EC, February 19, 2008, No M.4726

THE COMMISSION OF THE EUROPEAN COMMUNITIES,

Having regard to the Treaty establishing the European Community, Having regard to the Agreement on the European Economic Area, and in particular Article 57 thereof, Having regard to Council Regulation (EC) No 139-2004 of 20 January 2004 on the control of concentrations between undertakings (1), and in particular Article 8(2) thereof, Having regard to the Commission's decision of 8 October 2007 to initiate proceedings in this case, Having regard to the opinion of the Advisory Committee on Concentrations, Having regard to the final report of the Hearing Officer in this case, Whereas:

I. INTRODUCTION

1. On 3 September 2007, the Commission received a notification of a proposed concentration pursuant to Article 4 of Council Regulation (EC) No 139-2004 ("the Merger Regulation") by which the undertaking the Woodbridge Company Limited ("Woodbridge"), which is the Thomson Corporation ("Thomson") family holding company, Canada, acquires within the meaning of Article 3(1)(b) of the Council Regulation sole control of the undertaking Reuters Group PLC ("Reuters"), United Kingdom, by way of scheme/plan of arrangement involving a dual-listed company structure. Thomson and Reuters are collectively referred to in this decision as "the notifying parties".

2. After examination of the notification the Commission concluded on 8 October 2007 that the notified operation fell within the scope of the Merger Regulation and that it raised serious doubts as to its compatibility with the common market and the EEA Agreement. The Commission therefore initiated proceedings in accordance with Article 6(1)(c) of the Merger Regulation.

3. On 6 December 2007, the notifying parties offered initial commitments with a view to rendering the concentration compatible with the common market. Following several weaknesses highlighted in the market test conducted by the Commission, these initial commitments were improved and the final version of the commitments was submitted on 21 December 2007.

4. The Commission has concluded that the commitments entered into by the notifying parties remove the serious doubts as to the compatibility of the notified operation with the common market. The concentration is therefore to be declared compatible with the common market and the EEA Agreement pursuant to Articles 8(2) and 10(2) of the Merger Regulation and Article 57 of the EEA Agreement.

II. THE PARTIES AND THE OPERATION

5. Thomson is a global provider of value-added information, integrated with software tools and applications, to professionals in the legal, tax, accounting, financial services, scientific research and healthcare markets. Thomson is controlled by Woodbridge, the Thomson family holding company.

6. Reuters is a global provider of information tailored for professionals in the financial services, media and corporate sectors. The company sources, aggregates and disseminates market data content including real-time market data, and provides analytics, trading and messaging capabilities needed by financial professionals. Moreover, Reuters is best known as the world's largest international multimedia news agency. It supplies news in the form of text, graphics, video and pictures to media organisations' websites, financial institutions and other businesses around the world.

7. The entire issued share capital of Reuters will be acquired by Thomson-Reuters PLC (or a subsidiary of Thomson-Reuters PLC), a newly established UK holding company, currently wholly owned by Thomson, in consideration for the payment of cash and the issue of shares in Thomson-Reuters PLC to the shareholders of Reuters through an English law scheme of arrangement and a Canadian law plan of arrangement. Upon implementation of the transaction, Thomson (which will be renamed Thomson-Reuters Corporation) and Thomson-Reuters PLC will enter into a series of contractual agreements, notably adopting identical boards of directors and combining their management and operations whilst retaining their separate legal identities. As a result of this transaction, Woodbridge will have a 53% voting control over Thomson-Reuters Corporation and Thomson-Reuters PLC and as a consequence will exercise sole control over Reuters.

8. The notified operation consists of the acquisition of sole control by Woodbridge of Thomson-Reuters PLC, and thereby Reuters, and therefore constitutes a concentration within the meaning of Article 3(1)(b) of the Merger Regulation.

III. COMMUNITY DIMENSION

9. The undertakings concerned have a combined aggregate worldwide turnover of more than EUR 5 billion (Thomson: EUR [...]*; Reuters: EUR [...]*). The aggregate Communitywide turnover of each of at least two of the undertakings concerned is more than EUR 250 million (Thomson: [...]*; Reuters: [...]* for 2005. Neither Thomson nor Reuters achieve more than two thirds of their Community wide turnover in any Member State. The notified transaction therefore has a Community dimension.

IV. RELEVANT MARKETS

10. The proposed transaction relates to the financial information industry, that is to say, the provision of financial information, analytics and trading capabilities. The main users of the notifying parties' products are customers in the financial service industry, such as banks, traders, funds, corporate customers.

IV.A. GENERAL CHARACTERISTICS OF THE MARKET

11. Financial information products include market data which represents one of the key components and consists of indicative or tradable prices for various types of financial instruments such as equities, corporate and government bonds, currency and traded commodities. Financial professional and organizations around the world involved in the trading of financial instruments use market data to make effective investments decisions as well as to provide advice and monitor and validate transactions after they are executed.

12. Financial information products may also include real time information, decision support tools, historical and reference data and news and analysis. Such information is designed to assist end-users to monitor the markets, reach investments decisions, manage investment risks, advise clients, etc. Some providers of financial information products also provide trading and messaging capabilities to enable customers to communicate and execute trades. In the financial information industry, customers range from global and very large institutions including central banks, financial institutions, government offices and agencies, traders and brokers operations, as well as hedge funds and private equity funds. These customers often consume a range of items within the markets for financial information products and tend to be sophisticated organizations.

13. Customers in the financial services industry can be broadly categorised into those engaged in "on-trading floor" activities and those engaged in "off-trading floor" activities. Ontrading floor users are those involved in the sales and trading of financial instruments and trade execution. Off-trading floor users are those involved in research, providing advice and asset management. Many large financial institutions have both types of users in different parts of the organisation.

14. Another distinction often drawn is between the "sell-side" (that is to say, customers whose primary business is selling or trading financial products) and the "buy-side" (customers whose primary business is investing in financial products).

15. Example of on-trading floor users include "sell-side" salespeople, sales traders and back-office personnel at broker dealers, as well as "buy-side" traders in asset management firms, mutual funds and hedge funds. Further, users operating in this sector are much more focused on real time data than is the case in other segments.

16. Examples of off-trading floor users include investment or portfolio managers, wealth managers, investment bankers, research analysts and corporate executives who make complex financial decisions outside the sales and trading environment. Given their longer-range focus, their needs tend to focus more on historical and reference data and analytics than on extensive real time data and information.

17. According to the parties, off-trading floor users can be further categorised into four main customer segments: Wealth Management ("WM"), Investment Management ("IM"), Investment Banking ("IB") and Corporates.

18. The Wealth Management segment primarily includes financial advisers, private banks and retail brokers whose activities are geared towards serving private individuals and advising them on the management of or directly managing their assets. This often also includes providing online services for the user's customers. Customers in the Wealth Management segment typically need access to basic market data, news and basic research products that also provide, or interoperate with client management functionality, allowing them to plan and manage individuals' wealth as well as enhance the client/advisor relationship. As the value of the assets involved is smaller than for professionals operating in other segments, wealth managers often use less sophisticated investment and analytical tools

19. Investment Management users include portfolio managers and research analysts at hedge funds, asset management firms, mutual funds and insurance companies. Their common characteristic is their management of investments on behalf of third parties such as institutional investors or collections of retail investors (e.g. policy/unit holders). Typically this segment has the most sophisticated needs and as such requires a much richer mix of historical and reference data (such as company filings and reports) than is the case for customers operating in the other segments.

20. Within the Investment Banking segment, the role of investment bankers is to advise corporate clients on possible mergers and acquisitions ("M&A") activity, and on raising finance in the equity and debt capital markets. Their primary needs are for a combination of market and reference data, deal data and ownership information about the companies they serve. Private equity and venture capital professionals are also included within this segment, as their needs are similar to investment bankers. Also included are sell-side research analysts, who produce research reports on investment opportunities.

21. In addition to users in the financial services industry, financial information products are purchased by corporate customers. The Corporates segment relates to tasks carried out internally within companies themselves, and includes two user sub-segments: (i) Investor Relations ("IR"); and (ii) Corporate Treasury. Corporate customers require basic market awareness (and trading in case of corporate treasury), research and advisory services, workflow tools and certain specialised capabilities such as ongoing and special event web-casting and web-hosting services (2).

22. Figure 1 below provides a graphical summary of these main distinctions and user groups of financial information products.

<table>

Source: notifying parties

IV.A.1. The parties' activities in the industry

23. Thomson Financial is a business unit of the Thomson Corporation. Thomson Financial is organised into six customer segment divisions, covering the supply of financial information products to each of six user group segments, together with a Customer Enterprise Solutions division and a number of other strategy and operations support divisions. The six customer segment divisions are: (i) Wealth Management; (ii) Investment Management; (iii) Investment Banking; (iv) Corporates; (v) Fixed Income; (vi) Institutional Equities. Approximately [70-80]*% of Thomson Financial's 2006 revenues were generated by its four off-trading floor divisions (3) and [10-20]*% was generated by Fixed Income and Institutional Equities divisions (4).

24. Reuters specialises in the provision of financial information products mainly to financial services professionals, as well as in the provision of multi-media news and related services to media organizations (5). Reuters is organized in four divisions: (i) Sales & Trading; (ii) Research & Asset Management, (iii) Enterprise; and (iv) Media. Sales & Trading is Reuters' largest business division accounting for [50-60]*% of revenues in 2006. It provides real time data, news and trading tools primarily to on-trading floor users (6) through its financial information products and trading systems (7). Research & Asset Management is smaller, accounting for [20-30]*% of revenues in 2006 and operates in the following segments: (i) Wealth Management; (ii) Investment Management; (iii) Investment Banking; and (iv) Corportates (8).

IV.A.2. Features of the markets (datafeeds / desktop products)

25. In the financial information business, there are different methods for the supply of financial information products to customers. Vendors can make products available across two or three of these delivery models, and customers, indeed, purchase financial information products using a combination of all of the main delivery mechanisms.

26. Relevant content/information can be delivered by suppliers to customers as datafeeds and Application Programming Interface (API), where customers obtain their content in a direct or "raw" format, from which they build their own internal applications or portals.

27. Apart from datafeeds, relevant content/information can be integrated into retail products, which contain a "front end" window that enables the user to access the content and functionalities contained in the product on the screen. Such front-end windows can take the form of either a web-delivered solution or a deployed/physical solution. Both of these are described as "desktop solutions" or "workstations" which are offered by all financial information providers and contain at least some third parties sources of content. The vast majority of suppliers supply their content using both methods as outlined above (directly to end-users as a datafeed and/or as a desktop product) (9).

IV.A.3. Distribution models in the industry

28. As regards relevant content set delivered as datafeed, both Thomson and Reuters supply content sets directly to end users as well as via redistributors (many of whom are also competitors at the desktop product level). Some of these redistributors offer broad desktop solutions while others offer more niche solutions or incorporate content sets supplied by Thomson and Reuters into proprietary portals or "free to access" websites.

29. Bloomberg has traditionally been the exception amongst financial information providers for not supplying its data to redistributors, and instead offers a single "closed" desktop product to end users containing a very broad range of content sets and data: "the Bloomberg Terminal". These terminals are not designated to interface with any other terminal.

30. Thomson and Reuters - as well as other suppliers in the financial information industry - operate two different types of distribution: (i) the "sell-through model and (ii) the royalty model.

The "sell-through" distribution model

31. It is the model that is most commonly used by Thomson and Reuters, particularly in the field of research and asset management. It is a kind of non-exclusive licence to use and market particular types of content sets. Generally, no money is passed between the "supplier" (that is to say, Thomson or Reuter, as the case may be) and the redistributor; the end user will contract with the supplier and will pay the supplier directly for a licence to use the particular content set through the desktop platform of the redistributor. Typically, the contract between the supplier and the redistributor will provide for the supplier to make available defined content sets to the redistributor, and for the redistributor to market the content sets to end users over the redistributor's own platform (that is to say, its front end desktop product). The redistributor will be entitled to integrate the supplier's content set into its own platform, but will need the approval of the supplier in order to release the content set to the end user.

32. It is important to note that in the "sell-through" model, the redistributor acts as an "agent" for the content suppliers (e.g. Thomson and/or Reuters) in selling the content set to its own desktop customer in that it sells the third party content together with its desktop product. In this connection, the redistributor views the content "sold through" its platform as a key element in its retail desktop offering and the redistributor remains the point of reference / main interlocutor of the customer for all queries and support which would arise for its desktop product including the content set stemming from the third party supplier.

The "royalty" distribution model

33. Under this model, the supplier of the data will receive a fee from the redistributor but will not receive any fee from the end users (albeit that the end user will need to sign up to the supplier's standard licensing terms). Typically, the contract with the redistributor will allow the redistributor to integrate the supplier's content sets into its platform in return for either a flat fee, or a variable fee (that is to say, a royalty based on the number of end users who sign up for the service). The choice of flat fee versus variable fee (that is to say, royalty) will depend on the redistributor's business model.

IV.B. RELEVANT PRODUCT MARKETS

IV.B.1. Product market definition as proposed by the parties

34. In a preceding case concerning the financial information business (Reuters/Telerate10), the Commission delineated the relevant market along the following lines: (i) the supply of real-time market data ("RTMD"); (ii) the supply of price and reference data to middle and back-office functions (historical data); (iii) the supply of market data platforms; and (iv) the supply of Foreign Exchange order management software. This delineation was based upon the observation that some customers need real-time information to sell and purchase securities, whereas other customers need historical data to conduct financial analyses. This delineation was also driven by the specific activity of Telerate in the market data platform area.

35. In the present case, however, the notifying parties proposed a different delineation of the relevant product markets. They maintain that this alternative market definition mirrors the internal organisation of the undertakings concerned and also their corporate product maps. They further suggest that it also reflects the demands of their customers and therefore better replicates the reality of the markets for financial information.

36. In particular, the parties delineated relevant markets and shares along the lines of the two groups: (a) Sales & Trading and (b) Research and Asset Management, broadly corresponding to customers engaged in, respectively, "on-trading floor" and in "offtrading floor" activities. These groups were further broken down as follows: Sales & Trading: (a-1) Fixed Income; (a-2) Institutional Equities; (a-3) Foreign Exchange; and (a-4) Commodities & Energy. Research and asset management: (b-1) Investment Management, (b-2) Wealth Management, (b-3) Investment Banking, and (b-4) Corporates:

Table 1

(A) Sales & Trading (alternatively named the "On-trading floor")

(a-1) Fixed Income

(a-2) Institutional Equities

(a-3) Foreign Exchange

(a-4) Commodities & Energy

(B) Research & Asset Management (alternatively named the "Off-trading floor")

(b-1) Investment Management

(b-2) Wealth Management

(b-3) Investment Banking

(b-4) Corporates

37. Sales & Trading: users include traders in various sectors. These users need real-time data and transaction services to monitor the markets and execute trades. The notifying parties have therefore submitted that the supply of information and the supply of transaction capacity (which are generally sold by the same companies such as Bloomberg or Reuters) should not be considered as separate markets. Traders traditionally specialise in one asset class and the notifying parties have submitted that the following categories are relevant in terms of product market definition as each corresponds to a specific category of customer which has a specific requirement: (a-1) Fixed Income; (a-2) Institutional Equities; (a-3) Foreign Exchange; and (a-4) Commodities & Energy. According to the notifying parties each of the above segments can be further divided according to their functions: pre-trade, trade and post-trade.

38. Research and Asset Management: users generally need historical data to evaluate investment opportunities and do not conduct transactions on a permanent basis as traders do. They can be further categorized into four main customer segments: (b-1) "Investment Management" includes portfolio managers and research analysts in asset management firms and insurance companies; (b-2) "Wealth Management" includes primarily financial advisers and private banks; (b-3) "Investment Banking" includes investment banks advising their customers on possible mergers and acquisitions, private equity and venture capital professionals; and (b-4) "Corporates" which includes corporate customers conducting investor relation or treasury activities.

39. Based on these definitions, at the EEA/worldwide level, the parties would overlap in six of the product segments: (a-1) Fixed Income; (a-2) Institutional Equities; (b-1) Investment Management; (b-2) Wealth Management; (b-3) Investment Banking; and (b- 4) Corporates, but their combined market share would not be particularly high

40. The information provided by the notifying parties however only partly supports this market definition. Although the notifying parties are organised within business units mirroring such market segmentation, the product map of Reuters clearly indicates a deviation from the proposed market definition. Reuters' best selling product is the Reuters 3000 Xtra, which accounts for [50-60]*% of Reuters sales in the EEA (in the above mentioned relevant markets). This product is sold across the user segments. Similarly, the second best selling Reuters product, Reuters Trader, accounts for [10- 20]*% of Reuters sales in the EEA and is also sold across all user segments. Thomson's product map could appear to be closer to the market segmentation proposed by the notifying parties, principally because Thomson ONE, the flagship product of Thomson, is sold to end-users with a product base (common for all customers) in association with specific functionalities which would appear to best fit/tailor each segment (e.g. the functionality Historical Mean Estimate is associated to the Thomson ONE Investment Management and principally sold to users in the Investment Management segment). However, Bloomberg, one of the major financial information providers, proposes a unique solution to all customers which encompasses all segments: "the Bloomberg Terminal". These terminals are standalone terminals which are designed not to interface with any other systems.

IV.B.2. Market investigation

41. Several respondents (among customers and competitors) raised serious objections against the proposed transaction; considering that the merged entity would have a very strong position in some relevant markets, the Commission does not reveal the exact identity of respondents in this proceeding. Respondents (both competitors and customers) have been allocated a random number, which is used consistently in the proceedings whenever such respondents are the source of the information provided to the Commission.

42. The market investigation has largely confirmed the distinction between Sales & Trading and Research & Asset Management fields. This distinction broadly corresponds to the overall segments "on-trading floor" and "off-trading floor". Real time data products are mostly used in sales & trading, whereas they are normally only used for specific fields of application in research & asset management (in particular for research purposes). Referential (past) data and related products are instead mostly used in research & asset management (11).

IV.B.2.1. Discrete content sets

43. The investigation has indicated that discrete content sets represent the appropriate antitrust markets for assessing the impact of the present merger. In particular, the investigation suggests that the value of the products in this industry lies in their functionalities and content rather than in the end user's corporates area. To this extent some discrete elements such as data, content , functionalities, transaction platforms (and executable prices/trades), datafeeds, real-time market data, non-real time data (historical/archival) (12), analytics, News, Broker Research, Fundamentals, Estimates, economic data, time series have been indicated as the most relevant fields for competitive interaction by respondents.

44. As far as competitors are concerned, the market investigation has shown that a significant majority of the respondent competitors (65%) considered that discrete content sets cannot be substituted for one another since they respond to different and well defined needs of final customers and therefore constitute accurate product market delineation under which to assess the impact of the notified Thomson/Reuters merger.

45. The view expressed by the remaining respondent competitors is not inconsistent with that of the majority. These competitors considered that discrete content sets do not belong to separate markets given that as a matter of fact they offer regularly discrete content sets to end-users in a bundled proposition (desktop products). This is notably the case for Bloomberg who has chosen not to sell its product in a standalone formula, but in seamlessly integrated desktop solution (Bloomberg Professional). Desktop solutions constitute another level in which competition takes place between the players in the financial information sector. In this respect, another submission by Competitor 4 sheds light on this issue. According to this respondent discrete content sets do not belong to distinct markets, but on the other hand they clearly determine in the first place the value of the final bundled proposition (13).

46. Moreover, for the same reasons a significant majority of the respondent customers (75%) believes that separate content sets cannot be substituted between each other. As in the case for competitors, the view of the customers who have expressed a different opinion is explained by the fact that they usually purchase content sets in a bundled proposition (desktops). In this respect, the submission of Customer 3 can be quoted as a relevant example. Customer 3 purchases financial information in a bundled solution (desktop), except for Earning Estimates (current/historical), Fundamentals, Broker Reports (real time/historical), and Time Series (economics); therefore, Customer 3 concluded that discrete content sets do not belong to separate product markets, except for Earning Estimates (current/historical), Fundamentals, Broker Reports (real time/historical), Time Series (economics) and Ownership (14).

47. In addition to customers and competitors, the internal documents provided by the parties have largely confirmed the Commission's view that discrete content sets cannot be considered substitutable for one another and that these content sets would represent the relevant areas for competition and therefore can be considered to be the relevant product markets.

48. In several Reuters' documents called [...]*, Reuters presented its competitive position in relation to competitors in line with discrete content sets particularly in the areas of Earning Estimates, Broker Reports and Fundamentals (15). Furthermore, a significant number of internal documents points to specific areas within the financial information sector. For instance, several recent Reuters' documents called [...]* showed the existence of a competitive relation between the notifying parties in the earning Estimates area (16). Another document called [...]* campaigns in favour of Reuters' Broker Research database in order to win customers from Thomson (17). In the area of Fundamentals, an internal document emphasized the competition which takes place between Reuters' Fundamentals offer and that of the competitors such as Thomson (IBES/Worldscope), JCF (Factset) and S&P (Compustat) (18). Lastly, an internal document called [...]* contains a comparison between Reuters' and competitors' offer in the Time Series-Economics segment (19).

49. As far as Thomson is concerned, a significant number of internal documents confirm the Commission's view that the relevant product market definition should be along discrete content sets. In this context, one might mention several Thomson's documents called [...]* which describes the company's offering in the areas of Estimates, Fundamentals, Times Series- Economics and Broker Research (20) and the competitive landscape for each of these products.

50. It has to be borne in mind that the various content sets may be sold/delivered by suppliers (e.g. Thomson, Reuters, and other vendors) in various modalities: (i) discrete products (databases or direct datafeeds) sold to customers (ii) content set bundled with analytics in desktop products offered by vertically integrated providers (such as Thomson and Reuters) (iii) content set bundled with analytics in desktop products offered by non-vertically integrated providers (such as FactSet etc.) who act then as "redistributors" as regards the standalone content set.

IV.B.2.2. User segments

51. As regards the segmentation proposed by the parties based on different user-groups (especially relevant in the research & asset management space), the investigation showed that, as regards standalone content areas, this segmentation did not take into account the cross-cutting needs of customers active in different segments and requesting similar products/content sets (21).

52. In particular, it is true that the different user groups broadly correspond to the way the parties segment their sales departments as well as how (at least) the large banks segment their operational business functions, however, the investigation showed that (i) a product designated for a certain type of user may be used by a user active in a different area (with some modification to take account of the job task and responsibility, especially in the off-trading floor space where a lot of commonalities exist between the various segments), and (ii) that a user group segment might use several content sets which are not substitutable for each other. To give an example: databases for company Fundamentals or the collection of earning Estimates by analysts are, on the one hand, used across different segments, in particular (but not only) in the fields of Investment Management and Investment Banking. On the other hand, these content sets are, from a user point of view, not substitutable, but an end-customer (e.g. an internal division of a bank) may need both content sets. The fact that both content sets are used within the same user group, or even by the same customer (a bank), therefore does not allow to conclude that they belong to the same product market.

53. The distinction between user groups may be more relevant at the level of desktop products (workstations) where the various types of content sets and software functionalities are bundled together by suppliers in order to customise their product offerings and respond to the varying requirements (in terms of mix and match of the information, data and functionality needs) of customer groups who are focusing their activities in specific segments of the financial industry (e.g. Investment Banking, Investment Management, Wealth Management, Corporates). However, at least some of the products are sold across the different segments, as already set out above (see paragraph 40).

IV.B.2.3. Upstream/downstream interplay between suppliers of discrete content sets and vendors of desktop solutions (workstations)

54. The market investigation revealed that in the research & asset management (off-trading floor) space, which combines the areas of Investment Management, Investment Banking, Corporates and Wealth Management, two different types of products can be distinguished: firstly, certain discrete content sets, such as databases comprising fundamental data of companies, earnings Estimates of different analysts, reports of analysts of broker houses on companies; secondly, "desktop products" (or "accesses") which bundle different types of data and content, such as the content sets mentioned before, with analytical tools. At the level of "desktop solutions", a significant customisation is operated in the context of the bundled products offerings in order to reflect the different needs of customer groups, hence the competitive assessment will have to take this into account.

55. A decisive difference exists in the way these two types of products may be sold. Whereas the desktop products (including content sets) are sold to the downstream final customers, such as the Investment Banking or Investment Management division of a bank, the relevant content sets can be sold on a stand-alone basis directly to final customers, but also via third parties who propose their own desktop products to customers. The market investigation has therefore revealed that specific business interactions take place between the compilers/aggregators of standalone data and the vendors of desktop products who use these data in their final products

56. As a preliminary point, it is worth mentioning that Thomson and Reuters are active in both types of products. Besides desktop products, Thomson and Reuters therefore also sell standalone databases (sometimes referred to as "datafeeds"). Different content sets can also be combined in a single database. For instance, Thomson (through Worldscope) and Reuters (through Multex) offer products that combine inter alia earning Estimates, broker reports and Fundamentals.

57. In addition, Thomson and Reuters are aggregators/compilers of "raw" data which are necessary for the production of discrete content sets.

58. The market investigation indicated that with respect to some discrete products and content sets such as earning Estimates, broker reports and Fundamentals, Reuters and Thomson supply companies which are their competitors at the desktop products level (downstream). More precisely, when selling such databases to their final customers, these competitors mostly act as agents and redistributors for Thomson and Reuters ("sell-through relationship), but sometimes also directly as wholesalers ("royalty relationship"). In this particular case, the customers can use the competitors' applications together with Thomson or Reuters' databases/content. This scenario may occur when these competitors do not have their own content database in this field, but offer predominantly an analytical tool.

59. It is worth reiterating that in the "sell-through" model, while the redistributor acts as an "agent" for the content suppliers (e.g. Thomson and/or Reuters) as regards the content set that its own desktop customer procures from the third party content supplier, the redistributor views the content supplied to its customer as a key element in its retail desktop offering (in order to attract customers). In addition, the redistributor remains the point of reference / main interlocutor of the customer for all queries and support which would arise for its desktop product including the content set stemming from the third party supplier.

60. Competition therefore takes place at two different levels: (i) upstream between aggregators/compilers of the discrete content sets and (ii) downstream between vendors of desktop products vis-à-vis the final customers. The market investigation revealed that the merger gives rise to significant horizontal overlaps between the parties (especially) at compiler level and to significant vertical effects in respect of the desktop products level. In particular, the merger will increase the dependency of downstream competitors on the merged entity as regards certain critical inputs (including historical referential market data) necessary for them to be competitive downstream (at retail level).

IV.B.3. Identification of relevant product markets

61. The market investigation revealed that individual content sets are not substitutable for one another since they respond to different and well defined needs of customers and are often traded separately; therefore these content sets can be considered on a standalone basis. At the same time, the investigation also revealed that, at the downstream level, different bundles of content sets, data and functionalities/analytics are traded in a packaged format in the marketplace in the normal course of business as desktop products or workstations. In addition, the investigation provided indications that transaction platform/services and market data platforms offered by suppliers in the market are also to be considered on a standalone basis.

62. In light of the above, in the following paragraphs, various content sets, desktop products including content and functionalities, transaction platforms, market data platforms etc. have been identified as separate relevant product markets.

63. Some of the content sets/products at stake may ultimately be more relevant for the sales & trading field than for the research & asset management area and conversely, but it cannot be excluded that - for some of those - their usage could stretch between ontrading and off-trading floor.

Sales & Trading

(i) Real-time market data sold through desktop products/workstations

64. Thomson and Reuters (as well as other vendors such as Bloomberg, IDC/Comstock, Telekurs/Fininfo) supply desktop products, which include the necessary real-time market data and financial information as well as analytics for traders in all or specific assets classes (Institutional Equities, Fixed Income and other relevant asset classes (22)). These desktop products which include real time market data as a component feature represent for instance the vast majority of the revenues obtained by the main players (Reuters and Bloomberg) in the sales & trading space. These desktop product which include real-time data are traded separately in the market and hence are considered as a separate product market.

(ii) Real time datafeeds

65. Thomson and Reuters (as well as other vendors such as IDC/Comstock, Telekurs/Fininfo, individual Exchanges) supply real-time market data on a standalone basis by way of data feeds (that is to say, they supply "raw" real time market data content). The parties and the market investigation have confirmed that stand-alone datafeeds are typically not comparable with desktop products which include real-time market data.

66. There are two types of standalone real time datafeeds: (i) "consolidated feeds", which involve the aggregation of feeds from various sources (exchanges) into a single source of data (23); and (ii) "direct feeds", which involve a more direct connection from an individual exchange to a customer. Direct feeds permit reduced delay in the transmission of pricing data (that is to say, lower latency).

67. There are a number of providers of (consolidated) real-time datafeeds in the marketplace (Reuters, Thomson, Bloomberg, IDC/Comstock, and Telekurs/Fininfo). These vendors typically offer different ranges of choices in terms of breadth and depth of exchange coverage. The number of customers purchasing a full service with the maximum amount of global coverage available is anyhow limited. Exchanges also increasingly commercialise their own real-time datafeeds directly to end customers. Therefore, realtime datafeeds represent a standalone content set which is traded separately in the marketplace and, therefore, are considered as a separate product market.

(iii) Market data platforms

68. Market data platforms are "middleware" that receive datafeeds from multiple sources (including customer sources) as inputs and standardise these datafeeds into a format so that they can be accessed through a single API (application protocol interface). A customer can then write its own applications using this standard API, enabling those programmes to access the source datafeeds. Likewise, third party application vendors can write their applications to this API, so that a customer can "plug" the inputs from the MDP into third party applications in the same way that it plugs them into its own applications.

69. Reuters (but not Thomson) supplies a market data platform (middleware platform), called Reuters Market Data System or RMDS, used to integrate all the different data sources supplied by different suppliers to one customer (e.g. a bank) and then to redistribute this data again in the customer's own organisation. In the Reuters/Telerate decision (24), a separate product market for market data platforms was defined. The investigation in this case confirmed that Market data platforms along with the emerging technologies which facilitate the integration and distribution of datafeeds stemming from multiple sources (that is to say, feedhandlers and similar) form part of a separate product market.

(iv) Instrument codes (e.g. Reuters' instrument codes, RICs)

70. Instrument codes or securities identifiers are widely used devices in the on-trading floor space. Reuters has its own proprietary instrument code, called RIC (Reuters' instrument code). RICs are alphanumeric codes that act as an identifier for individual financial instruments or products and are used as a navigation tool within Reuters' products. The RIC structure is global and covers all financial instruments held in Reuters' systems and databases (as opposed to all financial instruments). RICs are one of various symbologies used in the financial information markets. Several providers that supply information gathered from multiple sources maintain their own systems of identification, navigation and retrieval of financial instruments within their own services (Bloomberg, S&P, IDC, Telekurs, for example, each have their own coding systems). In addition, there are industry identification codes, supplied on a standalone basis, which are not supplied by commercial vendors, but rather by National Numbering Agencies (NNA), which are responsible for issuing an ISIN (that is to say, International Securities Identification Number set by the International Standards Organisation) that uniquely identifies exchange-traded instruments in that country. Examples of local exchanges identifiers issued by NNAs for those securities traded on the exchanges include: UK - SEDOLs, US - CUSIPs, Switzerland - Valoren and France - SICOVAMs.

71. While it is not evident that instrument codes, such as RICs, represent a separate product market, the Reuters' instrument codes (RICs) are analysed, in the competitive assessment part of the decision, in the context of a claim from some third parties about possible conglomerate effects which the merger would give rise to through these RICs.

(v) Transaction platforms/services for Fixed Income securities (e.g. Tradeweb)

72. Fixed Income instruments are less standardised compared to Institutional Equities and (unlike equities) are not traded over exchanges, therefore users are looking for pools of liquidity for their trading activities. Thomson offers transaction services through Tradeweb (and Tradeweb Retail in the US). Tradeweb is an Internet-based multi-dealer to customer trading system for Fixed Income securities and derivatives, which links dealers with buyside institutions. It distributes indicative market data (e.g. indicative price information, not data in relation to trades which have taken place) to a number of redistributors, including Reuters, as well as directly to dealers and institutional investors (e.g., asset managers, hedge funds, Central Banks and pension funds) (25). Reuters offers Reuters Trading for Fixed Income (RTFI). There are various providers of such transaction platforms/services in the marketplace: The main competitors to Thomson in this sub-segment are Bloomberg and MarketAxess. Interdealer brokers such as MTS/Bondvision, Eurex Bonds, ICAP/BrokerTec and eSpeed are also active in this space, along with exchanges. On this basis, transaction platforms/services for Fixed Income securities are a specific service responding to a unique and well identified need, hence they are considered as a separate product market.

(vi) News

73. Both Thomson and Reuters are present in the News area. Financial and business-related news constitute a financial content that is considered a key separate asset in the context of especially the on-trading floor, but also in respect of the off-trading floor space. News is considered a standalone product by customers and suppliers. This is the case also when the news service is delivered to customers in bundle with other content or functionality, such as data feeds, analytics, etc. Such financial news may be comprehensive or more limited as it may relate only to a single country, certain asset classes or particular economic sectors. News services can be provided in English or other languages.

Research & Asset Management

(vii) Broker Research reports

74. Both Thomson and Reuters supply Broker Research reports. A typical Broker Research report would contain an analysis of a security, an industry or a market and provide information that may be helpful to investors in making investment decisions. More specifically, a company research report may provide an overview of a company or stock, an analysis of its financial statements, a discussion of any significant events that may impact the performance of the company, an analysis of recent stock performance, peer analysis (i.e., comparison with like companies) or across an entire sector. In addition, most research reports also contain some quantitative analysis (e.g., valuation using discounted cash-flow analysis) and the author's estimate of a company's future performance on key metrics, such as quarterly earnings per share. Research reports may also include trading and direct access to analysts via phone calls and meetings.

75. Broker Research reports are authored by sell-side research firms (brokerage firms), as well as a small number of independent research firms; they represent a component of the service that a brokerage firm provides to its buy-side customers. Buy-side customers do not pay for research directly - a sell-side broker produces such research in the hope that it will earn additional revenues in the form of commissions by attracting instructions to buy or sell securities as a result of the report.

76. The market investigation has shown that Broker reports can be distinguished according to real-time research and Aftermarket Research. Aftermarket Research consists of Broker Research reports that are withheld from the public until these customers have had first access to the valuable insight. In the first period, broker firms make available this research only to their clients. After the initial embargoed period (around 14 days), the information in the research report becomes less useful with respect to trading decisions and many brokers are willing to sell the research they produce to anyone wishing to pay for it (which would typically include investment banks, companies and consulting firms). Aftermarket Research is often sold on a pay-per-view basis. Aftermarket Research reports are used by investors in order to help them to make better business decisions. In effect, investors can find in them forecast data, strategic insight, industry trends and market conditions, analysts' insight and product information.

77. The market investigation has also shown that customer bases for Broker Research are different: while real time research are addressed to buy side customers, Aftermarket Research are needed in the sell side investment/banking areas.

(viii) Earning Estimates

78. Both Thomson and Reuters have products containing data on earning Estimates. An earning estimate is a prediction made by an analyst that a company will achieve a certain amount of profit/loss in the future, typically with respect to key metrics such as annual or quarterly earnings per share.

79. Estimates expectations are the most important determinant of equity prices since investors buy stocks at prices that the market determines based on corporate earning potential. The financial media and investors compare the mean estimate of a company's earnings against the actual reported earnings; the difference between the two constitutes the earning surprise. Therefore, companies that fail to meet mean expectations are usually sanctioned by drops in their share price whereas companies that beat expectations see their stock rise.

80. Estimates are typically generated by sell-side research analysts as a component of an analyst research report or valuation model and they are usually included within Broker Research reports.

81. A distinction is sometimes drawn between current Estimates and historical Estimates. Current Estimates are earning Estimates that have been released in the past 12-18 months. 'Historical' Estimates are earning Estimates that date back to more than two years in history and can cover a period of up to 20 or even 30 years of history. Usages of current and historical Estimates are very often mixed. Indeed, analyses of historical Estimates are commonly used to better assess the impact, accuracy or trend of current Estimates. For this reason, financial institutions purchase bundled products that combine both.

(ix) Fundamentals

82. Both Thomson and Reuters supply databases for Fundamentals. Fundamentals data consists of various company-specific data, such as financial statement data (both interim and annual), calculated financial ratios (e.g. annual and five-year averages for growth rates, profitability, leverage, liquidity, asset utilisation), per share data (e.g. earnings/book value/cash flow per share), security and market identifiers, product and geographic segment information, and textual company profile data (including officers, address, web address, major shareholders, and business descriptions). It should be mentioned that Fundamentals are publicly available on an individual basis (e. g. from companies' websites), but some providers aggregate them in databases. For instance Reuters and Thomson's databases cover [50,000-60,000]* (for Thomson) and [50,000-60,000]* (for Reuters) public and private, active and inactive companies.

