Processing the Intel Judgment – A More Economic Approach?
Sales, Discounts, Rebates – some of you might have been lucky with some bargains during Winter Sales 21/22. So who would say that rebates are a bad thing? You guessed it: The Competition Law community, albeit – as always in law – it depends. From a Competition Law perspective, rebates schemes are ordinarily unproblematic or might even be beneficial consumer welfare. If used by dominant companies, however, certain types of rebates, in particular exclusivity or loyalty rebates, can be abusive under Article 102 TFEU.
The General Court's recent Intel judgment, handed down on 26 January 2022, offers the latest update on the controversy as to when such rebates are unlawful. The General Court in essence holds that the Commission, in Article 102 TFEU decisions relating to exclusivity or loyalty rebates, cannot simply take a formalistic view and presume competitive harm if the dominant undertaking has presented evidence to rebut that presumption. Rather, a detailed effects-based assessment has to be carried out, effectively strengthening dominant companies' possibilities to defend their case by allowing them to submit evidence on the absence of anti-competitive effects themselves.
Background of the case
The General Court's judgment follows a long series of events surrounding a 2009 Commission decision – the endpoint of a long investigation following complaints by competitor AMD as early as 2000 – and imposing a EUR 1.06 billion fine on Intel for having abused its dominant position on the worldwide market for x86 processors. The Commission found that Intel has granted exclusivity rebates for selected customers and made use of so-called "naked restrictions", i.e. payments to computer manufacturers to stop, cancel or postpone the launch of products using competitor's (in particular AMD) processors.
The judgment, handed down on 26 January 2022, is the latest part of the Intel saga. Initially, the General Court dismissed Intel's first challenge and upheld the decision in 2014. However, following Intel's appeal, the Court of Justice referred the case back to the General Court in 2017, resulting in the recent judgment.
The Court of Justice's 2017 landmark judgment
In 2017, the Court of Justice reaffirmed the basic – and rather formalistic – principle (based on the Hoffmann-La Roche case law) that exclusivity rebates granted by a dominant undertaking are by their very nature capable of restricting competition. However, the Court of Justice added an important tweak to the Hoffmann-La Roche case law, allowing the dominant undertaking to demonstrate, on the basis of supporting evidence, that rebates were in fact not capable of restricting competition and in fact did not produce foreclosure effects. The Court left however open what level of detail such supporting evidence must reach and what standards have to be adhered to.
If rebutting evidence is submitted, it is necessary and the Commission is obliged to analyze all the circumstances of the case and to carry out an effects-based analysis in such dominance cases. According to the Court, the legal test to be applied by the Commission for the analysis of foreclosure effects has to take into account five factors:
- the extent of the company's dominant position on the relevant market;
- the share of the market covered by the challenged practice;
- the conditions and arrangements for granting the rebates in question;
- their duration and their amount;
- the possible existence of a strategy aiming to exclude competitors that are at least as efficient as the dominant undertaking from the market;
In its 2009 decision, the Commission had resorted to the so-called as-efficient-competitor test (or AEC test) to substantiate the foreclosure effects of Intel's exclusivity rebates, even though it had primarily relied also on a prima facie presumption of harm (based on the Hoffmann-La Roche case law). Since the AEC test played an important role in the Commission’s assessment, the Court of Justice concluded that the General Court was required to examine all of Intel’s arguments concerning that test and how the Commission had applied it. Since the General Court had failed to do so, the Court of Justice set aside the initial 2014 judgment and referred the case back for a detailed examination of the AEC test.
The General Court's judgment annulling the Commission's decision
In its judgment of 26 January 2022, the General Court, based on the criteria set out by the Court of Justice, engages in a very extensive and painstakingly detailed economic analysis of the potential anticompetitive foreclosure effects of the conditional rebates. In particular, the General Court focused on the Commission's (incorrect) application of the AEC test.
