The European, Middle Eastern and African Antitrust Review 2017

EU: Economics

AlixPartners LLP

Unilateral public price announcements – cartel substitute?

In many markets firms unilaterally disclose their future pricing plans through public statements, press releases, websites or through generic or targeted letters to customers.

Since 2013 several European competition authorities have investigated whether such announcements may facilitate anticompetitive coordination. Markets investigated have included mobile telecommunications by the Dutch Authority for Consumers and Markets (ACM); aggregates and cement by the UK Competition Commission (CC); retail energy markets by the UK Competition and Markets Authority (CMA); liner shipping by the European Commission (the Commission); and interest rates by the Norwegian Competition Authority.

It is perhaps unsurprising that public price announcements have attracted regulatory scrutiny given that the Commission’s guidelines on horizontal cooperation agreements (the Horizontal Guidelines)1indicate that private disclosure by an undertaking of its future pricing intentions to a competitor ‘would normally be considered and fined as cartels’2 under European competition law. Accordingly, the disclosure of similar information publicly potentially poses competition concerns. Indeed, Mlex reported Juliette Enser, Director of Cartel Enforcement at the CMA, as stating that: ‘Conceptually speaking, the damage of sharing future pricing intentions [...] seems to be the same whether exchanges are in private and public.’3 However, an OECD report on ‘Unilateral Disclosure of Information with Anticompetitive Effects’ of October 2012 (the OECD report) observed that:

There can be considerable uncertainty surrounding the issue of whether/when competition law should intervene against unilateral disclosures of information (as opposed to reciprocal exchanges of information). There is little guidance from enforcement authorities and courts and the economic literature is at times equivocal. Making a distinction between purely unilateral conduct, such as unilateral communications (or signalling) which falls outside of the reach of competition laws, and co-ordinated conduct which could in principle fall within these laws, can be a difficult task for competition enforcers.4

This chapter considers the potential anticompetitive and pro-competitive effects of public announcements of future prices under a number of headings:

  • the policy and legal issues regarding the treatment of public announcements;
  • the different types of public price announcements;
  • the economics of coordinated effects and how price announcements may facilitate coordination;
  • pro-competitive effects, the counterfactual and broader issues; and
  • conclusions as to the object and effect of price announcements and the challenges of providing robust compliance advice.

Policy and legal issues

When are public price announcements agreements or concerted practices?

The first issue to be addressed is whether a unilateral public announcement may constitute an agreement or concerted practice between undertakings such that article 101 of the Treaty on the Functioning of the European Union or the domestic equivalent prohibitions in individual member states may apply. On the one hand, the Horizontal Guidelines indicate that: ‘Where a company makes a unilateral announcement that is also genuinely public, for example, through a newspaper, this generally does not constitute a concerted practice within the meaning of Article 101(1).’5 An important caveat, however, is noted in a footnote: ‘This would not cover situations where such announcements involve invitations to collude.’6As discussed further below, this footnote appears highly relevant to the ACM’s investigation and potentially the Norwegian Competition Authority’s ongoing investigation. Such invitations to collude have been investigated in the United States in cases where companies have made public statements to analysts that effectively invite a rival to increase its prices and suggesting reciprocity if it does so.7

Moreover, the Horizontal Guidelines indicate that unilateral announcements may nevertheless be anticompetitive concerted
practices:

However, depending on the facts underlying the case at hand, the possibility of finding a concerted practice cannot be excluded, for example in a situation where such an announcement was followed by public announcements by other competitors, not least because strategic responses of competitors to each other’s public announcements (which, to take one instance, might involve readjustments of their own earlier announcements to announcements made by competitors) could prove to be a strategy for reaching a common understanding about the terms of coordination.8

This exception can be expected to apply generally as firms will typically respond to changes in their rivals’ competitive offerings, whether announced or actual. Accordingly, unilateral public announcements of future price intentions may risk leading to anticompetitive price coordination in a wide range of scenarios.

Price announcements can also be considered in the context of UK market investigation references

Price announcements may also be investigated under the market investigation provisions of the UK Enterprise Act 2002.9 Indeed, the relevant guidelines indicate that price announcements are one way in which information may be obtained or exchanged between competitors and thus facilitate coordination by increasing market transparency.

In particular, in its aggregates, cement and ready-mix concrete market investigation, the CC prohibited generic unilateral price announcements10 and the exchanging of information relating to current market shares, as well as imposing divestments.

