[Home] [Databases] [World Law] [Multidatabase Search] [Help] [Feedback]
	England and Wales Court of Appeal (Civil Division) Decisions
You are here: BAILII >> Databases >> England and Wales Court of Appeal (Civil Division) Decisions >> Royal Mail PLC v Office of Communications & Anor [2021] EWCA Civ 669 (07 May 2021) URL: http://www.bailii.org/ew/cases/EWCA/Civ/2021/669.html Cite as: [2021] WLR(D) 266, [2021] EWCA Civ 669, [2021] Bus LR 1045

[New search] [Printable PDF version] [Buy ICLR report: [2021] Bus LR 1045] [View ICLR summary: [2021] WLR(D) 266] [Help]

IN THE COURT OF APPEAL (CIVIL DIVISION)
ON APPEAL FROM THE COMPETITION APPEAL TRIBUNAL
Peter Freeman CBE QC (Hon), Tim Frazer and Prof David Ulph CBE
[2019] CAT 27

		Royal Courts of Justice Strand, London, WC2A 2LL
		7 May 2021

Covid-19 Protocol: This judgment was handed down remotely by circulation to the parties' representatives by email, release to BAILII and publication on the Courts and Tribunals Judiciary website. The date and time for hand-down is deemed to be at 10:30am on 7 May 2021

Lord Justice Arnold:

Introduction

1. This is an appeal by Royal Mail plc ("RM") against a judgment of the Competition Appeal Tribunal (Peter Freeman CBE QC (Hon), Tim Frazer and Prof David Ulph CBE) ("the Tribunal") dated 12 November 2019 [2019] CAT 27 ("the Judgment") dismissing RM's appeal against a decision of the Office of Communications ("Ofcom") dated 14 August 2018 ("the Decision") finding RM guilty of an abuse of its dominant position in the wholesale market for bulk mail delivery services contrary to section 18 of the Competition Act 1998 and Article 102 of the Treaty on the Functioning of the European Union ("TFEU") by issuing Contract Change Notices ("CCNs") which introduced discriminatory prices, although those CCNs were never implemented. Ofcom imposed a fine of £50 million for that conduct. RM appeals on two closely-related grounds concerning the treatment by Ofcom in the Decision and by the Tribunal in the Judgment of an "as efficient competitor" ("AEC") test relied upon by RM. The Respondents to the appeal are Ofcom and Whistl UK Ltd ("Whistl"), which was the target of the conduct complained of and an intervener in the proceedings below.

The facts

2. Until the Postal Services Act 2000, RM held a statutory monopoly in the handling and delivery of the great majority of letters. After a process of gradual liberalisation, the UK postal service was fully opened up to the possibility of competition in 2006.
3. "Bulk mail" refers to high volume mailings, such as bank statements, utility bills and advertisements, which are sent to addresses across a substantial part of the UK. Suppliers on the retail market compete to offer customers a bulk mail service (which encompasses the collection, sortation and final delivery of bulk mail). Because RM has historically been the only supplier which undertakes the final delivery of bulk mail to individual addresses itself, it is able to offer a complete end-to-end ("e2e") service. All other suppliers on the retail market for bulk mail services contract with RM, at a wholesale level, to provide the "final leg" of their e2e offering i.e. the delivery of bulk mail to individual addresses. The wholesale market for bulk mail delivery is the market on which RM was found to have a dominant position, with a market share at the relevant time of at least 98%.
4. Wholesale bulk mail delivery services are known as "access services", and the purchasers of those services as "access operators" or "AOs". AOs purchase access services from RM in accordance with the terms of the Access Letters Contract ("ALC"). Under the ALC, AOs can choose from three price plans.
5. AOs on National Price Plan One ("NPP1") pay a nationally-averaged and uniform price, and are required to have a geographic posting profile similar to that of RM across all parts of the UK, which for these purposes is divided up into 83 areas known as Standard Selection Codes ("SSCs"). Deviation from that profile results in a surcharge broadly reflecting the increased costs to RM. A second national price plan ("APP2") operates on the same principle, save that AOs using this plan are required to post mail in line with RM's posting profile across four broad zones (London, urban, suburban and rural) rather than at SSC level. An AO on APP2 is not required to post to all areas of the UK, as long as it achieves the appropriate split between zones. A third, zonal, price plan ("ZPP3") does not require AOs to commit to any posting profile. AOs pay a price per item that reflects the cost of delivery in the zone in which the item is delivered (calculated pursuant to a metric called the "zonal tilt"). AOs can combine ZPP3 with either NPP1 or APP2.
6. Whistl (formerly called TNT Post UK Ltd and at the relevant time a wholly-owned subsidiary of PostNL NV) is an AO which purchases bulk mail delivery services from RM on APP2 and ZPP3. By 2013 Whistl was the largest AO in the UK, involved in the distribution of around 3.8 billion addressed letters in the UK annually. In 2012 Whistl began to develop its own final delivery service, which it rolled out incrementally into certain SSCs, enabling it to provide an e2e service in those SSCs. Whistl continued to purchase access services from RM in respect of the remaining SSCs. Whistl intended to expand the geographical coverage of its e2e service.
7. In the course of 2013 RM formulated plans to respond to what it described as "the threat of Direct Delivery" competition. As the Tribunal found, Royal Mail knew about Whistl's intentions in sufficient detail to plan against them and clearly had Whistl in mind when preparing its plans. As one of its options, RM considered whether to cut prices as a competitive response, but decided not to do so because of the adverse impact on its revenues and profits. The Tribunal found that RM was most reluctant to engage in direct price competition with AOs.
8. Instead, RM set out to devise a strategy that would limit direct delivery competition from Whistl by deterring Whistl from expanding its own direct delivery operations. The mechanism which RM employed for achieving this purpose was to introduce a price differential between NPP1 and APP2. Whistl could not avoid purchasing a substantial portion of its delivery needs from RM following the launch of its own competing service because it could only practicably enter the delivery market in a gradual way, and only in certain parts of the country. The price differential would therefore face Whistl with an invidious choice. If it wished to roll out a competing delivery service at any scale, it could not do so on NPP1, given the commitment to a national profile of mail and the penalties applicable for failing to comply with that. But if it used APP2, which would allow it to launch a rival service in particular areas, it would have to pay higher prices to RM to deliver mail for it everywhere else. RM's intention was that Whistl would be forced to switch to NPP1, a move that would have constrained its ability to operate as an e2e competitor.
9. At the time that RM was formulating its plans to address the competitive threat posed by Whistl, RM did not at any point conduct an AEC test to see whether a competitor with its own costs could survive the price increase. Instead, RM modelled Whistl's costs, in order to work out the plan that would be most effective at curtailing Whistl's entry into the e2e market.
10. RM's internal documents dating from shortly prior to the notification of the price increase identified "significant legal and competition law risks should Royal Mail take commercial action to respond to the threat" posed by Whistl. While formulating its plans, RM worked with its economic consultants, Oxera, to prepare justifications for its proposed conduct, on the assumption that it was prima facie abusive, in order to be able to defend the proposals in the event of a regulatory or competition investigation. The Tribunal found that these justifications had all the hallmarks of an ex post facto exercise, and were unsustainable.
11. On 6 December 2013 RM informed AOs that it had decided in principle to introduce a price differential between NPP1 and APP2. On 10 January 2014 RM announced its intention to apply a price differential of 1.2% to APP2, amongst other pricing changes. It did so by issuing the CCNs which, after a notice period of over two months, would operate to alter the terms of the ALC. The ALC provided that any CCNs would be automatically suspended if Ofcom decided to open an investigation into RM's changes.
12. On 28 January 2014 Whistl complained to Ofcom about the package of pricing changes announced in the CCNs, including the price differential. On 21 February 2014 Ofcom announced that it was opening an investigation into the pricing changes. On that date the CCNs were automatically suspended in accordance with the ALC. The CCNs were ultimately withdrawn in March 2015. Thus the price differential was never introduced and had no effect on the prices actually paid by AOs.
13. Although the notified price increase was suspended before it took effect, the Tribunal found that it nonetheless caused Whistl's funding to be suspended and limited its ability to recruit customers. It was forced to put its market entry plans on hold. Whistl withdrew its bulk mail delivery service in June 2015. RM has remained in a position of near-monopoly ever since.