83. Fundamentals are used by investors in order to evaluate the companies' performance over the past years, to draw an opinion whether this performance is likely to be maintained or improved and, by comparison with similar data from other companies, to reach investment decisions.

(x) Time Series of economic data (TS/E)

84. Both Thomson and Reuters are present in the Time Series content space. Time Series content represents a range of end of day (EOD) securities prices, volume and other data across equity, fixed-income, index, derivative and trust instruments as well as interest/exchange rate and Economic Time Series data.

85. More particularly, Economic data (Time Series-Economics, hereinafter also referred to as "TS/E") consist of data on macroeconomic variables, such as GDP, unemployment, money supply, balance of trade figures and inflation rates. Such information is represented as individual data points and tracked over a period of years (time series). The ultimate sources of such data are typically governmental bodies (national and supranational statistical offices or central banks), but such series also include proprietary data of, e.g., research institutes.

86. Economic data are used to quantify the general business conditions to which financial markets respond and in which governments, industries and individuals are operating; economic expectations are a prime determinant of the course of equity and bond markets; economic data have a central impact on financial futures interest and exchange rate contracts.

87. Economic data are principally used by sell side equity researchers and economists. This includes the Investment Management community, brokerage firms, and corporate finance departments from the Investment Banking community and to lesser extent business consultants, academics and media professionals.

(xi) Ownership

88. Both Thomson and Reuters are present in the Ownership content area. Ownership content data provide a researched view of global security ownership available in the financial marketplace. Ownership content data cover principally global securities ownership, current equity and Fixed Income ownership information.

(xii) Deals

89. Both Thomson and Reuters are present in the Deals content area. Deals content data provide a comprehensive insight into global capital market activity in different areas to bring institutional researchers the decision support tools they need when investigating companies. Deals data cover areas such as equities, bonds, medium-term note programs (MTN), syndicated loans, public finance, project finance, mergers and acquisitions, corporate finance, corporate restructuring etc.

(xiii) Other content sets: profiles, public filings, other times series (equity, fixed income, futures, options, warrants, etc.)

90. Other content sets data in which the parties are present comprise data such as profiles, public filings and time series other than economic data (TS/E).

91. Professional profiles content is composed generally of a global compilation of in-depth details on buy-side and sell-side institutions and their personnel including biographical, contact and related insight. This particular content set is tightly integrated with other content for in-depth views of security ownership, fund management and analyst coverage.

92. Public filings contain a collection of companies' filings documents comprising annual reports, insider trading filings and other data.

93. Time series other than economic data (TS/E) include time series on Equities, times series on Fixed Income securities, time series on futures, options, warrants and other data.

(xiv) Desktop products (workstations) in research & asset management: downstream level

94. Thomson and Reuters supply "umbrella" workstations/desktop products, namely the family of "Thomson One" and of "Reuters 3000Xtra" products, which are the flagship offerings of both parties. Another type of desktop solution available in the marketplace is the Bloomberg's terminal, which, unlike other workstations, may have the peculiarity of involving a dedicated physical infrastructure installed at the customer premises. These desktop products include real-time data and financial information as well as, possibly, analytics. In the research & asset management space (comprising the segments: Investment Banking, Investment Management, Wealth Management, Corporates), these desktop products essentially include a set of referential (that is to say, past) data. The exact mix and match between real time financial information and referential data and analytics depends on the ultimate tailoring that suppliers may want to provide to their desktop products in order to reflect the varying granular needs (in terms of functionalities and content) of downstream customer groups (user segments). This is particularly important in the research & asset management (off-trading) spaces.

(a) Desktop products for Wealth Management

95. Both Thomson and Reuters supply desktop products for Wealth Management. The investigation indicated that while there is an overlap between desktop products which are purchased and used by investment managers and investment bankers (26), the desktop products targeted and used by wealth managers show specific characteristics, which respond to the specific business logic of Wealth Management professionals. Wealth managers help individual clients (private persons) to achieve their financial goals by devising a suitable asset allocation and investment strategy and managing their savings/investment accordingly. The wealth manager uses analytical tools to monitor performance against the agreed goal on an ongoing basis. Their geographic focus is significantly geared towards narrower (national) boundaries given inter alia different investment patterns of consumers, the influence of national tax and other regulatory provisions concerning private investments (pension schemes, capital gain taxation, nation specific investment vehicles, significant product differentiation).

96. For example, in Wealth Management Reuters offers products with a clear geographic focus: "Reuters Trader UK" which is targeted at financial professionals who are focused primarily on managing individual/private client investors with an interest in UK equities. It offers similar solutions targeted at other member states or regions, for example central and Eastern Europe, Germany and the Nordic countries. For wealth managers, purchasing decisions are thus sometimes taken at a sub-regional level, and therefore competition conditions are not homogeneous across different countries, but mainly reflect national or regional logics.

97. The characteristics of the desktop products related to Wealth Management and the narrower geographic focus of demand in this user segment are sufficient indications that desktop products for Wealth Management must be regarded as a separate product market with national/regional geographic scope. In this respect, the data provided by the parties reveal that the parties have horizontal overlaps giving rise to horizontally affected markets especially in Belgium, Finland, and the UK with combined market of shares above 15%.

(b) Desktop products for other areas in research & asset management

98. As regards the other research & asset management segments, bearing in mind the common use of certain desktop products between (i) Investment Management and (ii) Investment Banking, it is not appropriate to split these segments from one another from a relevant market standpoint. In any event, the horizontal overlaps between the parties as regards desktop products in each of those user segments would give rise to (unproblematic) affected markets only as regards Investment Banking (at the worldwide and EEA level).

(c) Vertical relationships in research & asset management

99. More generally, given the interplay between vertically integrated and less vertically integrated suppliers (which is of particular importance in research & asset management) and the existence of an undeniable degree of customisation of desktop products at the downstream level across the segments, the competitive assessment will have to take into account the potentially different impact of the merger on competition in those (offtrading floor) user segments (Investment Management, Investment Banking, Wealth Management, Corporates) when analysing the vertical effects of the merger.

IV.B.4. Customer requirements (global coverage of content sets)

100. The market investigation also gave some indications that, besides the discrete products mentioned above, the relevant markets could be influenced by the different needs of customers of varying types/sizes. Larger global customers (or "Tier 1" firms) were described by some respondents as having buying patterns different from the smaller "Tier 2" and "Tier 3" customers. In particular one respondent featured these different patterns of demand in the following way: larger customers would tend to maintain two primary (full scope) data suppliers (using one as a backup) as well as a number of providers to fill the gaps and ensure the most accurate coverage (27), whereas Tier 2 users who have similar requirements to Tier 1 firms would tend to use one primary supplier for reasons of cost but will use additional suppliers to fill out information gaps as necessary. The smaller Tier 3 customers would typically require a more limited set of local market data and therefore may use a single national or regional data supplier.

101. While it is neither possible nor necessary to come to a definitive conclusion on whether and how to identify (if at all) a meaningful cut-off point between the various categories of these possible customer segments, on a more qualitative viewpoint it is true that the varying requirements of customers play a role in how the market dynamics work especially when global players require global scope and coverage of products. In this respect, the market investigation has shown that global customers need content sets with global coverage. It should be borne in mind that global content sets cannot be substituted to the sum of local/regional content sets without being aggregated and standardised by one particular aggregator/compiler. Therefore, it can be concluded that there is a standalone demand for global content sets as opposed to local/regional content sets.

IV.B.5. Suppliers of financial information who do not cater for the needs of professional customers

102. The market investigation has almost unanimously indicated that financial information products delivered on the Internet via companies such as Yahoo or Google cannot be considered suitable for business professionals in the financial information industry and are not considered credible alternatives to the content/products/functionalities offered by financial data compilers such as Reuters or Thomson. The market investigation has in particular indicated that these providers cannot supply real time data or transaction facilities and that the historical data they can provide lacks the depth and width of equivalent data provided by established financial information providers. Several competitors and customers for instance suggested that Yahoo and Google financial products are well adapted for individual investors requiring basic information but are poorly adapted to professionals requiring highly sensitive and accurate information (in addition the terms of use they post often specify that data may be used for noncommercial purpose only). The same applies for products offered by IT companies, which are rather viewed at most as complements or add-on features to the functionalities at stake at the desktop product level.

103. All of the respondent customers have submitted that the free-of-charge online financial information providers (e.g. Google, Yahoo) cannot be compared to Thomson and Reuters and those online free-of-charge providers are not reliable competitors. Although these undertakings may provide additional financial information, they still lack quality, precision and reliability to be viewed as fully fledged suppliers.

104. Almost 70% of the respondent competitors explained that certain free-of-charge financial information products are not substitutable to those of Thomson and Reuters. It was put forward that such free of charge products lack the sophistication and functionality of commercially available products required by professional clients and therefore are mainly used by private and not professional users.

IV.C. RELEVANT GEOGRAPHIC MARKETS

105. In Reuters/Telerate, the Commission concluded that the market investigation provided indications that the markets at stake in that case were at least EEA-wide in scope. However the geographical definition of the market was left open.

106. As regards the separate content sets markets in this case, the market investigation indicated that these markets are to be regarded as global in scope, but it provided some indications that differences may exist between the very large investment (tier-1) banks, which normally source on a global scale, and the smaller tier-2 or tier-3 financial institutions, which would source on an EEA-wide or even national level. In any case, the geographic scope of the global data would appear to be at least EEA-wide and is probably worldwide in scope.

107. As regards the desktop products (workstations), the notifying parties submitted that the relevant geographic scope for each of the markets based on user segments is global or at least EEA-wide. This assumption is notably based on the observations that products are not generally customised for sale in a specific country (28), price lists are generally similar in the EEA, and competitors and customers are generally present all over Europe. According to the notifying parties, it is only the Wealth Management segment, which, to a certain extent presents characteristics of national markets. This is because wealth managers are generally focussed on one country only and have a natural interest in funds or equities in their own country (due to the impact, in particular, of national tax law and investment preferences).

108. The market investigation revealed that most customer segments are regarded as global - or at least EEA-wide - to the exception of Wealth Management (within the research & asset management space) for which the investigation has shown that demand and supply are much influenced by national tax and pension legislation, which may restrict the geographical scope of such segment, unlike the other ones, to national boundaries.

IV.D. CONCLUSION ON RELEVANT MARKETS

109. In the light of the above considerations, the Commission is of the view that the product market delineation should be based on several discrete product sets since the different discrete content sets, the desktop products and the various infrastructure and/or software related services available in the marketplace cannot be considered as substitutable for one another and respond each to different and well defined needs and use.

110. Therefore, for the purpose of this case, the Commission has carried out its investigation in the following areas: (i) real-time market data sold through desktop products/workstations; (ii) real-time datafeeds; (iii) market data platforms; in the context of real-time datafeeds and market data platforms, the role of (iv) Reuters Instrument Codes (RICs) among securities identifiers is analysed in connection with a claim that the merger would give rise to concerns due to these Reuters' security identifiers; (v) transaction platforms for Fixed Income securities (vi) News (vii) research - broker reports, with a distinction between (a) Real-time Research and (b) Aftermarket (embargoed) Research; (viii) earning Estimates; (ix) Fundamentals; (x) Times Series/economic data (T/S); (xi) Ownerships; (xii) Deals; (xiii) other content sets (profiles, public filings, other time series); (xiv) desktop products in research & asset management, in particular Wealth Management desktop products.

111. The geographic dimension of the relevant global content sets markets would appear to be at least EEA-wide and is probably worldwide in scope. At the level of desktop products in the research & asset management space (off-trading floor), the market investigation showed that in respect of Wealth Management demand and supply of products are much influenced by national tax and pension legislation which may restrict the geographic scope of this segment, unlike the other ones (Investment Banking, Investment Management, Corporates), to national and/or regional boundaries. As regards other segments it can be left open whether the geographic dimensions is EEA or world-wide, as the assessment would not change since the merger would not raise competition problems.

112. In addition, there are competitive interactions involving different (vertically integrated and less vertically integrated) desktop product vendors, addressing the needs of the specific downstream segments especially in the research & asset management space (off-trading floor), such as investment bankers, investments managers, wealth managers and Corporates. In these segment-markets, Thomson and Reuters compete with competitors which (or the customers of which) may procure some of Thomson's and Reuters' content sets as input for their desktop products. The interaction between content sets data aggregators/compilers should be borne in mind while assessing the related vertical effects of the merger in the downstream desktop solution markets.

V. COMPETITIVE ASSESSMENT

113. According to the market definition along the lines of individual user groups, as proposed by the parties, based on the shares provided by the parties, the parties' combined shares would be as follows:

<table>

114. On that basis the parties argue that the transaction is unlikely to give rise to any competition problem.

115. However, the results of the market investigation suggest adopting a product-per-product approach in order to assess the effects of the merger, and to also factor in as appropriate the interplay between the upstream data supply level (done by compilers) and the downstream desktop level (where both compilers and redistributors compete for the same customers). On this basis, the effects of the merger are different between the sales & trading (on-trading floor) space and the research & asset management (off-trading floor) space.

116. In this respect, at the outset it can be stated that the merger does not give rise to concerns in the sales & trading space. Conversely, in the research and asset management space, for some specific databases/content sets, the merger will lead to significant overlaps, will remove competition between Thomson and Reuters and, given the position of the parties, would lead to competition concerns of both horizontal and vertical nature.

117. This conclusion has been confirmed by an internal document from Thomson supporting the theory that thanks to the merger, the newly formed company could generally increase its prices. [A consulting firm]* notably concluded that the merger could increase the pricing power of Thomson/Reuters, allowing the newly formed company to grow [less than 2%]* faster than the market, this additional [less than 2%]* growth only derived from price increases (30).

118. In the following chapters the assessment is done per each relevant market.

V.A. SALES & TRADING

V.A.1. Real-time market data sold through desktop products/workstations

119. Some respondents to the Commission's market investigation have put forward the argument that the merger would reduce the number of full-scope providers of real-time market data from three to two, thereby giving rise to adverse effect on competition. The inquiry showed that Reuters and Thomson derive the vast majority (respectively [70- 80]*% for Reuters and [70-80]*% for Thomson in 2006) of their global revenues in the sector of sales of real-time market data through desktop products, the remaining being sales of real-time market data through datafeeds.

120. In respect of real-time market data sold though desktop products, the worldwide market structure is already pre-merger a near-duopoly between Reuters ([30-40]*% in 2006) and Bloomberg ([40-50]*% in 2006), as it was already concluded by the Commission in the context of the Reuters/Telerate case. Fringe competitors with very limited shares ([less than 2%]* each at the worldwide level in 2006) are IDC/Comstock and Telekurs/Fininfo (31). The presence of Thomson in respect of real-time market data through desktop products is very limited, given the historical focus of Thomson on the research and asset management space. Thomson's share is less than 1% ( [less than 1%]* worldwide, [less than 1%]* in the EEA in 2006) (32). Therefore, the pre-merger market structure is not impacted by the merger between Reuters and Thomson to any significant extent.

V.A.2. Real time datafeeds

121. Third parties have claimed that the transaction could bring about adverse effects as regards real-time datafeeds. In particular, it was considered that the merger would remove Thomson as the No 3 supplier of comprehensive, global, real-time datafeeds, behind Bloomberg and Reuters, whereas competitors were often only active in the supply of limited datafeeds.

122. The investigation has provided clear-cut indications which dispel these concerns. Reuters is the predominant supplier of "consolidated" real time datafeeds with a market share of [60-70]*% in 2006 (worldwide). Thomson does not appear to be the No 3 provider for real-time consolidated datafeeds, on the contrary it seems to have a market share of around [0-10]*% in 2006 (worldwide). Other players such as Telekurs and IDC/Comstock appear to be better placed than Thomson to exercise competitive constraints on Reuters with much higher market shares (around [10-20]*% each for consolidated datafeeds in 2006 at the worldwide level). Bloomberg is a small player in the field of consolidated datafeeds with a share of [0-10]*% in 2006 (worldwide) (33).

123. In addition, the deployment of effective feed-handler solutions by technology vendors in the marketplace contribute to create competition constraints stemming from "direct feeds" from Exchanges vis-à-vis "consolidated datafeeds" offered by traditional vendors, such as Reuters. Feed-handlers are technology that facilitates the delivery of real time market data from raw data generators (e.g. exchanges) to end users. Such feed-handlers allow customers to obtain multiple low-latency direct feeds in timelier manner compared to consolidated feeds (that is to say, at lower latency) (34). Therefore the merger does not change the competitive situation in respect of the real-time datafeeds to any significant extent.

V.A.3. Market data platforms

124. In this market Reuters is the largely dominant player with around [90-100]*% market share (of its RMDS middleware). In respect of market data platforms (MDPs), third parties had claimed that, although Thomson does not supply a market data platform, the merger would remove the incentives of Thomson to enter such market. This could have seemed plausible since Thomson had developed a partnership with a provider of a competitive product for MDPs, namely a feed-handler software provider (Wombat).

125. It is true that Reuters has already pre-merger a largely dominant position as regards MDPs, with emerging competitors being far behind Reuters' incumbency (Wombat, Activ Financial). The investigation has nevertheless indicated that, the role of Thomson in the context of the partnership with one of Reuter's main competitors in the MDP space (Wombat) was to provide Thomson's consolidated real-time datafeed to customers through Wombat feed-handler (thereby strengthening Wombat's offering of its feed-handler software). In this respect, it appears that the possible disruption of Thomson's partnership with Wombat will not adversely affect the market dynamics as regards the MDP and feedhandler space as compared to the pre-merger situation. Wombat can find a suitable replacement for the ex-Thomson's real-time consolidated datafeed among other real-time datafeeds providers, should the partnership end post-merger. Therefore, the merger does not appear to significantly change the competitive situation in respect of MDP to any significant extent.

126. Still in the context of real-time datafeeds and MDP, a third party competitor has claimed that the merger, due to the leading and/or dominant position of Reuters as regards datafeeds and MDP, would give rise to adverse effects (namely foreclosure) as regards access to the so-called "contribution data". "Contribution data" are financial data generated or computed by banks and brokers which own the corresponding intellectual property rights. These "contribution data" are mostly used for transactions regarding bonds, derivatives, structured products, forex and interest rate information, since these securtities are not traded in a stock exchange. As such, Reuters and Thomson have no role as regards the production of "contribution data", but post-merger could, allegedly, foreclose third party competitor's acces to these "contribution data" since these data are normally stored on Reuters' MDP (RMDS) used by the originator to stock such data (that is to say, at the bank or at the broker' premise). This third party competitor states that the merger would strengthen the incentive of the parties to impede access to "contribution data", worsening the present situation, whereby Reuters would already undertake an exclusionary behaviour vis-à-vis competitor in this space. This concern cannot be considered as related to the concentration between Thomson and Reuters. Given the absence of overlap between the parties as regards MDP, the merger does not change the competitive situation to any significant extent in respect of the accessibility of "contribution data" by third party competitors.

V.A.4. RICs

127. RICs are securities identifiers that are widely used in the sales & trading space (on-trading floor). As mentioned in the section on relevant markets, instrument codes are not to be seen as relevant markets for the purpose of this decision, nevertheless RICs have been investigated in the context of a claim that the merger would give rise to concerns due to these particular instruments codes. A competitor has claimed that Reuters' RIC is considered to some extent a de facto industry standard. Allegedly, RICs became an industry standard because they are an integral part of the way traders identify financial instruments and indices, due to the dominant position of Reuters as regards real-time datafeeds.

128. This competitor has put forward that the merger could bring about some adverse conglomerate effects, because Reuters (who claims intellectual property rights on its proprietary RICs) does not license these RICs to competing datafeeds vendors. The combination of Reuters and Thomson would create an entity that has a dominant position in datafeeds in the front office as well as in the middle and back office (35). Therefore, the combination of Reuters and Thomson could give the merged entity an incentive to embed/ extend the use of RICs throughout Thomson's products also in the research and asset management (off-trading floor) space, thereby foreclosing competitors from access to Thomson's customer base. In particular, vendors (also active in the off-trading floor space) lacking access to RICs would become less attractive to customers, because of the customers' extra burden when processing non-RIC identified securities which would artificially inflate the costs of doing business with any vendor other than Thomson- Reuters.

V.A.4.1. Effects in Sales & Trading

129. As regards the use of RICs in the sales & trading space, the investigation has revealed that these RICs may represent a possible bottleneck, especially due to the very strong position of Reuters in respect of the underlying real-time datafeeds. Given that Reuters claims copyrights over RICs, this may entail a certain lock-in effect for customers and a barrier to entry for competitors willing to displace Reuters' datafeeds, also in view of the growing importance of seamless integration throughout the on-trading floor space (straight through processing, STP, that is to say, the progressive integration between front-middle and back office in the sales & trading space).

130. Nevertheless, this concern cannot be considered as related to the concentration between Thomson and Reuters. The claim that RICs are an industry standard does not appear to relate to the merger at stake either (36). The presence of Thomson in the real-time datafeeds space is very limited, therefore the addition of RICs on Thomson's sales & trading products would not change the situation to any significant extent compared to pre-merger.

V.A.4.2. Effects in Research & Asset Management

131. As indicated above, RICs are predominantly used in market awareness products that serve front office on-trading floor environments and are used primarily for price discovery purposes. That is to say, RICs are used to find out a real time price for a specific instrument before a decision to execute a trade on it - as such, they are used by traders, sales traders, sales and other front office personnel who are time-sensitive and need to know instrument prices in real time. RICs are less important in the products that serve the research & asset management space (off-trading floor) applications. In particular, it appeared that in the off-trading floor space the inclusion of additional securities identifiers is not a decisive means of competitive differentiation (37).

132. As regards the use of RICs in the research & asset management space, the investigation did not bring about any convincing element indicating that this possible action from the merged entity would foreclose competitors and impede the customer to switch to non- RICs content and desktop product providers. Therefore, also supposing that the merged entity would extend the use of RICs to Thomson's research & asset management products, given the nature of the products in that space (see further), the effects of this element alone is unlikely to represent a significant barrier to entry for competitors which would then render competitors' products less attractive for Thomson's customer base in the research & asset management space.

V.A.4.3. Conclusion on RICs

133. Therefore, the transaction does not give rise to conglomerate effects as regards the use of RICs by the merged entity throughout the sales & trading and the research and asset management space.

V.A.5. Transaction platforms/services for Fixed Income securities (e.g. Tradeweb)

134. Thomson and Reuters both offer Fixed Income transaction services. However, the activity of Reuters in this area, via Reuters Trading for Fixed Income, is marginal and generates almost no revenue. Only Thomson, via Tradeweb, has a significant activity. The notifying parties estimate the market share of Thomson/Reuters in the Fixed Income trading area at [10-20]*% at EEA level and [10-20]*% worldwide. A number of competitors are active in this market, including Bloomberg, MarketAxess, eSpeed or Exchanges which facilitate the trading of some bonds. As there is no horizontal overlap, the merger does not raise any horizontal competition concerns.

135. Indicative market data is generated via transactions on transaction platforms, which constitute the raw material for financial information providers such as Bloomberg or Reuters. A share of indicative market data generated on Tradeweb is distributed by Reuters (38). As such, the concentration will not substantially change the access to indicative market data for the main Fixed Income financial information providers.

136. Although the market share of the notifying parties does not reflect a dominant position on the market for Fixed Income transaction platforms, Tradeweb is generally referred to as the main platform in this area. Tradeweb has notably achieved this position by gaining support from liquidity providers (banks, funds, etc). Thomson announced on 11 October 2007 to have reached an agreement with nine of the largest financial institutions (39), where the latter would (a) take a stake of [...]* in Tradeweb trading platform and derivative, and (b) set up a new entity together with Thomson (where they will have an interest of [...]*) to develop trading in new asset classes.

137. A third party raised a concern with regard to the future importance of Tradeweb. This competitor expects that Tradeweb will become in the near future the far dominant Fixed Income transaction platform, because it has gained support of many liquidity providers, and will continue to attract more financial institutions whose interest is to buy and sell on most dynamic market places. The third party complains that when Tradeweb concentrates the bulk of Fixed Income transaction, competitors will be foreclosed from the market for distribution of Fixed Income financial information as Thomson/Reuters will not make accessible to its competitors indicative market data generated on Tradeweb. It can be noted that indicative market data generated by some other transaction platforms, such as Bloomberg's platform, is not distributed via Thomson or Reuters, and conversely.

138. Nevertheless, the Commission considers that this concern cannot be considered as related to the concentration between Thomson and Reuters. Firstly, Reuters is currently not active (or marginally) in the Fixed Income transaction platform; secondly indicative data generated by transactions on Tradeweb is not distributed by main competitors pre-merger; and thirdly Reuters currently already redistributes indicative market data generated on Tradeweb and as such the transaction does not substantially modify the access to such data. It can also be added that the future success of Tradeweb still has to be confirmed and that the strategy developed by Thomson/Tradeweb to secure relations with liquidity providers was initiated before the merger and could still be replicated by competitors.

139. In addition, it is not obvious that the indicative market data generated via transactions on Tradeweb will only be made available to Thomson/Reuters customers. For instance one can consider that Tradeweb will have an interest in selling and making available its indicative market data to Bloomberg considering the leading position of the latter in the Fixed Income area ([40-50]*% market share according to the notifying parties - 2006 worldwide).

V.A.6. News

V.A.6.1. Introduction

140. Both Thomson and Reuters are present in the financial and business-related News area (hereafter 'News'). News is a financial content that provides investors with the information they need to make swift and intelligent investment decisions. Such financial news may be comprehensive or more limited as it may relate only to a single country, certain asset classes or particular economic sectors. Customers are business professionals active in Research and Asset Management such as investment managers, fund managers but also professionals involved in Sales and Trading.

V.A.6.2. Distribution channels

141. A distinction should be drawn between non-integrated financial News content suppliers such as Dow Jones, Informa, AP who deliver their content through platform providers (e.g. Bloomberg, FactSet, Thomson, Reuters) and integrated News providers who have developed their proprietary News content and are able to distribute them to financial institutions through their terminals (e.g. Reuters, Thomson and Bloomberg). Thomson, also distributes its News product as a standalone data feed to competing third party distributors, whereas Bloomberg and Reuters sell their proprietary News only on their platforms to final customers.

142. These three integrated News suppliers also distribute financial News from third party financial News providers (e.g. Dow Jones) in addition to their own offerings. They typically charge a royalty fee to third party news providers for the distribution of the News service which corresponds to a portion of the revenue generated by the third party providers.

V.A.6.3. Thomson and Reuters' products

143. Thomson Financial News (TFN, formerly AFX) is a blend of content created by its employed journalists and content sourced from partners such as Agence France Presse (AFP), Associated Press, Deutsche Presseagentur (DPA), Amalthia, Radiocor and others and rebranded as Thomson. Thomson has only recently - after its acquisition of AFX from Agence France Press in 2006 - started to commercialize its own News products. Thomson's TFN is distributed though its terminal products (e.g: Thomson One) or as standalone data feeds.

144. Reuters News enjoys strong reputation and credibility in the market and therefore is viewed as a leader in the News segment. It boasts one of the world's largest international news agencies competing with Bloomberg and Dow Jones. Reuters News is only distributed through its terminal products (e.g. Xtra 3000).

V.A.6.4. Competitors

145. Thomson and Reuters compete against two categories of competitors: global financial news providers, that is to say, providers offering news covering of companies in all the regions (e.g. Dow Jones) and local financial news providers offering news covering one specific country or region (eg. Nikkei for Japan).

146. Several competitors are active in the market for the provision of global financial News: Bloomberg, Dow Jones Newswires, S&P and the FT Group. Bloomberg and Dow Jones are considered together with Reuters the principal News sources the market relies on for global overall coverage. TFN/AFX which also offers global coverage is however more focused on European news and is not considered to belong to the same league. The same is true for /smaller global smaller players such as S&P, and the FT Group.

147. There are also numerous News suppliers focusing on certain countries or particular economic sectors or asset classes and competing with Reuters and Thomson news services to a certain degree. These include Nikkei, Interfax or Deutsche Pressagentur.

V.A.6.5. Market Shares

148. The table below provides the notifying partiers' estimates of the main competitors' total market shares in News segment in 2006.

Table 3

Global % EEA %

Thomson-Reuters [20-30]*% [30-40]*%

Thomson [0-10]*% [0-10]*%

Reuters [20-30]*% [30-40]*%

Bloomberg [30-40]*% [30-40]*%

Dow Jones [10-20]*% [10-20]*%

FT Group [0-10]*% [0-10]*%

Nikkei [0-10]*% [0-10]*%

S&P [0-10]*% [0-10]*%

DPA [0-10]*% [10-20]*%

Radiocor [0-10]*% [0-10]*%

Source: Notifying parties

V.A.6.6. Competitive assessment - horizontal effects

149. Regarding the horizontal effects of the merger, internal documents from the parties have shown that Thomson was serious in its attempt to enter the News market, by investing in the acquisition of AFX, and that this was done with the purpose of competing more directly with Reuters' News and to feed Thomson desktop products. The merger could eliminate rivalry between Thomson and Reuters in the provision of News services on the downstream level and create an incentive to discontinue TFN as a competitor to Reuters, leaving Dow Jones Newswires as the last major financial newswire that is not part of Bloomberg or Thomson/Reuters.

150. The parties however rightly pointed that the horizontal overlap between Thomson and Reuters in the supply of financial news is very limited since TFN still remains a small player which has not much traction on the market and does not constitute a competitive constraint to Reuters. As a result, the impact of Thomson/Reuters merger on the competitive structure of the market is likely to remain essentially unchanged post-merger.

V.A.6.7. Competitive assessment - vertical effects

151. In the course of the market investigation concerns have been expressed with respect to two potential anti-competitive vertical effects: i) input foreclosure for distributors competing on the downstream market and ii) foreclosure of competing News providers.

152. In relation to the first aspect, several third party redistributors have voiced that with the proposed merger, Dow Jones would be the last major financial newswire that is not part of Bloomberg or Thomson-Reuters (TFN and Reuters News). In the same way as Reuters has not distributed its News product to re-distributors, the merged entity would no longer distribute Thomson's News offering via competitors. In turn, Dow Jones could become the only newswire available to redistributors downstream. Dow Jones could take advantage of this situation to increase its prices.

153. It should be recalled that on the distribution market, TFN has a market share of [0-10]*% in the EEA and of [0-10]*% worldwide. The market investigation showed that the impact of TFN in the market as a constraint on Dow Jones is currently considered to be very limited. This is notably due to the more comprehensive coverage offered by Dow Jones News.

154. Furthermore, it is unclear whether Dow Jones would have an incentive to increase its prices following the disappearance of TFN since it would have an interest in maintaining several channels of distributions open for its News services and keep the redistributors competitive against the integrated News offer of Thomson/Reuters and Bloomberg terminals.

155. Secondly, concerns have also been voiced that the merger would result in reduced access to the marketplace for third party News suppliers. As a result of the proposed merger, a combined Reuters-Thomson could impose uneconomic charges for access or push third party News providers (such as Dow Jones) off its terminals. Despite having a large number of distributors, Dow Jones generates a significant share of its revenues from the sale of its data through Thomson and Reuters desktops. The parties' incentives to exclude Dow Jones from their platform would derive from their interest in foreclosing one of their main competitors to access their customers.

156. In view of Dow Jones' brand recognition and of its wide acceptance in the financial community such a move appears unlikely as customers are likely to continue to ask for Dow Jones services. A decision by the parties to cease distribution of Dow Jones could face strong customer resistance and would put Thomson/Reuters at a disadvantage if compared with its competitors (for instance Bloomberg and Factset distribute Dow Jones as an optional service). This probably explains why Reuters, even in the on-trading floor where it faces little competition from Thomson, distributes Dow Jones News services on its desktop products, and this despite the fact Dow Jones directly competes with Reuters' News. In addition, Reuters generates significant revenues from the distribution of Dow Jones News services since it typically charges a royalty fee between [...]* of the revenue generated by News providers. The new merged entity would have an interest in keeping this stream of revenues alive especially if there is strong demand for Dow Jones News products from customers. It therefore does not seem likely that the merger will significantly change the distribution of Dow Jones News services.

V.A.6.8. Conclusion on News

157. For the reasons developed above, the concentration is not likely to lead to a significant impediment of effective competition in the field of financial news.

V.B. RESEARCH AND ASSET MANAGEMENT

V.B.1. Research - broker reports

V.B.1.1. Introduction

158. Broker reports are documents containing an analysis of a security (unique firm) or a market (several firms active in the same business area). They provide information helpful to investors in making their investment decisions. Such information generally includes recent stock performance, comparison of performance with peer companies and industry trends. Broker reports can also contain Estimates, e.g. forecast of a company future performance. They generally conclude with an investment recommendation, typically rating a security as "buy", "hold" or "sell".

159. Broker reports are principally (40) prepared by "brokerage" firms. Brokerage firms are sellside research firms who provide their buy-side customers with such reports with an objective to generate demand for transaction. Such transactions will generate commissions and contribute to additional revenues for the brokerage firm. Brokerage firms are generally large banks (e.g. Crédit Suisse, Société Générale, Goldman Sachs, Lehman, etc).

160. Buy side customers are obviously interested in receiving such reports as they constitute an important component of their investment decision process. Customers can be banks, investment managers, fund managers, etc.

V.B.1.2. Real-time vs. Aftermarket Research

161. The principal objective of broker reports is to generate transaction from buy-side customers. The latter make their investment decision at very short term and need constantly refreshed information. As such, the information conveyed in broker reports is short-lived and presents the highest value for brokerage firms and buy-side customers at the date of release. Broker reports are referred to as Real-time Research in the first days or weeks after their release, typically around 14 days. During this period, brokerage firms specifically control who is permitted to access their research and avoid disclosing their reports to sell-side competitors.

162. Conversely, other actors in the financial market are interested in the information contained in broker reports at a later stage, although a share of this information is outdated. Investment banks for instance can have an interest in having an industry insight even one year after a report was published as the global functioning of the market and the competitive position of firms is likely to have remained stable. Such broker reports are referred to as Aftermarket Research (alternatively named "embargoed" or "delayed" research).

V.B.1.3. Distribution channels

163. Distribution of real-time and Aftermarket Research have to be distinguished according to the market definition set above.

V.B.1.3.1. Real-time Research

164. Real-time research can be distributed via: (i) direct distribution (email) from the brokerage firm to its customers; (ii) downloading of reports on the brokerage firm's website; and (iii) distribution via a data provider offering an aggregated database of research reports (typically Thomson, Reuters, Bloomberg). Such databases aggregate reports from multiple brokers.

165. Direct distribution and the downloading of reports on the brokers'website are free of charge for brokers' customers. Distribution via a data provider is generally accessible via a desktop solution or via a flat fee.

166. In all modes of distribution, the brokerage firms conserve the ownership of the reports and data providers are only permitted to distribute research to those customers identified by report owners. In this respect, a distinction should be drawn between two categories of brokers: (i) those brokers who give an individual prior approval for each customer for the distribution of their reports, and according to the notifying parties are the largest brokers; and (ii) those brokers who exercise a lower control by granting access to all customers by default, who are according to the notifying parties smaller brokers (having for example a geographical focus). Prior-approval brokers account for [20-30]*% of the total number of brokers distributed by Reuters and [10-20]*% of those distributed by Thomson.

167. Although brokerage firms have an interest in largely distributing their analyses, they can also impose limits on the distribution of Real-time Research reports as they contain valuable and tradable information. Nevertheless, the information provided by the notifying parties shows that clients of Reuters and Thomson purchasing a Real-time Research offering generally have access to the large majority of brokers.

168. [10-20]* Reuters' top 20 customers have access to the publications of more than [70- 80]*% of brokers distributed by Reuters (access to [80-90]*% of brokers on average). [10-20]* out of the Thomson top 20 customers have access to the publications of more than [70-80]*% of brokers distributed by Thomson (access to [60-70]*% of brokers on average for the top 20 customers). These proportions are however substantially lower when considering only those brokers who give a prior approval for the distribution of their reports, where top 20 clients of Reuters and Thomson have on average access to respectively [30-40]*% and [30-40]*% of brokers.