The different types of public price announcements

It is important to appreciate that the competitive effects of price announcements depend on their nature, with the Horizontal Guidelines considering three different types of public price announcements:

  • public announcements of future pricing intentions which are predominantly targeted at competitors that are ‘invitations to collude’;
  • public announcements of future actual prices, which are quite different from the announcement of future pricing intentions; and
  • public announcements of future pricing intentions which are received by customers, but that might facilitate anticompetitive coordination.

Invitations to collude

Invitations to collude may be made in public as well as private, whereby firms disclose their future pricing intentions to rivals with the aim of setting out a collusive strategy.11 Accordingly, such invitations raise analogous issues to private exchanges of pricing information.

In the recent ACM antitrust probe in the mobile telecommunications sector, operators suggested changes to their commercial terms publicly prior to their implementation. These changes were disclosed even before the operator had made a final internal decision to adopt them. The ACM specifically refers in its press statement to a director announcing at a trade conference the operator’s consideration of reintroducing connection charges.12 Shortly thereafter, two rivals internally proposed a similar policy, which led to all operators reintroducing connection fees within half a year. This raises the obvious question of whether this parallel behaviour would otherwise have been observed.

While the probe found no factual evidence of price fixing, the chairman of the board of the ACM, Mr Fonteijn, made it clear that the observed behaviour raised a question of motivation: ‘Why would operators disclose their plans, if not to reach a common understanding about future market conduct?’13 The ACM also emphasised that such announcements did not involve any commitment to actually implement the changes, leaving operators leeway to see how their competitors react.

In January 2014, the three major operators, KPN, Vodafone and T-Mobile, accepted binding commitments to refrain from making verbal or written statements in the public domain about future, non-finalised conduct. They also committed to include this obligation in their compliance programmes and to organise staff compliance training.

It would also appear that similar concerns are motivating the Norwegian Competition Authority’s ongoing investigation of the way banks publicly signal interest rate changes and disclose other financial information.14

Announcements of future actual prices

A footnote to the Horizontal Guidelines indicates that the exchange of information regarding future prices that the company is committed to sell at, as opposed to intended prices, will not normally be considered a restriction of competition by object.15 Commitment is defined in terms of firms being ‘fully committed to sell in the future at the prices that they have previously announced to the public (that is to say, they can not revise them)’.16 In fact, firms’ commitment to their future prices is typically a question of degree (eg, for a period of time, while stocks last, etc), and rivalry might force subsequent changes to prices after announcement.17

Competition authorities will not wish to discourage normal price advertising, which inherently involves unilateral announcement of ongoing and thus future prices. Such price advertising is generally pro-competitive since it enables consumers to shop around for the best offer and thus brings prices down.18

The summary of the roundtable discussion in the OECD report indicated that:

In terms of policy, Prof. Motta suggested that unilateral public announcements of future prices or output with commitment value should be generally viewed as pro-competitive. On the other hand, private announcements or public announcements without commitment value are normally anticompetitive and should be viewed with suspicion.19

The Guidelines also add that even announced actual future prices may have an anticompetitive effect by committing a firm not to reduce prices.20 However, there are two counterpoints to this hypothetical concern. First, such a commitment may be difficult to achieve as customers tend not complain about price reductions, although they might complain about price cuts being offered to other customers. Second, any such commitment will also increase the profitability of ‘cheating’ by preventing competitors from responding to price-cutting by rivals.

Announcements of future pricing intentions

Absent any commitment, announced future prices may act as a low-cost price signalling device. For example, if a firm wishes to signal that it would like to charge higher prices, it could announce a future price increase, and then only increase its actual prices if other competitors follow by increasing their own prices. In such a scenario it will lose no sales if its rivals do not increase their prices as it would simply withdraw its announced price increase. However, this depends on customers disregarding announced future price increases until they are implemented. Announcing future price increases might be costly if customers respond by seeking quotes from alternative suppliers and switch suppliers.

It should be noted that, notwithstanding that the Horizontal Guidelines indicate that ‘Information exchanges between competitors of individualised data regarding intended future prices or quantities should therefore be considered a restriction of competition by object’,21 they also add a footnote to this sentence: ‘that public announcements of intended individualised prices may give rise to efficiencies and that the parties to such exchange would have a possibility to rely on article 101(3).’22

The assessment of announcements of future pricing intentions is therefore more complex, but is key to the recent Liner Shipping and Cement cases.

Liner shipping

On 22 November 2013, the Commission opened formal proceedings against 14 container liner shipping companies for potentially breaching article 101.23 The Commission expressed concern over the alleged concerted practice of liner shippers regularly communicating their future price intentions through press releases and articles in the trade press.