The proceedings below

14. Following a four-and-a-half year investigation, Ofcom concluded in the Decision that in issuing the CCNs RM had abused its dominant position on the bulk mail delivery market by introducing discriminatory prices in the form of the price differential between NPP1 and APP2. Ofcom found that the infringement lasted from the date of the issuance of the CCNs on 10 January 2014 until at least 21 February 2014, the date on which the CCNs were suspended. The Decision runs to 332 pages.
15. On appeal to the Tribunal, which was a full appeal on the merits, RM attacked the Decision on six main grounds. One of those grounds, ground 3, was that Ofcom was wrong to conclude that the price differential resulted in a competitive disadvantage for competitors, and in particular that Ofcom had wrongly failed properly to take into account an AEC test relied upon by RM. The issues of fact and law raised by RM were investigated by the Tribunal during a hearing which included evidence from both factual and expert witnesses and lasted some six weeks. The members of the Tribunal included a distinguished professor of economics, Prof Ulph, as well as two legally qualified members highly experienced in competition law. Much of the expert economic evidence was given concurrently in a session led by Prof Ulph. The Judgment runs to 813 paragraphs and 232 pages.

The law

16. Article 102 TFEU (previously Article 82 of the Treaty establishing the European Community and before that Article 86 of the Treaty of Rome) prohibits "any abuse by one or more undertakings of a dominant position within the internal market or in a substantial part of it… in so far as it may affect trade between Member States". The Article sets out a non-exhaustive list of types of abusive conduct, including (in paragraph (c)) "applying dissimilar conditions to equivalent transactions with other trading parties, thereby placing them at a competitive disadvantage". Section 18 of the Competition Act 1998 (referred to as "the Chapter II prohibition") is in materially the same terms as Article 102 TFEU, save that it applies to conduct that may affect trade within the UK.
17. The classic articulation of the concept of abuse is to be found in the judgment of the European Court of Justice, now the Court of Justice of the European Union, in Case C-85/76 Hoffmann-La Roche & Co AG v Commission of the European Communities [1979] ECR 461 at [91]:

"The concept of abuse is an objective concept relating to the behaviour of an undertaking in a dominant position which is such as to influence the structure of a market where, as a result of the very presence of the undertaking in question, the degree of competition is weakened and which, through recourse to methods different from those which condition normal competition in products or services on the basis of the transactions of commercial operators, has the effect of hindering the maintenance of the degree of competition still existing in the market or the growth of that competition."

18. The concept of "normal competition" (or, as it is more usually termed nowadays, "competition on the merits") means competition on price, quality, choice and innovation. Thus there is nothing wrong with a dominant undertaking competing with other undertakings on price, and a dominant undertaking may maintain or even increase its market share by doing so. But it is unlawful for dominant undertakings to adopt pricing practices which are anti-competitive, and in particular to adopt differential prices which place other undertakings at a competitive disadvantage.
19. It is settled law that, where it is alleged that an undertaking in a dominant position has abused that position by a pricing practice, it is necessary to consider all the circumstances and to investigate whether the practice tends to remove or restrict the buyer's freedom to choose its sources of supply, to bar competitors from access to the market, to apply dissimilar conditions to equivalent transactions with other trading parties, thereby placing them at a competitive disadvantage, or to strengthen the dominant position by distorting competition: see Case 322/81 NV Nederlandsche Banden-Industrie Michelin v Commission of the European Communities [1983] ECR 3461 at [73]; Case C-95/04 British Airways plc v Commission of the European Communities [2007] ECR I-2331 at [67]; Case C-280/08 Deutsche Telekom AG v European Commission [2010] ECR I-9555 ("Deutsche Telekom") at [175]; Case C-52/09 Konkurrensverket v TeliaSonera Sverige AB [EU:C:2011:83] ("TeliaSonera") at [28], [68]; Case C-209/10 Post Danmark A/S v Konkurrencerådet [EU:C:2012:172] ("Post Danmark I") at [26]; Case C-549/10 Tomra Systems ASA v European Commission [EU:C:2012:221] at [71]; Case C-23/14 Post Danmark A/S v Konkurrencerådet [EU:C:2015:651] ("Post Danmark II") at [29], [68]; Case C-525/16 MEO – Servicos de Comunicoes e Multimedia SA v Autoridade de Concurrencia [EU:C:2018:270] at [28], [31]; and Case C-165/19 Slovak Telekom as v European Commission [EU:C:2021:239] at [42].
20. Relevant considerations identified in the case law include: (i) the structure of the market; (ii) the extent of the dominant position; (iii) the nature of the conduct; (iv) evidence as to the dominant undertaking's intent; (v) the extent of the likely impact on the market, assessed at the time of the conduct; and (vi) the evidence as to any actual effects which eventuated.
21. It is common ground that there is no obligation on a competition authority considering whether a dominant undertaking has abused its position by a pricing practice to test the effects of that practice by reference to a notional competitor which is as efficient as the dominant undertaking and thus has the same costs ("an AEC test"). If the authority does rely upon an AEC test to establish that the pricing practice is anti-competitive, however, then it must carry out and apply the test correctly, and hence the conduct and application of the test by the authority can be reviewed for any error of law.
22. It is also common ground that, where the authority does not itself rely upon an AEC test, but the undertaking under investigation relies upon an AEC test as rebutting the contention that the pricing practice in issue is anti-competitive, the authority must fairly evaluate that evidence. RM contends that in such a case the authority must, unless it concludes for justifiable reasons that the AEC test had not been properly carried out by the undertaking, treat the AEC test as either determinative of, or at least highly relevant to, the question of whether the pricing practice results in a competitive disadvantage to competitors, depending on how much "headroom" the AEC test shows there is for a less efficient competitor to enter the market. (Obviously the undertaking would not rely upon the AEC test if it did not purport to show that an AEC would be able to compete at the prices in question.) Ofcom and Whistl dispute this contention.
23. The CJEU has considered the relevance of an AEC test in a number of cases. The earliest such case to which we were referred is Case C-62/86 AKZO Chemie NV v Commission of the European Communities [1991] ECR I-3359, but attention focussed on a series of five cases decided since 2010.
24. In Deutsche Telekom the dominant telecommunications undertaking in Germany was found to have abused its position by pricing which amounted to a "margin squeeze" on competitors generated by an inappropriate spread between wholesale charges for local loop access services and retail charges for end-user access services. The Commission had analysed this conduct by means of an AEC test. On appeal to the General Court Deutsche Telekom argued, among other things, that the Commission was wrong to rely on the AEC test, but the General Court rejected that argument. On appeal to the CJEU Deutsche Telekom argued, among other things, that the General Court had misapplied the AEC test to the instant case because Deutsche Telekom was not subject to the same regulatory and material conditions as its competitors, but the CJEU rejected that argument.
25. In that context the Second Chamber of the CJEU stated:

"196. As to whether [Deutsche Telekom's] complaint is well founded, … the as-efficient-competitor test used by the General Court in the judgment under appeal consists in considering whether the pricing practices of a dominant undertaking could drive an equally efficient economic operator from the market, relying solely on the dominant undertaking's charges and costs, instead of on the particular situation of its actual or potential competitors.