169. It can therefore be concluded that only prior-approval brokers exercise a tight control over the distribution of their reports, whereas non prior-approval brokers generally allow a large distribution. Although these prior-approval brokers are generally large banks and as such most influent brokers (benefiting from their reputation of major financial actors), it should be recalled that they only account for a small proportion of brokers the reports of which are distributed by Reuters and Thomson (respectively [20- 30]*% and [10-20]*%).

170. Brokerage companies have a strong interest in distributing reports to their customers in order to generate commissions. As such, the direct distribution, which is not impacted by the merger, is a channel that is actively used. The notifying parties have notably submitted that direct distribution exercises a significant competition constraint on the activity of Real-time Research aggregators as their customers could easily cancel the purchase of a Real-time aggregated offer to receive Real-time Research directly from the brokers.

171. However, this assumption has been contradicted by the market investigation. Almost all customers have indicated that there is a specific added value in receiving real-time research via an aggregator and that direct distribution by brokers, either via email or consultation of brokers' web sites, would not be a satisfactory substitute to an aggregated offer (41). Direct collection of broker reports is time consuming and would incur high costs for the customers. It would also require building a large number of interfaces with different brokers, and would imply a risk of missing and losing information.

172. In addition, as indicated above, the large majority of brokers distributed via Thomson and Reuters have adopted a non prior-approval system. Although a majority of these brokers are not the main suppliers in terms of reputation or number of reports, they bring specific added value to their customers because they have a special focus on a sector or a geographical area. Several hundreds of such brokers (Reuters: [500-600]*; Thomson: [800-900]*) are distributed by aggregators. For these brokers it would be clearly impossible to access customers via direct distribution as it would not be possible for these customers to manage so many direct feeds.

173. As such, the direct distribution of Real-time Research cannot be considered an appropriate substitute for the Real-time aggregated offers. As quoted by Customer 2, it can be used as a back-up, or to collect additional specific information, but cannot replace an integrated offer.

V.B.1.3.2. Aftermarket Research

174. Aftermarket Research is not useful to generate trading and as such less directly distributed by brokers, although the latter can make their research accessible to their customers via their websites. Embargoed reports are therefore largely distributed via data providers and are principally sold on a pay-per view basis. They are not subject to prior-approval by brokers and are accessible to all aggregators' customers.

175. As users of Aftermarket Research are generally financial operators seeking historical or contextual information on a company or a sector, they have an interest in gathering as many reports as possible on the subject they are working on, and therefore a database aggregating a large number of reports from different brokers, allowing to sort the reports and covering different periods, is a particular asset. As such the need for an offer aggregating multiple sources is even more prominent than for real-time research.

176. Not surprisingly, and similarly as for real-time research, customers have indicated that direct collection of broker reports on brokers' websites would not be a substitute to an aggregated offer as it would imply a high collection cost (both in terms of staff and time) and customers would certainly miss reports of many smaller broker firms.

V.B.1.4. Thomson & Reuters' products

Thomson

177. Thomson Real-time Research distribution service is marketed under the FirstCall (42) brand. Customers can access this service via (i) a data feed either direct to customer or via a third party distributor (e.g. Factset); or (ii) a core or add-on module of a Thomson desktop. FirstCall Research distributes broker reports from [1,000-1,100]* brokers and has a worldwide coverage. In 2006, estimated revenues generated by FirstCall Research were EUR [10-100]* million at worldwide level and EUR [0-10]* million in the EEA (43).

178. Thomson Aftermarket Research is named Thomson Research (formerly Investext). This product is principally delivered as an identifiable product (or included as a component within Thomson One Banker). Thomson Research distributes broker reports from [500- 600]* brokers (in 2007) and has a worldwide coverage. This product generated in 2006 a worldwide revenue of EUR [10-100]* million and an EEA revenue of EUR [0-10]* million.

179. Thomson database enjoys a strong reputation on the financial market. It is the only competitor having signed deals with [40-50]* brokers to have the exclusive right to distribute their Aftermarket Research (44). Reuters notably indicates in an internal document that in Aftermarket Research Thomson leading position has given leverage to secure exclusive deals. "Thomson also commands a dominant market position in delayed research with Investext, where their market share has given them leverage to secure exclusive distribution contracts with key content distributors" (45).

Reuters

180. Reuters' Real-time Research offering is distributed via Reuters Knowledge (46) as well as via the Reuters desktop, Reuters 3000 Xtra Reuters. This service is also accessible via third party distributors who build their own interface. Reuters distributes real-time broker reports from [700-800]* brokers and has a worldwide coverage. In 2006, revenues generated by the real-time research offering of Reuters were EUR [10-100]* million at worldwide level and EUR [0-10]* million in the EEA (47).

181. Reuters' Aftermarket Research offering is distributed via direct sales and third party redistributors (only in the US, in the EEA there is no redistribution of Aftermarket Research). It is also available through Reuters Knowledge. Reuters distributes broker reports from [400-500]* brokers and has a worldwide coverage. This product generated in 2006 a worldwide revenue of EUR [0-10]* million and an EEA revenue of EUR [0-10]* million.

182. Internal documents from Thomson illustrate that Thomson and Reuters research products are close to each other and that the two companies actively compete one against each other in the market. [statement by a Thomson sales representative on Reuters research product]* (48).

V.B.1.5. Competitors

183. As indicated above, broker reports belong to "brokerage" firms who have generated them. Data providers do not own the content, however it should be distinguished between those competitors who compile a database and manage the relationship with brokers, and those who only redistribute the database aggregated by a competitor. Some competitors also compile a part of the database they commercialize and complete it by redistributing the research offering of another competitor.

184. Some competitors only distribute real-time research, some both real-time and Aftermarket Research. It is worth noting that there is no competitor distributing (or compiling) only Aftermarket Research without real-time research. Some competitors also only cover a geographical area, whereas others compile broker reports from different areas (North America, Europe, Asia). Some competitors compile contribution of a small number of brokers (contributors). The following table summarises the characteristics of the main providers of aggregated offers of real-time or Aftermarket Research.

<table>

185. Thomson and Reuters are the main providers of Real-time research, and the only providers of an aftermarket offer with a global coverage. Bloomberg, Factset and Capital IQ have recently launched their own-compiled real-time research offering, whereas TheMarkets.com entered the market in 2000. Zacks restricts its compilation to US companies.

Bloomberg

186. The company has recently started to operate its own-compiled real-time research offering. [...]*. The Bloomberg offer is not yet comparable to the one of Thomson and Reuters (number of brokers, reputation), and several customers have indicated that they thought Bloomberg would need a few years (one to two years) to be fully comparable to Thomson and Reuters in that respect. The size and reputation of Bloomberg should facilitate the signing of agreements with brokers. Nevertheless, Bloomberg does not redistribute its real-time research offer via competitors and is not expected to do so in the near future. Bloomberg does not compile or redistribute Aftermarket Research. Factset

187. It started its own-compiled real-time research offering in 2006, and is not yet recognised as a strong player in this area (49). It should be noted that Factset continues to advertise Thomson and Reuters' real-time product on its own web site (50) to be in a position to offer a credible offer to its customers. Factset, in December 2007, could only propose access to broker reports of eight global contributors (51). Customers have indicated that they would expect Factset to propose a comparable offer as the real-time research offer of Thomson and Reuters in a few years (52) although Factset is only an imperfect substitute to Thomson and Reuters offers in the very short term. Factset has not started to develop an aftermarket offer and depends on Thomson in this regard.

Capital IQ

188. It is in a similar situation as Factset. It currently redistributes Reuters offer but has started to develop its own-compiled real-time research offering. It started to develop its own offer by signing contracts with brokers and building the software infrastructure in 2005. In December 2007, Capital IQ still needed to sign agreements with some large contributors to complete its offer. Capital IQ does not distribute its own-compiled Aftermarket Research and acts as a redistributor of Reuters' offer.

TheMarkets.com

189. It is a consortium created by 11 major brokers who all contribute to the real-time and Aftermarket Research by making their broker reports available. TheMarkets.com can propose a global coverage only via the distribution of Reuters research offering. In an internal document, Thomson states that TheMarkets.com lacks the breadth of Thomson products[...]* (53). In this document it is worth noting that only TheMarkets.com and Reuters are referred to as credible competitors in the research area. Although it cannot be excluded that TheMarkets.com develops its own compiled database for real-time and Aftermarket Research it cannot currently constitute a credible alternative to Thomson and Reuters' offerings without Reuters data. In addition, it is not obvious that those brokers who do not have contributed to the creation of TheMarkets.com would have an incentive to contribute to an offer controlled by competitors.

190. The notifying parties have also submitted a list of 10 additional competitors active in the distribution of real-time or Aftermarket Research. All of these competitors are either redistributors of Thomson and Reuters, or compile broker reports in a limited geographical area (e.g. IFIS compile real-time broker reports in Japan from 30 different sources).

191. Real-time research: Factset and Capital IQ, as well as other competitors such as IDC, who directly compete with Thomson and Reuters in the area of providing desktops to end-customers, all depend on Thomson or Reuters in terms of real-time research. However, the emergence of Factset and Capital IQ in the real-time research might challenge this dependency on the notifying parties offer within one or two years. Bloomberg has recently developed its own real-time research offer, but it would not constitute a credible supplier as its business model is not to redistribute its products via competitors. TheMarkets.com is a credible competitor only via the redistribution of Reuters offering, and Zacks only covers the US.

192. Aftermarket Research: There is no competitor providing a global Aftermarket Research offer, but TheMarkets.com which however redistributes Reuters Aftermarket Research and does not own-compile such a product. Zacks only covers the US and its offer is significantly less reputable than that of Thomson and Reuters.

V.B.1.6. Market shares

193. The notifying parties have provided an evaluation of the market shares of Thomson, Reuters and their main competitors concerning the market for the sales of Real-time Research and Aftermarket Research to end-customers (banks, fund managers, etc). It should be noted that a share of the revenue generated by each company's desktop solution has been attributed to research and therefore such revenue is a theoretical construction and does not correspond to real transactions on the market. Although such market shares must be interpreted with care, they can provide an insight of the market for the distribution of real-time research to end-customers.

Table 5 Real-time Research

REAL-TIME RESEARCH - MARKET 2006

Company World (mEUR) World (%) EEA (mEUR) EEA (%)

Thomson-Reuters [10-100]* [20-30]*% [10-100]* [10-20]*%

Thomson [10-100]* [10-20]*% [0-10]* [0-10]*%

Reuters [10-100]* [0-10]*% [0-10]* [0-10]*%

Bloomberg [100-200]* [50-60]*% [10-100]* [60-70]*%

Factset [10-100]* [0-10]*% [0-10]* [0-10]*%

Capital IQ [10-100]* [0-10]*% [0-10]* [0-10]*%

TheMarkets.com [10-100]* [10-20]*% [0-10]* [0-10]*%

Zacks [0-10]* [0-10]*% [0-10]* [0-10]*%

Total Market [200-300]* 100% [10-100]* 100%

Source: Notifying parties

194. [The market shares were calculated by the notifying parties under a number of reasonable assumptions to respond to the Commission requests. However, the Commission decided to use this data with great care, due to the unsatisfactory results of the exercise.]*

195. The market shares as proposed by the notifying parties suggest that Thomson/Reuters would not be in a position to materialize any market power in the distribution of realtimer research. However, such presentation of the market is not illustrative of the competition problems raised by the merger as many of the main competitors such as Factset, Capital IQ and TheMarkets.com still depend to some extent (or completely) on Thomson/Reuters' offering. Their revenue notably includes revenue generated by the redistribution of Thomson and Reuters' offers.

196. The notifying parties submit that 50% of users receive reports directly from the brokers via email or via brokers' portals. Nevertheless, the market investigation has indicated that customers make a clear distinction between direct transmission from multiple brokers and a unique transmission of an aggregated offer, emphasising that direct transmission is not a substitute for an aggregated offer.

Table 6

Aftermarket Research

AFTERMARKET RESEARCH - MARKET 2006

Company World (mEUR) World (%) EEA (mEUR) EEA (%)

Thomson-Reuters [10-100]* [70-80]*% [0-10]* [60-70]*%

Thomson [10-100]* [60-70]*% [0-10]* [50-60]*%

Reuters [0-10]* [0-10]*% [0-10]* [0-10]*%

Bloomberg 0 - 0 -

Factset [0-10]* [0-10]*% [0-10]* [0-10]*%

Capital IQ [0-10]* [0-10]*% 0 -

TheMarkets.com [0-10]* [10-20]*% [0-10]* [20-30]*%

Zacks [0-10]* [0-10]*% [0-10]* [0-10]*%

Total Market [10-100]* 100% [10-100]* 100%

Source: Notifying parties

197. Only TheMarkets.com is presented by the notifying parties as a credible competitor in the market for Aftermarket Research. Nevertheless, although TheMarkets.com compiles the broker reports of 11 broker companies, it depends on Reuters to propose a complete offering in Aftermarket Research. TheMarkets.com revenue includes revenue generated by the redistribution of Reuters' offer.

198. In an internal document, Reuters also estimates that the market for on demand research on a pay per view basis (Aftermarket Research) will increase from US$ [0-100]* million to more than US $[400-500]* million by 2009-2010 and that "Thomson dominates the market today with a 90% share in dollar terms. Our [Reuters'] share of the market is 10%" (54).

199. Only Thomson and Reuters have a complete offering for Aftermarket Research in the form of a searchable database. As indicated above, access to embargoed research on the brokers' websites it not a substitute to such a searchable database, as customers are looking for an access to a large number of reports on a specific subject, with the possibility to sort reports from multiple sources.

V.B.1.7. Adverse horizontal effects (non coordinated)

Real-time Research

200. As Factset and Capital IQ are expected to propose a credible Real-time Research product within a few years, [...]*, it is unlikely that the notifying parties would have the ability nor the incentive to increase their prices or to stop distributing their real-time research offers.

201. Such a strategy would rather speed up the switch of redistributors to Factset or Capital IQ, would imply a loss or revenue for Thomson/Reuters, and would notably not result in a decrease in the quality of the desktops sold by these redistributors as they would have the ability to replace the notifying parties' offer by another equivalent one.

202. It can therefore be concluded from the above that the concentration is unlikely to lead to a significant impediment of effective competition in the distribution of Real-time Research.

Aftermarket Research

203. Conversely, with regard to Aftermarket Research, Thomson and Reuters propose the only complete offer of own-compiled offer. The newly merged company could have the ability and the incentive to either increase its prices or to limit the access to its Aftermarket Research offering to its own customers (those customers purchasing a complete desktop solution from Thomson/Reuters). Such adverse horizontal effects would impact on both the direct customers of the notifying parties and on their redistributors.

204. If Thomson/Reuters increased its prices, the reaction of its customers would only be limited. TheMarkets.com is the only aggregator alternative on the market, and is dependent on Thomson/Reuters pricing strategy as its offer is only complete via the redistribution of Reuters' offer. There is therefore no comparable alternative but to contact each broker company which is a timely and cost-consuming exercise, and not considered as a viable alternative by customers. Such a price increase strategy would therefore be hardly opposable by customers and competitors.

V.B.1.8. Barriers to entry/expansion are high

205. Barriers to entry/ expansion are only analysed with regard to Aftermarket Research, considering that no adverse horizontal effects are foreseen in the Real-time Research area. However, examples of entry in the Real-time Research business are used to illustrate the necessary efforts to enter the Aftermarket Research business, as a number of common factors are observed between the two markets.

206. The notifying parties have submitted that building an Aftermarket Research offering is not difficult and that entry in this business is easily feasible as illustrated by the recent entry of Bloomberg, Factset and Capital IQ in the Real-time Research business, where they submit that conditions to entry are comparable to those prevailing in the Aftermarket Research area. However, although the Real-time Research offer from Bloomberg is expected to be rapidly comparable to the offers of Thomson and Reuters, Bloomberg has not entered the area of Aftermarket Research and has not announced any intention to do so. The capacity of Bloomberg to enter any new business area in the financial information business is however unique considering the financial capacity, customer base and reputation of this company. With regard to Factset and Capital IQ, their own-compiled Real-time offerings are nascent activity and the companies still depend on Thomson & Reuters at the present time.

207. Building a credible Aftermarket Research offering requires negotiating agreements with a large number of broker companies. Both Thomson and Reuters have agreements with several hundreds of broker firms. Nevertheless, the notifying parties have submitted that approximately [60-70]*% of aftermarket page downloads of Thomson and Reuters are for reports from their top 25 contributors. They have also indicated that [80-90]*% of Aftermarket Research downloads are for research reports not older than 12 months. In their view, this shows that building a credible Aftermarket Research offering would require a limited amount of efforts and agreements with brokers. Nevertheless, it is again not obvious that an Aftermarket Research offer limited to a small number of brokers would attract the customers, and signing agreements with the top brokers is a long process.

208. The nascent activity from Factset in Real-time research is quite illustrative in this respect. Although Factset claims that agreements have been signed with almost 400 brokers, its offer is still considered largely behind those of Thomson and Reuters. In addition, signing with the largest brokers is not immediate and may require long negotiations. In the case of Factset, it is worth noting that although the signing of agreements started in 2006, only eight global brokers had signed an agreement with Factset in December 2007 (55). It should also be recalled that Factset is among the financial information providers that already have regular contacts with brokers via its Estimates business, and that this privileged position has not enabled Factset to build a credible offer yet. Although this specific experience relates to Real-time Research only, it can be assumed that signing agreements with brokers in regards to an Aftermarket offer would face the same limitations.

209. In addition, although the majority of sales in Aftermarket Research are achieved with the most recent reports, a competitor who had not built a database including older reports would be at a disadvantage compared to Thomson and Reuters, as customers in this area will tend to favour the most complete offerings. Building a comparable Aftermarket Research offer would therefore require several years of activity as Brokers might not make available on their websites their oldest reports, notably because these reports might contain assumptions that were not confirmed after the publication of the reports.

210. Entry in the Aftermarket Research is also made more difficult by the fact that a number of brokers ([40-50]*) have signed exclusive agreements with Thomson. These exclusive deals account for [10-20]*% of the total download volume of Aftermarket Research of Thomson. If a new entrant entered the market, its offer would present a weakness face to Thomson's offer.

211. More generally, a competitor willing to build a strong alternative to Thomson/Reuters would have to sign agreements with a number of brokers superior to 100-120 which is already a time-consuming exercise, or its offer could be viewed as a basic package lagging behind. It should also be noted that in the area of Aftermarket Research, a competitor would not have access to those brokers having signed an exclusive deal with Thomson which would diminish the value of its offering.

212. It can also be argued that brokers could support the emergence of competitors like TheMarkets.com. Should such a support materialize, this would however require several years for a strong and reliable competitor to emerge. In addition, the economic interest of developing such a database for a totally new player is uncertain. It is for instance worth noting that TheMarkets.com, which was created in 2000, never developed its own compiling activity of reports from brokers not belonging to the 11 founding members. In this respect, it should be recalled that brokers are competitors and have diverging interests. As such, it is not clear whether brokers who did not contribute to the creation of TheMarkets.com have an interest in contributing to this research offering.

213. In regard of the current experience of Factset and Capital IQ to develop their owncompiled Real-time Research, it can be estimated that about three years are necessary to build a credible offer, notably to build its own infrastructure and to sign agreements with contributors. This performance is however accessible to companies already benefiting from a reputation in the financial information market. Factset and Capital IQ real-time research offers could present credible alternatives to Thomson and Reuters' offers within this time frame in this respect.

214. Should they decide to build their own-compiled Aftermarket Research, it can be estimated that they would need between three to five additional years. Two to three years would certainly be sufficient to build their infrastructure, sign agreements with contributors (generally aggregators sign separate agreements with brokers for the distribution of real-time and Aftermarket Research (56)). However they would likely need four or five years to collect an historical research database covering a sufficiently long period, and to be accepted by customers.

V.B.1.9. Adverse vertical effects

215. Thomson/Reuters will likely take advantage of its new position in the Aftermarket Research market to foreclose its competitors by either increasing the price for this content set distributed via redistributors (royalty based and sell-through), or by simply stopping providing its competitors with such data. Thomson/Reuters should have both the ability and the incentive to develop such a strategy.

V.B.1.9.1. Ability to foreclose

216. A way to measure the dependency of competitors on Thomson/Reuters Aftermarket Research offering is the share of competitors' desktops (workstations) which carry Thomson and Reuter's research offering.

Table 7

PERCENTAGE OF COMPETITORS' WORKSTATIONS WHICH CARRY THOMSON OR REUTERS AFTERMARKET RESEARCH OFFERING

Segment Aftermarket Research (%)

All segments [30-40]*%

Fixed Income -

Institutional Equity -

Investment Management [30-40]*%

Wealth Management [0-10]*%

Investment Banking [30-40]*%

Corporates 100%

Source: Notifying parties

217. In Research and Asset Management, [30-40]*% of all competitors' desktop offers carry Thomson or Reuters Aftermarket Research offerings. Only the Wealth Management segment is not dependent on Thomson and Reuters offering, because broker reports are not demanded by customers in this category. Wealth Management desktops commercialised by Thomson and Reuters for instance do not include this content set (57).

218. [The market shares were calculated by the notifying parties under a number of reasonable assumptions to respond to the Commission requests. However, the Commission decided to use this data with great care, due to the unsatisfactory results of the exercise.]*

219. In addition, as indicated above, there is no other competitor in a position to propose an equivalent Aftermarket Research offering on the market. Only TheMarkets.com is active in the market for redistribution of Aftermarket Research and redistributes the Reuters' offering. It is likely that almost 100% of competitors distributing an Aftermarket Research offer depend on Thomson or Reuters. The notifying parties will therefore clearly have the ability to foreclose competitors in this market.

V.B.1.9.2. Incentive to foreclose

220. Should Thomson/Reuters decide to increase its prices or stop distributing its Aftermarket Research offering via its competitors, (including TheMarkets.com) the latter could not find an alternative offer on the market. It can be envisaged that Bloomberg could build its own Aftermarket Research offering, firstly because it already has a real-time offering and contacts with brokers, secondly because it provides enough guarantees to brokers in terms of access to end-customers, and thirdly because it has the financial capacity to build such an offer. Nevertheless, it is very unlikely that Bloomberg would commercialize such a complete (real-time and aftermarket) research offering to its competitors' customers.

221. Other competitors could decide to build their own-compiled Aftermarket Research offering. However this would take several years according to the results of the market investigation (58), with no guarantee of success. Such foreclosing could be in the interest of Thomson/Reuters as it would strengthen its offering in comparison to those desktop solutions proposed by competitors. A share of customers would therefore likely abandon their original provider to purchase Thomson/Reuters desktop solutions.

222. Thomson and Reuters generated in 2006 a revenue of EUR [0-10]* million worldwide and EUR [0-10]* million in the EEA via the sales of Aftermarket Research data through redistributors ("royalty based" and "sell-through"). Should Thomson/Reuter lose this revenue by foreclosing its competing redistributors, the newly formed company would need to compensate this loss by capturing customers from its competitors generating at least the same revenue. As such customers would switch to Thomson/Reuters to purchase desktop products, generating substantially more revenues than the sales of content sets via redistributors. Based on the data provided by the parties, a foreclosure strategy would be profitable (in terms of revenue's stream) with the displacement of only a minimum number of (ex-redistributor's) customers (59).

223. An internal document from Reuters illustrates the disciplining effect of having both Thomson and Reuters on the market for distribution of broker reports. Reuters indicates that would Thomson pull out the distribution of its research offering from [a distributor]*, Reuters should take advantage of this move to capture [a distributor]* customers. It is worth noting that the strategy envisaged by Reuters is not to replace Thomson as a content set provider, allowing end-customers to continue purchasing a desktop solution from [a distributor]*, but rather to fully displace [a distributor]*. [analysis of the impact of a withdrawal by Thomson and Reuters of their distribution of Research and Estimates from a third party distributor]* (60).

224. The notifying parties also argue that broker companies would certainly oppose any attempt from Thomson/Reuters to increase prices or to foreclose competitors, notably via direct distribution of reports (email or own website) or via supporting such initiatives as TheMarkets.com. However, the capacity of brokers to oppose such strategies is unclear. The value of Thomson and Reuters offering is in the aggregation of reports from multiple sources and individual brokers cannot replace such an offer via direct sending of their own reports.

225. It could be envisaged that brokers refuse to grant to Thomson/Reuters the right to distribute their reports. However such assumption would be very speculative, as it is in the interest of brokers to have their offers distributed to the largest possible extent in real-time research (to generate commissions on trade) and because revenues from Aftermarket Research are achieved at no cost by brokers (pure margin). Therefore it is very likely that brokers will continue to distribute their reports via Thomson/Reuters as well as via Bloomberg for real-time research.

V.B.1.9.3. Conclusion on adverse vertical effects

226. Based on the above, it can therefore be concluded that Thomson/Reuters would have the ability and the incentive to foreclose competitors in the Aftermarket Research market.

227. In addition, such vertical foreclosure would be able to aggravate the horizontal effects raised by the concentration as the increase in price for content sold to/through redistributors and/or the degradation of the quality of the desktop products (containing Broker Research) distributed by the redistributors would ultimately increase the likelihood that Thomson/Reuters can profitably increase its prices for its own customers. As already indicated, no other competitor provides a product equivalent to Thomson or Reuters', and as a result these customers would not have to possibility to switch to a comparable provider, in terms of price or in terms of content quality.

V.B.1.10. Conclusion on Research- Broker Reports

228. The entry of Capital IQ and Factset in the Real-time Research is expected to discipline Thomson/Reuters. Conversely Thomson and Reuters also are the only providers of complete offer of Aftermarket Research. As such, the newly formed entity would have an incentive to increase its prices for Aftermarket Research products. In addition, the newly formed company would likely vertically foreclose its competitors by increasing prices for Aftermarket Research distributed via redistributors or by limiting the access to its Aftermarket Research offering to its own customers (those customers purchasing a complete desktop solution from Thomson/Reuters).

229. It can therefore be concluded from the above that the concentration is likely to lead to a significant impediment of effective competition in the distribution of Aftermarket Research.

V.B.2. Earning Estimates

V.B.2.1. Introduction

230. Estimates are predictions regarding the future performance of a company with respect to key metrics such as annual or quarterly earnings per share. Estimates expectations are the most important determinant of equity prices as investors buy stocks at prices that the market determines based on corporate earnings potential. Estimates are typically produced by sell-side independent analysts as an integral part of the analyst's Broker Research reports. As with the underlying research the Estimates remain the intellectual property of the sell side research firm who has control over the permission and use of the data.

231. The main customers of earning Estimates data are institutional investors, who are professionals who manage various categories of funds (mutual funds, pension plans, hedge funds) and are commonly called "the Buy Side". These professionals include portfolio managers who manage the funds and buy-side research analysts who identify for the portfolio manager stocks that have the best upside potential. Other smaller customers include corporations, Investment Banking groups and individual investors.

V.B.2.2. Current vs historical Estimates

232. A distinction is made between 'current' Estimates and 'historical' Estimates. Current Estimates are earning predictions that have been released in the past 12-18 months and that relate to future performance of a company. By contrast 'Historical' Estimates are earning Estimates that relate to past performances. They generally date back to more than 18 months in history and can cover a period of up to 20 or even 30 years of history.

233. Current Estimates are for instance used to derive valuation ratios (e.g: estimated Price Earning ratio), assess the relative value of a stock within a sector or compare results communicated by companies with consensus Estimates and adapt trading decisions accordingly. Historical Estimates are used for analysis of surprise, which is differences between historical Estimates and actual Estimates, analysis of correlation between revision of Estimates and change in stock value and assessment of analyst performance on a stock or in a sector.

234. Usages of current and historical Estimates are very often mixed. Indeed, analyses of historical Estimates are commonly used to better assess the impact, accuracy or trend of current Estimates. For this reason, financial institutions purchase bundled products that combine both.

V.B.2.3. Distribution channels

235. The channels of distribution for earning Estimates are identical to those described for real time research (see paragraph V.A.3) since Estimates are a subset of Broker Research reports although they may sometimes be distributed separately. In fact, agreements concluded with brokers for the distribution of real time research usually also cover Estimates. As an illustration, each of the agreements concluded by Thomson and Reuters with the [0-10]* largest brokers for real time research covers Estimates.

236. Estimates can be distributed via: (i) direct distribution (email) from the brokerage firm to its customers; (ii) downloading of reports on the brokerage firm's website; and (iii) distribution via a data provider offering an aggregated database of research reports (such as Thomson or Reuters). Direct distribution and the downloading of reports on the broker's website are free of charge for brokers' customers. Distribution via a data provider is generally accessible via a desktop solution for a flat fee. In all channels of distribution, the brokerage firms conserve the ownership of the reports and data providers are only permitted to distribute research to those customers identified by report owners.

237. The role of financial data providers like Thomson and Reuters is typically to compile Estimates data from hundreds of broker firms and offer them to customers on a consolidated basis. This work entails the extraction of data and the computation of various constructed financial ratios such Estimate Earning per Share (EPS) Long Term Growth, Book Value-Adjusted or EBITDA (earnigs before interest, tax, depreciation and amortisation).

238. Thus, on top of the detailed Estimates per broker, the compiler will provide consensus Estimates that is to say, a mean or a median value of all Estimates from Research departments relating to a series of constructed financial ratios. This set of consolidated data is viewed as a critical tool of analysis in investment decisions.

239. By contrast, Estimates provided by brokers directly to their customers, because they do not provide a consensus view of all brokers, cannot be used for analytics and modelling purposes and therefore are not considered substitutes to earning Estimates products offered by aggregators by a huge majority of respondents to the market investigation. Some respondents have also pointed that direct distribution would not be considered a satisfactory substitute, as recurring personnel and resources would be required to manage the aggregation and consolidation of data which would not be time and cost effective.

V.B.2.4. Global Estimates products

240. The market investigation has shown that financial institutions and corporations require global Estimates data with history, that is Estimates covering companies in all the regions of the world for a period of at least several years, and that in order to build a credible product for the global market, a financial data provider needs to collect Estimates from hundreds of different broker and analysts firms. The four bigger players on the market (Thomson, Reuters, Bloomberg and FactSet) all claim to have between 400 and 600 regular contributors to their database.

241. Entering into an agreement with a broker firm, especially a big one, is not necessarily easy and can be time consuming. Missing certain key broker Estimates may be damaging for the credibility of the consensus Estimates offered by the data provider. The number of contributors to the database as well as the presence of key Broker Research is viewed as critical element to determine the breadth and depth of an estimate product.

242. Moreover, the aggregation and consolidation of data is a sophisticated process. It requires significant investments in technologies and human resources including the presence of well trained analysts able to compare the raw data, build a consensus and adjust the consensus after events that change the market view. Several hundred people, including staff qualified in accounting, are typically dedicated to the management of an estimate database.

V.B.2.5. Thomson & Reuters' products

243. Both Thomson and Reuters offer earning Estimates products. Thomson Estimates offerings are incorporated in First Call and IBES. First Call mainly provides current earning Estimates data and consensus Estimates. IBES forecast are available as a detailed time series of individual Estimates or in a summary view that provides a consensus level of synopsis. IBES history dates back to 1981 for the US market and to the early 1990's for European companies in addition to some aggregated Estimates.

244. Thomson offers access to its earning Estimates either (i) via a data feed - which is distributed either directly to customers or via a third party redistributors (e.g. [...]*) or ii) as a core module integrated in a desktop solution (e.g. Thomson One).

245. Similar to Thomson, Reuters distributes earning Estimates via data feed and desktop solutions. Reuters' main products are Reuters Estimates and Reuters Estimates Forecast Pro. Reuters Estimates is an estimate database that contains current Estimates data, plus 18 months of historical Estimates. Reuters Estimates Forecast Pro delivers broker detail and consensus data for both current and historical Estimates. The depth of research and Estimates content made available over these desktop/application products varies on a product by product basis depending on the needs of the average user of those products.

V.B.2.6. Competitors

246. As indicated above, the raw information used to produce earning Estimates data is "extracted" from research reports authored by brokerage firms and independent analysts who have generated them. Two categories of competitors should be distinguished: those who compile an earning database and manage the relationship with brokers (such as Thomson and Reuters), and those who only redistribute the database aggregated by a competitor (e.g. IDC, SunGard, Mergent).

247. Within the first category, a limited number of compilers offer a comprehensive database of global Estimates. Apart from Reuters and Thomson, only FactSet and Bloomberg supply such a product. The situation of TheMarkets.com, a consortium of eleven investment banks will be shortly addressed. There are also a significant number of smaller data providers which own a database covering a more limited geographical area.

The most significant are Zacks (for US data) or Toyo Keizai (for Japan).

248. Institutional professionals typically make purchase decisions as regards Estimates based on the coverage (number of companies), quality and number of contributors, timeliness (speed to with which new data is added), accuracy (lack of mistakes), depth (number of items covered), history (number of years of information) and geographical scope. The following table 8 summarises the characteristics of the main providers of aggregated offers of Estimates with respect to certain of these key criteria.

<table>

249. The products of the parties have an advantage against their competitors in most if not all the areas. The parties have explained that distributing Estimates is simply a matter of obtaining the agreement of a sufficient number of brokers and point to the relatively recent entry of FactSet and Bloomberg and the planned entry and expansion by S&P/Capital IQ and TheMarkets.com. The competitive situation of each of the competitors to Thomson/Reuters is addressed below.

FactSet

250. FactSet is considered the number three player on the market, following the acquisition of JCF, a French Estimates compiler in 2004. FactSet is relatively strong in European content, but end customers would generally not consider FactSet proprietary global content to be an equal substitute for either Thomson's IBES or Reuter's Estimates. It is worth recalling that in addition to its self source Estimates data, FactSet also distributes Thomson IBES content and Reuters Estimates.

251. Competitor 21 has explained: 'our customers would not consider FactSet's proprietary content to be an equal substitute for either IBES or Multex'.(61)

252. According to internal document from Reuters, FactSet Estimates do not offer the same breadth and depth: [...]* (62) and do not compete in the same league as Thomson and Reuters: [...]* (63)

253. According to a Thomson document Facstet/JCF 'offers weak market coverage outside its key European markets (France, Germany)'.(64) Several respondents in the market investigation have also pointed to the lack of breadth of FactSet Estimates offer outside Europe and in historical data.

Bloomberg

254. Bloomberg is a recent player on the market. It has offered its own estimate database on its terminals since November 2006. In February 2007, Bloomberg cancelled its contract with Thomson IBES Estimates and now self source all their Estimates. In an internal document Reuters points to several weaknesses of Bloomberg Estimates offer including the limited coverage and the lack of history and concludes that the product is not competitive against Reuters:

'BBG (Bloomberg) has around 450 broker compared to 580 covered by Reuters. Took us around 14 years to grow this business, BBG where we were 10 years ago (...) cannot offer a reliable consensus. ' (...) 'Lack of history will be a problem. Their consensus calculations methodology has been constantly changing over the past 2 years. Impossible to backfill Estimates credibly especially with missing broker (...) Nowhere near the global company and broker coverage that Reuters offers (...) This is far removed from BBG's traditional premium quality content'.

255. Thomson also points to the same weaknesses of Bloomberg's offer: 'Estimates data are limited as is collection experience. Only limited history (...) Assuming no aggregate data' (65) and stresses a weak brand name for Estimates and limited channels of distribution.

256. Bloomberg, despite its size and reputation, is not expected to offer a competitive global offer on the market anytime soon. Bloomberg in any case does not redistribute its Estimates to third party distributors and thus offers it Estimates only on its terminals as it operates under a close model. Its market share is therefore unlikely to grow beyond its base in the future.

TheMarkets.com

257. According to the notifying parties, TheMarkets.com which is consortium of 11 well established broker firms, is planning an expansion of its Estimates offering and could exercise a competitive constraint on the parties. However the market investigation has revealed that TheMarkets.com heavily relies on Reuters [...]* for the collection of broker Estimates beyond its 11 members and has no global offer of its own. TheMarkets.com estimate offer therefore does not compare with the parties and cannot be considered as constituting a real competitive constraint to the parties in the foreseeable future. S&P/Capital IQ

258. The notifying parties have submitted that S&P/Capital IQ is in the process of building a global research and estimate database and that it could become a formidable competitor to Reuters and Thomson. The market investigation has shown that it would take years before Capital IQ be in a position to develop a credible global offer that could compete with Reuters and Thomson because of the costs induced by such an investment and the difficulty to reach agreements with the necessary contributors.

259. In conclusion, the market investigation has established that vendors such as Bloomberg, FactSet or S&P have the technology and resources to develop Estimates products that could compete with Thomson and Reuters but that for the moment their offer did not match those of the parties and that it would take them several years to develop products that could offer the same breadth and depth and history as Reuters and Thomson.