The public announcements in question were generic, that is, not targeted at a specific individual client, but instead informed the general customer base about a specific increased rate for cargo and containers on a particular trade route, as well as about the date of implementation. The Commission’s press release adds that the increases are ‘generally similar for all announcing companies’ and the ‘announcements are usually made by the companies successively a few weeks before the announced implementation date’. These announcements appear to relate to planned list price increases, rather than final actual prices.

This practice, the Commission argued, may allow shipping companies to signal to each other their future price intentions, which could eliminate strategic uncertainty and potentially allow companies to coordinate their behaviour in the market.24

Mlex has reported that the Commission has been discussing with the parties draft binding commitments that would bring the investigation to an end.25 These commitments would preclude the generic planned list price increases (eg, of say €250), but instead that the final actual price for specific shipments will increase (eg, from say €2,000 to €2,250).

Cement and aggregates

In January 2014, the CC published its final report on its aggregates, cement and ready-mix concrete market investigation.

The CC investigated the effects of price increase announcement letters which were sent by major aggregates and cement suppliers to their customers. In both markets, announcements usually take place at least once a year (around January, or more frequently), and are usually sent at least one month before the date of the planned increase. The CC only reached an adverse finding as regards anticompetitive coordination in relation to the supply of cement. On aggregates, the CC found that the evidence of price parallelism following the price announcements was mixed and the CC did not consider that there was evidence of anticompetitive coordination from market outcomes.

However, the CC found that coordination was occurring between the largest three cement producers, with this coordinating likely to evolve mainly around Great Britain market shares. The CMA also found that coordination:

...may be supplemented with information on prices which can be gathered through discussions with customers and in the context of cross-sales, as well as through price announcement letters. While we did not find coordination on price to be likely (as prices are individually negotiated), we considered that the price announcement letters facilitated price parallelism in the GB cement markets […] and also softened customer resistance to price increases.

This conclusion is discussed further in below.

The economics of coordinated effects and how price announcements may facilitate anticompetitive coordination

The previous section highlighted aspects of the theories of harm associated with unilateral price announcements. However, it is still necessary to consider whether any theory of harm is supported by the facts.

Since the fundamental concern with unilateral price announcements is that they may facilitate coordination, a natural starting point is to consider the economics of coordination. The mere existence of some common interest between firms (eg, increasing industry profits) may be insufficient for coordination. Instead, to form and maintain coordination, firms must reach a common understanding, align each party’s incentives, and they need to be able to detect and deter potential ‘cheating’.

These issues are also relevant to the assessment of coordinated effects in the context of mergers,26 and thus the Commission’s Horizontal Merger Guidelines’ analysis of coordinated effects is relevant.27

Reaching consensus and aligning incentives

The first potential impediment to coordination is the need for consensus to be reached as to the form of coordination firms will jointly adopt (eg, raising prices, allocating customers between firms, or cutting capacity), and the outcome of this coordination (eg, how high prices should be, how customers should be allocated, and/or what amount of capacity should be withdrawn). This is inherently difficult if the consensus required is complex, for example, where firms would need to agree on prices across many product variants, or where conflicts of interest between firms (eg, owing to differences in firms’ market shares, costs and capacities, etc) lead to firms having different incentives as to their preferred prices.

Obviously, one potential way in which consensus may be achieved is in the ‘smoke-filled room’ of a cartel meeting.28 However, even in the absence of direct private communications between firms, their competitive decisions may converge on some focal point, such as a certain level of prices or a particular discount structure. Unilateral price announcements may facilitate this convergence through the information they make publicly available, even if the announcements are not verifiable or binding.

Sustaining consensus – internal and external stability

The second potential impediment to coordination is the need to sustain adherence to the consensus outcome. In oligopolistic markets individual firms may have an incentive to depart or ‘cheat’ from the consensus (eg, charge lower prices) in order to increase their market share and profits, with this being a particular concern if adherence to the terms of coordination cannot be observed (eg, owing to a lack of market transparency).

The oligopoly dilemma of cheating is not necessarily removed by there being an explicit cartel agreement between firms, although this may provide a mechanism for reaching consensus. This is because even if a cartel agreement is reached, individual firms may have incentives to cheat on this agreement since the agreement will not be legally enforceable.

A collusive outcome may be more stable against cheating if the ‘game’ is repeated, in the sense that firms can frequently change their prices or increase capacity from one period to the next. In such circumstances, firms’ incentives to deviate depend on their profits from adhering to the consensus compared with the possible (temporary) gains to cheating followed by the lower profits that they may then experience once rivals detect and punish such cheating with any coordination collapsing.

The final issue is that any coordination may be rendered unsustainable owing to the threat of new entry and expansion or the exercise of buyer power.