The AEC test analyzes whether a competitor, which is considered to be as efficient as the dominant firm (pertaining to costs, offerings, product quality etc.) can effectively match that firm's offer. In case of exclusivity or loyalty rebates, the AEC test considers whether a customer would switch to such an as-efficient competitor in absence of the dominant firm's rebate scheme (or if the as-efficient competitor was to apply a similar scheme). Anticompetitive foreclosure effects arise if the dominant company's customer would have switched volumes of business to a competitor. According to the General Court, the Commission committed several errors when applying the AEC test to Intel's rebates:
- As a dominant company is in most cases an unavoidable trading partner, it has to be assessed under the AEC test which share of their demand customers can and would switch to alternative suppliers (the so called "contestable share"). When calculating the contestable share, the Commission falsely relied on only certain pieces of evidence (one spreadsheet that had been prepared by one of Intel's customers), while disregarding several other pieces of evidence that Intel had submitted and that pointed the other way.
- Furthermore, according to the General Court, the Commission also erred in assessing the value of the conditional rebates and of so-called non-cash advantages that these rebates allegedly had for Intel's customers. In particular, the Commission failed to establish the specific value of such rebates and how that value transposes into an anti-competitive foreclosure of Intel's competitors, thereby failing to substantiate the AEC test.
- The Commission also did not meet the proper standards for extrapolating results of the AEC test for one single-quarter period to the entire infringement period. Rather than using real data for the entire period, it claimed that only data for one specific single-quarter period and restricted to one OEM customer was available. According to the General Court, this data was not only no accurate basis for extrapolating the AEC test results for that customer, but the Commission also erred by simply assuming that these results are representative and can also be extrapolated, without any further evidence, for the AEC analysis of another company.
The General Court concluded that, in light of the foregoing, the "AEC analysis carried out by the Commission […] is vitiated by errors" (para. 482 of the General Court's 2022 judgment) and that the Commission did not establish that Intel's rebates were capable of having or likely to have anticompetitive foreclosure effects.
As the exclusivity rebates subject of this judgment constituted only part of the conduct found to be a violation of Article 102 TFEU in the 2009 decision, the question arose whether the General Court will uphold the fine regarding the naked restrictions while setting aside the exclusivity rebates-related fine. However, the General Court stated that it does not consider itself to be in a position to identify the amount of the fine relating solely to the naked restrictions, as the Commission has assessed both conducts as a whole. Thus, on the grounds set out above, the General Court annulled the EUR 1.06 billion fine in its entirety.
Implications for the future
The 2022 Judgment is a major victory for Intel and sets a higher bar for the Commission to bring abuse of dominance cases. The General Court agreed with Intel that regulators must prove that exclusivity and loyalty rebates have anti-competitive effects before sanctioning them, despite the tendency of regulators generally disliking such rebates used by dominant companies and presuming competitive harm. If the dominant company submits economic evidence during the administrative procedure suggesting a that its rebate scheme is not anti-competitive, the Commission will be required to carry out effects-based analyzes instead of merely relying on the presumption of (by nature) anti-competitive effects.
But the questions arises: What does supporting evidence mean in practice and what level of detail is demanded? Both the Court of Justice and the General Court remain silent in that respect and it is hard to draw general conclusions from such the highly specific facts underlying the Intel case. However, the Intel saga shows that every aspect of the Commission's AEC test will be highly scrutinized on appeal, starting with the evidentiary basis for the contestable share up to the conditions of extrapolating results to the entire infringement period. It remains to be seen what this entails for future abuse of dominance cases, but given the importance of submitting qualified and convincing rebuttal evidence, the judgment might not only result in a more economic approach, but also more work for economists. Their involvement is already crucial, but will – as the complexity of such abuse of dominance cases – certainly increase further.
Beyond the question of how to counter the Commission in abuse of dominance cases, both the recent judgment and the Court of Justice's 2017 landmark case provide an important clarification on the general assessment of exclusivity rebates. While it was advisable – from a compliance point of view even potentially mandatory – for dominant companies to abstain from exclusivity and loyalty rebates, Intel may open up the door a little more for "innovative" rebate schemes. Given the severe sanctions that the Commission can impose for infringements of Article 102 TFEU, dominant companies will however have to assess with great care and in line with the criteria set out by the courts whether there is sufficient economic evidence to demonstrate that a given rebate scheme does not have foreclosure effects.
Since the AEC test is also at issue in other dominance proceedings, such as in Qualcomm (T-235/18) and Google Android (T-604/18), the Commission can be expected to appeal the 2022 judgment.