On November 12, 2013, the European Commission published a summary of its December 20, 2012 decision (the “Decision”) accepting commitments offered by members of the Thomson Reuters group (“Thomson Reuters”). The Decision ended a three-year-long investigation that began in October 2009 into Thomson Reuters’ alleged abuse of a dominant position in a market defined as consolidated real-time data feeds (“CRTDF”).
Although Thomson Reuters did not agree with the Commission’s analysis, it offered commitments to address the Commission’s concerns, and the Decision made these commitments binding without formally finding an infringement or imposing fines. Ending an investigation through commitments may allow the Commission and the target of an antitrust investigation to address competition concerns while avoiding some of the time and expense involved in a full investigation. For practitioners, however, commitment decisions can be frustrating, because the Commission does not discuss its analysis in detail. Thus, commitment decisions offer limited guidance into the Commission’s thinking.
This post analyses the public information on Thomson Reuters in an effort to elucidate the Commission’s approach. Interestingly, it is not clear that the Decision is consistent with Microsoft v. Commission and other cases in which the Commission has found an abuse where a dominant company uses intellectual property (“IP”) in one market to deter competition in another.
The Thomson Reuters investigation concerned an alleged abuse of dominance in the worldwide market for CRTDFs by Thomson Reuters’ use of Reuters Instruments Codes (“RICs”), short alphanumerical codes that identify securities and their trading locations. Thomson Reuters developed RICs and claimed copyright over them. Thomson Reuters prohibited customers from using RICs to retrieve data from other providers’ CRTDFs and prevented third parties from creating and maintaining mapping tables incorporating RICs that would allow Thomson Reuters’ customers to interoperate with other providers’ CRTDFs. The Commission preliminarily concluded that these practices created substantial barriers to switching providers and constituted an abuse of Thomson Reuters’ dominant position in the worldwide market for CRTDFs.
Why the Commission thought Thomson Reuters’ RIC policies locked in customers was discussed in the Commission’s December 14, 2011 notice. Thomson Reuters explained that it did not prevent customers from cross-referencing RICs with the symbology of other providers as long as the resulting database or mapping table was not used to retrieve data from another provider (para 4, footnote 4). The Commission did not claim that customers had to use RICs to access CRTDFs of other providers, but found that “where RICs have become embedded in server-based applications, switching providers involves the removal of RICs and the recoding of those applications to replace RICs with an alternative symbology. This is an arduous and often excessively costly undertaking” (para. 6).
Between November 2011 and November 2012, Thomson Reuters submitted three sets of commitments, including the final commitments. Under those commitments, Thomson Reuters will offer licenses (“ERLs”) to CRTDF customers, giving those customers the possibility to license additional RIC symbology usage rights for the purpose of switching CRTDF providers by retrieving real-time financial data from competitors’ data feeds, as well as licenses to third-party developers to develop and maintain switching tools for sale to customers. ERLs will be granted for five-year terms, with an option to extend for two years.
The final commitments reflected a number of changes required as a result of the Commission’s market tests of the first two packages. The changes included reducing the ERL fees and simplifying the fee structure; broadening the ERLs’ geographic scope, which was made worldwide for customers with genuine business operations in the EEA; broadening the permitted uses to include single-source OTC instruments and human interfaces; and potentially lengthening ERLs’ terms by providing for a two-year option to extend. The possibility for third-party developers to take a license to develop and maintain switching tools for sale to customers was also added.
Market Definition and Dominance
The Commission defined the relevant market as the global market for CRTDFs. The Commission noted that CRTDFs compete only marginally with direct real-time data feeds and do not compete with integrated desktop services. The Commission concluded that Thomson Reuters held a dominant position in the global market for CRTDFs, noting that its market share is a multiple of the next largest competitor’s and that no customer represents a sufficient percentage of demand that it enjoys a substantial countervailing market power.
By contrast, in Case No COMP/M.4726 – Thomson Corporation/ Reuters Group (“Thomson/Reuters”), the Commission defined markets for real-time market data (“RTMD”) sold through desktop products/workstations and real-time data feeds, among others. In Case No COMP/M.3692 – Reuters/Telerate (“Reuters/Telerate”), the Commission analysed the provision of RTMD to end-users. The Commission discussed the possibility of segmenting this activity by asset class (para. 11) but ultimately left the market definition open.
The Commission also addressed the markets for financial data in a more recent case, Case No COMP/M.6166 – DEUTSCHE BÖRSE/NYSE EURONEXT, where the Commission discussed overlaps in the provision of “non-proprietary” market data. The notifying parties’ activities overlapped in RTMD and real-time data feeds, as well as other content sets. The Commission noted that the parties’ market shares were low and that these markets were characterised by strong players such as Thomson Reuters and Bloomberg, as well as other competitors, such as SIX Telekurs and IDC (para. 151).