197.     In the present case, … the appellant's costs were taken into account by the General Court in order to establish the abusive nature of the appellant's pricing practices where the spread between its wholesale prices for local loop access services and its retail prices for end-user access services was positive. In such circumstances, the General Court considered that the Commission was entitled to regard those pricing practices as unfair within the meaning of Article 82 EC, where that spread was insufficient to cover the appellant's product-specific costs of providing its own services.

198.     In that regard, it must be borne in mind that the Court has already held that, in order to assess whether the pricing practices of a dominant undertaking are likely to eliminate a competitor contrary to Article 82 EC, it is necessary to adopt a test based on the costs and the strategy of the dominant undertaking itself ….

199.     The Court pointed out, inter alia, in that regard that a dominant undertaking cannot drive from the market undertakings which are perhaps as efficient as the dominant undertaking but which, because of their smaller financial resources, are incapable of withstanding the competition waged against them ….

200.     In the present case, since … the abusive nature of the pricing practices at issue in the judgment under appeal stems in the same way from their exclusionary effect on the appellant's competitors, the General Court did not err in law when it held … that the Commission had been correct to analyse the abusive nature of the appellant's pricing practices solely on the basis of the appellant's charges and costs.

201.     As the General Court found, in essence, … since such a test can establish whether the appellant would itself have been able to offer its retail services to end-users otherwise than at a loss if it had first been obliged to pay its own wholesale prices for local loop access services, it was suitable for determining whether the appellant's pricing practices had an exclusionary effect on competitors by squeezing their margins.

202.     Such an approach is particularly justified because, as the General Court indicated, in essence, … it is also consistent with the general principle of legal certainty in so far as the account taken of the costs of the dominant undertaking allows that undertaking, in the light of its special responsibility under Article 82 EC, to assess the lawfulness of its own conduct. While a dominant undertaking knows what its own costs and charges are, it does not, as a general rule, know what its competitors' costs and charges are.

203.     Those findings are not affected by what the appellant claims are the less onerous legal and material conditions to which its competitors are subject in the provision of their telecommunications services to end-users. Even if that assertion were proved, it would not alter either the fact that a dominant undertaking, such as the appellant, cannot adopt pricing practices which are capable of driving equally efficient competitors from the relevant market, or the fact that such an undertaking must, in view of its special responsibility under Article 82 EC, be in a position itself to determine whether its pricing practices are compatible with that provision."

26. In TeliaSonera the dominant telecommunications undertaking in Sweden was alleged by the Swedish competition authority to have abused its position by applying a pricing policy under which the spread between the sale prices of ADSL products intended for wholesale users and the sale prices of services offered to end users was not sufficient to cover TeliaSonera's own costs. The Stockholm District Court referred various questions to the CJEU before reaching a decision.
27. The First Chamber of the CJEU said that the referring court should consider whether the pricing practice introduced by TeliaSonera amounted to a margin squeeze. As the Court explained:

"31. A margin squeeze, in view of the exclusionary effect which it may create for competitors who are at least as efficient as the dominant undertaking, in the absence of any objective justification, is in itself capable of constituting an abuse within the meaning of Article 102 TFEU ….

32.       In the present case, there would be such a margin squeeze if, inter alia, the spread between the wholesale prices for ADSL input services and the retail prices for broadband connection services to end users were either negative or insufficient to cover the specific costs of the ADSL input services which TeliaSonera has to incur in order to supply its own retail services to end users, so that that spread does not allow a competitor which is as efficient as that undertaking to compete for the supply of those services to end users.

33.       In such circumstances, although the competitors may be as efficient as the dominant undertaking, they may be able to operate on the retail market only at a loss or at artificially reduced levels of profitability."

28. The Court went on to consider the prices and costs that should be taken into account when making this assessment. In that context the Court stated:

"39. It must be recalled, in that regard, that the Court has already made clear that Article 102 TFEU prohibits a dominant undertaking from, inter alia, adopting pricing practices which have an exclusionary effect on its equally efficient actual or potential competitors ….

40.       Where an undertaking introduces a pricing policy intended to drive from the market competitors who are perhaps as efficient as that dominant undertaking but who, because of their smaller financial resources, are incapable of withstanding the competition waged against them, that undertaking is, accordingly, abusing its dominant position ….

41.       In order to assess the lawfulness of the pricing policy applied by a dominant undertaking, reference should be made, as a general rule, to pricing criteria based on the costs incurred by the dominant undertaking itself and on its strategy …..

42.       In particular, as regards a pricing practice which causes margin squeeze, the use of such analytical criteria can establish whether that undertaking would have been sufficiently efficient to offer its retail services to end users otherwise than at a loss if it had first been obliged to pay its own wholesale prices for the intermediary services ….

43.       If that undertaking would have been unable to offer its retail services otherwise than at a loss, that would mean that competitors who might be excluded by the application of the pricing practice in question could not be considered to be less efficient than the dominant undertaking and, consequently, that the risk of their exclusion was due to distorted competition. Such competition would not be based solely on the respective merits of the undertakings concerned.

44.       Furthermore, the validity of such an approach is reinforced by the fact that it conforms to the general principle of legal certainty, since taking into account the costs and prices of the dominant undertaking enables that undertaking to assess the lawfulness of its own conduct, which is consistent with its special responsibility under Article 102 TFEU, as stated in paragraph 24 of this judgment. While a dominant undertaking knows its own costs and prices, it does not as a general rule know those of its competitors ….

45.       That said, it cannot be ruled out that the costs and prices of competitors may be relevant to the examination of the pricing practice at issue in the main proceedings. That might in particular be the case where the cost structure of the dominant undertaking is not precisely identifiable for objective reasons, or where the service supplied to competitors consists in the mere use of an infrastructure the production cost of which has already been written off, so that access to such an infrastructure no longer represents a cost for the dominant undertaking which is economically comparable to the cost which its competitors have to incur to have access to it, or again where the particular market conditions of competition dictate it, by reason, for example, of the fact that the level of the dominant undertaking's costs is specifically attributable to the competitively advantageous situation in which its dominant position places it.

46.       It must therefore be concluded that, when assessing whether a pricing practice which causes a margin squeeze is abusive, account should as a general rule be taken primarily of the prices and costs of the undertaking concerned on the retail services market. Only where it is not possible, in particular circumstances, to refer to those prices and costs should those of its competitors on the same market be examined."