Context is all

Several factors may affect the ability of firms to reach consensus, and to detect and then punish cheating. These are commonly referred to as a ‘check list’, with the Horizontal Guidelines identifying a range of market characteristics that may facilitate coordination, such as high transparency, high market concentration, non-complexity, stability and symmetry between firms. The Guidelines also consider how these characteristics might be affected by information sharing.29

In our view, these criteria should not be considered in isolation but in the precise context of any obstacles to price coordination which would otherwise exist absent the price announcements, and whether behaviour consistent with coordination can be observed.

Is competitive coordination on price feasible?

Competitive coordination can potentially occur in a variety of forms, but price announcements will inherently at most facilitate coordination on price.

Accordingly, an initial question is to explore whether any competitive coordination is likely to occur on price given the nature of competition in the market in question. For example, in some markets prices can be too complex, bespoke or volatile for coordination on price to be feasible, and announcing intended list prices changes may not facilitate the reaching of any consensus on competitive coordination. Instead, competitive coordination might occur – if at all – in other dimensions, such as customer allocation or geographic territories.

In this regard, the CC’s conclusions on the impact of price announcement letters relating to cement supplies are difficult to follow as the CC ‘did not find coordination on price to be likely (as prices are individually negotiated)’ but instead that coordination was occurring around market shares. Coordination may occur in multiple dimensions, but the CC’s report did not appear to reach a clear finding to this effect.

What is the evidence of competitive coordination on price?

As Marshall and Marx observe,30it would be natural for at least one firm to announce a price increase following a cost or demand shock even in markets where firms are vigorously competing. This is because a public announcement provides reassurance for buyers that they are being treated fairly, rather than being singled out for a price increase which other customers are not facing.31

Scheffman and Coleman32 have considered various empirical analysis to assess whether coordination is presently occurring. In particular, they note that if all or a large proportion of sales are made at list prices, then such lists can in theory be used to reach consensus and monitor deviations from list prices. In such a scenario, to consider coordination on price it may be sufficient to assess whether there is parallelism in list prices in terms of the timing and amount of list price changes, such that there is stability in the relative prices charged by different suppliers to individual customers.

They note that the position becomes more complex if many customers do not pay list prices. Nevertheless, it is still possible that transaction prices may be highly correlated with list prices. In such circumstances, one would need to assess whether there is parallelism in list prices in terms of the timing and amount of list price changes and that this translates into parallelism in realised transaction prices, such that there is stability in the relative prices charged by different suppliers to individual customers.

Finding either low levels of parallelism as regards price announcements, or if the implementation of list price changes varies widely across customers in size or direction, or that the relative transaction pricing of firms varies materially over time would each suggest that coordination on price via price announcements is not occurring and that reaching consensus on price is likely to be difficult.

This pricing analysis is inherently a one-way test. This is because there is a counterfactual question as to whether a high degree of parallelism might occur in any event. Parallelism could arise owing to common cost or demand conditions. More generally, parallelism might be attributable to prices being transparent for other reasons. For example, the CC found that cement suppliers obtained information on prices through discussions with customers and in the context of cross-sales between suppliers, as well as through price announcement letters.33 (This is not to dispute that price announcements might increase transparency.)

Since (lack of) transparency is easy to assert, Scheffman and Coleman suggest that the regulator should compare firms’ internal documents covering rivals’ prices to individual customers, capacities, sales and so on with rivals’ actual data. If estimates are significantly different from reality this would suggest a lack of transparency, thus making consensus and detection of deviations more difficult.34

Against this background, a number of observations can be made as regards the CC’s assessment of price announcements in cement markets.

Cement: assessment of price parallelism

In its investigation of the cement market, the CC carried out four separate sets of pricing analysis.35

The first set of analysis carried out by the CMA focused on possible coordination on price announcements in terms of the timing, amount and identity of which supplier announced prices first. The CMA found that there was ‘clear parallelism’ across these dimensions, and that ‘the cement suppliers appear to be signalling that they will try to accommodate the other GB producers’ price increases in may cases’.

The second set of analysis focused on the extent to which the announced prices were achieved on average. The CC found that ‘in many cases’ the GB producers achieved an average price increase of more than half of the announced price increase, but in only a small number of cases did they achieve the full or greater increases than those announced and in others there was no increase.36 This does not seem determinative, particularly as focusing on averages is not particularly informative as to the price changes faced by individual customers.

The CC then considered parallelism at the level of individual customers:

The degree of dispersion varied between announcements rounds. On some occasions, most price increases were close to the announced price increase. On other occasions, customers faced a range of price increases. On other occasions yet, there was little dispersion but increases did not cluster around the announced increase. This suggests that announced price increases do not, on their own, provide clear information about each individual customer’s price increase.37

This finding suggests that price announcements did not facilitate price coordination around the prices that customers actually paid.