Smaller compilers

260. Apart from these five sources, there are smaller localized suppliers such as Zacks in the US, Tokyo Keizai in Japan and Ipreo/Hemscott in Europe. However, these suppliers do not really compete with global players but rather provide a complementary offer responding to a specific demand pertaining to local markets. As such, they are not considered direct substitutes to Thomson and Reuters. (66)

261. As regards the possibility for potential competitor to aggregate data from local providers to build a comprehensive data base, the market investigation has demonstrated that any company that would try to collect data from different sources would be confronted with the lack of consistency of the data and that consolidation of such data would be costly, time consuming and very resource intensive.

V.B.2.7. Market shares

262. The notifying parties have provided an evaluation of the market shares of Thomson, Reuters and their main competitors concerning the market for the sale of Estimates to end customers (banks, fund managers, etc). It should be noted that a share of the revenues generated by each company's desktop solution has been attributed to Estimates and therefore such revenue is a theoretical construction and does not correspond to real transactions on the market. Although such market shares must be interpreted with care, they can provide an insight of the market for the distribution of Estimates to end customers.

Table 9 Shares of supply in Estimates worldwide in 2006

ESTIMATES - MARKET 2006

Company World (m$) World (%) EEA (m$) EEA (%)

Thomson-Reuters [100-200]* [40-50]*% [10-100]* [30-40]*%

Thomson [100-200]* [30-40]*% [10-100]* [20-30]*%

Reuters [10-100]* [0-10]*% [10-100]* [0-10]*%

Bloomberg [100-200]* [30-40]*% [10-100]* [30-40]*%

Factset [10-100]* [0-10]*% [10-100]* [10-20]*%

Zacks [10-100]* [0-10]*% [0-10]* [0-10]*%

TheMarkets.com [0-10]* [0-10]*% [0-10]* [0-10]*%

Toyo Kezai [0-10]* [0-10]*% [0-10]* [0-10]*%

Ipreo Hemscott [0-10]* [0-10]*% [0-10]* [0-10]*%

Total Market [400-500]* 100% [100-200]* 100%

Source: Notifying parties

263. [The market shares were calculated by the notifying parties under a number of reasonable assumptions to respond to the Commission requests. However, the Commission decided to use this data with great care, due to the unsatisfactory results of the exercise.]*

264. It is worth noting that in an internal document, Reuters quantifies Bloomberg revenues generated from Estimates at $ [10-100]* million, far less than the $[100-200]* million suggested by the parties. Even more surprising is the share of revenue attributed to Reuters. In the same internal document, Reuters Estimates that its revenues for 2006 should reach $[100-200]* million while the notifying parties suggest only $[10-100]* million.

265. [...]*

266. Reuters' internal document provides an analysis of the competitive landscape more in line with the market power of the notifying parties. According to the document, the total market size for Estimates is around $ [400-500]* million a year and 'Thomson currently have approximately [40-50]*% of this market'. (67) Reuters would come second with [20- 30]*% of the market shares with $[100-200]* million generated from Estimates essentially from Estimates content sold through their desktop platforms. The combined market shares of the two companies would therefore reach [70-80]*% worldwide. Bloomberg would come third with $[10-100]* million of revenues and [10-20]*% market share followed by FactSet with revenues of $[10-100]* million and [0-10]*% of market shares.

267. Also respondents to the Commission's inquiry indicate an order of magnitude for the market shares of the parties as regards Estimates which is much higher than what the notifying parties have provided the Commission with and come to much higher market share Estimates for the notifying parties.

<table>

V.B.2.8. Thomson and Reuters are the closest substitutes

268. The market investigation has established that Thomson's IBES is the global market leader with respect to earning Estimates thanks in particular to the breadth and depth of its IBES database. Reuters which has invested heavily in Estimates since its acquisition of Multex in 2003 (68), is viewed by a significant number of competitors and customers as the only real alternative to Thomson. (69). One respondent has also stressed: Reuters is No 2 and has increased its presence in this segment to the point that they constitute today the only real alternative to Thomson's. (70)

269. Internal documents also show that the parties see each other Estimates offer as close substitutes. According to an internal document from Reuters 'in Estimates, only Reuters and Thomson are best in class and highly competitive' (71). Reuters would even score better in certain segments: 'Thomson are the incumbent in the Estimates market but Reuters has surpassed them in many areas: Quality -IBES relies on broker contributed spreadsheets for their Estimates data, rather than Broker Research. This has severe implication on the quality and timeliness of its data which is inferior to Reuters (...); Methodology -IBES doesn't us the same value-added methodologies and rigorous policies that differentiate Reuters Estimates; Timeliness -IBES are not considered in the same ballpark as Reuters for timeliness. Reuters will update broker Estimates typically within 2 hours of receiving the research, and no later than 24 hours after. (...) IBES are well known to take days to update their Estimates and actuals. (72)

270. According to an internal document from Thomson, Reuters Estimates offers 'comparable product offering to TF (offer real-time data, means on same accounting basis of Estimates/actuals, multiple product/delivery methods, etc)'. (73)

271. The parties have claimed that there is no justifiable concern with respect to historical Estimates as Thomson IBES which is the uncontested leader in this area, does not compete with Reuters. Reuters Estimates would be targeted at referential users, which do not consume historicals. This assertion is neither supported by the internal documents from the parties nor by the market investigation. First, although the parties have originally claimed that Reuters has no historical offering they have admitted that Reuters has a limited presence in historical Estimates sales via distributors (74). In addition, it is uncontested that Reuter's products include historical Estimates. In a recent submission the parties explained that as regards Reuters' Estimates offer: "Two main classes of products exist: a) products that provide a current view of Consensus Estimates or Detailed Estimates; b) products that provide a full time series (75) of Consensus Estimates or Detailed Estimates."

272. In an internal document dedicated to the analysis of the competitive landscape for Estimates, Thomson refers to Reuters/Multex Historical data as covering a period up to '1999 for North America, (and) 1996 International' that is 8 years for the US and 11 years for the rest of the world. (76)

273. Competitor 21 further clarified that 'Thomson and Reuters have broadly similar offerings, up to a certain degree of historical coverage, though Thomson has more history. This additional history makes Thomson stronger for customers who require longer history (especially quants). However, no other company has a comparable offering to either Thomson or Reuters for historical Estimates.' (77)

274. In addition, several customers have indicated that Reuters historical Estimates are used by traders, analysts, portfolio managers and investment bankers for quantitative arbitrages, research and modelling purposes. For the reasons developed above, Thomson and Reuters Estimates offer can be considered close substitutes.

V.B.2.9. Adverse horizontal effects (non coordinated)

275. The market investigation - including internal documents - has shown that in earning Estimates, Thomson and Reuters are respectively No 1 and No 2 and see themselves as each other as main competitors. Also, there is no other close substitute product on the market since no other company offers the breadth and depth of Reuters and Thomson databases. FactSet/JCF is a distant No 3 with a weak market coverage outside Europe and Bloomberg, which is a new entrant on the market, is not considered by customers and competitors as a solid competitor in the short term because of its limited coverage and its lack of historical Estimates.

276. As a result of the merger there is a strong likelihood that the parties will discontinue or merge their products because of the overlaps. Reuters' Estimates product line could be progressively discontinued to the profit of Thomson Estimates products. The likely effect of the merger will be the disappearance of one of the only players that offer credible global Estimates products. This would be a 2 to 1 situation. The merged entity would become dominant on the Estimate market with market shares exceeding [60- 70]*% and would even have a stronger position on historical Estimates with only limited competition from FactSet.

277. As the use of historical estimate data becomes more and more critical with the increasing number of funds based on quantitative models requiring historical data (78), there is a risk for the market that all the quantitative models would use the same source which could amplify the risks of volatility on the market.

278. Would Thomson/Reuters increase its prices, the reaction of the customers could only be limited. (79) With regard to global current Estimates, a customer could switch to a Bloomberg terminal. However such terminals are relatively expensive if compared with other desktop solutions in the market and switching would unlikely be profitable for the customer. Switching to FactSet would unlikely bring any advantage as FactSet redistributes Thomson/Reuters products and its own estimate global offer is considered weaker than the parties especially for US companies. With respect to historical data, Bloomberg would not even be an option and the weakness of FactSet offer would play against considering it as a credible alternative. Such adverse horizontal effects would impact both direct customers of the notifying parties and their redistributors.

V.B.2.10. Barriers to entry/expansion are high

279. The notifying parties have submitted that building an estimate offering is not difficult and that entry in this business is easy as illustrated by the recent successful entry of FactSet and Bloomberg.

280. The market investigation however does not support this assertion. It should first be recalled that in order to build a credible estimate product for the global market, a financial data provider needs to collect Estimates from hundreds of different broker and analysts firms and that this process is neither easy nor speedy. Starting from scratch would typically take more than 2 years to reach a critical mass of contributors.

281. Failing to secure a few significant contributors could damage the quality of the estimate offer and the competitiveness of the product. Besides, brokers are unlikely to share their research with a distributor which does not already have significant buy-side client base. This makes it extremely difficult for a distributor which is not already an established data supplier to enter the business. FactSet and Bloomberg, two significant data providers are still struggling to secure certain key contributors with a view to improving their coverage.

282. In addition, the aggregator will need to have developed the proper software infrastructure to collect reports from the hundreds of contributors in an automated way, and make them easily available for extraction so that they can be quickly uploaded in the system to update the database. Also, the computation of consensus Estimates for certain key data measures (e.g. EBITDA, price earning ratio) would typically require the involvement of several hundreds of skilled people including qualified accountants to process the data, consolidate and normalize them in order to calculate mean estimate ratios.

283. The costs associated with the development of a comprehensive database of global Estimates should also not be underestimated. They are indeed significant and must be borne during a period of several years as illustrated by the example of FactSet which in order to set up a footprint on the market acquired an existing company JCF in 2004 for $50 million (80) and subsequently heavily invested in the business in order to expand.

284. Lastly, building up an historical database can only be achieved by collecting current data over a sufficiently long period of time as it is not possible for a third party to acquire historical Estimates outside Reuters or Thomson since these data are simply not longer available or nor released anymore. It would therefore typically take several years for a new entrant before being in a position to offer a global Estimates offer with historical data.

285. As it has been demonstrated above, the assumption made by the notifying parties that the brokers can easily induce new entry by third party aggregators in the event that the existing distribution systems for Estimates did not operate to their satisfaction is not supported by the facts.

V.B.2.11. Adverse vertical effects

286. Thomson/Reuters could take advantage of its new position in the Estimates market to foreclose its competitors by either increasing the price for this content set distributed via redistributors (royalty based and sell-through), or by simply stopping providing its competitors with such data. Thomson/Reuters should have both the ability and the incentive to develop such a strategy.

V.B.2.11.1. Ability to foreclose

287. A way to assess the market power of the combined entity is to measure the dependency of competitors on Thomson/Reuters Estimate offering by looking at the share of competitors' desktops (workstations) which carry Thomson and Reuters' estimate offering as summarized in the table below.

Table 11: Penetration of Thomson and Reuters' estimate products of third party terminals

PERCENTAGE OF COMPETITORS' WORKSTATIONS WHICH CARRY THOMSON OR REUTERS ESTIMATES OFFERING

Segment Thomson (%) Reuters (%) Total Estimates (%)

All segments [40-50]*% [0-10]*% [50-60]*%

Fixed Income [60-70]*% - [60-70]*%

Institutional Equity - [40-50]*%

Investment Management [70-80]*% [10-20]*% [80-90]*%

Wealth Management [20-30]*% [0-10]*% [20-30]*%

Investment Banking [40-50]*% [10-20]*% [50-60]*%

Corporates [90-100]*% [0-10]*% [90-100]*%

Source: Notifying parties (Response to the 30-10-2007 Article 11 questionnaire, Annex 7)

288. [50-60]*% of all competitors' desktop offers carry Thomson or Reuters Estimates offering (current, historical or both). It must be noted that in Investment Management and Investment Banking, the two segments where the offer of the parties more significantly overlap, their combined penetration share is respectively [80-90]*% and [50-60]*%. [90-100]*% of competitors' offers in the Corporates segment carry Thomson or Reuters' Estimates offering. This clearly point to the potential market power that the combined entity could acquire after the merger.

289. [The market shares were calculated by the notifying parties under a number of reasonable assumptions to respond to the Commission requests. However, the Commission decided to use this data with great care, due to the unsatisfactory results of the exercise.]*

290. According to figures provided by the parties, Thomson/Reuters account for [70-80*]% of third party Estimates content redistributed on competitors' platforms. In addition [50- 60]*% of the competitors platforms include either Thomson or Reuters estimate offering or both. These figures show the degree of dependency of third party redistributors on the parties Estimates offer. Apart from the parties, FactSet is the only significant Estimates compiler to make its content available to third parties since Bloomberg does not redistribute its Estimates via competitors and is not expected to do so in the near future. The notifying parties will therefore clearly have the ability to foreclose competitors in this market.

V.B.2.11.2. Incentive to foreclose

291. As already stated, because of the high degree of overlap between Reuters and Thomson Estimates offer, the parties will have an incentive to discontinue one line of product, most likely Reuters. Redistributors which are highly dependent on Reuters could be severely affected by the disruption but also (and to a lesser extent) data compilers such as Zacks, FactSet and TheMarkets.com which in addition to their own estimate offer also propose Reuters estimates.

292. The merged entity would also have an incentive to increase its prices or stop supplying Estimates content to/through third party redistributors since Thomson and Reuters would no longer compete against each other on the upstream market for the distribution of Estimates contents to redistributors and that redistributors are competitors of the parties. This would result in a significant impediment of effective competition on the market for the supply of Estimates to third party distributors as redistributors would be left with no option than to source from suppliers proposing lower standard products, such as FactSet, TheMarkets.com or smaller regional providers.

293. Several aggregators that depend heavily on Thomson's data feed would be deprived of their Estimates, or should increase their prices, which would lessen the competitiveness of their platform. Indeed, Estimates are viewed by end customer - especially in the Research and Asset Management segments- as critical element in the desktop offer. This would, in turn, limit the ability of independent data vendors from acting as a competitive constraint to Reuters/Thomson on the downstream market of data feeds and workstations.

294. Thomson and Reuters generated in 2006 revenue of EUR [10-100]* million worldwide and EUR [0-10]* million in the EEA via the sales of Estimates data through redistributors ("royalty based" and "sell-through). Should Thomson/Reuter lose this revenue by foreclosing its competing redistributors, the newly formed company would need to compensate this loss by capturing customers from its competitors generating at least the same revenue. As such customers would switch to Thomson/Reuters to purchase desktop products, generating substantially more revenues than sales of content sets via redistributors. Based on the data provided by the parties, a foreclosure strategy would be profitable (in terms of revenue stream) with the displacement of a minimum number of (ex-redistributor's) customers (81).

295. An internal document from Reuters illustrates the disciplining effect of having both Thomson and Reuters on the market for distribution of Estimates. Reuters indicates that would Thomson pull out the distribution of its Estimates offering from [a distributor]*, Reuters should take advantage of this move to capture [a distributor]* customers. It is worth noting that the strategy envisaged by Reuters is not to replace Thomson as a content set provider, allowing end-customers to continue purchasing a desktop solution from [a distributor]*, but rather to fully displace [a distributor]*. "Rumours have circulated that Thomson may pull some or part of its distribution from [a distributor]* [Estimates and Broker reports]". " [Reuters should] capitalise on the opportunity to switch [a distributor]* users [and] accelerate efforts to reduce the number of Reuters content set distributed via [a distributor]* and encourage clients to consider Reuters offerings [] versus [a distributor]*" (82).

296. The parties have suggested that if the merged entity would stop making available its estimate content to redistributors other estimate compilers such as FactSet, Zacks, or Tokyo Keizai could easily expand their sale to compensate for such a disruption. This assumption overlooks the fact that FactSet currently does not offer a close substitute product to Thomson and Reuters and would therefore not allow the redistributors to offer an as competitive Estimates product as before. This would be even truer for Zacks and Toyo Keizai who only have a regional coverage.

297. The parties have also argued that supply via redistributors is a very minor channel accounting for only 13.2% of all sales. Assuming it was correct, this element would have no impact on the competitive analysis since all the redistributors (including IDC, Edgar Online, LSE) and several third party compilers (such as FactSet, Capital IQ, TheMarkets.com) are highly dependent on the parties for the sourcing of their estimate offer. This figure would only show that - contrary to what the parties suggestredistributors do not currently exercise a strong competitive constraint on the parties and that competition mainly takes place between estimate data compilers. The remarks developed above as regards the negative horizontal effects of the merger would then be fully applicable.

V.B.2.11.3. Conclusion on adverse vertical effects

298. Based on the above, it can therefore be concluded that Thomson/Reuters would have the ability and the incentive to foreclose competitors in the Earning Estimates market.

299. In addition, such vertical foreclosure would be able to aggravate the horizontal effects raised by the concentration as the increase in price for content sold to/through redistributors and/or the degradation of the quality of the desktop products (containing Earning Estimates) distributed by the redistributors would ultimately increase the likelihood that Thomson/Reuters can profitably increase its prices for its own customers. As already indicated, no other competitor provides a product equivalent to Thomson or Reuters', and as a result these customers would not have to possibility to switch to a comparable provider, in terms of price or in terms of content quality.

V.B.2.12. Conclusion on Estimates

300. Thomson and Reuters are the only providers of comprehensive databases of global Estimates available to third party distributors and end customers. As such, the newly formed entity would have an incentive to increase its prices for Estimates products. In addition, the newly formed company would vertically foreclose its competitors by increasing prices for Estimates distributed via redistributors or by limiting the access to its Estimates offering to its own customers (those customers purchasing a complete desktop solution from Thomson/Reuters).

301. It can therefore be concluded from the above that the concentration will lead to a significant impediment of effective competition as regards Estimates.

V.B.3. Fundamentals

V.B.3.1. Introduction

302. Thomson and Reuters are active in the field of Fundamentals data. As described above, Fundamentals consist of various company-specific data, such as reportable pro forma financial statement data (including balance sheet, cash flow and income statements, both interim and annual), calculated financial ratios (e.g. annual and five-year averages for growth rates, profitability, leverage, asset utilisation), per share data (e.g. earnings per share, book value per share, cash flow per share), security and market identifiers, product and geographic segment information, and textual company profile data (including officers, address, web address, major shareholders, and business descriptions). Vendors do not create Fundamentals data (since the primary data is publicly available in annual and interim company accounts and publicly reported events), but they may calculate certain financial ratios from primary data and may standardise data that is reported by companies.

303. In the context of Fundamentals, a distinction is relevant between compilers of Fundamentals data, such as the parties, who collect the raw data from the primary source, and redistributors of content sets produced by the compilers, who operate at the next level down in the supply chain. The parties are both active in the compilation of Fundamentals data and both Thomson and Reuters supply databases for Fundamentals.

304. From a product standpoint, an important difference within Fundamentals is between "as reported" versus "standardised" data. "As reported" are (often electronic) data corresponding to the original documents (that is to say, the original balance sheet filed by a company) without further treatment done by any compiler, while "standardised" data are (mainly electronic) Fundamentals on which the compiler has done some adjustment to "translate" the original data into categorizations to conform to a standard, based on documented methodologies specific to the compiler and/or relevant to the users. The main databases in the marketplace may contain "as reported" data, "standardised" data or both. The significance of standardized data is in the fact that users can avoid manipulation of data to conform them to their business calculation needs, while "as reported" data may be still necessary in order to allow users to back64 track the original classification of a certain item in the company's accounts and to further drill down into the company's specific accounting methodology.

V.B.3.2. Compilers of Fundamentals

305. According to the parties there are several viable compilers of Fundamentals in the marketplace, besides Thomson and Reuters. The main compilers include S&P's Compustat and Capital IQ offerings, FactSet, Bloomberg, Mergent, Hemscott/Ipreo, Hoovers, and Edgar Online, WorldVest Base.

Box 1: Compilers of Fundamentals (raw data)

According to the parties, compilers of raw data that utilise public sources to populate a Fundamentals database include S&P's Compustat and Capital IQ offerings, Bloomberg, Mergent, Hemscott/Ipreo, Hoovers, and Edgar Online, WorldVest Base.

S&P's Compustat is said to be the leading provider of Fundamentals data in the world, covering over 65,000 companies back to 1950.

Compustat offers a database of aggregated Fundamentals data and also supplies third parties (the parties believe both compilers and redistributors) with data. S&P/Compustat has been continuously upgrading its offering to keep a best-in-class status (including by delivering a new product called "backtester", which is suited for quantitative analysis.

S&P's Capital IQ: S&P's Capital IQ offers coverage of private companies, S&P's Capital IQ's Fundamentals coverage includes over 60,000 public companies and 685,000 private companies, with coverage going back 10 years.

Bloomberg offers a comprehensive set of Fundamentals data provided over its desktop products, and recently over a datafeed for customers that use its terminals. Its database is believed to cover 150,000 companies (of which approximately 40,000 are public companies) with coverage data back to 1989 or earlier.

World'Vest Base (WVB) covers 51,000 companies in more than 130 countries, along with private companies coverage for 6 countries.

World'Vest may be obtained over the WVB web-site, through FTP, or through various WVB strategic partners, including Bureau van Dijk, Hugin, OneSource and RIMES.

Mergent also collects Fundamentals data and maintains a database containing information on 35,000 companies globally, as well as a collection of industry reports that describe the market environment and industry landscape for companies, industries, and regions. For US companies, Mergent's Fundamentals coverage would extend back to 1970.

Hemscott/IPREO also collect and provide company Fundamentals data with global coverage of companies. Its database would cover over 15,000 companies dating back to 1987. Hoovers, a Dun & Bradstreet company collectscompany Fundamentals data. The parties consider it only as a recent entrant in this space. It does not have much of a historical set of Fundamentals.

Edgar Online collects and maintains a database of Fundamentals data, with over 30,000 companies globally

V.B.3.3. Re-distributors of Fundamentals

306. Reuters and Thomson currently both supply Fundamentals data to redistributors or for incorporation in other products. They contend that in addition to the above compilers of Fundamentals information, there are numerous redistributors of Fundamentals data (that is to say, companies that source from those who aggregate with other information and distribute in the marketplace). The main redistributors of Fundamentals include FactSet, (considered both a compiler of raw source data and an aggregator/redistributor of third party data), Fame, a SunGard company, InfoUSA/OneSource, RIMES, Bureau Van Dijk, Hoovers, Broker portals and applications such as Morgan Stanley ModelWare.

Box 2: Redistributors of Fundamentals

According to the parties redistributors of Fundamentals (that is to say, they do not populate their own databases but source from those who do, aggregate with other information and distribute in the marketplace) include the following:

FactSet is both a compiler of raw source data and an aggregator/redistributor of third party data. The parties believe that FactSet is rapidly building up its own Fundamentals database, but it continues to source Fundamentals from a number of other providers in the marketplace.

Fame, a SunGard company, sources its Fundamentals data from third parties and provides aggregated Fundamentals in its product offerings. Fame's coverage is as per the sources it aggregates, which include Thomson and Compustat.

InfoUSA/OneSource also is an aggregator/redistributor of company Fundamentals sourced from over 25 third parties. It provides coverage of approximately 17 million public and private companies, with history as per the source.

RIMES is an aggregator/redistributor of company Fundamentals sourced from the Parties and third party compilers. Its coverage is as per the source it aggregates.

Bureau Van Dijk is an aggregator/redistributor of company Fundamentals sourced from third parties.

It provides global coverage of approximately 35 million companies (of which 48,000 are public companies) dating back to 1997.

Hoovers, is an aggregator/redistributor as well as a compiler.

V.B.3.4. Global Fundamentals products

307. The market investigation has shown that financial institutions and other customers require global fundamental data with deep history, that is fundamental data covering companies in all the regions of the world for a period of at least several years, and that in order to build a credible product for the global market, a financial data provider needs to collect fundamental data from hundreds of different sources (including companies and repositories). Moreover, it has emerged beyond doubt that the compilation and treatment of data to conform to codes and standards meaningful to users is a sophisticated process, which requires significant investment in technologies and also in specialised human capital able to extract relevant raw data, manipulate it in order to create adjusted and normalised database and to present a one-stop-shop solution to customers. Several hundreds people, including qualified accountants, are typically dedicated to the management of a Fundamentals database.

308. Institutional professionals typically make purchase decisions as regards Fundamentals based on the breadth and depth of the product (database) offered in terms of coverage (number of companies), timeliness (speed of adding relevant information e.g. press releases), accuracy (lack of mistakes), depth (number of different items covered), history (number of years of information), "as reported" and standardized information, geographical scope. The following table summaries the characteristics of the main compilers of fundamental data, which will be also examined individually in further sections.

<table>

V.B.3.5. Thomson & Reuters' products

309. Thomson offers Fundamentals data under the Worldscope brand, which it acquired in 2000. The Worldscope database was established in the mid 1980's and offers coverage of [30,000-40,000]* active companies and [10,000-20,000]* inactive companies in more than 70 established and emerging markets around the world with full history back to 2001 and partial history back to 1980. Thomson's Worldscope data feed is sold both directly to end-users and indirectly to end-users through third party redistributors such as [...]* and [...]*.

310. Reuters offering of Fundamentals data is commonly referred to as Reuters Fundamentals. Reuters, which acquired its Fundamentals database with the acquisition of Multex in 2003, provides a global offering, containing global data on [30,000- 40,000]* active companies ([50,000-60,000]* including inactive ones) with roughly 20 years of data for US companies and 7-10 years of data for non-US companies. Reuters provides limited Fundamentals data in many of its desktop applications, including Reuters 3000 Xtra, Reuters Station, Reuters Trader, Reuters Knowledge, and Reuters Wealth Manager.

311. The parties' offerings have a very good reputation in the market and are considered by nearly all respondents to the investigation the best in class and as very close products in terms of breadth and depth of content set covered. Broadly speaking, Reuters' product is more recent and less rich in history compared to the Thomson's one.

V.B.3.6. Competitors

312. Reuters and Thomson currently both compile databases of Fundamentals and supply Fundamentals data to redistributors or for incorporation in other products. According to the parties' own submission (83), Reuters' database covers 53 000 (public and private) active and inactive companies, whereas Thomson database covers 53 836 (public and private) active and inactive companies. These databases cover history back to 10 years (and partially back to more than 20 years). Based on the parties submission, other providers may appear, on a face value basis, to have a similar offerings and coverage as the parties, but, at a closer look, they do not have offerings comparable to the parties' ones.

313. In particular, across their various submissions, the parties identify some nineteen compilers of fundamental data in the marketplace: Thomson, Reuters, S&P Compustat, S&P Capital IQ, Bloomberg, Mergent, Bisnode, Fitch, Value Line, Ipreo/Hemscott, Fininfo, Tokyo Keizai, SNL Financial, Edgar Online, World/Vest Base, Economatica, Jordans, Findata, Zacks. Apart from Bloomberg, all of these compilers are said to redistribute their fundamental data trough third parties, notably through Alacra, Bureau Van Dijk, IDC, FactSet, Rimes, SunGard, Telekurs, etc.

V.B.3.6.1. Smaller compilers (scope of database/regional players)

314. Based on the parties' submission (84), regardless of the sheer number of companies which are covered by their relevant database offerings, some compilers can immediately be put in a clearly different category compared to other compilers based on the geographic or business sector coverage of their fundamental data offerings. This first selection criterion based on coverage and scope of the databases would involve the following eleven companies: (i) Bisnode has European and Scandinavian focus only, (ii) Fitch covers insurance and banks only and, being a rating agency, has a very small business in respect of Fundamentals, (iii) Value Line has a US focus only (7 600 active companies), (iv) Ipreo/Hemscott has a UK, Ireland and US companies focus (15 000 companies), (v) Fininfo covers Scandinavia and France, (vi) Tokyo Keizai is Japan-only (3 500 companies), (vii) SNL Financial has a US-only coverage (in Energy, Financial, Media and Real Estate sectors), (viii) Economatica has a small number of companies covered (2 170 companies) and only in the Americas, (ix) Jordan is UK-only database, (x) Findata is Scandinavia-only, (xi) Zacks is US focussed only (16 000 entities covered).

315. Even if the parties credit some of the above suppliers with sizable market shares (Bisnode [0-10]*% worldwide, Fitch [0-10]*% worldwide, Ipreo/Hemscott, Fininfo, Tokyo Keizai all with [0-10]*% worldwide, see further) (85), the above companies are at best regional suppliers which can compete in respect of certain sectors data, on niches and/or on specific regions. They cannot take on the parties at the level of their global offering, which as mentioned above are largely requested in the marketplace. Therefore these suppliers are not considered by customers and/or by redistributors as valid substitutes to the parties' offerings, but rather, possibly at best, as complementary sources (in case of specific niches/regions requirements) and/or when seldom applicable, as very distant, limited alternatives only for those marginal customers who are price sensitive and have much less rigid requirements in terms of accuracy, standardization and timeliness of data (which instead is normally the rule in this market).

316. The eight outstanding compilers which, according to the parties, would have comparable offerings in terms of global coverage, depth and breadth of content scope and history would be: Thomson, Reuters, S&P Compustat, S&P Capital IQ, Bloomberg, Mergent, World/Vest Base, and to a much lesser extent Edgar Online. These compilers (apart from the parties) are considered below:

V.B.3.6.2. Larger compilers (global scope)

Compustat

317. Belonging to the S&P group of companies, the parties consider Compustat as the incumbent and the "gold standard" for North America fundamental data, with the deepest history and depth". The parties state that Compustat Global dataset is broadly similar to the parties' in terms of coverage depth and history (86). According to the parties, the S&P group (that is to say, S&P/Capital IQ and Compustat) is by far the leading player worldwide and will remain so post-merger with a share of [20-30]*%.

318. The investigation did not confirm the parties' contention that Compustat is the world leader and the incumbent for global Fundamentals. While it is true that Compustat is historically considered as the initiator of this Fundamentals business in the US, the investigation suggested that its strengths, although significant, are all confined to the US and the North American region. The "Compustat North America" (87) database would be a "standardized" database with no "as reported" data in it and would still be a kind of "legacy" benchmark for Fundamentals related to US companies, but "Compustat Global" (that is to say, the product which is in more direct competition with Thomson Worldscope and with Reuters Fundamentals) would lag far behind the parties' offerings, given its poor coverage of non- US companies in terms of both depth and breadth of database scope (88).

319. Respondents to the investigation have confirmed that Compustat is a significant provider of fundamental data in the marketplace with a solid legacy base and reputation in the US and one who makes available its database through redistributors, which then include that in their own platform offerings. The investigation has in any event also indicated that Compustat is not a close substitute to the offering of Thomson and Reuters, especially when it comes to the requirements of global data and with specific reference to the needs of quantitative investment managers, who are used to building models and scenarios based on very deep and precise coverage of history and performance of companies and of sectors across the globe (89).

320. Several redistributors have also indicated that Compustat, Thomson and Reuters are the three largest compilers, and that the Compustat database does not have the same characteristics of the merging parties' databases, which instead stand out for their uniqueness in the marketplace in terms of breadth and depth of global coverage. (90)

321. This is largely confirmed by Reuters' internal documents, where the following is stated: "Reuters competes with Compustat mainly in the US market. Compustat has a strong content offering for US data, particularly longer history, coverage of [...] companies. On the other hand, their non-US database is very weak in quality, timeliness and coverage, and Reuters is usually preferred by users who need one vendor for global data. In addition, Reuters Fundamentals benefits quantitative user by providing original filling dates, exact fiscal period end dates, restatement/reclassification dates and period length [...] Compustat [...] captures only period end month." (91).

322. Based on the above, Compustat can be considered a significant player in the field of Fundamentals but certainly not a contender and a substitute to Thomson and Reuters in respect of Fundamentals with global coverage in the off-trading floor space.

Capital IQ

323. Still within the S&P group of companies, the parties consider that "Capital IQ has assumed a leading market position for US Fundamentals in a few years since foundations" and has an expanding international presence (92). In particular, Capital IQ would have built a database of Fundamentals data within 18 to 24 months since its acquisition of Simply Stocks in 2003 and gained international presence. Nevertheless, the parties do not claim that Capital IQ has either the same breadth and depth of Thomson's or Reuters' databases, or of Compustat's offering.

324. The investigation did not indicate that Capital IQ is a compiler of fundamental data comparable to the parties, and not even to its affiliate Compustat. Capital IQ is a relatively recent player in Fundamentals and several customers receive Thomson and Reuters fundamental content on Capital IQ platform (that is to say, desktop products). Therefore, Capital IQ appears to be a redistributor in the first place while as a compiler, although it must be recognised that it is steadily growing, at this stage of its business development, it is still a player with significantly smaller scale of operation compared to Thomson and Reuters.

325. In this respect, it can be noted that Thomson's internal Sales Aid document (released in third quarter 2007) does not mention Capital IQ in the section developing and analysis of competitors as regards Fundamentals, while other players are considered in much depth (Compustat, Reuter/Multex, Tokyokeizai, Nikkei, Bloomberg) (93).

326. In another internal document from Thomson which is instead only dedicated to the analysis of the strengths and weaknesses of Capital IQ, Thomson declares: "Capital IQ has, in a short time, undoubtedly come to be the most serious threat to existing aggregators of fundamental and other content, in the North American market, and has its eyes firmly fixed on the prospects for international growth also ". Nevertheless, the same document further states the following: "Capital IQ Weaknesses: Geographical Focus: Still heavily US in focus and content, although attempting to expand internationally. In the international markets, Worldscope has an excellent reputation for coverage and quality, and it will take Standard & Poor's quite some time before it can challenge in these markets[.]" (94).

327. Capital IQ database of Fundamentals has less depth and breadth compared to Thomson and Reuters. Compared to Compustat, Capital IQ product has more extensive geographical coverage, but has a quite limited history compared to Compustat and also to the parties' databases. Capital IQ can be considered a growing player in the field of Fundamentals, but certainly not as a close contender and a substitute to Thomson and Reuters in respect of Fundamentals with global coverage in the off-trading floor space.

Bloomberg

328. The parties claim that "Bloomberg has worked extensively to build up its global Fundamentals offering as part of its comprehensive suite of content and functionality since the 1990s. Its breadth and depth of coverage is equal to either of the Parties, and probably exceeds them" (95). The parties attribute [10-20]*% worldwide market share to Bloomberg as regards Fundamentals.

329. The investigation did not support the parties' view that Bloomberg is a major supplier of fundamental data content sets, capable of constraining the parties in this field. Firstly, Bloomberg offering of Fundamentals is part of the "Bloomberg terminal" suite solution without any availability of a fundamental-only offer for customers. To this extent, all respondents to the investigation, both customers and above all redistributors, did not mention Bloomberg as an alternative to the parties and/or to Compustat for their standalone needs of Fundamentals, although Bloomberg includes Fundamentals data in its suite terminal solution available to end-customers (96).

330. In any event, even considering Bloomberg's footprint as regards fundamental data more in general, it must be observed that within its terminal solution, Bloomberg Fundamentals appear to be targeted to the requirements of sales & trading usage rather than be suited to the research & asset management area, since the fundamental data are a complement to the strong suite offering of Bloomberg terminal primarily targeted at the on-trading, real-time data oriented space. This is confirmed by a Reuters' internal document, where it is clearly stated that Bloomberg is not seen as a competitor outside the desktop products since it does not have a standalone datafeeds offering (97). To corroborate this conclusion, Reuters also states that "Bloomberg Fundamentals data is mainly targeting at traders' referential usage, not so much at in-depth investment analysis" and that "as their target market is sales and trading not research and asset management, their data is not designed for analytical purpose". The same conclusion stems from Thomson's internal documents (98).

331. Based on the above, Bloomberg cannot be considered a credible contender of Thomson and Reuters in Fundamentals in the off-trading floor space, where the bulk of the parties' overlap is. Mergent

332. The notifying parties claim that "based on a strong legacy of historical databases from Moody's Financial Services and Ford Equity, and paired with strong collection technology from Praedea, Mergent is a viable global competitor with the backing of Xinhua Financial" (99). The parties contend that Mergent with its 15 000 US entities and 20 000 non- US entities assures the global coverage needed by customers and that Mergent has deep history (back to 1970 for the US and back to 1987 for non-US). The notifying parties state that the commercial focus of Mergent is the US (100) and they credit Mergent with a [0- 10]*% market share worldwide ([0-10]*% in the EEA).