29. In Post Danmark I Post Danmark enjoyed a statutory monopoly over a large part of the postal market in Denmark, but competed with other undertakings in the unaddressed mail market. The Danish competition authority found that Post Danmark had abused its dominant position in the unaddressed mail market by charging low prices to certain former customers of a competitor. Post Danmark challenged that finding before the Danish courts. The Danish Supreme Court referred questions as to the circumstances in which such a practice could amount to an abuse of a dominant position.
30. In the course of recapitulating the applicable principles the Grand Chamber of the CJEU stated:

"21. … It is in no way the purpose of Article 82 EC to prevent an undertaking from acquiring, on its own merits, the dominant position on a market …. Nor does that provision seek to ensure that competitors less efficient than the undertaking with the dominant position should remain on the market.

22.       Thus, not every exclusionary effect is necessarily detrimental to competition …. Competition on the merits may, by definition, lead to the departure from the market or the marginalisation of competitors that are less efficient and so less attractive to consumers from the point of view of, among other things, price, choice, quality or innovation.

23.       … a dominant undertaking has a special responsibility not to allow its behaviour to impair genuine, undistorted competition on the internal market ….

25. Thus, Article 82 EC prohibits a dominant undertaking from, among other things, adopting pricing practices that have an exclusionary effect on competitors considered to be as efficient as it is itself and strengthening its dominant position by using methods other than those that are part of competition on the merits. Accordingly, in that light, not all competition by means of price may be regarded as legitimate ….

26. In order to determine whether a dominant undertaking has abused its dominant position by its pricing practices, it is necessary to consider all the circumstances …"

31. The Court noted at [28] that, in order to assess the lawfulness of a low price policy practised by a dominant undertaking, it had "made use of criteria based on" an AEC test "as well as on the [dominant undertaking's] strategy". It went on to discuss a price-cost comparison which had been employed by the Danish competition authority, and to conclude at [44] that "a policy by which a dominant undertaking charges low prices to certain major customers of a competitor may not be considered to amount to an exclusionary abuse merely because the price that undertaking charges one of those customers is lower than the average total costs attributed to the activity concerned, but higher than the average incremental costs pertaining to that activity".
32. In Post Danmark II Post Danmark retained its statutory monopoly, which at the relevant time extended to over 70% of the bulk mail market. In 2003 Post Danmark had implemented a rebate scheme in respect of direct advertising mail at a time when there was no competition in the bulk mail market. In 2007 Bring Citymail entered the market for bulk mail, but it withdrew from the market in 2010 after suffering heavy losses. On a complaint by Bring Citymail, the Danish competition authority found that Post Danmark had abused its dominant position in the bulk mail market in 2007-2008 by applying rebates in respect of direct advertising mail which had the effect of tying customers and foreclosing the market. The authority held, contrary to Post Danmark's submission, that it was not appropriate to base the assessment of the anti-competitive exclusionary effect on the market caused by the rebate scheme on the AEC test. Post Danmark challenged the decision before the Danish Maritime and Commercial Court which referred questions to the CJEU asking for clarification of (among other things) the relevance of the AEC test in assessing a rebate scheme.
33. The Second Chamber of the CJEU addressed this issue after holding at [21]-[50] that, in order to determine whether a rebate scheme such as that in issue was capable of having an exclusionary effect contrary to Article 82 EC, it was necessary to examine all of the circumstances of the case. The Court stated:

"55. The as-efficient-competitor test has been specifically applied by the Court to low-pricing practices in the form of selective prices or predatory prices (see, in respect of selective prices,…  Post Danmark, …, and in respect of predatory prices, … AKZO v Commission … and France Télécom v Commission, C-202/07 P, EU:C:2009:214 …), and margin squeeze (… TeliaSonera …).

56.       As regards the comparison of prices and costs in the context of applying Article 82 EC to a rebate scheme, the Court has held that the invoicing of 'negative prices', that is to say, prices below cost prices, to customers is not a prerequisite of a finding that a retroactive rebate scheme operated by a dominant undertaking is abusive (... Tomra Systems and Others v Commission, C-549/10 P, EU:C:2012:221 …). In that same case, the Court specified that the absence of a comparison of prices charged with costs did not constitute an error of law ….

57.      It follows that, as the Advocate General stated in points 61 and 63 of her Opinion, it is not possible to infer from Article 82 EC or the case-law of the Court that there is a legal obligation requiring a finding to the effect that a rebate scheme operated by a dominant undertaking is abusive to be based always on the as-efficient-competitor test.

58.       Nevertheless, that conclusion ought not to have the effect of excluding, on principle, recourse to the as-efficient-competitor test in cases involving a rebate scheme for the purposes of examining its compatibility with Article 82 EC.

59.       On the other hand, in a situation such as that in the main proceedings, characterised by the holding by the dominant undertaking of a very large market share and by structural advantages conferred, inter alia, by that undertaking's statutory monopoly, which applied to 70% of mail on the relevant market, applying the as-efficient-competitor test is of no relevance inasmuch as the structure of the market makes the emergence of an as-efficient competitor practically impossible.

60.       Furthermore, in a market such as that at issue in the main proceedings, access to which is protected by high barriers, the presence of a less efficient competitor might contribute to intensifying the competitive pressure on that market and, therefore, to exerting a constraint on the conduct of the dominant undertaking.

61.       The as-efficient-competitor test must thus be regarded as one tool amongst others for the purposes of assessing whether there is an abuse of a dominant position in the context of a rebate scheme.

62.       Consequently, the answer to the third and fourth subparagraphs of Question 1 is that the application of the as-efficient-competitor test does not constitute a necessary condition for a finding to the effect that a rebate scheme is abusive under Article 82 EC. In a situation such as that in the main proceedings, applying the as-efficient-competitor test is of no relevance."

34. In Intel the Commission found that Intel had abused its dominant position in the semiconductor industry through a combination of conditional rebates and of payments to customers intended to cause them to cancel or delay orders from Intel's main competitor AMD. In its decision the Commission held that the rebates in issue were by their very nature capable of restricting competition so that an AEC test was not necessary in order to find an abuse of a dominant position, but nevertheless carried out a very detailed analysis of the AEC test which led it to conclude that this supported the finding that the rebates were exclusionary. On appeal to the General Court Intel argued that the Commission's analysis of the AEC test was flawed. The General Court held that it was not necessary to consider whether the Commission had carried out the AEC test correctly.
35. The Grand Chamber of the CJEU allowed Intel's appeal. Having repeated at [133]-[136] what it had said in Post Danmark I at [21]-[23] and [25] (quoted above), the Court went on:

"138. [The Court's earlier] case-law must be further clarified in the case where the undertaking concerned submits, during the administrative procedure, on the basis of supporting evidence, that its conduct was not capable of restricting competition and, in particular, of producing the alleged foreclosure effects.

139.     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 ….

140.     The analysis of the capacity to foreclose is also relevant in assessing whether a system of rebates which, in principle, falls within the scope of the prohibition laid down in Article 102 TFEU, may be objectively justified. In addition, the exclusionary effect arising from such a system, which is disadvantageous for competition, may be counterbalanced, or outweighed, by advantages in terms of efficiency which also benefit the consumer …. That balancing of the favourable and unfavourable effects of the practice in question on competition can be carried out in the Commission's decision only after an analysis of the intrinsic capacity of that practice to foreclose competitors which are at least as efficient as the dominant undertaking.

