient competitors. According to the Court, at paragraph 181, the capability of rebates to foreclose an as-efficient competitor “must be assessed, as a general rule, using the AEC test” (underlining added). Although the test is “merely one of the ways of assessing whether an undertaking in a dominant position has used means other than those that come within the scope of ‘normal’ competition”, it is specifically about assessing whether an as-efficient competitor can reproduce the conduct of the undertaking in a dominant position. Consequently, it assesses whether that conduct comes “within the scope of normal competition, that is to say, competition on the merits”.
By contrast, the draft Guidelines do not mention the AEC Test at all in the section on exclusive dealing (which includes exclusivity rebates), even stating that proving the possible existence of a strategy aimed at excluding actual or potential competitors of the dominant firm is not legally required to establish the conduct’s capability to produce exclusionary effects. The draft Guidelines only recognise the relevance of a “price-cost test” in their section on conditional rebates that are not subject to exclusive purchase or supply requirements.
This is another area where the judgment should give the Commission pause to think about whether the draft Guidelines hit the right note. While the Court seems to be willing to entertain other methods of assessing foreclosure of an as-efficient competitor, it is clear that it is expecting “as a general rule” that the AEC Test is applied when the conduct consists of rebates.
Where do we go from here?
From when AMD filed its complaint against Intel, the case has now run for almost 25 years. It is not over yet, as Intel’s appeal against the Commission’s re-adopted decision relating to “naked abuses” is still pending. Intel’s liability for abuse of dominance in that case is final, but Intel is challenging the size of the fine that the Commission imposed on it in 2023. There are also two cases before the General Court in which Intel is seeking compensation from the Commission. The Intel saga is far from over and may well make it to 2030 at this rate.
As for the draft Guidelines, the situation is not so dramatic. The Commission’s draft Guidelines are laudable in their attempt to clarify one of the most difficult areas of competition law. The Commission can now take on board this judgment in the final version of the Guidelines. The broader question is whether these Guidelines serve the purpose for which they appear to have been written (at least in part), namely the streamlining of Article 102 enforcement. For that, the Commission should perhaps focus more on reforming its procedures, but that is a topic for another blog post.