As a final set of analysis the CC also considered the extent of correlations in average prices, and it found a ‘very high’ level of correlation among the three main GB cement producers of between 70 and 95 per cent even on a de-trended basis.38 This analysis is difficult to interpret for the following reasons:

  • correlations measure the existence of a linear relationship in prices, not whether there is a constant percentage differential in prices. Accordingly, average relative prices could have changed over time despite high correlation coefficients – if one supplier has become relatively cheaper over time this could reflect price competition;
  • average prices do not capture what is happening to prices at the individual customer level at which any coordination on price would need to occur, otherwise customers would switch to cheaper suppliers; and
  • moreover, the CC made no attempt to link these correlations to the announced prices, which is problematic given that the CC found that the announced prices did not ‘provide clear information about each individual customer’s price increase’.

Pro-competitive effects, the counterfactual and broader assessments of the risks of anticompetitive effects

The ECJ’s judgment of 1993 in Wood Pulp II is an important case as it focused on how price announcements by competitors should be assessed.

By way of background, pulp producers commonly concluded with their customers long-term supply contracts for up to five years. Under such contracts, the producer guaranteed its customers the possibility of purchasing each quarter a minimum quantity of pulp at a price which was not to exceed the price announced by it at the beginning of the quarter. These prices were communicated to customers and agents in advance, and were generally published in the trade press. The customer was free to purchase more or less than the quantity it reserved and could negotiate reductions in the announced price.39 The Commission asserted that this practice had the effect of making the market artificially transparent by enabling producers to obtain a rapid and accurate picture of competitors’ price quotes.

The ECJ’s assessment of this system of price announcements focused on two questions. First, whether the system constitutes in itself an infringement of article 101, but the ECJ rejected the Commission’s arguments that a concerted practice could simply be inferred.40

Second, the ECJ assessed whether that system was ‘merely evidence of concertation on announced prices which took place at an earlier stage’.41 The ECJ went on to consider whether the system of price announcements, the simultaneity or near simultaneity of the announcements and the fact that the announced prices were the same constitute ‘a firm, precise and constituent body of evidence of prior concertation’.42

Before addressing the substance of this issue, the ECJ held that:

[...] parallel conduct cannot be regarded as furnishing proof of concertation unless concertation constitutes the only plausible explanation for such conduct. It is necessary to bear in mind that, although Article [101] of the Treaty prohibits any form of collusion which distorts competition, it does not deprive economic operators of the right to adapt themselves intelligently to the existing and anticipated conduct of their competitors.43

In order to explore these issues, the ECJ appointed two economic experts. Their key findings can be summarised as follows.

First, while the ECJ did not go as far as considering whether an exemption was warranted, it did emphasise the views of the experts as to the importance of the context in which price announcements were made.

In particular, price announcements needed to be viewed in the context of long-term customer–supplier relationships, with there being multi-year contracts, and where there are switching costs to changing the pulp mixture paper manufacturers used and the market was cyclical. Since market pulp constitutes between 50 and 75 per cent of the cost of the paper, purchasers wished to ascertain these costs as soon as possible to estimate their own costs and fix their own paper prices and that customers. Given this, customers had demanded the introduction of price announcements. They announced a price ceiling below which transaction prices were subsequently negotiated.

This raises an important point, namely that greater certainty as to a supplier’s future prices may be valuable to customers and smaller competitors in a variety of ways:

  • it enables customers to compete more actively in the markets they serve by providing greater certainty as to their costs. Even if announced prices are merely a price ceiling, knowledge as to the evolution of maximum prices may be helpful to customers;
  • it may provide customers with an opportunity to seek an alternative source of supply; and
  • it may assist smaller suppliers with (imperfect) information as to trends in prices, and potentially to target the customers of larger suppliers.

 In this regard, it is appropriate to emphasise that any claims as to pro-competitive effects should be supported by hard evidence, and any benefits need to offset any anticompetitive effects.

Second, the ECJ emphasised that the experts held that the simultaneous price announcements were owing to high market transparency, not the communication by suppliers of their announced prices to the trade press. Accordingly, the ECJ emphasised that this transparency ‘does have to be described as artificial’.