333. The investigation does not support the view that Mergent is a viable alternative to Thomson and Reuters' Fundamentals. Mergent's Fundamentals are not considered capable of matching the parties' offerings, especially since their database would be lacking (inter alia) standardized data. As such they would not fit the requirement of most customers in the offtrading floor space, where a common need exist for off-the-shelf utilisable fundamental data, on which these customers would not have to perform further difficult, timeconsuming and mission-critical adjustment work on the data provided.

334. Competitor 20 does not consider Mergent a direct competitor of the merging parties, both at desktop and datafeed levels, and states that its products are not considered substitutable or close to the parties' ones from their own customers' perspective (101). Based on the above, it is unlikely that Mergent could come closer to the parties in the foreseeable future as regards Fundamentals, bearing in mind "the level of expertise and [the] difficulty to create and standardise, combined with the costs, especially with regard to creating historical data" (102). The notifying parties also do not appear to consider Mergent as particular competitive threat, since in internal documents they qualify it as a small actor in the marketplace. (103)

WorldVest/Base

335. The notifying parties claim that "World'Vest Base has been active in the collection of global fundamental data as long as either of the Parties, and its stated coverage and functionality is now the equal or superior to them. A comprehensive strategy of offshoring to reduce costs while constantly expanding coverage and depth has made WVB an attractive alternative for redistributors in particular" (104). The notifying parties recognise that WorldVest/Base core product is a "as reported" database (WVB Master), while it is unclear in the parties' submission what quality/strength they would attribute to WorldVest/Base "standardized" database.

336. In the market investigation, customers (especially those having expressed concerns in respect of fundamental data) do not mention WorldVest/Base as a valid/possible full alternative provider to the merging parties (105). As regards compilers/redistributors, many respondents do not consider WorldVest/Base a full-blown substitute to the database procured from the merging parties (106). Even if some redistributor (Competitor 36, Competitor 6) do (or may in the future consider to) source data from WorldVest/Base, this is generally for complementing their portfolio of fundamental data alongside the data of a main supplier (being normally Thomson, Reuters or Compustat - the latter for US data), and not as a substitute to the offerings of Thomson and Reuters. Almost unanimously, the parties' products are considered as the most complete and deep Fundamentals databases in the marketplace (107). This is consistent with the parties' own view, as set out in internal documents (108).

337. In addition it is worth noting that, unlike several other players (Compustat, Capital IQ, Bloomberg, FactSet, Hemscott/Ipreo, Edgar Online, Tokyokeizai), WorldVest/Base is not even mentioned by Reuters in an internal document in which Reuters carries out a thorough review of the main competitors in the field of Fundamentals (109). Edgar Online

338. The parties claim that "EDGAR Online, while historically strong in North American filings and Fundamentals, has in recent years earned considerable visibility through its pioneering work on the XBRL standard, both in developing the consortium's taxonomies as well as releasing XBRL-ised products such as I-Metrix to the market. Its work with IRIS to make XBRL Fundamentals available for China's Shanghai and Shenzhen exchanges has brought it considerable attention, and it is well positioned to take advantage of interactive data as more markets move to adopt the standard, such as Japan, Korea and Singapore in 2008, as well as several European markets where XBRL pilot projects are underway. The Parties do not yet have significant product capability in this area" (110)

339. In the market investigation, customers (especially those having expressed concerns in respect of fundamental data) (111) do not appear to consider Edgar Online as a valid alternative to the merging parties. No competitors having expressed concerns in the field of Fundamentals (112) either mention, or rank Edgar Online along the same line as the parties as full or close substitute for Fundamentals content. It is true that Edgar Online could elect to partner with some other smaller compilers and/or redistributors, but given the large distance between these various providers and the parties, this move would not allow Edgar Online to offer a database/product capable of matching the offering of Thomson and/or Reuters in the foreseeable future without major investments.

340. Edgard Online provides access inter alia to SEC filings and other corporate fundamental data in the form of desktop and of datafeed products. Competitor 89 acknowledges that Edgard Online has a strong coverage of some geographical areas (especially the US and China), nevertheless its offering cannot be compared to that of the parties since the parties have access to "far more global breadth in coverage" as "they provide access to data from over 40 countries around the world". In particular in order to compete with the parties' breadth of coverage, Edgard Online would need to invest significant amounts of money, partner or acquire third party companies (113). In addition, the notifying parties do not consider Edgar Online a particular threat to their business, since in a Reuters' internal document, Edgar Online is credited with a market share of [0-10]*% (114).

V.B.3.7. Market shares

341. The parties have provided market shares on Fundamentals, in response to a Commission's request for information, at global and EEA level, at compilers level and separately for "distribution" sales only. Based on such shares of sales, they contend that the impact of the merger on Fundamentals would not be significant, both at the Worldwide and the EEA level, and both as regards compilers and at the distribution level.

342. As regards compilers, the parties' computation of market shares would be the following on a worldwide basis: Thomson&Reuters [10-20]*%, Bloomberg [10-20]*%, Compustat [10-20]*%, Capital IQ [0-10]*%. The parties indicate that S&P Compustat is the incumbent and the market leader in Fundamentals. Similarly, at the EEA level, the parties' computation of market shares would be the following: Thomson&Reuters [10- 20]*%, Bloomberg [10-20]*%, Compustat [10-20]*%, Capital IQ [0-10]*%. Bloomberg would then be the EEA market leader.

<table>

343. Based on the above shares, the parties contend that their overlap would not be significant, and should not cause competitive concerns due to the large number of global, regional and national vendors of Fundamentals data which ensures that the supply of Fundamentals in Europe and globally will remain competitive after the transaction.

V.B.3.8. Considerations on the parties' market shares

344. [The market shares were calculated by the notifying parties under a number of reasonable assumptions to respond to the Commission requests. However, the Commission decided to use this data with great care, due to the unsatisfactory results of the exercise.]* In particular, Bloomberg market share is nearly at par with the parties, but the investigation as well as the parties' internal documents clearly give the indication that Bloomberg is not considered as a strong player in the off-trading floor area. Bloomberg use a closed-off model, where all content sets and functionalities are bundled together into the Bloomberg terminal solution, while it does not make any content sets accessible to its competitors (via royalty) or to the competitors' customers (via a sell-through solution). [...]*

345. As regards the notifying parties, in an internal document, Reuters estimates its revenues for Fundamentals as being around [10-100]* million dollars, while it has declared to the Commission that its global revenues from Fundamentals were approximately [10-100]* million dollar (115) (EURO [10-100]* million) in 2006. In addition, as regards sales through redistributors, the notifying parties' shares appear to be significantly understated. In particular, in the same internal document from Reuters, FactSet (who is a redistributor of Thomson's and Reuters' Fundamentals) is credited a revenue from Fundamentals of approximately [10-100]* million dollars, which is a much higher figure than the entire sales from redistribution that Reuters has declared to the Commission as regards Fundamentals ([10-100]* million dollar in 2006, of which approximately [0-10]* million dollar through FactSet) (116). In this connection, it's worth mentioning that FactSet is the largest "sell-through" distributor of fundamental data for Thomson and Reuters.

346. Factset declared that a significant percentage of its customers source Fundamentals from Reuters through Factset's platform (117). Applying the same split to the figure ([10-100]* million dollar) contained in the mentioned Reuters' internal documents as being the revenue to be attributed to FactSet, the redistribution sales of Reuters through Factset would be around [10-100]* million dollar, that is to say, more than the total distribution sales of Reuters as declared to the Commission ([10-100]* million dollar) and more than double the revenue declared by Reuters to the Commission as "sell-through" revenues via FactSet ([0-10]* million dollars, see Annex 13 to the response by the parties dated 26 October 2007 to a Commission's request for information).

347. Based on the above, it seems that the market share picture as supplied by the parties is not the right angle under which to assess the impact of the present transaction.

V.B.3.8.1. Market shares in Internal Document

348. In internal documents (118), Thomson estimates Compustat's revenues as half the revenues provided by the parties to the Commission and Bloomberg's revenues in Fundamentals as around 20% of the revenues indicated in their submissions to the Commission (119). A Reuter's internal document (120) shows a completely different picture of the market as seen by the parties in the representations to the Commission. In particular, the following market shares are discussed at the global level. It has to be noted that these shares include both compilers and redistributors:

Table 14

Fundamentals Market Shares USD m %

Reuters [10-100]* [10-20]*%

Thomson [10-100]* [10-20]*%

Compustat [10-100]* [10-20]*%

CapitalIQ [10-100]* [0-10]*%

FactSet [10-100]* [0-10]*%

Hemscott [10-100]* [0-10]*%

Edgar Online [10-100]* [0-10]*%

Bloomberg [10-100]* [10-20]*%

Tokyo Keizai [0-10]* [0-10]*%

Others (Mergernt, BVD, etc) [10-100]* [0-10]*%

TOTAL [400-500]* 100%

Source: "Reuters Fundamentals Business Proposal" April 2007, Internal document No RTRSE-CLAA-00020356

349. According to the above table, Reuters has [10-20]*%, Thomson [10-20]*% (in another part of the same document, Thomson is attributed [20-30]*%), Compustat [10-20]*%, Capital IQ [0-10]*%, Factset [0-10]*%, Bloomberg [10-20]*%. These shares significantly understate the parties' market shares as Factset is a redistributor of Fundamentals (mainly of Reuters and Thomson Fundamentals) and does not have its own offering, while Bloomberg has only a one-size-catch-all solution encompassing all content, and data sets and analytics.

350. Also respondents to the Commission's inquiry indicate an order of magnitude for the market shares of the parties as regards Fundamentals which is much higher than what they have provided the Commission with and come to much higher market share Estimates for the parties (121). The table 15 below provides an example of such representations:

<table>

V.B.3.9. Thomson and Reuters are the closest substitutes

351. Based on the above considerations, a more qualitative analysis appears to fit the case at stake. On this basis, the market investigation showed that there are generally three providers of Fundamentals (in the global format which is required by the global customers) which are also supplied via third parties, Thomson, Reuters and S&P Compustat. As regards S&P Capital IQ, its policy vis-à-vis redistribution is not clearcut, in the sense that it would appear to the Commission that, similar to a few other players in the marketplace, Capital IQ may or may not elect to redistribute its content on a case-by-case basis.

352. Respondents to our investigation indicated that the offering of Compustat database would not be comparable to that of Thomson and Reuters, since it is US-only focussed. Capital IQ, as a compiler of global Fundamentals, appears to be an emerging competitor but at significant distance from the big three and certainly not a suitable substitute to the parties' databases.

353. This is confirmed by a Reuters's internal document and by the extensive analysis carried out before. For customers, Compustat, Bloomberg and Capital IQ would not be an alternative to Thomson&Reuters, but rather a complement, given the different (lower) breadth and depth of these alternative databases. Bloomberg does not re-distribute its database and in a Reuters' internal document, it was clearly stated that Bloomberg Fundamentals are rather suitable for traders than for analysts (122). The products of the parties would in particular be the only ones which fit the requirements of those customers who are active in quantitative Investment Management and who need "global" Fundamentals, with depth and back history (especially the buy-side customers such as hedge funds and investment managers).

V.B.3.10. Adverse horizontal effects (non-coordinated)

354. Based on the foregoing, the merger will likely lead to a reduction from three to two of the full-blown providers of Fundamentals which generally also supply redistributors. Taking into account the fact that the focus of Compustat's offering is to the US and North America, the merged entity would be the only supplier of Fundamentals with a global coverage and deep history available in the marketplace.

355. This is likely to determine an outright reduction of choice for customers since it is highly likely that the two databases will be merged and that the two independent product lines will be discontinued in the foreseeable future. Being unconstrained, Thomson/Reuters will be in a position to increase the outright retail license fees for Fundamentals databases to the detriment of end-customers. Since it is time and capitalconsuming to reconstruct such databases, entry would not be easy and timely (see further). Such adverse horizontal effects would impact both direct customers of the notifying parties and their redistributors.

356. The notifying parties insist that the merged entity will be far from being dominant. Allegedly, the S&P group (that is to say, S&P/Capital IQ and Compustat) is by far the leading player worldwide and will remain so post-merger, other compilers would be present and Bloomberg would have a share approximately equivalent to that of the parties (123). There would also be several smaller players which would compete effectively (124). In addition, the notifying parties stated that other vendors collect Fundamentals data regionally or nationally and that Exchanges are also beginning to supply company Fundamentals data to the market. According to the notifying parties, therefore, "the Parties would not have the ability to increase prices to end customers: end-users can obtain Fundamentals by plenty of other established providers which ensure effective competition".

357. Despite the contentions set out by the notifying parties, the investigation has clearly indicated that the merger combines the leader (Thomson) and the close contender (Reuters) in the global fundamental space, Compustat being particularly weak in non- US Fundamentals. The investigation did show that other compilers said to provide global coverage cannot match the offerings of the parties (not even that of Compustat) and that all these players (including Compustat) are not suitable to satisfy the full-scope requirements of global customers to the same level as the notifying parties and, hence, are not credible competitive constraints vis-à-vis the notifying parties.

358. Thomson/Reuters have also stated that "the Parties would not have a de facto monopoly at "wholesale" level: redistributors can obtain Fundamentals content from global suppliers or they can mix and match to meet their requirements from a vast number of Fundamentals suppliers who are independent of Thomson and Reuters" (125). Despite these contentions, the investigation demonstrates that the merger brings about the two most credible providers of fundamental data with global coverage and the closest substitutes in this market also in respect of content which is redistributed (whether via "sellthrough" or "royalty") downstream by third parties to their own customers. In particular, post-merger redistributors cannot procure a comparable fundamental content (than the parties) from any other compiler in the marketplace for the same reasons as applicable to end-customers as illustrated above. In this situation, the merged entity will have the ability and the incentive to undertake price increases (and price discriminate) in particular vis-à-vis those customers who access the parties' fundamental content through desktop platforms (accesses) of competitors.

V.B.3.11. Barriers to entry/expansion are high

359. The parties contend that barriers to entry and expansion are low and diminishing in Fundamentals data and entry is sufficiently easy and quick to ensure effective competition in general and to protect redistributors in particular. Fundamentals products would be highly commoditised while barriers to entry have virtually disappeared with the emergence of electronic availability of Fundamentals data, adoption of electronic coding standards such as Extensible Business Reporting Language (XBRL), the convergence of international accounting standards, International Financial Reporting Standards (IFRS) (126), and the internet as a communication medium both to receive primary materials with Fundamentals data and to distribute fundamental products (127).

360. These reporting standards (in conjunction with the common taxonomies imposed by the adoption of XBRL) would be creating consistency across industries and regions and rapidly diminishing the difference between "as reported" and "standardised" Fundamentals. As an example of successful entry, the parties mention the case of the company Simply Stocks (which was subsequently acquired by Capital IQ). The notifying parties estimate that S&P/Capital IQ built a robust Fundamentals offering in roughly 18 months to two years. Now, S&P/Capital IQ is considered by the parties as one of the strongest US suppliers of Fundamentals data and has expanded into international markets, taking market share from established providers across the board (including from Thomson and Reuters). Moreover, the emergence of highly skilled, low cost personnel, notably in India, has reduced the costs even when manual intervention is required.

361. Contrary to the notifying parties' view, the Commission is of the opinion that barriers to entry are significant in this market. A number of years and substantial investment are needed for reconstructing a database comparable to the one of Thomson or Reuters and to be able to offer a credible product which would be acceptable by the market. The hurdles come, first, from the need to collect fundamental data with a global coverage and second, to collect fundamental data going back in time several years.

362. The added value of a Fundamentals database also lies in the fact that financial accounts are made comparable by vendors of such databases by standardisation, that is to say, by making accounts of companies reporting under different accounting rules comparable and responding to codification systems relevant to users. As mentioned before, the significance of standardized data is in the fact that users can avoid manipulation of data to conform them to their business calculation needs, while "as reported" data may be still necessary in order to allow users to back-track the original classification of a certain items in the company's accounts and to further drill down into the company's specific accounting methodology.

363. High barriers to entry exist for constructing a database rich in historical Fundamentals: since commoditization of data (XBRL filings) and standardisation of reporting techniques (IFRS, GAAP) are rather recent and are not largely extended practices. Anyhow, these facilitating elements would not allow going back in history to any sufficient extent. Building a database with historical Fundamentals would therefore have to be done manually, while aggregating data from tens of thousands of companies worldwide over a period of at a minimum the last ten years would be very resourceintensive and costly. In addition, simply tracking "as reported" data (however difficult) does not warrant a complete product, since standardisation and adjustments are needed to build up a thorough database (where normally "as reported" and "standardised data" are offered at the same time) able to rival the parties' ones. This also involves significant and particularly skilled manpower.

364. If these operations (collection and standardisation) have to be carried out for tens thousands of companies, for the entire spectrum of economic sector and for the widest possible geographical scope, it is self-evident that - unless the business is already in place and established - this process requires several years and significant investment and hence represents a huge barrier to entry.

365. Respondents in the market investigation have confirmed that it "is impossible to assemble historical versions of offerings that will be satisfactory to many users because the raw materials needed to create these databases are simply unavailable at any price. This is the case with international fundamental data, as well as historical earnings Estimates." (Competitor 21). Competitor 31 estimated that it would take five years and cost EUR 60 million to build a competitive offering in this field (128).

366. The notifying parties mention Capital IQ as the example that barriers to entry are low. Quite apart from the fact that Capital IQ cannot necessarily be considered a totally standalone company, independent of its affiliate Compustat (both being controlled by S&P), the investigation and the internal documents of the notifying parties (also in respect of Capital IQ) have shown that despite the investment and high quality technology, Capital IQ did not reach the same level as the notifying parties and still would need time and investment to fill-in the outstanding significant gaps as regards history which characterise its fundamental offering. For that very reason the parties have stated that [Capital IQ] would still take quite some time before it can challenge [Thomson] in these markets (129)"

367. Unlike Capital IQ, other players are not even in a position to approach even at some distance any time soon the notifying parties in the Fundamentals area. In this respect it needs being reiterated that that, even when a supplier could believe to have created and tested a theoretically reliable product, a major hurdle in the marketplace is the reputation and credibility that any product has to gain vis-à-vis highly demanding and non-price sensitive customers. This creates a further very significant time-consuming hurdle to bypass before being a competitive constraint into the market.

368. For the reasons above, it can be concluded that barriers to entry/expansion are significant in the marketplace and will prevent any third party to timely offsetting the adverse effects brought about by the merger.

V.B.3.12. No timely repositioning of competitors

369. In the market situation as described in the preceding paragraphs, the other players (including S&P and Capital IQ) do not have the possibility of successfully repositioning in a timely fashion: (i) filling the gaps in their respective offerings may not be possible at all (e.g. outright unavailability of deep historical data from any sources, coverage issues) in the short term (ii) being accepted by the market as credible suppliers will in any event entail time-consuming processes and massive investment before being able to overcome the threshold of credibility and market acceptance required in the industry especially for global fundamental content. Therefore, it can be concluded that timely repositioning for existing players will not be possible and will not offset the adverse effects brought about by the merger.

V.B.3.13. Adverse vertical effects

370. Thomson/Reuters will probably take advantage of its new position in the Fundamentals market to foreclose its competitors by either increasing the price for this content set distributed via redistributors (royalty based and sell-through), or by simply stopping providing its competitors with such data. Thomson/Reuters should have both the ability and the incentive to develop such a strategy.

V.B.3.13.1. Ability to foreclose

371. Third party redistributors of Fundamentals data (integrated into their own desktop products) are highly dependent on the notifying parties for fundamental content having global coverage, whether they act as "sell-through" or "royalty" redistributors. In fact their desktop customers needs of global Fundamentals can only be satisfied if they (the customers/redistributors) are able to receive through the redistributor' platform that type of content procured through/from Thomson and/or Reuters. No other competitor can provide a comparable product in terms of depth and breadth.

372. The notifying parties indicated that they would not have the ability to foreclosure for those redistributors whose current contracts are unexpired (130). However, such contracts generally cover a short period [...]* and most renegotiations are already under way[...]*. Therefore, the notifying parties will have the ability to foreclose competitors in this market.

V.B.3.13.2. Incentive to foreclose

373. The notifying parties also submitted that they would have no incentive to foreclose redistributors as a result of the merger, as redistributors could simply switch to any of the numerous other providers of Fundamentals data that already make their content available for redistribution. Furthermore, they submitted that foreclosure would not be a profitable strategy, and would result only in a net loss of revenue for the notifying parties: In case of a sell-through distributor, there would be no assurance that the endcustomer would continue to purchase its Fundamentals from the merged entity, even post-foreclosure. In the case of a redistributor using a royalty based arrangement the net result would be a loss of revenue with no corresponding upside. In addition, foreclosure would be a high risk strategy, given that many customer relationships span a variety of different product lines. Customers could discipline the notifying parties' behaviour by switching their business in other areas of financial information products. Some redistributors are suppliers to Thomson and/or Reuters and could therefore retaliate in other product areas as well (131).

374. However, as it stems from internal documents (132), already pre-merger the notifying parties had a clear incentive to reduce their dependency on redistributors to reach customers. This incentive will be amplified post-merger, when, given the lack of any sufficient competition constraint to the merging parties in the global fundamental area, the notifying parties will also have the ability to implement such a strategy successfully. The merger will in particular give the notifying parties the ability and the incentive to successfully offering to customers a formidable wealth of products and content sets in a one-stop-shop solution, displacing competitors by pricing the standalone content offer (when accessible through competitors' desktop) at par with their bundled offer (which will include content and analytics).

375. Thomson and Reuters generated in 2006 revenue of EUR [10-100]* million worldwide and EUR [0-10]* million in the EEA via the sales of Fundamentals data through redistributors (royalty based and sell-through). Should Thomson/Reuter lose this revenue by foreclosing its competing redistributors, the newly formed company would need to compensate this loss by capturing customers from its competitors generating at least the same revenue. As such customers would switch to Thomson/Reuters to purchase desktop products, generating substantially more revenues than sales of content sets via redistributors. Based on the data provided by the parties, a foreclosure strategy would be profitable (in terms of revenue stream) with the displacement of a minimum number of (ex-redistributor's) customers (133).

376. Therefore, it is likely that the merger entity will either undertake price increases for its Fundamentals content when this is to be made available through the platform of third parties redistributors, or stop selling Fundamentals data to/via redistributors, without being disciplined by any countervailing power in the market.

V.B.3.13.3. Conclusion on adverse vertical effects

377. Based on the above, it can therefore be concluded that Thomson/Reuters will have the ability and the incentive to foreclose competitors in the Fundamentals market.

378. In addition, such vertical foreclosure would be able to aggravate the horizontal effects raised by the concentration as the increase in price for content sold to/through redistributors and/or the degradation of the quality of the desktop products (containing Fundamentals) distributed by the redistributors would ultimately increase the likelihood that Thomson/Reuters can profitably increase its prices for its own customers. As already indicated, no other competitor provides a product equivalent to Thomson or Reuters', and as a result these customers would not have to possibility to switch to a comparable provider, in terms of price or in terms of content quality.

V.B.3.14. Conclusion on Fundamentals

379. The market investigation carried out has shown that, despite the contentions of the parties, the merger will bring about adverse effects on competition as regards Fundamentals. The investigation has indicated that Thomson and Reuters currently possess the most credible databases for Fundamentals, with global coverage encompassing both the US and EEA data, with a depth of history going back 15-20 years. This type of Fundamentals databases is particularly important for customers (essentially on the buy-side but also to a more limited extent on the sell-side) who need to develop quantitative trading based on sound modelling and on analysis of historical trends.

380. As regards the horizontal effects, the merger is likely to have significant adverse effects. The merger would eliminate rivalry between the two leading suppliers of Fundamentals with global coverage in breadth (US and EEA-wide coverage at the least) and depth, plus with sufficient history. In the same vein, the merger will likely entail rationalising of product lines and hence is likely to reduce the choice currently available to customers of fundamental products.

381. In addition, the newly formed company would likely vertically foreclose its competitors by increasing prices for Fundamentals distributed via redistributors or by limiting the access to its Fundamentals offering to its own customers (those customers purchasing a complete desktop solution from Thomson/Reuters).

382. It can therefore be concluded from the above that the concentration will lead to a significant impediment of effective competition as regards Fundamentals.

V.B.4. Times series/Economic data (TS/E)

V.B.4.1. Introduction

383. Economic data largely represent macroeconomic variables, such as GDP, unemployment, money supply, balance of trade figures and inflation rates. Such information is collected as individual data points and tracked over a period of years (time series). Their ultimate sources are typically governmental bodies (national and supranational statistical offices or central banks), but such series also include proprietary data of, e.g., research institutes.

V.B.4.2. Market shares calculated by the parties

384. On the basis of revenue estimates, the parties attempted to calculate their market shares and that of their competitors for the supply of TS/E.

<table>

385. While the notifying parties indicate that their estimates of the revenues achieved from TS/E is somehow imprecise due to the difficulty to precisely isolate TS/E turnover from package offerings, many competitors who responded to the Commission's market investigation had also difficulty to provide precise turnover figures on TS/E, though few competitors were able to give such information. For the data which the Commission was able to collect, it appears that the notifying parties estimated higher revenues than the competitors actually provided hence their market share would be lower compared to the notifying parties' market share. A further difficulty was to determine precisely the total market volume, as TS/E offerings are often coupled with or integrated in other data products for which the vendor charges a global price but rarely a price on an individual basis.

386. For instance, Bloomberg offers all its data products within the Bloomberg Professional Services, a terminal-based solution, for which the end user pays a single fee per month per terminal irrespective of data use. Hence, Bloomberg was unable to provide any turnover information isolated for TS/E.

387. An internal document from Reuters (134) showed that Reuters estimates its own market share for TS/E to about [20-30]*%. In the same document, Reuters also provides market shares of its competitors, namely for Thomson with a [30-40]*% market share, Global Insight about [10-20]*%, Feri about [0-10]*% and all others about [0-10]*%. Also the underlying market volume, which the document mentions, is significantly different.

388. It therefore can be concluded that in order to provide an accurate picture of any party's position in the market the estimated market shares would have to be complemented by other criteria.

V.B.4.3. Comparison of main suppliers' offerings

389. Most customers who responded to the market investigation judge the credibility of any TS/E offer on the basis of the underlying breadth and depth of the content set.

390. The parties provided a comparison of Thomson and Reuters TS/E products with those of their competitors in which they outlined in detail the breadth of content, the depth of content and the geographic focus both in terms of the content set and the focus of activity of the company.

<table>

391. As the details about Thomson and Reuters suggest, both companies provide the greatest depth and breadth of economic data and also provide a wide geographic coverage. Reuters has by far the largest data set counting about [2-3]* million compiled data which cover almost equally US and European markets.Thomson's offer is similar to Reuters in its depth, breadth and geographic coverage, though Thomson has less data.

392. Bloomberg, for instance, has by far less data and they date back only 10 to 15 years. Moreover, Bloomberg does not re-distribute their TS/E data. Furthermore, Bloomberg only offers TS/E data as part of its integral terminal solution. As within the terminal solution Bloomberg does not provide a price for the individual TS/E part, an end user who wants only to use TS/E would have to buy the complete package. Even if an end user goes for the terminal solution, the Bloomberg terminal is pre-dominantly used in dedicated businesses e.g. Fixed Income trade, a user community that is less interested in TS/E for their daily work. Hence, Bloomberg appears to be no viable competitive constraint to the parties. Such conclusion is also supported by the market investigation in which competitors and customers made no mention of Bloomberg to be a substitute for the parties' TS/E products.

393. As to Global Insight, the company has a TS/E content set that is more US centric, but also offers data from other world regions. Global Insight sees themselves somehow in competition with Thomson's Datastream and Reuters' Ecowin but hardly with any other competitor. However, Global Insight recognises its dependency on re-distribution via other desktop providers, even more so, since alternative desktop providers tend to favour their own TS/E offer. Thus, Global Insight would be a less strong competitor compared to Thomson/Reuters.

394. Haver Analytics is a very US centric vendor of compiled and re-distributed TS/E data. It provides its services mainly to US customers. With respect to Haver's European customers, Haver's data coverage suggests that these customers tend to focus on US information only and may use Haver as back-up solution on US data, a feature that is widely used in the financial services industry. However, Haver Analytics would not be a viable content set provider on a global basis.

395. CEIC provides its data through re-distribution and its data set covering in particular Europe is rather small as compared to Thomson and Reuters.

396. Economy.com, belonging to Moody's, has a large collection of data however only providing almost exclusively US data. Like Haver, their sales focus lies in the US.

397. Furthermore, the parties also mentioned Factset as competitor. However, Factset is largely a re-distributor of the TS/E content set, mainly from Reuters, Global Insight and others, while they have only a small content set of their own. If Reuters Ecowin product was to disappear from the market, Factset would have to find alternatives and possibly as a first step to re-distribute Global Insight's data set. On this basis, it can be concluded that Factset can hardly be considered a viable competitor post-merger as the company depends largely on re-distribution mostly from the parties.

398. Finally, other competitors are relatively small in terms of depth and breadth of their content sets and often focus on particular niches.

V.B.4.4. Thomson and Reuters are the closest substitutes.

399. The market investigation revealed that many competitors and customers source their TS/E needs from either Thomson or Reuters while they see only few alternatives. For many large customers, Thomson and Reuters are almost equal in their offer. Some customers also mention Global Insight rather frequently whereas other TS/E were referred to as being for backup or providing better local and complementary coverage rather than a full-scope alternative to Thomson or Reuters.

400. Also the parties see themselves to be close competitors, as internal documents from Thomson and Reuters confirm (136). Furthermore, the market investigation revealed that over the past years Reuters TS/E product has come closer to Thomson's TS/E and only for few years both products have been closely matching offers in terms of depth and breadth of data.

V.B.4.5. Adverse horizontal effects (non coordinated)

401. Many customers source TS/E from Thomson and Reuters, and often they use even both Datastream and Ecowin in parallel. Depending on the customer, they source in aggregate around 50% to 80% of their needs from Thomson and Reuters hence they heavily depend on them. After the merger, it appears likely that either Thomson's or Reuters' product would become redundant, as the parties provided no evidence or any incentive argument from which it can be concluded that they would continue to run both products. Only few customers consider having viable alternatives, and they see at global level possibly Global Insight only. However, Reuters' internal documents seem to interpret Global Insight to be less of a competitive threat than Thomson (137).

402. Smaller TS/E vendors exist but it appears doubtful whether they would be able to enhance quickly enough their offer in terms of depth and breadth similar to the second not such a close competitor in order to be considered an equal player to Thomson/Reuters or Global Insight.

403. On the basis of the above, the merger would leave on this market only two competitors. Smaller competitors would represent no viable alternative as they are likely to continue as niche players. Furthermore, post-merger it would eliminate the closest competitor in the market for TS/E.

404. In such scenario, it appears likely that prices for TS/E data increase. Internal documents of Reuters confirm that in the past, when Reuters fiercely competed with Thomson, they offered Ecowin to its customers with lower prices or in a bundle as part of Reuters' desktop offer. Post-merger, such competition would come to an end and hence prices would likely move up. Furthermore, customers moving away from Thomson/Reuters would likely go only to Global Insight. However, Global Insight would have hardly any incentive to fiercely compete with Thomson/Reuters knowing that potential competitors would need a long time to become a creditable alternative. Such adverse horizontal effects would impact both direct customers of the notifying parties and redistributors.

405. Due to the fact that competition takes place at different levels, that is to say, at desktop products level and upstream, fierce competition from Global Insight appears unlikely at the level of the desktop TS/E as Thomson/Reuters largely own the desktops sold to their customers whereas Global Insight depends on other financial information providers' desktops (that is to say, redistributors) to reach its customers. As to the upstream level, distribution of TS/E data via datafeed represents smaller turnover compared to desktop distribution and hence it appears unlikely that it would be a competitive threat to the parties' desktop offer.

V.B.4.6. Barriers to entry/expansion are high

406. The notifying parties argue that entry barriers are low in the market for TS/E. To support their argument, the notifying parties presented a list of publicly available sources from which to receive macroeconomic data. The notifying parties appear to assume that any competitor who wishes to enter or expand in TS/E could obtain such data and to build a competitive offer.

407. Many customers, who responded to the market investigation, typically work with TS/E in normalised or compiled format but hardly with raw data. The provision of raw data from multiple sources, by which the end user would have to go through a multitude of individual websites, represents no viable option for customers. The notifying parties further argue that end users would be able to by-pass the "middleman" that is to say, the data compilers. However, as customers need to work with compiled data and raw data represent no alternative, as they don't consider becoming a data compiler themselves they need to continue sourcing from compilers.

408. Building-up from raw data a compiled content set requires many years of effort and resources. The market investigation indicates that at least two to three years appear necessary to construct compiled TS/E data and substantial investments are needed in personal and infrastructure to collect a wide scope of raw data and to normalise them into meaningful compiled data. By reference to Reuters' internal documents, the history of Reuters' Ecowin perfectly illustrates and confirms that building-up a TS/E offer is a resource intensive operation, for which even Reuters as a large player needed many years and substantial investments to get Ecowin to a level to effectively compete in the market place.

409. The market investigation furthermore indicated that a reputation on high quality data delivery is vital for a vendor. Customers tend to choose only data providers from which they expect complete and reliable content sets. A vendor, which has a reduced content set or lacking up-to-the-standard compiled data, is no option for customers in the financial services industry. However, the time to build-up credibility that is to say, to produce a complete data set which potential customers consider a viable alternative to Thomson/Reuters' complete content set can take many years, about two to three years. It therefore appears doubtful whether smaller competitors would have the economic incentive to invest in such work, in particular in view of the fact that during the build-up period no revenue stream can be expected from selling TS/E data out of a reduced content set as customers appear to decide for a complete offer or no offer at all.

410. The parties provided their estimate on how much investment is needed to enter the TS/E space. They calculated as basic requirement roughly EUR [0-10]* million for content collection and operation of the database in the start-up period which they expect to last about one year. Continuing operating costs expect the parties to be much lower. However, some competitors who responded to the market investigation expressed much higher investment costs, that is to say, in the range between EUR 10 to 30 million. They also expect the start-up period to last for at least two to three years to construct a complete content set with which they would be able entering into competition with Thomson/Reuters. This longer period would be needed to make the content set up to a credible level of standard which potential customers expect from a viable vendor of TS/E. It appears that potential vendors of TS/E, who have had experience with constructing similar content sets, view entry into the TS/E space as cost intensive and time consuming and as such representing significant entry barriers.

V.B.4.7. Adverse vertical effects

411. Thomson/Reuters will likely take advantage of its new position in the TS/E market to foreclose its competitors by either increasing the price for this content set distributed via redistributors, or by simply stopping providing its competitors with such data. Thomson/Reuters will have both the ability and the incentive to develop such a strategy.

V.B.4.7.1. Ability to foreclose

412. The notifying parties are both active in re-distribution of TS/E content sets for Datastream and Ecowin products. Although their revenues from such activity is smaller than from distribution of TS/E data via their desktop products, the distribution of Thomson and Reuters data is a vital add-on for any re-distributor's product offer. For instance, Factset offers within its desktop product TS/E data from Reuters, a fact with which for Factset promotes the sales of its desktop as the market perceive Reuters' TS/E data to be robust and credible. Along the same line that is probably true for Thomson re-distribution partners.

413. As other competitors, including Global Insight, would not be in a position to propose an equivalent product on the market, the notifying parties will have the ability to foreclose competitors in this market.

V.B.4.7.2. Incentive to foreclose

414. Post-merger, the incentive to re-distribute would fall apart. Thomson and Reuters are no longer in competition and customers would perceive the combined TS/E product to be the best offer, so it would make the Thomson/Reuters' desktop more attractive. If the parties continue to re-distribute their combined TS/E data via competitors who also have desktop offers, it would somehow leverage the parties' desktop selling argument. Whereas if the parties begin to offer their TS/E solely on their desktop products, competitors would have difficulty to convince customers to choose their desktop offers as the most reliable TS/E content set is not available.

415. Thomson and Reuters generated in 2006 a revenue of EUR [0-10]* million worldwide and EUR [0-10]* million in the EEA via the sales of Economic Time Series data through redistributors (royalty based and sell-through). Should Thomson/Reuter lose this revenue by foreclosing its competing redistributors, the newly formed company would need to compensate this loss by capturing customers from its competitors generating at least the same revenue. As such customers would switch to Thomson/Reuters to purchase desktop products, and they would generate substantially more revenues than sales of content sets via redistributors. Based on the data provided by the parties, a foreclosure strategy would be profitable (in terms of revenue stream) with the displacement of a minimum number of (ex-redistributor's) customers (138).