141.     If, in a decision finding a rebate scheme abusive, the Commission carries out such an analysis, the General Court must examine all of the applicant's arguments seeking to call into question the validity of the Commission's findings concerning the foreclosure capability of the rebate concerned.

142.     In this case, while the Commission emphasised, in the decision at issue, that the rebates at issue were by their very nature capable of restricting competition such that an analysis of all the circumstances of the case and, in particular, an AEC test were not necessary in order to find an abuse of a dominant position …, it nevertheless carried out an in-depth examination of those circumstances, setting out … a very detailed analysis of the AEC test, which led it to conclude … that an as efficient competitor would have had to offer prices which would not have been viable and that, accordingly, the rebate scheme at issue was capable of having foreclosure effects on such a competitor.

143.     It follows that, in the decision at issue, the AEC test played an important role in the Commission's assessment of whether the rebate scheme at issue was capable of having foreclosure effects on as efficient competitors.

144.     In those circumstances, the General Court was required to examine all of Intel's arguments concerning that test.

145.     It held, however, … that it was not necessary to consider whether the Commission had carried out the AEC test in accordance with the applicable rules and without making any errors, and that it was also not necessary to examine the question whether the alternative calculations proposed by Intel had been carried out correctly.

146.     In its examination of the circumstances of the case, carried out for the sake of completeness, the General Court therefore attached no importance … to the AEC test carried out by the Commission and, accordingly, did not address Intel's criticisms of that test.

147.     Consequently, … the judgment of the General Court must be set aside, since, in its analysis of whether the rebates at issue were capable of restricting competition, the General Court wrongly failed to take into consideration Intel's line of argument seeking to expose alleged errors committed by the Commission in the AEC test."

36. It should be noted that the Court did not refer to Post Danmark II, although it was mentioned in the Advocate General's opinion, three members of the Chamber in Post Danmark II were members of the Grand Chamber in Intel and Judge da Cruz Vilaça was the rapporteur in both cases. If the Grand Chamber in Intel had considered that the Chamber in Post Danmark II had been in error in ruling that it was not necessary to carry out an AEC test in order to find that a pricing practice was abusive, it is probable that the Grand Chamber would have said so, particularly given the convention that earlier decisions of the CJEU should only be overruled by the Grand Chamber.
37. It is clear from this case law that an AEC test may be relied upon by a competition authority to establish that a pricing practice is anti-competitive, in particular in cases where it is alleged that the practice amounts to selective pricing, predatory pricing or a margin squeeze. It is also clear that one of the advantages of an AEC test in such circumstances is that it can provide legal certainty for the dominant undertaking, in particular because the dominant undertaking will know its own costs, but may well not know the costs of any competitor.
38. In my judgment, however, the case law does not establish that an AEC test which is relied upon by the undertaking under investigation must be treated as highly relevant to, let alone determinative of, the question of whether a pricing practice is anti-competitive. On the contrary, it is clear from Post Danmark II at [61] that the AEC test is one tool among others for the purposes of assessing whether there is an abuse of a dominant position. It is also clear from that case at [59]-[60] that there may be circumstances in which carrying out an AEC test is either impracticable or inappropriate. I do not consider that those statements are only applicable to rebate schemes, in particular because the statements at [59]-[60] are consistent with what the CJEU said in the context of a margin squeeze in TeliaSonera at [45]-[46].
39. I do not accept the submission of counsel for RM that it is only legitimate to disregard an AEC test where the emergence of an AEC is practically impossible, which is contradicted by what the Court said in TeliaSonera at [45] and Post Danmark II at [60].
40. Nor do I accept the submission of counsel for RM that Post Danmark II has been silently overruled or qualified by Intel. Not only did the Grand Chamber in Intel not cast doubt on Post Danmark II, but also there is no inconsistency between the two decisions. The essence of Intel is simply that the General Court was wrong not to consider whether or not Intel's criticisms of the AEC test carried out by the Commission were well founded.
41. Above all, as the CJEU has consistently held, all of the circumstances of the case must be considered. There may be other evidence which establishes that a pricing practice is anti-competitive even if an AEC test relied upon by the dominant undertaking appears to show otherwise.

The Decision

42. In the Decision Ofcom dealt with the AEC test relied upon by RM as follows. First, it held at [7.191]-[7.195] that, as matter of law, an AEC test was not a pre-requisite for a determination that a pricing practice amounted to an abuse of a dominant position.
43. Secondly, it held at [7.196]-[7.198] that the conduct complained of in this case was not low pricing or a margin squeeze, but rather discriminatory pricing targeted at potential scale entrants, and that no AEC test was needed to establish that RM's conduct was abusive.
44. Thirdly, it held at [7.199] that the relevant market was characterised by high barriers to entry, particularly given that RM was overwhelmingly dominant, benefitted from significant economies of scale and scope and was an unavoidable trading partner, and thus an AEC test was not relevant.
45. Fourthly, it held at [7.200] that the analysis relied upon by RM did not appropriately reflect economic reality given the prevailing features and conditions of the bulk mail delivery market at the time the price differential was introduced for the following reasons (footnotes omitted):

"a) The EEO Test: The EEO [equally efficient operator i.e. AEC] test advanced by Royal Mail is based on Royal Mail's costs, which its own advisers appear to acknowledge are not likely to be similar to those of an entrant, and it assumes a conversion rate of 100%. In their report, Compass Lexecon note that the sensitivity analysis that was carried out, which made certain adjustments to the inputs to the modelling for the base case EEO model (as discussed in (b) below), 'may be considered closer to the position of a new entrant'. It is therefore clear that their EEO test approach is not a realistic basis for assessing the impact of a pricing practice in the context of an overwhelmingly dominant undertaking responding to nascent competition in the market.

b) The sensitivity analysis: The sensitivity analysis conducted by Royal Mail's advisers: (i) assumes a roll out profile based on Royal Mail's estimates of the likely operating costs of a new entrant; and (ii) assumes an initial conversion rate of 60%, rising to 80%. However, each of the scenarios examined by Royal Mail's advisers is still based on Royal Mail's downstream costs (using an adjusted version of Royal Mail's LRAIC model, see paragraph 5.46 of the FTI report). Royal Mail does not seek to model the actual costs of a new entrant to assess the impact of the price differential on a competitor in that position, despite the fact that Royal Mail had developed a 'Direct Delivery Operating Cost Model' as a 'proxy [for] the likely costs of an efficient entrant'.

c) Other relevant factors are not considered: Royal Mail's assessment of a notional as-efficient entrant also fails to capture a number of other factors which are relevant to an access operator's decision as to whether to enter:

i) A potential entrant (and its investors) would take into account risk as well as expected profitability. The price differential reduced the upside potential for higher profits from entering into bulk mail delivery and increased the downside in the event that entry proved unsuccessful.

ii) As discussed in Section 6, Royal Mail had a number of advantages unrelated to costs, such as reputation and experience, and VAT status. These would make it more difficult to attract customers even if an entrant could match retail prices."