Finally, the ECJ noted that the experts found that the price trends were attributable to suppliers responding to the evolution of demand or capacity. The ECJ also be observed that the market did not appear particularly susceptible to coordination as other firms which were not addressees of the Commission’s decision had a combined market share of some 40 per cent. Professor Motta has concurred with this view, notwithstanding his concerns as to the announcement of future price intentions.44

Conclusions: the object and effect of price announcements and the challenges of providing robust compliance advice

The object and effect of price announcements

There is a substantial debate as to the appropriate scope of object infringements of European competition law.45 In principle, it is easy to distinguish between two particular types of price announcements:

  • invitations to collude where there may be a clear anticompetitive object to induce competitors to follow price increases. However, as the US Federal Trade Commission observed in Valassis, competition authorities should not over-interpret public statements as invitations to collude so as ‘not to unduly chill legitimate speech’; and
  • announcements of actual future prices (as opposed to future pricing intentions), as in GB energy markets, which are not being used for signalling purposes. It would seem inappropriate to treat these as presumptively facilitating coordination.

However, public announcements of future pricing intentions are much more nuanced and this assessment will depend on the facts. If such announcements facilitate anticompetitive coordination, then they may seriously prejudice customers’ interests. It will be very interesting to read why the Commission has considered that issues arise in the liner conferences case should it accept commitments.

The challenges of providing robust compliance advice as regards announcements of future intended prices

We would suggest considering the following factual matters as regards the assessment of unilateral public announcements of future intended prices:

  • Are the products or services homogenous and predominantly supplied by four or fewer large suppliers, with stable market shares?46
  • Why do suppliers announce future intended prices? Are there genuine customer benefits?
  • Is there strong parallelism in announced prices?
  • How do suppliers set prices to individual customers (for example, do they closely track announced prices or not?) and what impact do price announcements have on actual transaction prices?
  • What degree of transparency is there as to competitors’ prices to individual customers suppliers?
  • Are the relative prices of different suppliers stable over time?

This should include a careful review of the firm’s internal documents covering these matters, not least because any competition authority may review these documents.

