Intel and the AEC Test


On 26 January 2022, the General Court (“GC”) annulled the €1 billion fine imposed on Intel for alleged anticompetitive rebates. (Case T-286/09 RENV, Intel v. Commission). 

A flashback: In 2009, the European Commission (the “Commission”) imposed a fine of €1 billion on Intel for the alleged abuse of its dominant position. The Commission found Intel to be dominant on the worldwide x86 CPU market and considered that Intel had abused this position through two practices:

  • First, Intel paid retailers money if they only stocked computers with Intel’s x86 CPUs, and it granted hidden rebates to computer manufacturers and other customers on condition that they bought virtually all of their x86 CPUs from Intel. These practices fall in the category of (open or hidden) “rebates”.
  • Second, Intel made payments to computer manufacturers to halt or delay the launch of products containing competing x86 CGUs. These payments were not related to purchases of Intel products and therefore constituted “naked restrictions”.

The Commission found that these practices had the ability to foreclose competing manufacturers of x86 CPUs and to harm consumers by reducing the availability of alternatives. 

In 2014, the GC upheld the Commission’s decision on appeal. The GC distinguished three categories of rebates offered by dominant companies:

  • Pure quantity rebates, which are permissible; 
  • Exclusivity rebates, which are as such abusive irrespective of their actual effects;
  • Other loyalty enhancing rebates, which have to be assessed on a case-by-case basis, taking into account all circumstances of the individual case.

The GC put the Intel rebates into the "exclusivity" category and therefore declined to examine the Commission’s effects analysis, notwithstanding Intel’s challenge of the Commission’s effects analysis in its appeal.

In 2017, the Court of Justice (the “CEJC”) overturned the GC's 2014 judgment and held that the abusive character of exclusivity rebates cannot be presumed but must be established based on all the circumstances of the case, which include the extent of Intel’s dominant position on the relevant market, the market coverage of the rebates and the conditions and arrangements for granting the rebates in question, as well as their duration and value. 

The CJEU also confirmed that any category of rebates may be objectively justified by advantages and efficiencies, which benefit the consumer. Favourable and unfavourable effects must be carefully balanced to reach a conclusion on the legality of rebates. Moreover, the Commission’s assessment may include a so-called as efficient competitor (“AEC”) analysis (as it did in the case at hand).

The economic analysis carried out in the AEC test assesses to what extent equally efficient competitors can still compete despite the rebates offered by the dominant firm. In the case at hand, the AEC analysis seeks to establish at what price an AEC facing the same costs as Intel would have had to offer processors in order to compensate an OEM or retailer for the loss of the rebates offered by Intel, while still covering its costs. The Commission had carried out such an analysis, but the GC had declared it unnecessary and therefore declined to review it, erroneously as the CJEU found in 2017.

The CJEU referred the matter back to the GC, which adopted its judgment on 26 January 2022.

The 2022 Intel Judgment

The GC invalidated the finding of an abuse and annulled the fine. 

The initial GC judgment was affected by a single error: the failure to take into account Intel’s challenge of the Commission’s AEC analysis. Hence, when revisiting the Commission decision, the GC limited its re-assessment to this issue. Consequently, the GC did not revisit its earlier factual findings, its distinction between rebates and naked restrictions, nor the finding that the naked restrictions were abusive.

Step 1 of the GC’s analysis

The GC set out the method defined by the CJEU for assessing whether a system of rebates has the capacity to foreclose competitors. According to the Court the Commission must:

  • Analyse the extent of the undertaking’s dominant position.
  • Assess the share of the market covered by the contested practice, together with the conditions and arrangements for granting the rebates in question, their duration and their amount.
  • Assess the possible existence of a strategy to exclude AECs.
  • Where the Commission has carried out an AEC test, that test is one of the factors that must be considered in assessing possible foreclosure effects.

Step 2 of the GC’s analysis

The GC reviewed whether the Commission covered the above criteria. 

In the GC’s view, the Commission had erred when concluding that the AEC test, which it had carried out sua sponte, was not necessary to conclude that the rebates had the capability to foreclose.

Step 3 of the GC’s analysis

The GC set out the rules on the burden and required standard of proof. The presumption of innocence requires the Commission to establish the existence of an infringement. 