46. Fifthly, it held at [7.201]-[7.202] that RM had not conducted an AEC test at the time and that its ex post analysis was not persuasive in circumstances where its conclusions were inconsistent with (a) the contemporaneous evidence as to what RM considered to be the likely impact of the price differential, (b) Ofcom's assessment that the price differential was reasonably likely to give rise to a competitive disadvantage and (c) the consequences of the introduction of the price differential for Whistl.

The Judgment

47. The Tribunal dealt with ground 3 of the appeal at [450]-[618]. It considered that this ground raised three broad questions. The first related to the use of an AEC test: whether Ofcom was required to apply such a test, and if so in what form, and whether it had properly considered the AEC test advanced in the administrative proceedings by RM. The second concerned the way in which Ofcom had assessed the competitive disadvantage arising from the abusive conduct it had found; in particular, whether it had placed too great a reliance on the actual fate of the market entrant, Whistl, and whether the effect of the abusive conduct was sufficiently material to amount to an infringement. The third was whether Ofcom had correctly assessed the issues of anti-competitive foreclosure and competitive disadvantage in the light of all the circumstances.
48. The Tribunal addressed the first of those questions at [470]-[590]. It considered that it raised five issues. The first was whether there was a requirement on Ofcom to establish anti-competitive foreclosure by means of an AEC test in all pricing cases. The second was whether there was a clear class of cases whose potential anti-competitive foreclosing effects should be investigated by means of an AEC test, and, if so, whether this case fell into that class. The third was whether, in the particular circumstances of this case, the concept of a competitor equally as efficient as RM was appropriate. The fourth was whether the particular AEC test submitted by RM had any problematic features, and, if so, how helpful the test was. The fifth is whether the consideration that Ofcom gave to the AEC test provided by RM was adequate.
49. The Tribunal considered issue 1 at [471]-[522]. It concluded that it was not necessary, whether as a matter of law or of economics, to conduct an AEC test in all cases.
50. The Tribunal considered issue 2 at [523]-[531]. It concluded that the price differential in this particular case could not readily be put into any of the existing categories of pricing practice in which an AEC test had been considered to be an appropriate method for determining whether conduct was anti-competitive.
51. The Tribunal considered issue 3 at [532]-[548]. It identified two reasons for thinking that "the concept of an AEC is in any event inappropriate in this case". The first was that no e2e competitor would attempt to set up its own direct delivery operations in all 83 SSCs, but only in some of them. The second was that RM's special status as the designated universal service provider gave it certain advantages (such as exemption from VAT) and disadvantages (such as the need to comply with the universal service obligation ("USO")), which would not apply to any entrant. The Tribunal therefore concluded that "the concept of an AEC is highly problematic in the context of this case".
52. The Tribunal considered issue 4 at [549]-[578]. It identified four "major concerns" with the AEC test that RM had carried out. First, it was concerned as to whether the long run average incremental costs ("LRAIC") assigned to the AEC by RM's expert Mr Dryden represented the correct level of cost.
53. Secondly, it was concerned at the approach adopted by Mr Dryden of treating bulk mail as a purely incremental activity and assigning all common costs to the USO, when in reality an entrant to bulk mail delivery would have to incur such costs. The Tribunal accepted that some allowance should be made for common costs, but noted that there was a lack of guidance as to how to do this and expressed the view that at the very least it would have been helpful to have some sensitivity analysis of this.
54. Thirdly, RM's AEC test ignored the fact that Whistl's customers would have to make their own investments in order to convert to Whistl's e2e service, and thus more customers would be likely to convert the more SSCs Whistl covered, which would affect the profitability of delivery in SSCs which Whistl had already entered. This meant that RM's AEC test had "some serious limitations".
55. Fourthly, the Tribunal considered that there were three crucial features of the investment environment which made the application of a simple net present value (NPV) calculation of an AEC's future profitability as Mr Dryden proposed inappropriate.
56. For these reasons the Tribunal concluded that RM's AEC test "may be less robust and hence less informative" than RM's expert suggested.
57. The Tribunal considered issue 5 at [579]-[590]. It held that Ofcom had considered the AEC test put forward by RM "in enough detail to satisfy itself that carrying out such a test would serve no useful purpose in this particular case" and that it did give adequate consideration to the AEC test.
58. Having considered the second and third questions identified in paragraph 46 above, the Tribunal's overall conclusion on ground 3 at [681] was that "Ofcom was correct in finding that an AEC test was neither appropriate nor necessary in this case and that its analysis of the likely effects of the conduct in question and its findings on competitive disadvantage were fully justified". Accordingly, it rejected this ground of appeal.

The role of this Court

59. RM's appeal is brought under section 49 of the 1998 Act, and is confined to points of law. It is therefore important to bear in mind what Sir John Dyson giving the judgment of the Supreme Court said in MA (Somalia) v Secretary of State for the Home Department [2010] UKSC 49, [2011] 2 All ER 65 at [43]: 

"Before we examine these two criticisms, we need to make some general points about the proper role of the Court of Appeal in relation to appeals from specialist tribunals to it on the grounds of error of law. Although this is not virgin territory, the present case illustrates the need to reinforce what has been said on other occasions. The court should always bear in mind the remarks of Baroness Hale of Richmond in AH (Sudan) v Secretary of State for the Home Department [2007] UKHL 49, [2008] 1 AC 678 at para 30:

'This is an expert Tribunal charged with administering a complex area of law in challenging circumstances… [T]he ordinary courts should approach appeals from them with an appropriate degree of caution; it is probable that in understanding and applying the law in their specialised field the Tribunal will have got it right… They and they alone are judges of the facts… Their decisions should be respected unless it is quite clear that they have misdirected themselves in law. Appellate courts should not rush to find such misdirections simply because they might have reached a different conclusion on the facts or expressed themselves differently.'"

60. Baroness Hale's observations have been applied to tax tribunals: see in particular Procter & Gamble UK v Revenue and Customs Commissioners [2009] EWCA Civ 407, [2009] STC 1990 at [11]. They apply with at least equal force to the Competition Appeal Tribunal, which is both a specialist and an expert tribunal.