Notes

  1. ‘Guidelines on the applicability of article 101 of the Treaty on the Functioning of the European Union to horizontal co-operation agreements’, OJ C 11, 14 January 2011, pp. 1–72.
  2. Section 74 of the Horizontal Guidelines. Moreover, section 62 of the Horizontal Guidelines adds that ‘mere attendance at a meeting where a company discloses its pricing plans to its competitors is likely to be caught by article 101, even in the absence of an explicit agreement to raise prices.’
  3. Ms Enser indicated that her remarks during the ABA Antitrust Spring Meeting were made in a personal capacity. Mlex also reported Ms Enser as adding that it was also relevant to consider the reasons for the public announcements, the context coming out of the announcement, and that there may be legitimate reasons for disclosing seemingly sensitive information.
  4. ‘Unilateral Disclosure of Information with Anticompetitive Effects’, OECD (2012) , p. 19. Available at www.oecd.org/daf/competition/Unilateraldisclosureofinformation2012.pdf.
  5. In the UK, the Chapter I prohibition of the Competition Act 1998.
  6. Section 63 of the Horizontal Guidelines.
  7. Section 63 of the Horizontal Guidelines.
  8. See, for example, the description of the US Federal Trade Commission’s cases against Valassis Communications and U-Haul in Collusion and Unilateral Price Announcements, by Antonio Capobianco, Senior Expert on Competition Law, OECD Competition Division, published by Competition Policy International 2012.
  9. Section 63 of the Horizontal Guidelines.
  10. ‘Guidelines for market investigations: Their role, procedures, assessment and remedies’, Competition Commission (2013), ‘Part 3: Section 3 – The Competitive assessment, 3. Coordinated conduct by firms’. Price announcements are referred to at section 258.
  11. The CC decided to impose behavioural remedies which restrict the use of generic price letters. Firms can only send customer-specific price announcement letters that specify the effective date of the price change, the current and new price, and details of any other changes that contribute to the overall price paid.
  12. An interesting example of such an alleged invitation to collude arose in relation to the 2006 US Federal Trade Commission (FTC) charges that Valassis had violated section 5 of the Federal Trade Commission Act by inviting its only rival, News America Marketing, to collude. The FTC found that in Valassis’ earnings call it communicated to its proposed rival with ‘extraordinary specificity’ that ‘Valassis would cease competing for News America’s customers, providing that News America likewise ceased competing for Valassis’ customers. In addition, Valassis proposed that prices should be restored to the pre-price war level of US$6.00 per page and US$3.90 per half page per thousand booklets and described how business with shared customers and outstanding bids to News America’s customers would be handled’. The FTC observed that there was no legitimate business justification for disclosing such information, and considered it irrelevant that there was no evidence that News America had accepted Valassis’ invitation to collude. See www.ftc.gov/sites/default/files/documents/cases/2006/03/060314ana0510008.pdf.
  13. ACM (7 January 2014): ‘Teoezeggingsbesluit mobiele operators’.
  14. ‘Dutch probe into telecom price signalling sends broader warning, Fonteijn says’, Mlex interview, 31 January 2014.
  15. Christine Meyer, director general of the Norwegian Competition Authority, was also reported by Mlex as stating at the ABA 63rd Antitrust Spring Meeting in April 2015: ‘What we have been looking at, and what we need to investigate more is: is there a form of signaling practice in the financial industry which is detrimental to competition?’
  16. Section 74 of the Horizontal Guidelines.
  17. Section 74 of the Horizontal Guidelines.
  18. An example of the assessment of announced future actual prices is provided by the CMA’s consultative working paper on ‘Coordination in the retail market facilitated by price announcements’ of March 2015 in the context of its ongoing energy market investigation (available at https://assets.digital.cabinet-office.gov.uk/media/54f8765de5274a1414000001/Coordination_retail_pricing.pdf).
    The CMA found that the six large energy suppliers make public statements, in advance of implementation, of intentions to change the price of their standard variable energy tariff. These announcements will typically give a headline rate change and an implementation date. The headline rate is typically an average across regions for a dual fuel domestic customer, paying by monthly direct debit with typical consumption.
    The CMA observed that since mid-2011 changes in headline rates were publically announced at most around 10 days before the six large energy suppliers started to notify domestic customers of the price change, with such advance notifications of price changes being required under energy suppliers’ supply licences.
    One of the key issues which the CMA considered was whether suppliers were using these announcements to signal their intentions to rivals and for rival suppliers to be in a position to adjust their behaviour accordingly. However, bar one slight error in an announcement by one supplier in 2007 (which was then subsequently corrected), there was no evidence of any of the six large energy suppliers modifying the level or timing of price changes between announcement and implementation. Accordingly, the CMA concluded that: ‘At this stage, we do not have evidence of suppliers using price announcements as a mechanism to signal their intentions in relation to the pricing of their standard variable tariff to rival suppliers to determine their prices accordingly.’
  19. Restrictions on competitors’ ability to advertise their prices have been held to be anticompetitive in a range of contexts. For example, the first infringement decision of the newly formed CMA on 19 March 2015 related to local agreements involving a trade association (Three Counties), three of its estate agent members (Castles, Hamptons International and Waterfords) and a local newspaper (the Surrey & Hants Star Courier). In particular, these agreements prevented members of the trade association and also non-members from advertising their fees or discounts in the Surrey & Hants Star Courier. The CMA’s press release indicated that: ‘The CMA had launched the investigation over concerns that these arrangements had the object of reducing the competitive pressure on estate and lettings agents’ fees in the local area in and around Fleet in Hampshire. In addition, the restrictions may have made it harder for potential competitors to enter the market by using the level of their fees to attract new customers. The CMA alleged that these practices potentially limited consumers’ choice and obstructed their ability to compare prices and assess value for money.’ (Available at www.gov.uk/government/news/companies-fined-over-775000-in-cma-investigation-into-advertising-of-agents-fees).
  20. OECD report, p. 223.