The principal difference between the Commission’s approach to market definition in this case lies in the definition of CRTDFs as a narrower market compared to RTMD. In Thomson/Telerate, the Commission found that Reuters’ acquisition of Telerate would not lead to single-firm dominance, noting that Bloomberg’s share of the global RTMD market was significantly higher than Reuters’ (35-45% compared to 24-35%) and that there were a number of other, smaller providers, including individual exchanges (para. 17). In a market limited to CRTDFs, however, Thomson Reuters’ market position was apparently much stronger.
Interestingly, the Commission’s decision does not discuss Thomson Reuters’ position in RICs. In Thomson/Reuters, the Commission discussed instrument codes as a potential relevant market. The Commission stated that although it “is not evident that instrument codes, such as RICs, represent a separate product market, [RICs] are analysed, in the competitive assessment part of the decision” (para. 70). The Commission did not take a position on the market in which RICs compete, but it concluded that the combination of Thomson Corporation and Reuters Group would not give rise to conglomerate effects as regards the use of RICs (para. 133).
The Commission concluded that Thomson Reuters’ use of RICs created a barrier to customers’ ability to switch to alternative CRTDF providers and constituted an abuse of Thomson Reuters’ dominant position. The Decision does not explain the Commission’s legal analysis, and the Commission did not make a formal finding of an infringement.
It is not clear how the Commission’s approach squares with Microsoft and other established law in this area. In Microsoft, the Court of Justice held that a dominant company’s refusal to license IP would constitute an abuse of its dominant position where the refusal:
- relates to a product or service that is indispensable to the exercise of an activity on a neighbouring market;
- excludes any effective competition on that neighbouring market;
- prevents the emergence of a new product for which there is potential consumer demand; and
- is not objectively justified.
If the Commission had followed the Microsoft approach, it might have defined separate markets for RICs and CRTDFs. Assuming the Commission concluded that Thomson Reuters had a dominant position in RICs (instead of or in addition to CRTDFs), the Commission could then have determined whether the use of RICs was indispensable for other companies to compete in CRTDFs and whether Thomson Reuters’ conduct excluded any effective competition in CRTDFs and prevented the emergence of a new product for which there is potential consumer demand.
If the Commission had followed this approach, however, the Commission might have concluded that Thomson Reuters’ conduct was not abusive, for at least four reasons.
First, the Commission previously found that it is not clear that RICs constitute a market separate from CRTDF or RTMD.
Second, the Commission admits that access to RICs is not indispensable for customers to use alternative CRTDF providers, merely that customers choosing to embed RICs in server-based applications might find it “excessively costly” to switch CRTDF supplier. The Decision doesn’t indicate why the cost was considered excessive or what an acceptable cost level might be.
The situation of Thomson Reuters’ competitors seems analogous to that of Bronner in Case C-7/97, Bronner v. Mediaprint. In that case, Bronner sought access to Mediaprint’s nationwide newspaper home-delivery service, arguing that nationwide home-delivery services constituted a separate market to the publication of daily newspapers and that it would be “entirely unprofitable” for Bronner to develop its own nationwide delivery service in view of its smaller number of subscribers. The Court rejected Bronner’s argument, noting that for an abuse to arise Bronner would need to show “at the very least” that it would not be economically viable for Bronner to develop a nationwide home-delivery service if its circulation was equal to Mediaprint’s (para. 46).
Third, although the Commission found that Thomson Reuters enjoyed a dominant position in CRTDFs, the existence of at least one well-resourced competitor — Bloomberg – suggests that it would be difficult to show that Thomson Reuters’ RIC practices excluded any effective competition.
Fourth, even assuming that Thomson Reuters’ use of RICs constituted a barrier to switching, this barrier did not prevent the development of a new product, since both Thomson Reuters and its competitors already offered CRTDFs.
Because the Commission’s procedure for making commitments binding does not require the Commission to adopt a full-fledged infringement decision, the Decision leaves many open questions. In particular, the legal basis for the Commission’s conclusion that Thomson Reuters’ practices constituted an abuse of its dominant position in the global CRTDF market is unclear. Based on the information published by the Commission, it seems unlikely that the Microsoft criteria for requiring a dominant company to license IP to third parties to facilitate competition in CRTDFs would have been met.
Whether the Commission applied the Microsoft criteria or developed a new theory of abuse with broader implications will remain unclear unless and until the Commission adopts an infringement decision applying the legal theory underpinning the Decision. The uncertainty is regrettable, because the application of EU competition rules in the financial sector is a key competition policy concern in the wake of the financial crisis. The situation is even more regrettable since these questions intersect with regulatory requirements in relation to market data, including in particular the reform of the Markets in Financial Instruments Directive.