Ground 1

61. RM's first ground of appeal is that the Tribunal erred in law in finding that it was irrelevant, when assessing whether the prices contained in the CCNs would give rise to likely anti-competitive effects if implemented, that those prices would not foreclose an AEC, and further erred in concluding that it was unnecessary to take into account or give any weight to extensive evidence submitted by RM during the administrative process which showed that an AEC would not be foreclosed.
62. In considering this ground, I shall begin with three preliminary points. The first is that in my view the ground is premised on an inaccurate characterisation of the Tribunal's reasoning. I have summarised the Tribunal's reasoning with respect to the AEC test in paragraphs 48-57 above. In short, what the Tribunal held was that: (i) it was not necessary to conduct an AEC test in all cases; (ii) the present case could not readily be put into any of the categories in which an AEC test had been considered appropriate; (iii) the concept of an AEC was highly problematic in this case; (iv) the Tribunal had major concerns about the AEC test relied upon by RM; and (v) Ofcom had given adequate consideration to RM's AEC test. Leaving aside point (v) for the moment, the essence of the Tribunal's reasoning was that it did not find the AEC test relied upon by RM persuasive that the differential pricing complained of was not anti-competitive.
63. The second point is that many of the submissions made by counsel for RM under this heading were not in truth arguments that the Tribunal had erred in law, but rather took issue with the Tribunal's assessment of the economic evidence. It suffices to give two examples of this.
64. First, counsel for RM argued that the pricing practice complained of in this case either was a margin squeeze or at least was closely analogous to a margin squeeze. The Tribunal's assessment of the expert evidence, however, was that "the access pricing proposals in the CCNs - particularly the price differential" were "particularly difficult to classify" ([530]), and accordingly "the price differential in this particular case cannot be readily put into any of the existing categories of pricing practice" in which an AEC test had been held to be "an appropriate method for determining whether or not it is anti-competitive" [(531]). That was an assessment falling within the Tribunal's specialist expertise.
65. Secondly, counsel for RM took issue with the four points made by the Tribunal in its issue 4. Each of those four points was based on the expert evidence before the Tribunal, however. Again, they represent assessments falling within the Tribunal's specialist expertise.
66. Absent a submission that there was no evidence which entitled the Tribunal to reach those conclusions, which was not a submission advanced by counsel for RM, those assessments cannot be questioned in this Court.
67. The third point is that counsel for RM criticised certain statements in the Judgment which were peripheral to the Tribunal's reasoning. The prime example of this is the Tribunal's suggestion at [520(1)] that there "may be some deterrence value in having a degree of legal uncertainty" and that it was "at least arguable that any margin of error in the assessment of abuse should fall on the side of compliance with the law". I would not endorse those statements as correctly representing the law, but this is immaterial. They were no more than passing observations made by the Tribunal en route to its conclusion on issue 1 that it is not necessary to conduct an AEC test in all pricing cases, a conclusion which RM does not take issue with.
68. So far as the substance of the Tribunal's reasoning as to the persuasiveness of the AEC test relied upon by RM is concerned, counsel for RM argued that the Tribunal's conclusions that the concept of an AEC was highly problematic in the context of this case and that RM's AEC test was neither robust nor informative were ones that were not open to the Tribunal as a matter of law. For the reasons given in paragraphs 38-41 above, however, I do not accept this. Accordingly, the Tribunal made no error of law when it concluded that abuse of RM's dominant position was established by other evidence although the AEC test relied upon by RM purported to show that even a less efficient competitor would be able to compete with RM despite the pricing differential which RM proposed to introduce.

Ground 2

69. RM's second ground of appeal is that the Tribunal erred in law in finding that Ofcom had given adequate consideration to the AEC analysis put forward by RM during the administrative process.
70. I do not accept this contention either. As can be seen from my summary and quotation in paragraphs 42-46 above, Ofcom did consider the AEC test relied upon by RM in the Decision. In short, Ofcom concluded that: (i) an AEC test was not a pre-requisite for a determination that a pricing practice amounted to an abuse of a dominant position; (ii) given that the complaint in the present case was not of low pricing or a margin squeeze, no AEC test was needed to establish that RM's conduct was abusive; (iii) an AEC test was not relevant given the characteristics of the market; (iv) RM's AEC test did not reflect economic reality for a number of reasons; and (v) the AEC test was unpersuasive given the other evidence of abuse. For the reasons explained above, Ofcom was not required as a matter of law to treat the AEC test as either determinative or highly relevant. In those circumstances Ofcom gave adequate consideration to the AEC test, and the Tribunal did not err in law in so concluding.

Conclusion

71. For the reasons given above I would dismiss this appeal.

Males LJ:

72. I agree that this appeal must be dismissed for the reasons given by Lord Justice Arnold. I add some further observations concerning the role of an "as efficient competitor" or "AEC" test in cases where the alleged abuse of a dominant position arises from the dominant undertaking's pricing practices.
73. The purpose of such a test is to determine by what is essentially a hypothetical economic analysis whether the pricing practices of a dominant undertaking could drive an equally efficient economic operator from the relevant market (or could prevent it from entering that market) by making it very difficult or practically impossible for that operator to offer its goods or services in the market at a profit.
74. The role of an AEC test in assessing the pricing practices of a dominant undertaking has been considered in a number of European cases. The principal cases to which we were referred were: Deutsche Telekom, Post Danmark I, TeliaSonera, Post Danmark II and Intel.
75. It is important, in considering what these cases decide, to have in mind in each case the nature of the pricing practice in question and what the issue was. At the risk of travelling over some of the ground already covered by Lord Justice Arnold, I should explain what, in my view, those cases do decide about the use of an AEC test.
76. Deutsche Telekom was a margin squeeze case. In reaching its conclusion that the prices charged by the dominant undertaking were abusive, the General Court had relied on an AEC test carried out by the Commission which was based on the dominant undertaking's own costs. That is to say, the test assumed that the costs which the competitor would incur in order to offer the relevant services were the same as the dominant undertaking's costs. The dominant undertaking challenged this approach, saying that its competitors were subject to different legal or material conditions, and that the test should therefore have been based on their costs. The CJEU rejected this challenge, holding that the AEC test was properly carried out by reference to the dominant undertaking's own costs. This was, in part at least, because a focus on the dominant undertaking's own costs promotes the principle of legal certainty: the dominant undertaking knows what its own costs are, and can therefore assess the lawfulness of its proposed conduct by carrying out such a test, but does not as a general rule know what its competitors' costs are. Thus the CJEU held at [200] that "the General Court did not err in law when it held … that the Commission had been correct to analyse the abusive nature of the appellant's pricing practices solely on the basis of the appellant's charges and costs". Accordingly the case holds that, in an appropriate case, the regulator is entitled to rely upon an AEC test for a finding of abuse and that, when such a test is carried out, it should be based on the dominant undertaking's costs.
77. TeliaSonera shows, however, that this is not an inflexible rule. It was also a margin squeeze case, which came to the CJEU on a preliminary reference. The question arose whether an AEC test had to be based solely on the costs of the dominant undertaking or whether it could be legitimate to base the test on the costs incurred by competitors. The CJEU held at [41] that the test should as a general rule be based on the costs of the dominant undertaking, but at [45] that there could be circumstances where the costs of competitors were relevant. It gave as examples cases where (1) the cost structure of the dominant undertaking is not precisely identifiable, (2) the dominant undertaking uses an infrastructure whose production cost has already been written off, and (3) particular market conditions mean that the level of the dominant undertaking's costs is attributable to the competitively advantageous situation in which its dominant position places it. In the first example the dominant undertaking could not frustrate the carrying out of an AEC test by failing to provide relevant information. In the second and third examples, to use the dominant undertaking's own costs would be possible, but to do so would produce a result which was not "economically comparable to the costs which its competitors have to incur". Accordingly, although the judgment goes on at [46] to say that it is only where use of the dominant undertaking's own costs is not possible that the costs of competitors should be used, the examples given demonstrate that this is too narrow a view. Thus TeliaSonera holds that, although in general an AEC test should be carried out by reference to the dominant undertaking's own costs, there are exceptions to that principle which are necessary to ensure that the test produces an economically valid comparison between the dominant undertaking and the notional "as efficient competitor". The case demonstrates that what matters is not the carrying out of a test according to rigid rules, but the extent to which the test provides useful and relevant information.
78. In Post Danmark I the dominant undertaking offered price reductions to customers of a competitor for reasons unrelated to its own costs. One issue was whether the mere fact that the price charged to these customers was lower than the average total costs attributed to the relevant business activity meant that this was an abuse under Article 102. The CJEU held that it did not. It was necessary for the national competition authority to consider "all the circumstances" of the case. One relevant consideration could be an AEC test, but the CJEU did not suggest that this would necessarily be decisive or even that it was a factor to which particular weight had to be attached. Another consideration was the dominant undertaking's strategy in introducing the price reductions.
79. The issue in Post Danmark II was rather different. That was a case where the dominant undertaking applied a retrospective rebate scheme which was conditional on customers sending certain quantities of direct advertising mail during the relevant period, which therefore had the effect of tying customers to the dominant undertaking. The national competition authority concluded that this scheme was abusive, but it reached this conclusion without carrying out an AEC test. The dominant undertaking did not carry out such a test either. The issue was whether a finding of abuse could be made in the absence of such a test. Expressly approving the Opinion of Advocate General Kokott, the CJEU held at [55] to [57] that although an AEC test had previously been applied in pricing cases, this was not a necessary condition for a finding of abuse. Rather, it was necessary to consider all the circumstances of the case, with an AEC test being merely "one tool amongst others" (see [61]). As the Advocate General pointed out at [67] and [68] of her Opinion, the data used for an AEC test are not uncommonly open to different interpretations, while there are many other factors which may be relevant to a finding of abuse, so that it is always necessary to take into account all the relevant circumstances of each individual case. Indeed, the CJEU went further, affirming two important points made by the Advocate General. The first, at [59], was to recognise that in a situation where the dominant undertaking held a very large market share together with structural advantages which made the emergence of an "as efficient competitor" practically impossible, an AEC test would be "of no relevance". The second, at [60], was that even the presence of a less efficient competitor might contribute to intensifying the competitive pressure on the market to the advantage of consumers. These points confirm that an AEC test will not always be necessary, or even useful, and that the question whether such a test is useful in any given circumstances requires an exercise of judgment, having regard to the limitations of what such a test can show and the features of the particular market.
80. Finally, Intel was another rebate case where the rebates were conditional on customers purchasing all of their supplies from the dominant undertaking. The Commission carried out an AEC test, but maintained that this was not an essential part of the reasoning which led to its finding that the rebate scheme was abusive. That submission was accepted by the General Court, which held that the rebate scheme was inherently abusive, so that it was unnecessary to examine all the circumstances of the case. It was therefore unnecessary to consider the dominant undertaking's criticisms of the AEC test carried out by the Commission. That decision was reversed by the CJEU which held that the scheme was not inherently abusive. Accordingly it was necessary to examine all the circumstances of the case. Moreover, contrary to the Commission's submission, the CJEU concluded at [143] that the AEC test had played an important role in the Commission's assessment of abuse. It was therefore necessary to examine the dominant undertaking's criticisms of the test. For present purposes, therefore, Intel stands for two points. The first is a straightforward application of the principle of basic fairness: if the Commission had relied on an AEC test, the dominant undertaking had to be entitled to criticise the methodology of the test and to have its criticisms fairly considered. Second, even where there was an AEC test which supported the Commission's position, that was not necessarily decisive and it remained necessary to consider all the circumstances of the case. I would reject the submission by Royal Mail that Intel, a decision of the Grand Chamber, supersedes or qualifies or even overrules the decision in Post Danmark II. The two cases were dealing with different issues.
81. From these authorities I would derive the following propositions.
82. First, there is no obligation on a competition authority to carry out an AEC test before concluding that a pricing practice is an abuse (Post Danmark II).
83. Second, while an AEC test has been a useful tool for determining whether there has been abusive conduct in some pricing cases (Deutsche Telekom, Post Danmark I, TeliaSonera), such a test is not always relevant (Post Danmark II). That will be the position in particular where the dominant undertaking holds a very large market share together with structural advantages which make the emergence of an "as efficient competitor" practically impossible. However, I see no reason to conclude that this is the only situation in which such a test will be irrelevant. Whether the test is relevant depends on whether and to what extent it provides useful information (TeliaSonera). This is a matter of economic judgment rather than law.
84. Third, although there are rules which indicate how such a test should be performed, these rules must yield to economic reality where that is necessary to ensure information which is comparable to the cost which an efficient competitor would actually have to incur (TeliaSonera). Or as the Advocate General put it in Post Danmark II at [67], the issue of price-based exclusionary conduct cannot be managed simply by applying some form of mathematical formula based on nothing more than the price and cost components of the businesses of the undertakings concerned.
85. Fourth, where the competition authority relies on an AEC test as supporting a finding of abuse, it is open to the dominant undertaking to challenge the methodology and conclusions of that test. When it does so, the court (or in this jurisdiction, the CAT) must engage with those criticisms (Intel).
86. Fifth, the dominant undertaking may itself rely upon an AEC test, as in the present case, in support of an argument that its conduct is not abusive because it would not make it very difficult or practically impossible for an "as efficient competitor" to operate in the relevant market at a profit. In that event, the competition authority or the court must engage with that argument. But how it does so will depend on the circumstances. Such engagement may lead to a number of possible conclusions as a matter of economic judgment. One conclusion may be that the features of the market are such that an AEC test is irrelevant. Another may be that the methodology and conclusions of the test are flawed as a matter of economic analysis, so that the test carries little or no weight. A third possibility, however, is that the AEC test does support the dominant undertaking's argument.
87. Sixth, however, even when an AEC test does support the dominant undertaking's argument, that is relevant but not necessarily decisive in the dominant undertaking's favour. It is capable of being outweighed by other factors. That is because an AEC test is no more than one tool amongst others and it remains necessary to consider all the circumstances of the case (Post Danmark I, Post Danmark II, Intel). Competition law does not determine how much weight should be given to the AEC test. That is a matter for the judgment of the tribunal of fact, taking into account not only the robustness of the test itself, but all the other circumstances which point either towards or against a finding of abuse. Where the judgment is made by an independent and specialist tribunal such as the CAT, its judgment is entitled to considerable weight. This court will not interfere unless it is shown that the judgment of the CAT is clearly wrong.
88. In the present case the CAT made a number of important findings which bear on the issues relating to the AEC test on which Royal Mail relies, as Lord Justice Arnold has explained. To my mind these fully justify its conclusion that carrying out an AEC test in the circumstances of this case was not appropriate and that in any event the test carried out by Royal Mail was so seriously flawed that it provided no useful information in considering the issue of abuse. These were conclusions reached after detailed consideration of the expert evidence and involve no error of law.
89. But I would if necessary go further. The CAT found that the pricing changes announced by Royal Mail were intended and expected to restrict competition by excluding competitors (specifically Whistl) from the relevant market, that they did not constitute competition on the merits, and that they did have precisely the anti-competitive effect intended by causing Whistl to suspend its roll out of end-to-end bulk delivery services and causing its financial backers to withdraw their support. In those circumstances common sense would suggest, and it would not be surprising if the CAT had concluded, that a hypothetical AEC test conducted after the event would need (to say the least) to be particularly compelling in the dominant undertaking's favour in order to outweigh these considerations. For the reasons given by the CAT the test relied on by Royal Mail did not come close to doing so.

King LJ:

90. I also agree.