416. The parties' argument that competitors could easily switch to other data suppliers if they were to end re-distribution appears doubtful since alternative supply is rare, as discussed above. Alternative desktop providers anyway have lower penetration in the desktop space which makes them less attractive as re-distributors of TS/E data from a compilers' perspective.

V.B.4.7.3. Conclusion on adverse vertical effects

417. Based on the above, it can therefore be concluded that Thomson/Reuters will have the ability and the incentive to foreclose competitors in the TS/E market.

418. In addition, such vertical foreclosure would be able to aggravate the horizontal effects raised by the concentration as the increase in price for content sold to/through redistributors and/or the degradation of the quality of the desktop products (containing TS/E) distributed by the redistributors would ultimately increase the likelihood that Thomson/Reuters can profitably increase its prices for its own customers. As already indicated, no other competitor provides a product equivalent to Thomson or Reuters', and as a result these customers would not have to possibility to switch to a comparable provider, in terms of price or in terms of content quality.

V.B.4.8. Conclusion on TS/E

419. It can be concluded from the above that the proposed transaction will significantly impede effective competition in the common market. Post merger, one close competitor would disappear from the market for TS/E content sets and Thomson/Reuters would have the ability and incentive to increase its prices. Customers would be left with limited choices and alternatives which are less credible and have more limited scope in terms of depth and breadth of TS/E data. Entry into this market appears to be not timely and less likely.

420. In addition, the newly formed company would likely vertically foreclose its competitors by increasing prices for TS/E distributed via redistributors (royalty based and sell-through) or by limiting the access to its Economic Time Series offering to its own customers (those customers purchasing a complete desktop solution from Thomson/Reuters).

421. It can therefore be concluded from the above that the concentration will lead to a significant impediment of effective competition as regards TS/E.

V.B.5. Vertical foreclosure in Aftermarket Research, Earning Estimates, Fundamentals and TS/E

422. As indicated in the precedent chapters, Thomson and Reuters would have both the ability and the incentive to vertically foreclose their competitors in the Aftermarket Research, Earning Estimates, Fundamentals and TS/E. They could do so either by increasing the prices for these content sets distributed via redistributors (royalty based and/or sell through) or by stopping the sales of these content sets to their redistributors.

V.B.5.1. Ability and incentive to foreclose

423. Would such a strategy be implemented, the desktop offerings of a large number of competitors would be seriously affected, notably in the case Thomson/Reuters decide to foreclose its competitors with regard to the four content sets together. Within the market investigation it has been observed that among those competitors offering desktop solutions, [70-80]*% redistribute Thomson or Reuters content sets, although Thomson and Reuters might not be the only providers of such content sets. [60-70]*% of the competitors specifically redistribute Fundamentals, Estimates, Broker Research or Economic Time Series from Thomson or Reuters. It should be noted that the largest competitors, but Bloomberg, all redistribute one or several of these four specific content sets from the notifying parties.

424. In addition, it should be noted that for other content sets, such as Ownership, Deals, News, other time series, etc, Thomson/Reuters will compile internally its content sets and will not depend on other providers. Thomson/Reuters will broaden its offer and the position of the merged entity will be such that it will be less dependent on third parties to collect data. As such the merged entity will not fear retaliation from competitors, in the sense that no competitor will have the possibility to stop providing to Thomson/Reuters a content set that the notifying parties cannot compile internally.

425. As mentioned in various chapters of this decisions, the ability and incentive to foreclose competitors will particularly apply to Fundamentals, Estimates, Broker Research and Economic Time Series, but not to other content sets, where the merger has no adverse effects.

426. Customers are financial professionals in need for accurate, reliable and reputable information sources. The market investigation indicated that should their desktop providers stop including specific content sets, or should they seriously degrade the quality of specific content sets, provided within their desktop solutions, customers would envisage finding another desktop provider. Although, it is not possible to define a threshold below which the degradation of the content sets included in a desktop would trigger a change of provider, it is likely that without the best available content sets (as available pre-merger), a number of competitors would face serious problems for retaining their customers.

427. Customers changing desktop provider and willing to access to desktops providing complete and reputable content sets could principally switch to Thomson/Reuters or alternatively to Bloomberg. Bloomberg is however not as reputable as Thomson/Reuters in the research and asset management area (e.g. its estimate offer is recent and lacks historical information, Bloomberg does not offer Aftermarket Research), and in addition is viewed as particularly expensive, notably because it principally commercialises its Bloomberg Terminal which includes all financial products.

V.B.5.2. Profitability of foreclosure

428. A decision from Thomson/Reuters to stop commercialising its content sets via redistributors would have to be profitable. Such a move would concern the majority of downstream competitors and notably the largest ones, with the exception of Bloomberg. Bloomberg however only has a limited footprint in the research and asset management area, where these content sets are most needed.

429. Thomson/Reuters generated in 2006 a worldwide revenue of EUR [10-100]* million in 2006 (EEA: EUR [10-100]* million) via the redistribution (royalty based and sell-through) of its Fundamentals, Estimates, Aftermarket Broker Research and Economic Time Series content sets. To compensate for the loss of redistribution revenues for these four content sets, Thomson/Reuters should increase its revenues in the Research and Asset Management area of a similar amount (desktop including these content sets are primarily sold in this area) by displacing customers from its desktop competitors (redistributors).

430. Thomson and Reuters' competitors (redistributors) supplied by Thomson and Reuters with regard to one or several of the four content sets referred to above generated in 2006 a worldwide turnover of approximately EUR 2 billion in the Research and Asset management area. EUR [10-100]* million account for less than [0-10]*% of the turnover generated by these competitors.

431. In other terms, if the foreclosure strategy of Thomson/Reuters is sufficiently efficient to convince more than [0-10]*% of customers of desktop competitors currently supplied by the notifying parties with regard to one or more of the four content sets referred to above, this strategy will result in a net increase in revenue for the newly formed entity.

432. The displacement of customers necessary to compensate for the loss of redistribution revenue is therefore limited and it is likely that the degradation of the quality of competitors' desktops might lead to a sufficient transfer of customers to benefit of the notifying parties. As a result, it would likely be profitable for Thomson/Reuters to increase its price for redistribution of the four content sets via redistributors or to stop commercialising its content sets via redistributors, as even a small proportion of (exredistributor') customers switching to Thomson/Reuters desktop offer would make the foreclosure strategy profitable.

V.B.6. Ownership

433. Ownership content data provide a researched view of security ownership available in the financial marketplace. They cover principally global securities ownership, current equity and Fixed Income ownership information. They are used particularly in the offtrading floor. Ownership applications help users to see current and historical shareholder positions. Information can generally be sorted to look at different shareholders, different funds with the largest to smallest assets, the most important changes in positions from the prior report date and several other sort types of factors.

V.B.6.1. Thomson's and Reuters products

434. Thomson has a strong presence in the Ownership content segment. Thomson's Ownership data are principally marketed under the Thomson One brand products although other Thomson products such as Treasura, Sharewatch, Streetsight also contain Ownership data. Thomson compiles and aggregates its Ownership data which are sold subsequently to final customers or partly redistributed through third parties.

435. Reuters is a marginal player in the Ownership content segment. Reuters is not marketing an Ownership data specific product. However, some of its products such as Reuters Knowledge and Reuters 3000Xtra provide customers with Ownership information. Reuters sources principally its equity Ownership data from third parties (Factset/Lionshare) while it directly collects Fixed Income Ownership data. Its internally sourced Ownership data is not redistributed to final customers through third parties.

V.B.6.2. Competitive assessment

436. Within the market investigation, no significant concerns have been raised in relation to Ownership data. More particularly, the market investigation has shown that the merger will not impede effective competition in Ownership data segment. First, Reuters is a marginal player in this area as it sources its content principally from third parties ([...]* in equity ownership) and it does not therefore exercise a strong competitive constraint on Thomson. Second, Thomson will continue to face competition post merger from competitors significantly stronger than Reuters in the Ownership area such as Factset, Bloomberg and other smaller players which compile, aggregate and distribute Ownership data (Ilios, NAIC, Bureau van Dijk).

437. The fact that Reuters does not constitute a strong competitive constraint on Thomson is also confirmed by the internal documents of the notifying parties. In this respect, in one of the Thomson's documents called [...]* Factset appears to be the first serious competitor of Thomson in Ownership data (139). This is corroborated by a more recent Reuters' document called [...]* in which Facstet/Lionshare and Bloomberg, but not Reuters are presented as Thomson's main competitors in the Ownership data area (140).

V.B.7. Deals

438. Deals content data provide a comprehensive insight into global capital market activity in different areas to bring institutional researchers the decision support tools they need when investigating companies. Deals data cover areas such as equities, bonds, mediumterm note programs (MTN), syndicated loans, public finance, project finance, mergers and acquisitions, corporate finance, corporate restructuring etc. Deals content data are particularly used in the off-trading floor.

V.B.7.1. Thomson's and Reuters' products

439. Thomson has a strong presence in the Deals content area. Thomson's Deals data are not only marketed under the Thomson One brand, but also under other products such as Treasura, IFR and, more particularly, SDC Platinum. Thomson compiles and aggregates its Deals content data which are subsequently sold to final customers or are partly redistributed through third parties.

440. As in the case of Ownership data, Reuters is a marginal player in the Deals area. Reuters' Deals offer is included in its Reuters Knowledge and Reuters 3000Xtra products. Reuters compiles and aggregates its Deals content data which are sold directly to final customers, but unlike Thomson it does not redistribute Deals content data through third parties.

V.B.7.2. Competitive Assessment

441. Within the market investigation, no significant concerns have been raised in relation to Deals data. The market investigation has shown that the merger will not impede effective competition in Deals data segment. In effect, Reuters is only a marginal player unable to exercise a competitive constraint on Thomson. Conversely, the merged entity will still face competition post merger from an important player such as Dealogic which sells better and timelier company data across all asset classes, and other players including Capital IQ (S&P), Bloomberg, Mergerstat (Factset), or Merger Market.

442. This is confirmed by the internal documents provided by the parties. For instance, in one of Thomson's documents called [...]* (141), only Dealogic is presented as a main competitor of Thomson in Deals; some other significant niche content and lower pricing operators are also mentioned such as Merger Market, BvD Zephyr etc. Several Reuters' documents point to the fact that Reuters is a weak competitor, Thomson and Dealogic are the primary close competitors in the Deals area and that other competitors like Capital IQ and Factset are expected to develop strong alternative offers (142).

V.B.8. Other content sets (profiles, public filings and other time series)

V.B.8.1. Content sets description

443. Other content sets data in which the parties are present comprise data such as profiles, public filings and time series other than Economic Time Series (TSE) analysed above.

444. Professional profiles content is composed generally of a global compilation of in-depth details on buy-side and sell-side institutions and their personnel including biographical, contact and related insight. This particular content set is tightly integrated with other content for in-depth views of security ownership, fund management and analyst coverage.

445. Public filings contain a collection of companies' filings documents comprising annual reports, insider trading filings and other data.

446. Time series other than Economic Time Series (TSE) include time series on Equities, Fixed Income securities, futures, options, warrants and other data.

V.B.8.2. Competitive assessment

447. The market investigation has not pointed to any serious competition concern in profiles, public filings and other time series segments.

448. As far as profiles and public filings are concerned, only one merging party (Thomson) is marketing these data. Therefore, the merger does not give rise to any horizontal overlaps or increments of Thomson's market share. Moreover, the merged entity will continue to face competition from a significant number of competitors such as Big Dough, StreetSight, Bloomberg in the profiles segment and, respectively, Edgar Online, Perfect Information, 10K Wizard, Global Reports in the public filings segment.

449. In respect to other time series segment, within the market investigation no specific competition concerns have been raised. The market investigation has also shown that the merged entity will continue to face competition post merger in other time series from a significant number of competitors such as Factset, Bloomberg, RIMES etc.

V.B.9. Desktop products in research & asset management

V.B.9.1. Wealth Management desktop products in Belgium, Finland and the UK

450. As regards wealth management desktop products, the data provided by the parties reveal that the parties have horizontal overlaps giving rise to affected markets especially in Belgium ([20-30]*% combined market share, with an increment of around [0-10]*%), Finland ([30-40]*% combined share with increment of around [0-10]*%), and the UK ([30-40]*% combined market share with increment of [10-20]*%) (143).

451. In Belgium the parties' combined market share amounts to [20-30]*%. No concerns have been put forward to the Commission in the course of the market investigation, thereby indicating that this market is sufficiently competitive, also given the significant number of financial institutions operating in it.

452. Finland gives rise to an affected market (combined share of [30-40]*%), but it has to be noted that Thomson has revenues of approximately EUR [less than one million]* only.

453. For the UK, the combined share is approximately [30-40]*% with the highest overlap (Thomson pre-merger share is: [20-30]*%, Reuters pre-merger share is: [10-20]*%). Nevertheless, no concerns have been put forward to the Commission in the course of the market investigation, thereby indicating that this market is sufficiently competitive, also given the significant number of financial institutions operating in the UK wealth management space. In particular, the structure of the UK's Wealth Management industry reveals that there are a large number of independent financial advisers, and also in view of the UK's internationally oriented financial services industry, sufficient constraints stem from non-UK oriented Wealth managers as well.

V.B.9.2. Desktop products in other areas of research & asset management

454. As regards desktop products in Investment Management, Investment Banking and Corporates, the data provided by the parties reveal that the parties have horizontal overlaps giving rise to affected markets only in Investment Banking with the merged entity having a combined share of approximately [20-30]*% worldwide (Thomson [10- 20]*% and Reuters [0-10]*%) and [20-30]*% EEA-wide (Thomson [10-20]*% and Reuters [0-10]*%) in 2006 (144). No concerns have been put forward to the Commission in the course of the market investigation, while the combined share is not particularly high, thereby indicating that this market is sufficiently competitive, also given the number of financial institutions operating at the downstream level.

V.C. CONCLUSION OF THE ASSESSMENT

455. Based on the above the Commission has come to the conclusion that the proposed merger raises serious doubts that it significantly impedes effective competition in the markets for Broker Report (Aftermarket Research), Estimates (both current and historical Estimates), Fundamentals, and Time series/Economic data (TS/E) both at the worldwide and EEA level. The described adverse effects of the merger will take place both as regards the sales of standalone content sets and as regards the content sets when sold (embedded) in relevant desktop products containing such content sets in the Research & Asset Management space

VI. COMMITMENTS PRESENTED BY THE PARTIES

456. In order to remove the serious doubtsidentified by the Commission in the areas above, the parties have proposed commitments on 6 December 2007 ("the commitments"). The commitments are designed to enable the purchaser(s) of the assets to compete effectively in selling Estimates, research, Fundamentals and time series/economics.

VI.A. DESCRIPTION OF INITIAL COMMITMENTS

457. The initial commitments are based on the following main elements.

a) Thomson and Reuters will commit to sell a copy of the database for each of these content areas (Fundamentals, Estimates, Aftermarket Broker Research, and Time series/Economics), within a short time period after the Commission's decision.

b) Thomson and Reuters will provide transitional technical support services for a period of up to [...]* from the date of the divestiture to enable the purchaser(s) to integrate the database(s) into its/their own existing offering(s).

c) Thomson and Reuters will undertake reasonable best effort to assist the purchaser(s) with obtaining the content owner consents which are necessary to distribute the relevant databases and manage them on an ongoing basis and going forward. This is particularly relevant with respect to Estimates and research.

458. The notifying parties submit that these remedies represent a business which is capable, in the hands of a suitable buyer, to restore effectively competition on a lasting basis in the marketplace.

459. In all cases, the copy database would be up to date at the closing of the divestiture.

460. In addition, the Parties will license on a royalty free basis all IP rights, trade secrets, know how and technical information embodied in the databases and will provide software for collection, aggregation, normalisation and transmission so as to enable the purchaser of the relevant database to integrate and operate it on an ongoing basis. At the option of the Purchaser, the Parties may offer a list of Key Personnel to be hired. The parties specify that a Purchaser will receive a particular information, tool, software, personnel or assistance, when these assets are not available on the open market.

461. The Commitments do not include divestiture of the customer base attached to each of the database. They neither include the staff necessary to manage the databases. Key Personnel only account for a fraction of complete staff (145).

462. The purchaser criteria specify that any potential Purchaser should be an existing provider of financial information, have an incentive to distribute the relevant databases via third parties, have the financial resources, proven expertise to develop the database and should not raise prima facie competition concerns.

463. The notifying parties will appoint a monitoring trustee, independent of them. The monitoring trustee will oversee the on-going management and maintenance of the divested assets as long as they are not transferred to the purchaser, with a view to ensuring their continued viability and competitiveness (146).

464. In case the divestment of identified assets is not completed within the first divestiture period (147), the notifying parties will appoint a divestiture trustee which task will be to sell at no minimum price those assets that have not been already been sold (148).

VI.B. MARKET TEST OF THE INITIAL COMMITMENTS

465. The Commission market tested the initial commitments presented by the notifying parties. Although competitors and customers globally acknowledged that facilitating the rapid entry of a new competitor in the relevant fields is an adequate remedy, a number of weaknesses have been highlighted which would undermine the ability of a purchaser to fully exploit and develop a competitive product on the basis of initial commitments proposed by the notifying parties (149).

466. 81% of the respondent competitors considered that the proposed remedy is not effective to enable a purchaser to offer a competitive and viable product in the market place within a short period of time and is not sufficient to ensure a viable business on a going forward basis given that certain key assets (including personnel, intangible assets contributors and customer base) are not transferred to the potential purchaser. The following comments have notably been made on the key aspects of the remedies.

VI.B.1. Assets

467. The market test indicated that generally all assets necessary to run a business should be included in the divestment, whereas the initial commitments limit the divested assets to those that are "reasonably necessary". Similarly the initial commitments specify that assets available on the open market should not be transferred. Such criteria are very restrictive and leave a lot of room for interpretation. Considering that the tools necessary to run this business are generally available on the market, the criteria proposed by the notifying parties would principally limit the divested assets to in-house developed software. However the added-value of this business lies in the combination of instruments to be used, and limiting the divestiture to a portion of these instruments could limit the possibility that the purchaser acquires a viable and operational infrastructure. As such this could hinder the ability of the purchaser to correctly develop its own offering.

468. In the market test, a number of assets have also been identified as necessary to run the business, although they were not included in the initial commitments. These include updating procedures, source documents (from which data is extracted), mechanics to feed and distribute the databases, access to R&D projects, reporting systems, training material, intellectual property rights associated to index of databases, hardware. In addition the transitional period in which the parties may be required to provide support to the purchaser was suggested to be extended to take account of the inherent technical and other types of difficulties linked to the operation of the database on an ongoing basis for the purchaser.

VI.B.2. Personnel

469. The initial commitments limit the possibility for the purchaser to hire personnel from Thomson and Reuters to key personnel only. These key personnel consist of a limited number of executives having key roles in the construction and development of the four content sets concerned by the remedies. A number of competitors have stressed that only the transfer of all, or a majority, of the personnel involved on a daily basis in the construction and maintenance of the databases would make the commitments an efficient remedy to the competition concerns.

470. It was notably indicated that the know-how necessary to build and operate such services does not only lay in the description of procedures but also in the experience and practice of staff. Should a purchaser need to train its own staff to operate, normalise, and distribute the services, a serious risk of disruption or degradation of quality of the databases would arise. It was therefore suggested that the complete staff working on the Thomson and Reuters products would have to be transferred to the purchaser (150) to ensure a viable and credible entry on the market. Staff to be transferred should include all relevant personnel performing all tasks necessary to operate the divested assets, including management (key personnel), software development, data collection, data normalisation, or administration.

471. In addition, as the remedies do not consist in the divestment of an independent business unit, the personnel is not automatically transferred to the purchaser. Thomson and Reuters should therefore release their staff from all non-competition clauses in their contracts to allow the purchaser hiring the relevant persons.

VI.B.3. Contributors

472. A key element to operate an Estimates or an Aftermarket Research business is to obtain data from brokers, alternatively called the contributors. As such, the Estimates or Aftermarket Research assets can only be exploited by a purchaser if he has agreements with contributors to continue to fill and exploit the databases.

473. The initial commitments foresee that Thomson/Reuters will do its "reasonable best efforts" to help the purchaser to enter into such agreements. This point has raised a number of criticisms from competitors who hardly understand what the obligations of the notifying parties implied by this terminology are. In addition, the incentive of Thomson/Reuters to help a competitor to build an alternative offer is far from obvious. The market test therefore suggested that the notifying parties should have an obligation to update the database of the purchaser as long as the latter has not signed agreements with all, or a significant proportion of, the top brokers currently contributing to Reuters Estimates and Aftermarket Research offers.

474. Specifically with regard to Aftermarket Research, competitors and customers indicated that the exclusive contracts signed between Thomson and brokers would undermine the ability of the purchaser to propose a credible offering on the market (151). It was also indicated that Thomson/Reuters could notably leverage its position on the market to enter into new exclusive contracts. It can also be envisaged that Thomson/Reuters could sign exclusive agreements in the Estimates area. Such a move from Thomson/Reuters would make it very difficult for the purchaser to develop credible alternatives Estimates or Aftermarket Research products since it would be barred from collecting data from key contributors. It was therefore suggested in the market test that this issue should be addressed in the commitments.

VI.B.4. Customers

475. The initial commitments do not include the transfer of the customer base. The majority of competitors indicated that this could constitute a serious drawback. Operating a complete service in one of the divested content sets induces significant fixed costs. These costs are necessary to continue the collection, normalisation and distribution of data and if they are not compensated by a minimum level of revenue might seriously hinder the viability of a new offering.

476. Without a customer base, it may not be possible for a new entrant to attract a number of customers within a short period of time. The financial information business is characterised by the necessity for a supplier to benefit from a strong reputation on the market as customers are particularly adverse to risk and will not purchase a product to a firm without a proven track record.

477. As such, a minimum customer base is necessary for both ensuring the viability of the offer and to facilitate the entry of the purchaser. 86% of the respondent competitors considered that the transfer of customers attached to databases is necessary to provide a credible offer in the market place.

VI.B.5. Suitable purchaser criteria

478. The market test also permitted to validate the need for the purchaser to already be in the market for financial information, as it is likely that customers will not fear starting a business relation with a player whose reputation is already established. Similarly a purchaser already having a customer base on which he could amortise a part of the costs of its new offer, or having desktop customers to whom he could rapidly sell its new offer, would present a better guarantee of viability for the divested assets.

VI.C. FINAL COMMITMENTS

479. The Commission communicated the results of the market test to the notifying parties, who amended the initial commitments with an objective to respond to most of the concerns raised by competitors and customers. The final commitments were proposed on the 21 December 2007. They are presented below, where the signs (initial) and (amended) are inserted to signal which elements of the final commitments were or were not included in the initial commitments. On the basis of the improvements contained in the final commitments, the Commission found that these undertakings remove the serious doubts raised by the tansaction which can therefore be authorised in line with Article 10 (2) of the Merger Regulation.

480. The final commitments are based on the following main elements:

a) Thomson and Reuters will commit to sell a copy of the database for each of the content areas Fundamentals, Estimates, Aftermarket Broker Research, and Time series/Economics, within a short time period after the Commission's decision. In all cases, the copy database would be up to date at the closing of the divestiture (initial).

b) In order to operate the databases, the purchaser will be offered the opportunity to recruit key personnel (initial) and other personnel currently operating the databases on a daily basis (amended).

c) The notifying parties will license on a royalty free basis all IP rights, trade secrets, know how and technical information embodied in the databases and will provide software for collection, aggregation, normalisation and transmission so as to enable the purchaser of the relevant database to integrate and operate it on an ongoing basis (initial).

d) In addition to above mentioned elements, the notifying parties will provide relevant information on source documents, copies of training documents, versions of software in active developments and rights to use numerical identifiers currently necessary to identify companies in the databases. All software necessary to operate the databases will be transferred to the purchaser, unless such software is substantially used in connection with other operations by Thomson or Reuters (amended).

e) Thomson and Reuters will provide transitional technical support services for a period of up to [...]* from the date of the divestiture to enable the purchaser(s) to integrate the database(s) into its/their own existing offering(s) (initial). At the option of the relevant purchaser, this period can be extended to [...]* for the Thomson Fundamentals database and the Reuters Estimates database (amended).

f) Thomson and Reuters will provide regular updates to the databases for a period of up to [...]* from the date of the divestiture to enable the purchaser(s) to operate on the market immediately and compete effectively. At the option of the relevant purchaser, this period can be extended to [...]* for the Thomson Fundamentals database and the Reuters Estimates database (amended).

g) Thomson and Reuters will undertake reasonable best effort to assist the purchaser(s) with obtaining the content owner consents which are necessary to distribute the relevant databases and manage them on an ongoing basis and going forward (initial). This is particularly relevant with respect to Estimates and Aftermarket Research. If by the closing date the purchaser has not signed agreements with i) at least [...]* of the contributors to the relevant database and ii) [...]* out of the top [...]* contributors, the notifying parties will permit the purchaser to serve as a redistributor of the content of any such contributors with whom the purchaser does not yet have appropriate arrangements under Reuters normal redistributor commercial terms and conditions for a period of [...]*. Regular updates of the databases may also be provided for a period of [...]* (amended).

h) For a period of [...]*, Thomson/Reuters will not enter into exclusive agreements with new brokers for the provision of Aftermarket Research or Estimates. In case a competitor enters into significant exclusivity contracts, this obligation might however be waived by the Commission (amended).

i) Thomson and Reuters will assign to the relevant purchaser, to the extent assignable, customer contracts specifically for direct content datafeeds from the Thomson Fundamentals database and Reuters Estimates databases. To the extent that such contracts are not assignable, the notifying parties will allow such customers to terminate early, and without penalty, any obligation relating to such direct content datafeeds within 12 months. Thomson/Reuters will not interfere with the negotiation between the purchaser and such customers (amended).

j) The purchaser criteria specify that any potential purchaser should be an existing provider of financial information, have an incentive to distribute the relevant databases via third parties, have the financial resources, proven expertise to develop the database and should not raise prima facie competition concerns (initial). In addition, the purchaser should already serve institutional clients, and in respect of the Reuters Aftermarket database, be offering, or intending to offer, a real-time research product (amended).

481. The commitments comprise the following databases:

a) Fundamentals. The Parties will sell a copy of the database of Thomson Fundamentals as included in the Thomson's "Enterprise FX" product and, at the option of the purchaser, a copy of the Thomson's "WorldscopeFile2" product.

b) Estimates. The Parties will sell a copy of the collection of Reuters' Estimates database. This includes earning Estimates as included in Reuters "Knowledge Direct-Estimates" product (both detailed and "consensus" forms).

c) Research. The Parties will sell a copy of the collection of Reuters' Aftermarket Research as included in "Reuters Knowledge" product.

d) Time Series/Economics. The Parties will sell a copy of the database of Reuters' Economic database EcoWin.

VI.D. ASSESSMENT OF THE COMMITMENTS

482. The remedy proposed by the notifying parties is a carve out of existing businesses of Thomson and Reuters aiming at facilitating the rapid entry of a new strong player in the four content sets identified as raising competition problems.

VI.D.1. Global assessment

483. In the course of the investigation, the Commission has identified competition issues with four content sets. These content sets are not produced and supplied by distinct business entities that could simply be divested to remedy the Commission's concerns. They are usually combined with other content sets and functionality in terminal products or data feeds which are suited for the needs of various classes of customers. It reflects the attempt of Reuters and Thomson to organise their businesses to best capture the needs of customers in the Sales & Trading and Research & Asset management sectors.

484. For this reason full divesture of certain Reuters or Thomson businesses was not contemplated. Since these relevant content sets take the form of databases and can be copied and transferred, the parties offered to transfer copies of the databases together with tangible and intangible assets which are used in connection with the relevant databases.

485. The largest scope of the remedy package (in terms of assets, personnel, contributors and customers as applicable to each relevant market) as set out in the final commitments will enable the purchaser(s) to quickly establish itself (or themselves) in each relevant market concerned as a credible competitive force in the marketplace in competition with the merged entity.

VI.D.1.1. Assets

486. The objective pursued by the remedy is that all assets necessary to run a business in each of the four content set are included. The final commitments provide assurances that the divested assets (i.e.: the relevant database and related assets) are sufficient to allow a purchaser to develop within a reasonably short period of time a competitive offering.

487. In this respect, it is significant that the scope of the asset divested is not limited to the relevant databases but include key intangible assets such as i) a royalty free license to market and distribute products derived from the use of the database, ii) know-how and instructions which are used in connection with the database, iii) detailed descriptions of the data extraction, collection and normalisation processes and methodologies; and tangible assets such as proprietary software necessary for the integration, operation and update of the database.

488. The market investigation confirmed that the joint transfer of such tangible and intangible assets is necessary and sufficient to operate a new service in the affected content sets. Such commitments also leave complete freedom to the purchaser to fully exploit the new businesses without depending on the notifying parties for the exploitation of intellectual property rights associated to the divested databases. It should also be noted that hardware is not included in the divestment, considering that the systems and computers necessary to run the relevant businesses are standard and can be immediately purchased on the market. It may even be likely that the purchaser will use its own IT infrastructure.

489. The final commitments also foresee that technical support and regular updates will be provided to the purchaser in order to help it understanding the structure and content of the database for a transitional period of [...]* from the closing date and up to [...]* at the option of the purchaser for the Fundamental and Estimates databases. These two elements will ensure that the purchaser can almost immediately sell a competitive product on the market to restore competition. They also facilitate the inevitable transition period during which the purchaser will have to learn and adapt business practices associated to the exploitation and maintenance of divested databases.

VI.D.1.2. Personnel

490. In order to ensure that the purchaser will be able to operate the databases, the final commitments foresee the possibility for the purchaser to recruit Key Personnel from the parties that is staff who have critical expertise in how to operate and develop the databases, but also Additional Personnel from the parties including staff dedicated to the collection, normalisation and standardisation of the data. The parties are also committed to facilitate interviews between such Personnel and the Purchaser(s) and not to prevent or prohibit any employee from being hired by the purchaser.

491. These provisions combined with the possibility to use the notifying parties' technical support should allow the purchaser to hire the necessary staff to operate the database without a serious risk of disruption or degradation of the databases.A large number of employees is typically necessary to run the databases: [...]* people work at Thomson in the Fundamental area and [...]* at Reuters in Estimates.

492. It must be noted that the possibility to hire staff involved in the collection, normalisation and standardisation of the data is only offered for the Fundamental Database and the Estimates database since it was found that hiring such personnel to run the Economics and Aftermarket Research database was not essential. Operating an Economics database and an Aftermarket Research business does not require the use of a significant number of staff and key personnel are likely to be sufficient. This is notably due to the fact that these two services do not require massive manipulation of data nor normalisation of data. As regards Aftermarket Research, the requirement that purchaser must already be active in real time research implies that it will already have staff employed in the collection and processing of data who could also be used for Aftermarket Research (see suitable purchaser criteria for more details).

VI.D.1.3. Contributors

493. As explained above in section IV.B.3, a key element to operate an Estimates or an Aftermarket Research business is to obtain data from contributors as the Estimates and aftermarket databases can only be exploited by the purchaser if he has an agreement with contributors to use their data. In addition a purchaser may only be able to provide a competitive Estimates or Aftermarket Research products if it has signed a sufficient number of agreements with contributors and especially with the most important of them.

494. The final commitments foresee that the notifying parties will use their reasonable best efforts to procure the assignment of the contributors to the purchaser. If by the closing date the purchaser has not signed agreements with i) at least [...]*% of the contributors to the relevant database and ii) [...]* out of the top [...]* contributors, the notifying parties will permit the purchaser to serve as a redistributor of the content of any such contributors with whom the purchaser does not yet have appropriate arrangements under Reuters normal redistributor commercial terms and conditions for a period of [...]*. The purchaser will also benefit from updates of the database(s) for a period of [...]*.

495. Negotiations with hundreds of brokers may prove very time consuming since brokers generate little revenue from the agreements and do not consider them as a priority. Each agreement can typically take months to finalize. The commitments make it possible for a purchaser to start operating a competitive Estimates and Aftermarket Research product by the closing date that is in matter of months, and before having secured the necessary agreements with contributors. Considering that the purchaser has to be a recognised competitor in the financial information business, it is expected that a transition period of [...]* during which he could act as a redistributor or Thomson/Reuters offers and receive regular updates to the database(s) will be sufficient to negotiate with enough contributors to build a credible own-compiled offer.

496. In addition, the parties have committed not to sign exclusivity agreements with new brokers for Aftermarket Research and Estimates for a period of [...]*. This commitment should ensure that same market condition prevail as pre-merger with respect to the possibility for the purchaser to enter into agreements with contributors.

497. The final commitments also provide that as to any third party contract or consent that may be necessary to use the Fundamental or the Economic Data database, the parties will use their reasonable best efforts to procure the assignment to the purchaser of such contracts and will in the case of any third party unwilling to consent to assignment, to assist the purchaser in putting into place substitute arrangement directly with such third parties. Such a requirement should facilitate the task of the purchaser when negotiating agreements with contributors for the use of Economics and Fundamental data. The market test confirmed that the number of such necessary agreements is very limited and is not expected to raise any concern.

VI.D.1.4. Customers

498. The final commitments foresee that Thomson and Reuters will assign to the relevant purchaser, to the extent assignable, customer contracts specifically for direct content datafeeds from the Thomson Fundamentals database and Reuters Estimates databases. To the extent that such contracts are not assignable, the notifying parties will allow such customers to terminate early, and without penalty, any obligation relating to such direct content datafeeds within [...]*.

499. This requirement should facilitate the transfer of part of the customer base of the parties in Fundamentals and Estimates and ultimately increase the viability of the offer by ensuring streams of revenues to the Purchaser. Although the notifying parties could not provide a detailed evaluation of the revenue streams the purchaser could expect by capturing all customers currently purchasing direct content datafeeds from Thomson (Fundamentals) or Reuters (Estimates), they provided a high level evaluation. Thomson evaluates that revenues for Fundamentals will be in the region of US [10-20 million]*. For Estimates, Reuters estimates that revenues will be in the region of EUR [0-10 million]*. These evaluations account for approximately [50-60]*% of Thomson revenues in Fundamentals and [70-80]*% of Reuters revenues in Estimates when considering revenue stemming from distribution or sales of stand-alone contents.

500. It should however be noted that there will not be any automatic transfer of customers in the majority of cases. The purchaser will have to make a proposal to customers of Thomson (Fundamentals) and Reuters (Estimates) who will have the possibility to early terminate their contracts with Thomson or Reuters. If the purchaser is a competitor selling desktop solutions with Fundamentals from Thomson or Estimates from Reuters as direct content datafeeds, he will have the possibility to propose to its customers to migrate to its own-compiled solution.

VI.D.1.5. Suitable purchaser criteria

501. As explained above, the financial information business is characterised by the necessity for a supplier to benefit from a strong reputation on the market as customers are particularly adverse to risk and will not purchase a product from a firm without a proven track record. Without an existing customer base of institutional clients, a new entrant would require several years to attract a sufficient number of customers. This would not allow ensuring immediate restoration of effective competition on the market.

502. For this reason the commitments foresee that the purchaser should be an existing provider of financial information, have the financial resources, proven expertise to develop the database and already serve institutional clients. In order to maintain and restore pre-merger competitive structure, the commitments also foresee that the purchaser should have the incentive to distribute the relevant databases via third parties, and should not raise prima facie competition concerns.

503. Such suitable purchaser criteria are an additional guarantee that the remedies will be operational on the short term and that competition will be restored after completion of the merger. The largest scope of the remedy package (in terms of assets, personnel, contributors and customers as applicable to each relevant market) as set out in the final commitments will enable the purchaser(s) to quickly establish itself (or themselves) in each relevant market concerned as a credible competitive force in the marketplace in competition with the merged entity.

VI.D.2. Assessment per affected market

504. Most of the commitments described above apply to the four content sets and contribute to the restoration of effective competition in the four markets. In addition, certain commitments which are specific to one or two content sets create additional conditions for the emergence of new credible competitors on those markets.

VI.D.2.1. Aftermarket Broker Research

505. The commitments meet certain critical requirements that should facilitate the emergence of a new competitive offer on the aftermarket Broker Research market in particular with respect to the availability of database on the market, Suitable Purchaser, Contributors, and Exclusivity agreements.