  • The Commission may claim that the facts necessarily establish an infringement. However, this is insufficient to prove an infringement where the defendant produces an alternative plausible explanation of the facts.
  • Where a plausible explanation is provided, the Commission must produce a precise and consistent body of evidence that leaves no residual doubt about the existence of a foreclosure effect. 
  • Where the Commission produces such evidence, the defendant must demonstrate that the probative value of that evidence is insufficient.

Step 4 of the GC’s analysis

The GC reviewed Intel’s arguments as to the errors affecting the Commission’s AEC analysis in light of the above principles. 

The GC concluded that the Commission failed to establish to the requisite (but unidentified) legal standard that the rebates had the capacity to have a foreclosure effect.

  • As concerns the AEC test applied to Dell, the Commission could reasonably rely on data known by third parties (but not the defendant) for determining the ‘contestable share’, that is the share of demand which Intel’s customers were willing and able to switch to another supplier, given the nature of the product and the brand image and profile of the dominant supplier. However, the evidence put forward by Intel to challenge the Commission’s finding was capable of giving rise to doubt. Consequently, the GC concluded that the Commission’s evidence was insufficient to conclude that the rebates granted to Dell were capable of having a foreclosure effect. 
  • As concerns the AEC test applied to HP, the GC concluded that the foreclosure effect was not demonstrated for the entire infringement period. 
  • As concerns the AEC test applied to NEC and MSH, the GC found two errors: one affecting the value of the conditional rebates and the other relating to an unsubstantiated extrapolation of the results for one single quarter to the entire infringement period. 
  • As concerns the AEC test applied to Lenovo, the GC concluded that the Commission’s assessment of non-cash advantages accorded Lenovo was flawed. 
  • The GC also found that the Commission did not properly analyse the share of the market covered by the contested practice and the duration of the rebates.

In conclusion, the GC considered that the Commission’s AEC analysis was incomplete and did not meet the requisite legal standard for finding that the rebates were capable of having, or were likely to have, anticompetitive effects.

As the GC could not quantify the share of the € 1 billion fine affected by the annulment, it annulled the fine in its entirety.

Comments and Practical Outlook

A few quick comments on the 2022 Intel Judgment:

Size of the affected market

The Commission defined a broad chip market which included both home and business PCs.Because the rebates Intel’s targeted only the business PC of key customers Intel was able to argue that the rebates affected a relatively small percentage of the overall market. 

Contestable share

The GC overturned prior judgments by holding that the Commission can rely on evidence not available to the dominant firm.

Anti-competitive effect

It is not entirely clear to what extent evidence of anticompetitive intent (or the existence of a plan to eliminate a competitor) may be relevant to the Commission’s assessment of the defendant’s arguments rebutting the presumption of foreclosure effects. The GC could have taken the view that the plan to eliminate AMD meant that there was no need to engage in any further analysis, a position that the ECJ had accepted in the past. However, the fact that the CJEU referred the Intel case back to the GC probably precluded that result and suggests that intent may not be decisive for ruling out economic analysis in rebate cases.

The AEC test

The one clear point is that the Commission has to do a good job if it relies on the AEC test to establish an infringement. If the Commission decides not to carry out an AEC test for tactical reasons, the defendant can always produce its own AEC assessment, which then places on the Commission the burden of refuting that assessment. The GC’s a serious and strict evidentiary approach to the AEC test suggests that the Commission may be very reluctant to reply on the test in future decisions.

DISCLAIMER: Because of the generality of this update, the information provided herein may not be applicable in all situations and should not be acted upon without specific legal advice based on particular situations.

© Dentons | Attorney Advertising

Written by:


Dentons on:

Reporters on Deadline

"My best business intelligence, in one easy email…"

Your first step to building a free, personalized, morning email brief covering pertinent authors and topics on JD Supra:
*By using the service, you signify your acceptance of JD Supra's Privacy Policy.
Custom Email Digest
- hide
- hide

This website uses cookies to improve user experience, track anonymous site usage, store authorization tokens and permit sharing on social media networks. By continuing to browse this website you accept the use of cookies. Click here to read more about how we use cookies.