  21. Footnote 58 of the Horizontal Guidelines.
  22. Section 74 of the Horizontal Guidelines.
  23. Section 74 of the Horizontal Guidelines.
  24. Commission, ‘Antitrust: Commission opens proceedings against container liner shipping companies’, 22 November 2013.
  25. Accordingly to various articles on Mlex, including on 22 November 2013 and 7 March 2014. Mlex also reports a Commission spokesman as indicating that the price signalling ‘may constitute a more subtle way for competitors to collude and replace competition with practical cooperation’.
  26. ‘Liner shippers consider revised EU antitrust proposal’, MLex, 6 May 2015.
  27. The approach set out below closely follows the assessment of anticompetitive coordinated effects as regards mergers which is set out in Chapter 6 of UK Merger Control: Law and Practice, Second edition, Nigel Parr, Roger Finbow, and Matthew Hughes, Thomson/Sweet & Maxwell, 2004.
  28. ‘Guidelines on the assessment of horizontal mergers under the Council Regulation on the control of concentrations between undertakings’, OJ C 31, 5 February 2004, pp. 5–18. The Commission’s Horizontal Guidelines adopt the criteria established by the General Court in Airtours (2002): ‘Coordination is more likely to emerge in markets where it is relatively simple to reach a common understanding on the terms of coordination. In addition, three conditions are necessary for coordination to be sustainable. First, the coordinating firms must be able to monitor to a sufficient degree whether the terms of coordination are being adhered to. Second, discipline requires that there is some form of credible deterrent mechanism that can be activated if deviation is detected. Third, the reactions of outsiders, such as current and future competitors not participating in the coordination, as well as customers, should not be able to jeopardise the results expected from the coordination.’ (Section 41.)
  29. Indeed, section 73 of the Horizontal Guidelines states that: ‘Exchanging information on companies’ individualised intentions concerning future conduct regarding prices or quantities is particularly likely to lead to a collusive outcome. Informing each other about such intentions may allow competitors to arrive at a common higher price level without incurring the risk of losing market share or triggering a price war during the period of adjustment to new prices […]’
  30. Sections 77–85 of the Horizontal Guidelines. A good textbook which covers the economics of coordinated effects is Chapter 4 Competition Policy: Theory and Practice, by Massimo Motta, 2004. Mr Motta is, of course, Chief Economist at the Commission, which renders this book essential reading for those interested in European competition policy. There is also a large body of research as to the factors which influence the existence, duration, and impact of cartels, with an excellent summary being set out in the article on ‘What Determines Cartel Success?’, by Margaret C Levenstein and Valerie Y Suslow, Journal of Economic Literature Vol. XLIV (March 2006), pp. 43–95. An obvious caveat to this research is that it relates to cartels which have been detected. For example, cartels might be more likely to be detected in markets which are less concentrated since more lines of direct communication might be required to reach and sustain consensus between many cartelists, and the discovery of such communications will often be key to competition authorities proving the existence of a cartel.
  31. The Economics of Collusion – Cartels and Bidding Rings, Robert C Marshall and Leslie M Marx, The MIT Press, 2012.
  32. Ibid. This is not to dispute that there have been instances where price announcements have also been used as parts of cartels so as to soften buyer resistance to price increases, with Marshall and Marx indicating that this has been the case as regards a number of US and Commission cartel cases (see pp. 110–112).
  33. ‘Quantitative Analyses of Potential Competitive Effects from A Merger’, David T Scheffman and Mary Coleman, George Mason Law Review, 2003.
  34. CC, ‘Aggregates, cement and ready-mix concrete market investigation’, Final report, 2014, at section 8.208.
  35. Further questions which could be raised are:
    •   Whether prices comparisons should be based on averages or the prices faced by individual customers? Coleman and Scheffman observe rightly that since any effective coordination would be occurring at the level of prices to individual customers (otherwise they will switch suppliers), it is sensible to look at prices at that level. Averages are, by definition, heavily influenced by high and low prices, and if suppliers serve a different mix of customers this will lead to average prices varying even if similar customers face identical prices; and
    •   What degree of price parallelism is ‘high’ or ‘low’? They suggest using a figure of 5 per cent, reflecting that market definition questions are often formulated in terms of customers’ responses to 5 per cent price changes, but that this might also be tailored to the importance of small price differences to customers. For example, given that cement is a homogenous product, small price changes might affect customers’ choices of supplier.
  36. Much of the data have been redacted on confidentiality grounds, but the available data does not suggest a strong basis for concluding that price announcements materially contributed to any price parallelism at the level of individual customers, with this probably contributing to the CMA’s nuanced adverse finding.
  37. Section 7.196 of the CC’s Aggregates, Cement and ready mix concrete report.
  38. Section 7.210(a) of the CC’s Aggregates, Cement and ready mix concrete report.
  39. Section 7.211 of the CC’s Aggregates, Cement and ready mix concrete report. Correlation coefficients range from –100 per cent and +100 per cent. The closer the correlation coefficient is to +100 per cent, the more changes in one variable are associated with changes of the same sign in the other variable.
  40. ECJ judgment in Wood Pulp II (1993), section 13–14.
  41. In its decision in Wood Pulp I (1984), the Commission held that ‘the system of quarterly announcements, which the firms voluntarily chose, constituted in itself, at the very last, an indirect exchange of information on future market conduct’ (section 108(b)).
  42. ECJ judgment in Wood Pulp II (1993), section 57.
  43. ECJ judgment in Wood Pulp II (1993), section 70.
  44. ECJ judgment in Wood Pulp II (1993), section 71.
  45. See chapter 4 of Competition Policy: Theory and Practice, by Massimo Motta, 2004.
  46. The most recent contribution to this debate has followed the ECJ’s judgment in September 2014 that the General Court had been wrong to conclude that the pricing measures adopted by Groupement des Cartes Bancaires had as their object the restriction of competition.
  47. There is no hard and fast rule as to concentration levels at which coordinated effects are likely. 

Previous Chapter:EU: Chemicals

Next Chapter:EU: Energy

Interested in becoming a GCR author? Please contact our co-publishing manager.

Get in touch