506. First, the commitments foresee that in order to be a credible competitor for Aftermarket Research, the purchaser should be offering, or intending to offer, a real-time research product. Although this condition reduces the choice of potential competitors, this ensures that the potential purchaser will already have institutional clients and proven expertise to manage the database.

507. Second, the commitments make it possible for a purchaser to start operating an Aftermarket Research product by the closing of the divestiture that is within a short period of time and well before having secured the necessary agreements with contributors. Third, the prohibition imposed on the parties to enter into new exclusivity agreements ensures that the purchaser will evolve under the same competitive structure as pre-merger.

508. Reuters Aftermarket Research customers will not be transferred to the purchaser, nor will they be offered the possibility to early terminate their contracts with Reuters. The Commission has considered that such customer transfer would not be necessary in the case of Aftermarket Research as costs incurred by the Aftermarket Research business are much more limited than those incurred by Fundamentals and Estimates businesses and the lack of transfer of customers therefore does not hinder the viability of the new offering. Firstly a limited number of key personnel is sufficient as there is no process of data, and secondly as the purchaser will already be active in the real-time research area, the acquisition of an Aftermarket Research business will not substantially modify its cost structure. In addition, customers normally pay for the service on pay-per-viewbasis so that the transfer of a customer base would not seem to be possible. However, as the purchaser will already be active in the real-time research area, the purchaser will already have contractual relations with the relevant customer base.

VI.D.2.2. Estimates

509. The commitments set up favourable conditions to allow the restoration of effective competition on the market in particular with respect to availability of database on the market, Personnel, contributors, customers and exclusivity agreements.

510. First, they foresee the possibility for the purchaser not only to recruit key personnel but also additional personnel from the parties that is staff dedicated to the day to day management of the database. (152) This should enable a purchaser to hire a sufficient number of people with expertise to start running the database within a reasonable period of time. Second, the commitments make it possible for a purchaser to start operating Estimates product (via redistribution) by the closing of the divestiture that is within a short period of time and well before having secured the necessary agreements with contributors.

511. The commitments also foresee the possibility of transfer of the customer base of the parties as regards customer contracts specifically for direct content datafeeds. Finally, the prohibition imposed on the parties to enter into new exclusivity agreements ensures that the purchaser will evolve under the same competitive structure as pre-merger.

VI.D.2.3. Fundamentals

512. The commitments set up favourable conditions to allow the restoration of effective competition on the market in particular with respect to availability of database on the market, personnel, contributors and customers.

513. First, they foresee the possibility for the purchaser not only to recruit key personnel but also additional personnel from the notifying parties that is staff dedicated to the day to day management of the database. (153) This should enable a purchaser to hire a sufficient number of people with expertise to start running the database within a reasonable period of time

514. The final commitments also provide that the parties will use their reasonable best efforts to procure the assignment to the purchaser of third party contract or consent that may be necessary to use the Fundamental database. Finally, the commitments also foresee the possibility of the transfer of the customer base of the parties as regards customer contracts specifically for direct content datafeeds.

VI.D.2.4. Time Series/Economics

515. The commitments set up favourable conditions to allow the restoration of effective competition on the market in particular with respect to availability of database on the market and contributors as they provide that as to any third party contract or consent that may be necessary to use the Fundamental or the Economic Data database, the parties will use their reasonable best efforts to procure the assignment to the purchaser of such contracts.

516. Reuters EcoWin customers will not be transferred to the purchaser, and neither will they be offered an early termination right of their contracts with Reuters. The Commission has considered that such customer transfer would not be necessary in the case of TS/E as costs incurred by the TS/E business are limited compared to those incurred by Fundamentals and Estimates businesses and the lack of transfer of customers therefore does not hinder the viability of the new offering.. Few key personnel are sufficient to overview the data process. Furthermore, it is highly likely that the purchaser will already be active in the data processing area similar or identical to TS/E hence the acquisition of a TS/E business will not substantially modify its cost structure. Moreover, some competitors view the TS/E business to complement their existing data offerings with which they would possibly produce a new TS/E product offer anyway for which they would have to seek customers.

VII. CONDITIONS AND OBLIGATIONS

517. Under the first sentence of the second subparagraph of Article 8(2) of the Merger Regulation, the Commission may attach to its decision conditions and obligations intended to ensure that the undertakings concerned comply with the commitments they have entered into vis-à-vis the Commission with a view to rendering the concentration compatible with the common market.

518. The achievement of the measure that gives rise to the structural change of the market is a condition, whereas the implementing steps which are necessary to achieve this result are generally obligations on the parties. Where a condition is not fulfilled, the Commission decision declaring the merger to be compatible with the common market no longer stands. Where the undertakings concerned commit a breach of an obligation, the Commission may revoke the clearance decision in accordance with Article 8(5)(b) of the Merger Regulation. The undertakings concerned may also be subject to fines and periodic penalty payments under Articles 14(2)(d) and 15(1)(c) of the Merger Regulation.

519. In view of the foregoing, this Decision is conditional upon full compliance with the undertaking that the parties will divest the Divestment Assets to one or several purchasers pursuant to paragraph 1 sentence 1 and paragraph 2 of "Section B" of the Commitments. The Divested Assets are further described in paragraph 3 of "Section B" of the Commitments and "Schedules 1 and 2, and 3" attached as Annex to this Decision.

VIII. CONCLUSION

520. It is accordingly concluded that the Commitments as set out in the Annex modify the notified concentration to such an extent that the serious doubts of the Commission as to the compatibility of that concentration with the common market are removed. The concentration should, therefore, be declared compatible with the common market pursuant to Article 8(2) of the Merger Regulation and with the EEA Agreement pursuant to Article 57 thereof, subject to compliance with the commitments set out in the Annex which is an integral part of this decision.

Has adopted this decision:

Article 1

The notified operation whereby Woodbridge Company Limited, which is the Thomson Corporation family holding company, acquires sole control within the meaning of Article 3(1)(b) of Regulation (EEC) No 139-2004 of the whole of the undertaking Reuters Group PLC is hereby declared compatible with the common market and the EEA Agreement.

Article 2

Article 1 is subject to compliance with the conditions set out in Section B, paragraphs 1 and 3 (including Schedules 1 and 2) of the final version of the Commitments submitted by the parties on 21 December 2007 (attached as Annex to this Decision).

Article 3

Article 1 is subject to compliance with the obligations set out in the remaining provisions of Section B and in Sections C, D and E of the final version of the Commitments submitted by the parties on 21 December 2007 (attached as Annex to this Decision).

Article 4

This Decision is addressed to:

Notes:

1 OJ L 24, 29.1.2004, p. 1.

* Parts of this text have been edited to ensure that confidential information is not disclosed; those parts are enclosed in square brackets and marked with an asterisk.

2 In addition, corporate development professionals (who perform similar functions to external investment banker) use data, information and analytics similar to that used by investment bankers, although it is not as extensive or wide- ranging.

3 Wealth Management, Investment Management, Investment Banking and Corporates which together correspond to Reuters Research & Asset Management division.

4 Corresponding to a sub-set of Reuters Sales & Trading division.

5 Although Reuters is known for its news content, it should be noted that its Media division (which sell news to media organisations and online consumers) accounts for only [0-10]*% of Reuters' revenues.

6 Sell-side sales and traders and buy-side traders.

7 In particular, it operates in the following four segments: (i) Foreign Exchange; (ii) Commodities and Energy; (iii) Fixed Income; and (iv) Institutional Equities.

8 Other divisions within Reuters are as follows: Enterprise: the third largest division accounting for some [10- 20]*% of revenues in 2006. It provides data for enterprise-wide use, systems to manage such data and a secure messaging service allowing financial professionals to communicate and collaborate. Media is Reuters' smallest business division. The latter supplies news agencies and other media outlets, broadcasting customers and online audiences with general, political and financial news. It also provides news directly to customers through Reuters branded digital services across online, mobile and Internet protocol platforms.

9 For deployed solutions vendors install software on the customer's server or PC, which then allows users to access financial information products through a customised front-end interface, based on their normal operating system. Such interfaces are typically configurable, allowing content to be combined from several data suppliers. Hosted or Web-based solutions often use Internet browser technologies and network distribution over the Internet or Internet-based virtual private networks. Data hosting is primarily in the vendor's data centres although in some instances a small infrastructure footprint may be required within the customer's data centres.

10 COMP/M.3692 Reuters/Telerate.

11 Users in the on-trading floor segment will typically require real-time exchange or inter-dealer broker data, including depth of book, and instrument data that permit the complex analytics to execute trading. They will also require real-time news from multiple sources. Users in the off-trading floor segment typically require exchange data which may be delayed; they do not usually need depth of book and require fewer sources of news. In addition they can also consume time series data, which on-trading floor users seldom do.

12 The central distinction for users is whether information is supplied real-time or near-real-time/archival. Real-time information is required by security dealers, Foreign Exchange dealers and investment advisers for immediate consumption. Non-real-time/archival information cannot satisfy real-time demands and is primarily adapted to research and customer awareness demands. For non-real time information, timing is not as important as the completeness and reliability of the information.

13 See Competitor 4 response to Commission's request for information phase-II under Article 11 of the Merger Regulation, dated 16 November 2007 (question 7).

14 See Customer 3 response to Commission's request for information phase-II under Article 11 of the Merger Regulation, dated 15 November 2007 (questions 6 and 7).

15 Reuters' documents [...]*, reference MP-00005760; [...]*, reference MP-00012767;[...]*, reference MP- 00013311.

16 Reuters' documents[...]*, reference MP-00004684;[...]*, reference MP-00004668;[...]*, reference MP- 00004497.

17 Reuters' document [...]*, reference MP-00117892.

18 Reuters' document [...]*, reference MP-00043560.

19 Reuters' document[...]*, reference MP-00051761

20 Thomson's documents [...]* provided by the Department of Justice (DOJ).

21 See Customer 34, Customer 11 and Customer 21 response to a Commission's Art 11 request for information, phase I (question 6): "The proposed segmentation between on- and off-trading floor, with further segmentations, does not provide a transparent view for customer choice. Often, a product designated for a certain type of user (e.g. trading floor) may be cross-sold to a different area (e.g. Investment Management)... we would submit that a product's value lies in its functionality and content, rather than in the end user's Corporates area".

22 Apart from Fixed Income and Institutional Equities, other asset classes are Foreign Exchange and Commodities & Energy. Traders in large banks typically specialise in one particular asset class and therefore have specific requirements in terms of real time data/information relating to their trading environment. In Sales & Trading, Thomson offers products which should meet these specific real-time data requirements in respect of Institutional Equities (Thomson ONE Equity), and has negligible activities as regards real-time data relating to Fixed Income, where instead Thomson's main activities focus on transaction platform services (i.e. Tradeweb); in addition, Thomson is not present in Foreign Exchange and in Commodities & Energy. Reuters supplies a single "base" product (Reuters 3000 Xtra), which is configurable by the users depending on the asset class in which he trades. For the purpose of this case, it is not relevant to examine whether these asset classes represent separate antitrust markets.

23 The "consolidation" process of various separate real-time datafeeds may be operated by a provider through the so-called ticker plants, i.e. a provider's infrastructure which aggregates feeds from various sources transforming them into a single source of data. An alternative to such ticker plant infrastructure is the "feed handler" technology. A feed handler is a software translation engine: it translates the source data in the direct feed from its original format into the vendor's data format (normalisation) and delivers the translated data to the vendor's delivery infrastructure. The normalised format permits the user to access all data in the consolidated datafeed using a single application programming interface (API). It is this normalised and aggregated stream which is delivered to clients as what is commonly referred to as a consolidated real time data feed. A separate feed handler is needed for each direct datafeed.

24 Case COMP/M.3692 - Reuters/Telerate of 2005.

25 TradeWeb operates as an electronic request for quote system that allows authorised buy-side clients to request quotes (i.e. liquidity discovery) from, and execute trades with, sell-side dealers (i.e. market makers). TradeWeb also operates as an inventory based execution system for some securities in which clients view and select from an inventory of securities and prices that each individual dealer has provided to TradeWeb.

26 See Annex 30 to the Form CO.

27 As an indicative benchmark, one respondent suggested that Tier 1 customers would be ready to accept the cost of redundant data sets and more than one full-scope data suppliers, because information is "mission critical" for their operation as it is the need to get the expertise and software to compare multiple datasets, using different data representations and schemes.

28 The notifying parties maintain that product customisation for sales to specific countries is limited and where this takes place it generally occurs at the margins.

29 In (b-2), the Wealth Management segment, market shares would have also to be considered at the national level. The combined market share of Thomson-Reuters would reach [30-40]*% in the United Kingdom and [20-30]*% in Belgium. Nevertheless the notifying parties submit that this should not create any competition concern as they already face strong competition in both markets.

30 Thomson document - "BCG Presentation Thomson financial ,5 mai 2005" - reference Q1.0196.

31 This picture is confirmed at the EEA level, with Reuters having approximately [30-40]*% market share in 2006 and Bloomberg [40-50]*% (see Annex 15 to the Parties' response dated 26 October 2007 to an Article 11 request of the Commission).

32 The source of these market share data is Annex 15 to the Parties' response dated 26 October 2007 to an Article 11 request of the Commission.

33 The source of these market share data is the Parties's response dated 30 October 2007 to the Commission's Article 6.1(c) decision dated 8 October 2007. At the EEA level Reuters has a market share of [60-70]*% in 2006, while Thomson is not present; therefore there is no horizontal overlap in the EEA.

34 An example of providers of technology that facilitates the delivery of real time market data from raw data generators (e.g. exchanges) to end users is the company Wombat.

35 Front Office describes the dealing and support staff in a financial institution who are closest to the buying and selling operations of a market; Middle Office is the part of a financial institution's settlement process that most closely liaises with the front office, recording trades and trading positions (this office monitors and manages an institution's risk exposures); Back Office is the department in a financial institution that processes deals and handles delivery, settlement and regulatory procedures.

36 Other securities identifiers exist in the market place and customers can use systems for mapping and translating simultaneously these different securities identifiers.

37 In this respect the parties stated that users of Thomson's desktop products (primarily sold in the off-trading floor), in so far as they do require securities identifiers, are already familiar with using Thomson's symbology codes in that context and are likely to have those codes embedded in background Excel spreadsheets and applications. They contend that even if RICs were to be added to Thomson products, this would not necessarily cause even those customers to switch to use a new symbology.

38 Reuters receives composite data [...]*.

39 Credit Suisse, Deutsche Bank, Goldman Sachs, JP Morgan, Lehman Brothers, Merrill Lynch, Morgan Stanley, The Royal Bank of Scotland and UBS.

40 A few independent research firms also produce broker reports.

41 Real-time research: 67% of customers using such a service indicated that direct distribution would not be a substitute, 13% indicated that aggregated offers have some critical advantages however not taking a position on the substitutability issue, and 20% considered that direct distribution by brokers is an acceptable substitute to an aggregated offer. It should however be noted that those customers having considered that direct distribution is an acceptable substitute are all corporate customers, not directly active in the financial business. Aftermarket Research: respective numbers are 64%, 14% and 22%.

42 Thomson also commercialises estimates under the FirstCall brand.

43 It should be noted that a share of the revenue generated by the company desktop solutions have been attributed to this product and therefore this revenue is a theoretical construction.

44 As of the end of 2006, downloads from Thomson's exclusive contributors accounted for [10-20]*% of the total download volume of Aftermarket Research.

45 Reuters document [...]* - reference MP-00043713.

46 Reuters Knowledge is declined under different versions for Investment Management, Investment Banking and Corporates.

47 It should be noted that a share of the revenue generated by the company desktop solutions have been attributed to this product and therefore this revenue is a theoretical construction. It can also be noted that Reuters generates a substantial part of its revenue by distributing real-time research via TheMarkets.com (US $ [0-10]* million in 2006). This revenue is attributed to the US market, although TheMarkets.com redistributed Reuters offering worldwide.

48 Thomson document [...]* - reference Q4(B).0487.

49 Only one customer indicated that Factset offer is comparable to Thomson and Reuters. This customer is however a corporate customer not directly active in the financial information business. Other customers indicated that Factset offer either lacks the global coverage of Thomson and Reuters or that it is still a new service not yet assessed. For instance Customer 42 submitted that "Factset does not have a full global coverage that is adequate for our needs".

50 http://www.factset.com/data/content/data/content/brokerresearch.

51 http://www.factset.com/data/factset_content/data/factset_content/researchconnect.

52 All customers have considered that Factset would need a period between "more than six months" to three years to build a credible alternative offer. A "consensus" would be around two years. It should be noted that clients directly active in the financial business, by contrast with corporate customers, generally foresee a longer period to be able to propose a credible offer.

53 Thomson document - [...]* - reference Q3.0127.

54 Reuters document - [...]* - provided by the US Department of Justice.

55 http://www.factset.com/data/factset_content/data/factset_content/researchconnect.

56 [0-10]* top 10 contributors of Thomson research signed separate agreements for real-time and Aftermarket Research. [0-10]* top 10 contributors of Reuters research signed separate agreements for real-time and Aftermarket Research, [0-10]* signed an addendum to the real-time research agreement to cover Aftermarket Research, and [0-10]* directly signed an agreement covering both real-time and Aftermarket Research.

57 Thomson eXimius, Global TOPIC, Thomson Markets; Reuters Wealth Manager, Reuters Investor, Reuters Trades UK, Lipper.

58 Around 2-3 years for established competitors having started their own-compiled offer . It would take more time for another newcomer.

59 See the chapter "V.B.5. Vertical foreclosure in Aftermarket Research, Earning Estimates, Fundamentals and Economic Time Series" for a detailed evaluation of the profitability of the foreclosure strategy.

60 Reuters document [...]* - reference MP-00029132.

61 Competitor 21 response to the 15-10-2007 questionnaire to competitor, p 18.

62 [...]*, MP 00043978.

63 [...]*, MP 00043978.

64 Thomson document [...]*

65 Thomson document [...]*

66 Competitor 21 has explained that 'Our customers would not consider estimates data provided by smaller suppliers such as Zacks to be equivalent substitutes, because they do not offer the same breadth of coverage' response by a competitor, response to the Questionnaire 15-10-2007, p 25.

67 [...]* MP-00044822. A more detailed analysis follows where Reuters calculates that Thomson generates $[200-300]* million of revenues from estimates out of a total market of $ [400-500]* million giving it a [40- 50]*% market share.

68 Multex was a data compiler active in the supply of financial data including research and estimates.

69 67% of the respondents the market investigation consider that Reuters and Thomson provide close and substitutable estimates products.

70 Minutes of the meeting between the case team and Competitor 21 of 12 November 2007.

71 [...]*, MP-00012767.

72 [...]*, MP 00043978.

73 [...]*

74 Parties' response to the Commission's Article 6(1)c decision,.

75 The term time series is commonly associated with the coverage of a data over an historical period of time.

76 Thomson document [...]*

77 Competitor 21 response to the 15-10_2007 questionnaire to competitors, question 14, p 15

78 The importance of quantitative funds, which trade using statistical models (based on historical data) designed to identify patterns in financial markets is a well recognized phenomenon even if it is sometimes criticized. See for instance: 'Does not compute: How misfiring quant funds are distorting the markets', Financial Times, December 10, 2007.

79 67% of the respondents to the market investigation have indicated that they are unlikely to switch to another provider if the parties were to increase prices by 5 to 10%. Among the reasons given for staying with the same supplier the most commonly provided are i) there are no valid alternative on the market, ii) switching would be too costly, iii) switching would be too time consuming and resource intensive.

80 See FactSet's press release of 1st September 2004 at http://investor.FactSet.com/releasedetail.cfm?ReleaseID=138089.

81 See the chapter "V.B.5. Vertical foreclosure in Aftermarket Research, Earning Estimates, Fundamentals and Economic Time Series" for a detailed evaluation of the profitability of the foreclosure strategy.

82 Reuters document [...]* - reference MP-00029132.

83 Parties' response to the Commission's Article 6(1)c decision, Annex 5.

84 Parties' response dated 15 November 2007 to a Commission's Article 11 request decision, Annex 9.

85 Competitor 85 reiterated in their various responses to the Commission's request for information under article 11 of the Merger regulation that they do not consider themselves to be at all a competitor of Thomson and Reuters as regards fundamentals since they are a rating agency and has a completely different business focus.

86 Parties' response dated 15 November 2007 to a Commission's Article 11 request decision, Annex 9.

87 Compustat North America contains "standardized" data but no "as reported" data.

88 As regards Compustat database, Reuters mentions that it is weak for non-US data (no good reputation), therefore Reuters is preferred when global data are needed. See [...]*, Internal document No RTRSE-CLAA- 00020356.

The same document is also under MP-00044893 [...]*: "[Compustat - S&P] non-US database is very weak in quality, timeliness, and coverage...", p. 12. Conversely Thomson-Worldscope is identified as proposing a strong product, notably in non-US companies, p. 13. This documents also deals with Bloomberg to define it away as a direct competitive threat to Reuters. "As Bloomberg does not have a standalone datafeed product, Reuters Fundamentals rarely encounters them as a competitor.

89 See Customer 46 response to Commission's request for information phase-II under Article 11 of the Merger Regulation, dated 5 November 2007 (question 28): "On Fundamentals...it will be very difficult to find an alternative for Thomson-Reuters". See also Customer 11 response to Commission's request for information phase-II under Article 11 of the Merger Regulation, dated 5 November 2007 (question 21): "For Fundamentals, S&P Compustat is not as good as Thomson or Reuters". See also Competitor 36 response to Commission's request for information phase-II under Article 11 of the Merger Regulation, dated 26 October 2007 (question 31): "There are no other real competitors for the combination of Multex and First Call. There would be nowhere else to obtain the depth, quantity or quality of data offered by this combined entity". See also minutes of the conference call with Competitor 21 on 12 November 2007: "Up to now customers need to have Thomson and Reuters for global fundamentals. Other compilers of fundamentals are just not a good alternative for global investors".

90 See Competitor 31 response to Commission's request for information phase-II under Article 11 of the Merger Regulation, dated 29 October 2007 (questions 13-20-21): " [Competitor 31] believes that the vast majority of its client would not be able to switch from use of Thomson and Reuters Fundamentals data to any other Fundamental database. Clients generally do not consider the S&P data a substitute because of its lack of depth and detail in this content".

91 See [...]*, Internal document No RTRSE-CLAA-00020356. See also MP-00044893 [...]*.

92 Parties' response dated 15 November 2007 to a Commission's Article 11 request decision, Annex 9. See Thomson internal document Q4(B).0448 Request.004.044.151 Competitor [...]*.

93 See Thomson internal document Q3.0058 Request.003.003.247 [...]*

94 See Thomson internal document Q4(B).0448 Request.004.044.151 [...]*.

95 Parties' response dated 15 November 2007 to a Commission's Article 11 request decision, Annex 9.

96 See Customer 38 response to Commission's request for information phase-II under Article 11 of the Merger Regulation, dated 31 October 2007 (question 32): "For Fundamentals, S&P Compustat would be the only other credible but expensive provider". See also Customer 15 response to Commission's request for information phase-II under Article 11 of the Merger Regulation, dated 31 October 2007 (question 14) where Bloomberg is indicated as an alternative only for desktop. See also Customer 12 response to Commission's request for information phase-II under Article 11 of the Merger Regulation, dated 5 November 2007 (question 28): " There are only small competitors for Fundamentals". See also Competitor 31 response to Commission's request for information phase-II under Article 11 of the Merger Regulation, dated 29 October 2007 (question 22): "In the area of supply of raw data, there are no currently available datafeeds of similar depth, history and quality as those made available by each of Thomson and Reuters".

97 See [...]*, Internal document No RTRSE-CLAA-00020356. See also MP-00044893 [...]*: "As Bloomberg does not have a standalone datafeed product, Reuters Fundamentals rarely encounters them as a competitor" see page 9, 14 and 15 of this document.

98 See [...]* document No Q3.0040, [...]*: Thomson evaluates Fundamental revenues of Compusat and Bloomberg, its main competitors alongside with Reuters - p.7; Thomson acknowledges that main competitors are Compusat (in the US) and Reuters. Bloomberg is considered a relatively small player p.7.

99 Parties' response dated 15 November 2007 to a Commission's Article 11 request decision, Annex 9.

100 Parties' response dated 15 November 2007 to a Commission's Article 11 request decision, Annex 9.

101 See Competitor 20 response to Art 11 request for information, phase-II (response to question 13).

102 See Competitor 20 response to Art 11 request for information, phase-II (response to question 29).

103 See [...]*, Internal document No RTRSE-CLAA-00020356 "Mergent is not a major player in this market yet, however there is a great possibility that Mergent emerges as a global competitor in the future", page 15 of this document.

104 Parties' response dated 15 November 2007 to a Commission's Article 11 request decision, Annex 9.

105 See responses to Commission's request for information phase-II under Article 11 of the Merger Regulation, dated 16 October 2007, from Customer 15, Customer 25, Customer 19, Competitor 86, Customer 41, Customer 46, Customer 11, Customer 9 (passim and particularly across responses to questions 6, 7, 14, 15, 20, 21, 22, 32).

106 See responses to Commission's request for information phase-II under Article 11 of the Merger Regulation, dated 16 October 2007, from Competitor 31, Competitor 20, Competitor 88/Competitor12/Competitor 56, Competitor 4, Competitor 30, Competitor 6.

107 See inter alia the response to Commission's request for information phase-II under Article 11 of the Merger Regulation, dated 16 October 2007 by Competitor 4. Competitor 4 indicated (question 20) that "Thomson Worldscope competes very closely with Reuters Fundamentals (formerly MarketGuide). Worldscope has a stronger presence in Europe, while Reuters Fundamentals is stronger within the US. S&P Compustat is the only other major competitor in this space. It competes closely with both Reuters and Thosmon in the US, but is much weaker in Europe. Beyond the major players, Mergent and Zacks also offer fundamental data. However they do not offer the same scope of coverage, and our customers would not consider them to be complete replacements for Worldscope or S&P". See also the response to Commission's request for information phase-II under Article 11 of the Merger Regulation, dated 16 October 2007 by Competitor 30. Competitor 30 considers (inter alia question 20) that as regards fundamentals the possibility of switching fundamental suppliers is confined between Worldscope (Thomson) and Multex (Reuters).

108 See [...]*, Internal document No RTRSE-CLAA-00020356. here Reuters considers one of its strengths to be the fact that customers consider it as one of the best quality database for global fundamentals. Thomson Worldscope strength would inter alia reside in the fact that it covers non-US small caps and inactive companies as well, and used to be the only available database for non-US fundamentals.

109 See [...]*, Internal document No RTRSE-CLAA-00020356 passim in particular pages 8-16 of this document).

110 Parties' response dated 15 November 2007 to a Commission's Article 11 request decision, Annex 9.

111 See responses to Commission's request for information phase-II under Article 11 of the Merger Regulation, dated 16 October 2007, from Customer 15, Customer 9, Customer 25, Customer 19, Competitor 86, Customer 41, Customer 46, Customer 11, (passim and particularly across responses to questions 3, 6, 14, 15, 20, 21, 22, 32).

112 See responses to Commission's request for information phase-II under Article 11 of the Merger Regulation, dated 16 October 2007, from Competitor 66, Competitor 10, Competitor 31, Competitor 20, Competitor 88/Competitor 12/Competitor 56, Competitor 4, Competitor 30, Competitor 6, and Competitor 36 passim.

113 See Competitor 89 response to Commission's request for information phase-II under Article 11 of the Merger Regulation, dated 16 October 2007 (question 12).

114 See [...]*, Internal document No RTRSE-CLAA-00020356 passim in particular pages 9 of this document).

115 See [...]*, Internal document No RTRSE-CLAA-00020356 passim in particular pages 8).

116 See Parties' response dated 26 October 2007 to a Commission's Article 11 request, Annex 2, 13, 15.

117 According to Factset's estimates, significant percentage of the purchase of fundamentals made by Factset customers is from Thomson fundamentals, [significant percentage] are from Reuters fundamentals and [20- 30]% from others. See responses to Commission's request for information phase-II under Article 11 of the Merger Regulation, dated 16 October 2007 and in particular the follow-up response by Factset dated 21 November 2007.

118 See Thomson's Internal document Ref. no. Q4(B)0449: [...]*. See also Thomson's Internal document Ref no. Q4(A)0135:[...]*.

119 See Parties' response dated 27 October 2007 to a Commission's Article 11 request, Annex 15, Fundamentals.

120 See [...]*, Internal document No RTRSE-CLAA-00020356 passim in particular pages 8-10 of this document).

121 See Competitor 21 response to Commission's request for information phase-II under Article 11 of the Merger Regulation, dated 6 November 2007 (question 23). See also responses to Commission's request for information phase-II under Article 11 of the Merger Regulation from Competitor 31, Competitor 89, Competitor 30. As regards customers see also Customer 12 response to Commission's request for information phase-II under Article 11 of the Merger Regulation, dated 5 November 2007 (question 23).

122 See [...]*, Internal document No RTRSE-CLAA-00020356. See also MP-00044893 [...]*: "As Bloomberg does not have a standalone datafeed product, Reuters Fundamentals rarely encounters them as a competitor" see page 9, 14 and 15 of this document. "Bloomberg Fundamentals data is mainly targeting at traders' referential usage, not so much at in-depth investment analysis" and that "as their target market is sales and trading not research and asset management, their data is not designed for analytical purpose".

123 S&P/Compustat would be the market leader and has both broader and deeper coverage than either Thomson or Reuters. Compustat does make its fundamentals offering available via redistributors (e.g. FactSet, RIMES, SunGard, Alacra, CharterOak, Matlab and LIM all carry it); S&P/Capital IQ, although a relatively recent entrant, already offers substantial breadth of coverage (over 60 000 public and 685 000 private companies) on a global basis, with history going back over 10 years. It makes its products available over redistributors' platforms and has global coverage.

124 Mergent offers coverage on 35 000 companies, with US history back to 1970 and international history back to 1987. In July 2006, Mergent reached an agreement to purchase Praedea Solutions Inc., a company with technology that enables rapid, automated data extraction and database population. Mergent includes the International database from Ford Equity Research. Mergent distributes both directly to end-users and through redistributors, including Alacara, FactSet, and RIMES. IPREO/Hemscott offers coverage on 15,000 companies going back to 1987. IPREO/Hemscott distributes both directly to end-users and through redistributors, including Alacra, IDC, and Lexis-Nexis.Fundamentals data would also be available directly from government regulatory agencies to which the companies must file various disclosures. Many of these agencies maintain online repositories that are accessible to the public, including: (a) EDGAR-The US SEC's repository, (b) EDINET-The Electronic Disclosure for Investors' NETwork established by the Japanese Financial Service Agency, (c) SEDAR-The System for Electronic Document Analysis and Retrieval, Canada's financial filing system for public companies and mutual funds. Fundamentals data would also be available through regulatory news services such as BALO (France) and Regulatory News Service (RNS) (UK) and through official registries in Europe, such as Companies House (UK), Handelsregister (Germany) and Infocamere (Italy).

125 According to the parties, redistributors do aggregate fundamentals data from multiple sources, including national and regional suppliers. Redistributors can obtain fundamentals data from: (a) one-stop shops of other compilers of global fundamentals; and (b) regional and local suppliers of fundamentals, mixing and matching if and as necessary to obtain desired fundamentals coverage.

126 XBRL is a common industry standard that tags each data point electronically and making data collection for a database population easier. Barriers to entry and expansion have been further reduced by the emergence of reporting standards such as IFRS. In April 2001, the International Accounting Standards Board (IASB) consolidated then-current Internal Accounting Standards (IAS) into the new International Financial Reporting Standards (IFRS). IFRS are used in many parts of the world, including the EU, Hong Kong, Australia, Russia, South Africa, Singapore and Pakistan.

127 Fundamentals data (except for certain calculated ratios) can be obtained from the company itself. Companies in the more developed countries almost always release electronically the reports in which this information is contained. In the US (for example, the SEC has required public companies to file their disclosure statements in electronic form since the mid-1990s) and such data are available through company web-sites.

128 See Competitor 4 response to Commission's request for information phase-II under Article 11 of the Merger Regulation, dated 5 November 2007 (question 25):" For Fundamentals, new entrant would need several years to accumulate the necessary data". See also Competitor 21 response to Commission's request for information phase-II under Article 11 of the Merger Regulation, dated 6 November 2007 (question 24): "The challenge is even greater for historical data. For many users a complete history is required for 5 or 10 years, so the time period before a competitive product could be created would be substantial".

129 See Thomson internal document Q4(B).0448 Request.004.044.151 [...]*

130 Both globally and in the EEA, the largest redistributor of Thomson fundamentals is FactSet, which operates under a sell-through model. Its second largest redistributor of fundamentals is Moody's, the credit ratings agency (which operates under a royalty model). In the case of Reuters, its largest redistributor of fundamentals on both a worldwide and EEA basis is FactSet. Its second largest redistributor is OneSource, a general provider of business intelligence, and not viewed as a close competitor of the Parties.

131 Thomson obtains ratings data from Moody's, who is a redistributor of Thomson fundamentals using a royalty based model; it also obtains a variety of data from S&P which is a redistributor of Thomson fundamentals. According to the parties, if Thomson were to refuse to distribute its fundamentals data through Moody's or S&P, those organisations could retaliate and refuse to sell their data to Thomson.

132 See Reuter's Internal document, Ref. no. MP-00044822: [...]*: "We have built-up a significant dependency on third parties who now account for almost half of our revenue. It is important that in 2007 we take definitive steps that will allow us to reduce this dependency". The incentive to reduce such dependency on third parties in order to reach end-customers applies in Fundamentals as well. See also Thomson's Internal document Ref. no. Q1.0015[...]*: "Integrate Reuters content and eliminate third-parties sources". See also Reuter's internal document Ref. no. MP-00029132: [...]*, "Capitalise on the opportunity to switch Factset users, disgruntled by the inflexibility of the Thomson data on Factset to Reuters offering". "Accelerate efforts to reduce the number of Reuters content sets distributed via Factset and encourage clients to consider Reuters offering e.g. RKIM versus Factset".

133 See the chapter "V.B.5. Vertical foreclosure in Aftermarket Research, Earning Estimates, Fundamentals and Economic Time Series" for a detailed evaluation of the profitability of the foreclosure strategy.

134 [...]*, Doc. Ref. MP-00051761.

135 Source: Parties' response to Art. 11 request dated 30 October 2007, Annex 9.

136 [...]*, Doc. Ref. Q4(B).0355, Q4(B).0360; [...]*, provided via DoJ; [...]*, Doc. Ref. Vol. 12, Doc. 49 and Vol. 15, Doc. 10.

137 [...]*, Doc. Ref. Q4(B).0355, Q4(B).0360.

138 See the chapter "V.B.5. Vertical foreclosure in Aftermarket Research, Earning Estimates, Fundamentals and Economic Time Series" for a detailed evaluation of the profitability of the foreclosure strategy.

139 Thomson's document [...]*, reference Q(B).0437.

140 Reuters document [...]*, reference MP-00002480.

141 Thomson's document [...]*, reference Q(4)384.

142 Reuters' documents [...]*, reference MP-00005378; [...]*, reference MP-00043713; [...]*, reference MP- 00047710.

143 There are other countries where the merger gives rise to technically affected markets, but where or the overlap is de minimis or the analysis is pretty straightforward. In Austria, Reuters has pre-merger [80- 90]*% of the market, but the increment from Thomson's revenues is trivial ([less than 1%]*). In Ireland, the Parties' combined share amounts to around [10-20]*%, which is below the alert threshold set out in the Merger Regulation (25%). Italy also gives rise to an affected market, but the overlap from Thomson's revenues is less than 1% (combined [50-60]*%). In Luxembourg, the Parties' combined market share amounts to approximately [20-30]*%. However, when the Benelux countries are viewed as a whole, the combined market share is below 15%, which would therefore not give rise to an affected market. In Spain the Parties have combined share of [40-50]*%, but increment due to Thomson revenues is [less than 1%]*.

144 As regards desktop products for Investment Management, the parties' combined shares in 2006 were around [10-20]*% (EEA) and [10-20]*% (worldwide). As regards Corporates, the parties combined shares were around [0-10]*% (EEA) and [10-20]*% (worldwide) in 2006. Therefore, there is no technically affected market.

145 The number of such personnel is very limited according to the notifying parties (Fundamentals: [...]*; Estimates: [...]*; Economic time series: [...]*).

146 See Paragraph 22 of the Commitments.

147 See Part (1) Definitions of the Commitments.

148 See Paragraph 23 of the Commitments;

149 It should be noted that customers were less critical than competitors on the implementation issues raised by the commitments, which is understandable from their perspective as they do not operate such services as competitors do, and as such are less aware of business constraints.

150 According to the notifying parties, [...]* employees work at Thomson in the fundamental area, [...]* at Reuters work in Estimates, and [...]* in Research (including real-time and Aftermarket Research. In addition a number of employees active in the Estimates area also contribute to the research area), and [...]* in Economic time series.

151 93% of competitors and 82% of customers have indicated that the exclusive deals of Thomson would limit the viability of the purchaser.

152 Out of a maximum of [...]* employees working solely on Reuters Estimates database.

153 Out of a maximum of [...]* employees working solely on Thomson Fundamental database.

By hand and by fax: 00 32 2 296 4301

European Commission - Merger Task Force

DG Competition

Rue Joseph II 70 Jozef-II straat

B-1000 Brussels

Case M.4726 - Thomson Corporation/Reuters

COMMITMENTS TO THE EUROPEAN COMMISSION

Pursuant to Articles 8(2) and 10(2) of Council Regulation (EC) No. 139-2004 (the "Merger Regulation"), The Thomson Corporation ("Thomson") and Reuters Group PLC ("Reuters") (together the "Parties") hereby provide the following Commitments (the "Commitments") in order to enable the European Commission (the "Commission") to declare the acquisition of sole control of Reuters by Woodbridge Company Limited (the Thomson family holding company), by way of a scheme/plan of arrangement involving a dual-listed company structure (the "T-R Acquisition"), compatible with the common market and the EEA Agreement by its decision pursuant to Article 8(2) of the Merger Regulation (the "Decision").

The Commitments shall take effect upon the date of adoption of the Decision. For the avoidance of doubt, the Commitments shall cease to have effect if the Parties demonstrate to the satisfaction of the Commission that they have abandoned the T-R Acquisition.

The overall purpose of the Commitments is to restore effective competition in estimates, research, fundamentals and economic data by making available to the Purchaser(s) such assets and tools as to enable the Purchaser(s) to operate on the market immediately and compete effectively in the provision of Estimates, Aftermarket Research, Fundamentals and Economic Data as defined below.

This text shall be interpreted in the light of the Decision to the extent that the Commitments are attached as conditions and obligations, in the general framework of Community law, in particular in the light of the Merger Regulation, and by reference to the Commission Notice on remedies acceptable under Council Regulation (EEC) No 4064/89 and under Commission Regulation (EC) No 447/98.

Section A. Definitions and interpretation

(1) Definitions

For the purpose of the Commitments, the following terms shall have the following meaning: Additional Personnel: those individuals employed by the Parties dedicated to data collection, normalisation, standardisation and cleansing operations for the Thomson Fundamentals Database or the Reuters Estimates Database (but for the avoidance of doubt not including Key Personnel) as listed in Schedule 3.

Aftermarket Research: means reports prepared by sell-side and independent analysts that include an analysis of a security, company, or industry (including company-specific reports and industry-wide reports) and that are no longer restricted ("embargoed") as to recipients by the authoring firm and are generally available for sale.

Affiliated Undertakings: undertakings controlled by Thomson or Reuters and/or by the ultimate parents of Thomson or Reuters respectively, whereby the notion of control shall be interpreted pursuant to Article 3 of the Merger Regulation and in the light of the Commission Consolidated Jurisdictional Notice under Council Regulation (EC) No 139-2004 on the control of concentrations between undertakings.

Closing: the completion of the transfer of each set of Divestment Assets as contemplated by the sale and purchase agreement to be signed with each Purchaser.

Closing Date: in relation to each set of Divestment Assets, means the date on which Closing takes place.

Databases: the Reuters Estimates Database, the Reuters Aftermarket Research Database, the Reuters Economics Database or the Thomson Fundamentals Database, as the case may be, as defined in Schedule 1.

Direct Content Datafeeds: means datafeeds offering data within a discrete content set (i.e. the Thomson Fundamentals Database or the Reuters Estimates Database) for distribution on a standalone basis, where in each case the Direct Content Datafeed consists predominantly of Fundamentals or Estimates respectively, and delivered using FTP (file transfer protocol), CD or DVD media, or any other method which permits delivery using standard technology. For the avoidance of doubt this includes data delivered by or through redistributors.

Divestment Assets: copies of the Databases together with the other matters described in more detail in Section B and in Schedule 1, that the Parties commit to divest (each a "set of Divestment Assets"). Divestiture Trustee: one or more natural or legal person(s), independent from the Parties, who is approved by the Commission and appointed by the Parties and who has received from the Parties the exclusive Trustee Mandate to sell each of the sets of Divestment Assets to a Purchaser at no minimum price.

Effective Date: the date of adoption of the Decision.

Economic Data: means data on macroeconomic variables (e.g. GDP, unemployment, money supply, construction, national accounts, balance of trade figures and inflation) used as individual data points and tracked periodically over time.

Estimates: mean predictions as made over time by sell-side and independent analysts regarding the future financial performance of a company or security, typically with respect to key earnings metrics such as annual or quarterly earnings per share, that are not distributed in real time or used for current market awareness in the purchase and sale of securities.

First Divestiture Period: the period of [confidential] from the Effective Date.

Fundamentals: mean statistics over time pertaining to the core financial performance of a company, such as reportable financial statement data (e.g. balance sheet, cash flow and income statements), calculated financial ratios (e.g. annual and five-year averages for growth rates, profitability, leverage, asset utilization), and per share data (e.g. earnings per share, book value per share, cash flow per share), that are derived from company filings and financial statements.

Key Personnel: such personnel whose primary responsibility pertains to the Divestment Assets and who have proprietary subject matter expertise and are listed in Schedule 2.

Monitoring Trustee: one or more natural or legal person(s), independent from the Parties, who is approved by the Commission and appointed by Thomson, and who has the duty to monitor Thomson's compliance with the conditions and obligations attached to the Decision.

Personnel or Employees: the Key Personnel and the Additional Personnel.

Purchaser or Purchasers: the entity or entities approved by the Commission as acquirer(s) of the sets of Divestment Assets in accordance with the criteria set out in Section D.

Trustee: the Monitoring Trustee and the Divestiture Trustee.

Trustee Divestiture Period: the period of [confidential] from the end of the First Divestiture Period.

(2) Interpretation

Where an obligation in relation to the Divestment Assets is entered into by the Parties under these Commitments, insofar as such obligation pertains to the period prior to completion of the T-R Acquisition it shall apply to Thomson (and not Reuters) insofar as it relates to the Thomson Fundamentals Database, and to Reuters (and not Thomson) insofar as it relates to the Reuters Estimates Database, the Reuters Aftermarket Research Database or the Reuters Economics Database.

Section B. The Divestment Assets

Commitment to divest

1. In order to restore effective competition, the Parties commit to divest, or procure the divestiture of, the Divestment Assets by the end of the Trustee Divestiture Period to one or more Purchasers and on terms of sale approved by the Commission in accordance with the procedure described in paragraph 14. To carry out the divestiture, the Parties commit to find a Purchaser, and to enter into a final binding sale and purchase agreement, for the sale of each set of Divestment Assets within the First Divestiture Period. If the Parties have not entered into such an agreement for the sale of each set of Divestment Assets at the end of the First Divestiture Period, the Parties shall grant the Divestiture Trustee an exclusive mandate to sell the relevant set or sets of Divestment Assets in accordance with the procedure described in paragraph 23 in the Trustee Divestiture Period.

2. The Parties shall be deemed to have complied with this commitment if, by the end of the Trustee Divestiture Period, the Parties have entered into a final binding sale and purchase agreement in relation to each set of Divestment Assets, if the Commission approves the Purchaser(s) and the terms in accordance with the procedure described in paragraph 14 and if the Closing Date takes place within a period not exceeding [confidential], but in no event earlier than the completion of the T-R Acquisition.

Structure and definition of the Divestment Assets

3. The Divestment Assets, which are described in more detail in Schedule 1, include the following:

(a) those tangible and intangible assets which are used in connection with the relevant Database, as described in more detail in Schedule 1;

(b) a list of Personnel to enable the Purchaser(s) to make offers of employment or secondment with respect to which the provisions in paragraph 7 will apply; and

(c) for a transitional period of up to [confidential] after the Closing Date (or, in the circumstances described in paragraph 2(f) of Schedule 1, [confidential]) the benefit of support services to the Purchaser(s), including the provision of updates to the Databases, as detailed in Schedule 1, unless otherwise agreed with the Purchaser(s).

Section C. Related commitments

Preservation of Viability, Marketability and Competitiveness

4. From the Effective Date until Closing, the Parties shall preserve the economic viability, marketability and competitiveness of the Divestment Assets, in accordance with good business practice, and shall minimise as far as possible any risk of loss of competitive potential of the Divestment Assets. In particular each of the Parties undertakes:

(a) not to carry out any act upon its own authority that might have a significant adverse impact on the value, management or competitiveness of the Divestment Assets or that might materially alter the nature and scope of activity of the Divestment Assets;

(b) to make available sufficient resources for the development of the Divestment Assets, on the basis and continuation of the existing business plans; and

(c) to take all reasonable steps, including appropriate incentive schemes (based on industry practice), to encourage all Personnel to remain in the employment of the Parties (or, following completion of the T-R Acquisition, the merged entity).

Third Party Consents

5. The following shall apply in relation to third-party consents:

(a) the Parties shall use their reasonable best efforts to procure the assignment to the Purchaser(s) of contracts for the contribution of research and/or estimates between Reuters and brokers or other research contributors or, in the case of any broker or other research contributor unwilling to consent to assignment or whose contract cannot otherwise be assigned to a Purchaser, to assist the Purchaser(s) in putting into place substitute arrangements directly with such broker or other research contributor; provided that if by the relevant Closing the Purchaser does not have appropriate arrangements in place with:

(i) at least [confidential] of the contributors to the relevant Database (as it existed on the Closing Date), with respect to the obligation to procure assignment and/or substitution as to all such contributors generally; and

(ii) [confidential] of the top [confidential] contributors to the relevant Database (as it existed on the Closing Date), in the case of obligations to procure assignment and/or substitution with respect to such top [confidential] contributors; the Parties shall continue to use their reasonable best efforts to achieve the outcomes mentioned in sub-paragraphs (i) and (ii) above, up to the date which is [confidential] after the Closing Date. During such time, (A) Reuters will permit the relevant Purchaser(s) to serve as a redistributor of the content of any such contributors (with whom the Purchaser does not yet have appropriate arrangements) from the relevant Database as maintained by Reuters under Reuters normal redistributor commercial terms and conditions; and (B) upon appropriate arrangements being put into place between Purchaser and any such contributor the obligations in Schedule 1, paragraph 2(b)(i) (regarding the transfer of relevant data ) shall apply.

(b) as to any other third party contracts or consents which may be required for a Purchaser to receive and utilise a copy of a Database, including any contracts for the provision of Fundamentals data to the Thomson Fundamentals Database and any contracts for the provision of Economic Data to the Reuters Economics Database, the Parties will likewise use their reasonable best efforts to procure the assignment to the Purchaser(s) of such contracts or, in the case of any third party unwilling to consent to assignment or whose contract cannot otherwise be assigned to a Purchaser, to assist the Purchaser(s) in putting into place substitute arrangements directly with such third parties.

Exclusive broker contracts

6. Thomson currently has [confidential] exclusive contracts with brokers in relation to Aftermarket Research; Reuters has no such exclusive contracts, and neither Party has any exclusive contracts in relation to Estimates. For a period of [confidential] from the Effective Date the Parties shall not enter into any exclusivity contracts with any new brokers for the provision of Aftermarket Research or Estimates.

Personnel

7. The Parties shall use their reasonable best efforts to the extent permitted by law, to facilitate the transfer to the relevant Purchaser (if possible by collective transfer) of any Personnel desired by such Purchaser. The Parties shall provide relevant contact details for the Employees as desired by the Purchaser, or otherwise make the Employees available to the Purchaser subject to compliance with applicable laws. Prior to Closing, the Parties shall facilitate interviews between such Personnel and the Purchaser(s), shall not discourage such employee from participating in such interviews, and shall not interfere in employment negotiations between such Personnel and the Purchaser(s).

8. With respect to each Employee who receives an offer of employment from the Purchaser (conditional on or following the Closing), the Parties shall do the following: (i) not prevent, prohibit or restrict or threaten to prevent, prohibit or restrict the Employee from being employed by the Purchaser(s), and not offer any incentive to the Employee to decline employment with the Purchaser; (ii) if the Employee accepts such offer of employment from the Purchaser(s), the Parties shall cooperate with the Purchaser(s) in effecting transfer of the Employee to the employ of the Purchaser(s) and the Parties shall amend or waive the relevant provisions of employment agreements, stock options and other employee benefit arrangements of Personnel so that they do not suffer adverse consequences as a result of their negotiations with, or acceptance of an offer from, the Purchaser.

9. Each of the Parties undertakes, subject to customary limitations, that in relation to Personnel who are hired by (as opposed to seconded to) the Purchaser(s), it will not solicit and will procure that its Affiliated Undertakings do not solicit, such Personnel for a period of 12 months after Closing or after the date of termination of employment of such Personnel by the Parties (as applicable). Due Diligence

10. In order to enable potential purchasers to carry out a reasonable due diligence of the Divestment Assets, the Parties shall, subject to customary confidentiality assurances and dependent on the stage of the divestiture process, provide to potential purchasers sufficient information as regards the Divestment Assets (including customary technical due diligence on the relevant software and any configurations thereof).

Reporting

11. The Parties shall submit written reports in English on potential purchasers of the Divestment Assets and developments in the negotiations with such potential purchasers to the Commission and the Monitoring Trustee no later than 10 days after the end of every month following the Effective Date (or otherwise at the Commission's request).

12. The Parties shall inform the Commission and the Monitoring Trustee on the preparation of the data room documentation and the due diligence procedure and shall submit a copy of any information memorandum to the Commission and the Monitoring Trustee before sending such memorandum out to potential purchasers.

Section D. The Purchaser(s)

13. In order to be approved by the Commission, the Purchaser(s) must be able to ensure the immediate restoration of effective competition, which includes that the Purchaser(s) must:

(a) be existing providers of financial information products to institutional clients and be independent of and unconnected to the Parties;

(b) in respect of the Purchaser of the Reuters Aftermarket Research Database, be also offering, or intending to offer, a real-time research product;

(c) have the incentive to distribute the relevant Database(s) (or products derived from the relevant Database(s)) via third parties;

(d) have the financial resources, proven expertise and incentive to maintain and develop the relevant set(s) of Divestment Assets as a viable and active competitive force in competition with the Parties and other competitors; and

(e) neither be likely to create, in the light of the information available to the Commission, prima facie competition concerns nor give rise to a risk that the implementation of the Commitments will be delayed, and must, in particular, reasonably be expected to obtain all necessary approvals from the relevant regulatory authorities for the acquisition of the relevant set(s) of Divestment Assets (the before-mentioned criteria for the Purchaser(s) hereafter referred to as the "Purchaser Requirements").

14. The final binding sale and purchase agreement for each set of Divestment Assets shall be conditional on the Commission's approval. When Thomson and/or Reuters as the case may be has reached an agreement with a Purchaser, the Parties shall submit a fully documented and reasoned proposal, including a copy of the final agreement(s), to the Commission and the Monitoring Trustee. The Parties must be able to demonstrate to the Commission that the Purchaser meets the Purchaser Requirements and that each set of Divestment Assets is being sold in a manner consistent with the Commitments. For the approval, the Commission shall verify that the Purchaser fulfils the Purchaser Requirements and that each set of Divestment Assets is being sold in a manner consistent with the Commitments. The Commission may approve the sale of any set of Divestment Assets without one or more individual components thereof, if this does not affect the viability and competitiveness of the relevant set of Divestment Assets after the sale, taking account of the proposed Purchaser. For the avoidance of doubt, each set of Divestment Assets shall be sold to one Purchaser only, although it is possible that any individual Purchaser may purchase more than one set of Divestment Assets.

Section E. Trustee

I. Appointment Procedure

15. The Parties shall appoint a Monitoring Trustee to carry out the functions specified in the Commitments for a Monitoring Trustee. If Thomson or Reuters as the case may be has not entered into a binding sale and purchase agreement one month before the end of the First Divestiture Period or if the Commission has rejected a purchaser proposed by the Parties at that time or thereafter, the Parties shall appoint a Divestiture Trustee to carry out the functions specified in the Commitments for a Divestiture Trustee. The appointment of the Divestiture Trustee shall take effect upon the expiry of the First Divestiture Period.

16. The Trustee shall be independent of the Parties, possess the necessary qualifications to carry out its mandate, for example as an investment bank or consultant or auditor, and shall neither have nor become exposed to a conflict of interest. The Trustee shall be remunerated by the Parties in a way that does not impede the independent and effective fulfilment of its mandate. In particular, where the remuneration package of a Divestiture Trustee includes a success premium linked to the final sale value of a set or sets of Divestment Assets, the fee shall also be linked to a divestiture within the Trustee Divestiture Period.

Proposal by the Parties

17. No later than one week after the Effective Date, the Parties shall submit a list of one or more persons whom the Parties propose to appoint as the Monitoring Trustee to the Commission for approval. No later than one month before the end of the First Divestiture Period, the Parties shall submit a list of one or more persons whom the Parties propose to appoint as Divestiture Trustee to the Commission for approval. The proposal shall contain sufficient information for the Commission to verify that the proposed Trustee fulfils the requirements set out in paragraph 16 and shall include:

(a) the full terms of the proposed mandate, which shall include all provisions necessary to enable the Trustee to fulfil its duties under these Commitments;

(b) the outline of a work plan which describes how the Trustee intends to carry out its assigned tasks; and

(c) an indication whether the proposed Trustee is to act as both Monitoring Trustee and Divestiture Trustee or whether different trustees are proposed for the two functions.

Approval or rejection by the Commission

18. The Commission shall have the discretion to approve or reject the proposed Trustee and to approve the proposed mandate subject to any modifications it deems necessary for the Trustee to fulfil its obligations. If only one name is approved, the Parties shall appoint or cause to be appointed, the individual or institution concerned as Trustee, in accordance with the mandate approved by the Commission. If more than one name is approved, the Parties shall be free to choose the Trustee to be appointed from among the names approved. The Trustee shall be appointed within one week of the Commission's approval, in accordance with the mandate approved by the Commission.

New proposal by the Parties

19. If all the proposed Trustees are rejected, the Parties shall submit the names of at least two more individuals or institutions within one week of being informed of the rejection, in accordance with the requirements and the procedure set out in paragraphs 15 and 18.

Trustee nominated by the Commission

20. If all further proposed Trustees are rejected by the Commission, the Commission shall nominate a Trustee, whom the Parties shall appoint, or cause to be appointed, in accordance with a trustee mandate approved by the Commission.

II. Functions of the Trustee

21. The Trustee shall assume its specified duties in order to ensure compliance with the Commitments. The Commission may, on its own initiative or at the request of the Trustee or the Parties, give any orders or instructions to the Trustee in order to ensure compliance with the conditions and obligations attached to the Decision.

Duties and obligations of the Monitoring Trustee

22. The Monitoring Trustee shall:

(i) propose in its first report to the Commission a detailed work plan describing how it intends to monitor compliance with the obligations and conditions attached to the Decision;

(ii) oversee the on-going management and maintenance of the Divestment Assets with a view to ensuring their continued economic viability, marketability and competitiveness and monitor compliance by the Parties with the conditions and obligations attached to the Decision. To that end the Monitoring Trustee shall:

(a) monitor the preservation of the economic viability, marketability and competitiveness of the Divestment Assets, in accordance with paragraph 4 of the Commitments;

(b) in consultation with the Parties, determine all necessary measures to ensure that the Parties do not after the Effective Date use any business secrets, know-how, commercial information, or any other information of a confidential or proprietary nature relating to the Purchaser(s) or their use of the Divestment Assets which they may have gained as a result of the transactions and interactions contemplated in these Commitments including through any transitional support services they may offer the Purchaser(s);

(c) monitor the effective sale of the sets of Divestment Assets and the identification of Personnel;

(d) monitor the implementation of the transitional support services; and

(e) monitor the assistance provided in obtaining third party consents;

(iii) assume the other functions assigned to the Monitoring Trustee under the conditions and obligations attached to the Decision;

(iv) propose to the Parties such measures as the Monitoring Trustee considers necessary to ensure the Parties' compliance with the conditions and obligations attached to the Decision, in particular the maintenance of the full economic viability, marketability or competitiveness of the Divestment Assets;

(v) review and assess potential purchasers as well as the progress of the divestiture process and verify that, dependent on the stage of the divestiture process, (a) potential purchasers receive sufficient information relating to the Divestment Assets by reviewing, if available, the data room documentation, any information memorandum and the due diligence process, and (b) potential purchasers are granted reasonable access to the Personnel;

(vi) provide to the Commission, sending to each of the Parties a non-confidential copy at the same time, a written report within 15 days after the end of every month. The report shall cover the operation, management and maintenance of the Divestment Assets so that the Commission can assess whether the Divestment Assets are held in a manner consistent with the Commitments and the progress of the divestiture process as well as potential purchasers. In addition to these reports, the Monitoring Trustee shall promptly report in writing to the Commission, sending to each of the Parties a non-confidential copy at the same time, if it concludes on reasonable grounds that the Parties are failing to comply with these Commitments; and

(vii) within one week after receipt of the documented proposal referred to in paragraph 14, submit to the Commission a reasoned opinion as to the suitability and independence of the proposed purchaser(s) and the viability of the Divestment Assets after the sale and as to whether the Divestment Assets are being sold in a manner consistent with the conditions and obligations attached to the Decision, in particular, if relevant, whether the sale of the Divestment Assets without one or more of the elements included in the Divestment Assets affects the viability of the Divestment Assets after the sale, taking account of the proposed purchaser(s) in accordance with Section A(2) above.

Duties and obligations of the Divestiture Trustee

23. Within the Trustee Divestiture Period, the Divestiture Trustee shall sell at no minimum price those sets of Divestment Assets that have not already been sold to one or more purchasers, provided that the Commission has approved both the purchaser(s) and the final binding sale and purchase agreement in accordance with the procedure laid down in paragraph 14. The Divestiture Trustee shall include in the sale and purchase agreement(s) such terms and conditions as it considers appropriate for an expedient sale in the Trustee Divestiture Period. In particular, the Divestiture Trustee may include in the sale and purchase agreement(s) such customary representations and warranties and indemnities as are reasonably required to effect the sale. The Divestiture Trustee shall protect the legitimate financial interests of the Parties, subject to the Parties' unconditional obligation to divest at no minimum price in the Trustee Divestiture Period.

24. In the Trustee Divestiture Period (or otherwise at the Commission's request), the Divestiture Trustee shall provide the Commission with a comprehensive monthly report written in English on the progress of the divestiture process. Such reports shall be submitted within 15 days after the end of every month with a simultaneous copy to the Monitoring Trustee and a non-confidential copy to each of the Parties.

III. Duties and obligations of the Parties

25. The Parties shall provide and shall cause their advisors to provide the Trustee with all such cooperation, assistance and information as the Trustee may reasonably require to perform its tasks. The Trustee shall have full and complete access to any of the Parties' books, records, documents, management or other personnel, facilities, sites and technical information necessary for fulfilling its duties under the Commitments and the Parties shall provide the Trustee upon request with copies of any document. Each of the Parties shall make available to the Trustee one or more offices on their premises and shall be available for meetings in order to provide the Trustee with all information necessary for the performance of its tasks.

26. The Parties shall provide the Monitoring Trustee with all managerial and administrative support that it may reasonably request on behalf of the management of the Divestment Assets. This shall include all administrative support functions relating to the Divestment Assets which are currently carried out at headquarters level. The Parties shall provide and shall cause their advisors to provide the Monitoring Trustee, on request, with the information submitted to potential purchasers, in particular give the Monitoring Trustee access to the data room documentation and all other information granted to potential purchasers in the due diligence procedure. The Parties shall inform the Monitoring Trustee on possible purchasers, submit a list of potential purchasers, and keep the Monitoring Trustee informed of all developments in the divestiture process.

27. The Parties shall grant or procure Affiliated Undertakings to grant comprehensive powers of attorney, duly executed, to the Divestiture Trustee to effect the sale, the Closing and all actions and declarations which the Divestiture Trustee considers necessary or appropriate to achieve the sale and the Closing, including the appointment of advisors to assist with the sale process. Upon request of the Divestiture Trustee, the Parties shall cause the documents required for effecting the sale and the Closing to be duly executed.

28. The Parties shall indemnify the Trustee and its employees and agents (each an "Indemnified Party") and hold each Indemnified Party harmless against, and hereby agree that an Indemnified Party shall have no liability to the Parties for any liabilities arising out of, the performance of the Trustee's duties under the Commitments, except to the extent that such liabilities result from the wilful default, recklessness, gross negligence or bad faith of the Trustee, its employees, agents or advisors.

29. At the expense of the Parties, the Trustee may appoint advisors (in particular for corporate finance or legal advice), subject to the Parties' approval (this approval not to be unreasonably withheld or delayed) if the Trustee considers the appointment of such advisors necessary or appropriate for the performance of its duties and obligations under the Mandate, provided that any fees and other expenses incurred by the Trustee are reasonable. Should the Parties refuse to approve the advisors proposed by the Trustee the Commission may approve the appointment of such advisors instead, after having heard the Parties. Only the Trustee shall be entitled to issue instructions to the advisors. Paragraph 28 shall apply mutatis mutandis. In the Trustee Divestiture Period, the Divestiture Trustee may use advisors who served the Parties during the First Divestiture Period if the Divestiture Trustee considers this in the best interest of an expedient sale.

IV. Replacement, discharge and reappointment of the Trustee

30. If the Trustee ceases to perform its functions under the Commitments or for any other good cause, including the exposure of the Trustee to a conflict of interest:

(a) the Commission may, after hearing the Trustee, require the Parties to replace the Trustee; or

(b) the Parties, with the prior approval of the Commission, may replace the Trustee.

31. If the Trustee is removed according to paragraph 30, the Trustee may be required to continue in its function until a new Trustee is in place to whom the Trustee has effected a full hand over of all relevant information. The new Trustee shall be appointed in accordance with the procedure referred to in paragraphs 15-20.

32. Beside the removal according to paragraph 30, the Trustee shall cease to act as Trustee only after the Commission has discharged it from its duties after all the Commitments with which the Trustee has been entrusted have been implemented. However, the Commission may at any time require the reappointment of the Monitoring Trustee if it subsequently appears that the relevant remedies might not have been fully and properly implemented.

Section F. The Review Clause

33. The Commission may, where appropriate, in response to a request from the Parties showing good cause and accompanied by a report from the Monitoring Trustee:

(i) grant an extension of the time periods foreseen in the Commitments; or

(ii) waive, modify or substitute, in exceptional circumstances, one or more of the undertakings in these Commitments.

Where the Parties seek an extension of a time period, they shall submit a request to the Commission no later than one month before the expiry of that period, showing good cause. Only in exceptional circumstances shall the Parties be entitled to request an extension within the last month of any period.

In addition, it is particularly contemplated that the Commission may waive, modify or substitute one or more of the Commitments referred to in (a) and (b) below, without the need for the involvement of the Monitoring Trustee, in the specific circumstances mentioned in (a) and (b) below:

(a) The Commission may waive the requirement set out in paragraph 6 where the Parties demonstrate that the Purchaser(s) or any other provider of financial information products shall have entered into one or more significant exclusivity contracts for Aftermarket Research or Estimates with brokers.

(b) The Commission may waive the requirement to include "standardised" Fundamentals within the Thomson Fundamentals Database where the relevant Purchaser indicates that it does not need "standardised" Fundamentals to be included in the divestment package.

...................................................................

duly authorised for and on behalf of The Thomson Corporation

..................................................................

duly authorised for and on behalf of Reuters Group PLC

SCHEDULE 1

1. The Divestment Assets as operated to date has the following legal and functional structure:

The remedy proposed involves the sale of the Databases and related assets, as described in more detail in Section B. These are not stand alone businesses. As such, the Divestment Assets do not have a "legal and functional structure".

The Databases are defined as follows:

"Reuters Estimates Database" means a copy of the complete collection of Estimates as included in the Reuters Knowledge Direct-Estimates product, in both "detailed" and "consensus" forms, including a copy of any data used in support of such Database created by or in connection with the collection process, such as data relating to inclusions/exclusions of estimates from consensus or specific accounting/collection treatments for specific companies, as of the Closing Date.

"Reuters Aftermarket Research Database" means a copy of the complete collection of Aftermarket Research as included in the Reuters Knowledge product, including a copy of any data used in support of such Database created by or in connection with the collection process (although the Parties note that no such data currently exists) as of the Closing Date.

"Thomson Fundamentals Database" means a copy of the complete collection of "as reported" Fundamentals as included in the "Enterprise FX" product and a copy of the complete collection of "standardised" Fundamentals as included in the Thomson "Worldscope File2" product), together with a copy of any data used in support of such Database created by or in connection with the collection process, in particular such copies of the underlying source documents relating thereto as Thomson has in its possession, as of the Closing Date.

"Reuters Economics Database" means a copy of the complete collection of Economic Data as included in the Reuters EcoWin Economic database product, including a copy of any data used in support of such Database created by or in connection with the collection process, as of the Closing Date.

2. Following paragraph 3 of these Commitments, the Divestment Assets includes, but is not limited to:

(a) the following main tangible assets:

(b) the following main intangible assets: Tangible and intangible assets are dealt with together below.

With respect to each Database, the Divestment Assets will comprise:

(i) copies of such tangible embodiments of the relevant Database (including but not limited to electronic copies) as may be necessary or appropriate to enable the Purchaser to utilise the Database for the purposes set forth in these Commitments, to the extent the requisite broker, other research contributor or other third party approval or consents are obtained. This is subject to the proviso that if such approval or consents are obtained with respect to a broker, contributor or other third party after the Closing Date but during such period as the Parties are required to procure such approval or consents for such broker, contributor or other third party under paragraph 5, the relevant data for such broker, contributor or other third party will be transferred to the Purchaser as such data exists as of the date of such subsequent transfer;

(ii) a perpetual, worldwide, assignable, sub-licensable, transferable, royalty-free, nonexclusive licence to market, distribute, and prepare derivative works of the Database (and to manufacture, reproduce, and have reproduced such derivative works), without restriction and without further compensation to the Parties, to the extent Thomson (in the case of the Thomson Fundamentals Database) or Reuters (in the case of the Reuters Estimates Database, the Reuters Historical Research Database or the Reuters Economics Database) has such rights;

(iii) a perpetual, worldwide, assignable, sub-licensable, transferable, royalty free, nonexclusive licence of all IP rights, formulations, specifications, trade secrets, know-how, and technical information embodied in, pertaining exclusively or predominantly to, or necessary for the integration, operation or updating of, the Database (other than those excluded under paragraph 3 of Schedule 1) to the extent Thomson (in the case of the Thomson Fundamentals Database) or Reuters (in the case of the Reuters Estimates Database, the Reuters Historical Research Database or the Reuters Economics Database) has such rights, including (a) copies of all training and other manuals, know-how, workflow documents and instructions maintained by the owner of the Database which are used exclusively in connection with the Database, and copies of relevant parts of such training and other manuals, know-how and instructions which are used non-exclusively in connection with the Database; and (b) the unique numerical identifiers used by the owner of the relevant Database to identify companies in the Database and tables to map such identifiers to industry-standard third party identifiers (such as exchange ticker symbols and CUSIP numbers) but not including any additional proprietary identification coding which is not necessary to ensure the effective integration, operation or updating of the Databases; and

(iv) to the extent desired by the relevant Purchaser(s), copies of and a perpetual, worldwide, assignable, non-licensable, transferable, royalty-free non-exclusive license to use the following relevant Thomson (in the case of the Thomson Fundamentals Database) or Reuters (in the case of the Reuters Estimates Database, the Reuters Aftermarket Research Database or the Reuters Economics Database) software (including source code and all documentation relating thereto and including any subsequent version thereof in active development at the Closing Date), namely:

(A) software used exclusively or predominantly to integrate, operate and update any of the Databases, including without limitation:

I. software for collection, aggregation, normalisation and transmission so as to enable the Purchaser of the relevant Database to integrate and operate it on an ongoing basis;

II. software for click-through functionality (i.e. to access the underlying source document) in the Thomson Fundamentals Database; and

III. purchasing and billing software for Aftermarket Research; and

(B) software otherwise necessary for the integration, operation and updating of the Databases; provided that if any software described in sub-clause (iv)(B) above is substantially used in connection with other operations by Reuters or Thomson (such as permissioning software used in connection with research and estimates), Reuters or Thomson will instead grant a licence to, and provide, the components of such software, or a customised version of such software, or a licence to use such software solely for the purpose of integrating, operating and updating the Database, or such other elements of such software, in each case as desired by the relevant Purchaser to integrate, operate and update the Database, or if it is not technically feasible to provide such tailored components, versions or elements of such software, Reuters or Thomson shall provide such software or services related thereto to Purchaser on a transitional basis for up to 12 months to enable Purchaser to operate the Database (v) to the extent desired by the relevant Purchaser(s), a perpetual, worldwide, assignable, sub-licensable, transferable, royalty free, non-exclusive licence of all IP rights, formulations, specifications, trade secrets, know-how, technical information relating to any ongoing research and development efforts pertaining exclusively or primarily to the integration ,operation or updating of the Databases as of the Closing Date.

Additional specific included items are listed for each set of Divestment Assets below. A comprehensive technical description of the Divestment Assets will be provided to the Commission, potential purchasers and the Trustee.

• Fundamentals

(i) detailed descriptions of the data collection, normalisation and transformation process and methodologies;

(ii) detailed descriptions of the business logic used to map between as reported and standardised databases, including a guide to standardised data definitions.

• Estimates

(i) detailed descriptions of the data extraction, collection and normalisation process and methodologies; and

(ii) detailed information on processing electronic broker feeds and manual data entry applications.

• Aftermarket Research

(i) information on the various research feeds delivered by research providers, including the format, size and potential delivery mechanism; together with detailed documentation on industry best practice with reference to search, indexation and tagging e.g. RiXML standards and basic requirements from customers; and

(ii) description of reporting, billing and reconciliation obligations that are generally required by contributors with respect to royalties, payments, embargo periods, distribution arrangements and other core business operations.

• Economic Data

(i) detailed descriptions of each source of Economic Data used in the Reuters Economics Database; and

(ii) detailed descriptions of the collection, aggregation and delivery process and methodologies used for the Reuters Economics Database.

(c) the following main licences, permits and authorisations: No such licences, permits and authorisations are necessary to operate the Divestment Assets, but see paragraph 5 of the Commitments.

(d) the following main contracts, agreements, leases, commitments and understandings See paragraph 5 of the Commitments.

(e) the following customer, credit and other records: No customer, credit or other records will be included within the Divestment Assets, other than the following, to the extent desired by the relevant Purchaser(s):

(i) the Parties will assign to the relevant Purchaser, to the extent assignable, any contracts or assignable portions thereof specifically for Direct Content Datafeed delivery of data from the Thomson Fundamentals Database or the Reuters Estimates Database including any contracts for delivery to clients by or through redistributors;

(ii) to the extent that such customer contracts are not assignable, the Parties will allow such customers to terminate early (without penalty) any obligations specifically relating to such Direct Content Datafeed delivery (where the Direct Content Datafeed revenue is separately identifiable under the customer contract) in connection with such customer taking a competing service from the relevant Purchaser any time within 12 months after the Closing Date. The Parties will not interfere with the negotiations between the Purchaser(s) and such customers.

The Parties will provide the Purchaser(s) of the Thomson Fundamentals Database and the Reuters Estimates Database with a list of such customers as of the Closing Date. the following Key Personnel:

See the list set out at Schedule 2.

and

(f) the arrangements for the supply with the following products or services by the Parties or Affiliated Undertakings for a transitional period of up to [•] after Closing:

Pursuant to a transitional support services agreement on customary arm's-length commercial terms and conditions, to be approved by the Commission, the Parties will, at the request of the Purchaser(s), provide and where relevant procure the provision, for a period of up to [confidential] from the Closing Date (and, in the case of the Thomson Fundamentals Database and the Reuters Estimates Database, at the option of the relevant Purchaser, for one additional [confidential] period), of:

(i) consulting and support services to each Purchaser sufficient to give that Purchaser a full understanding of the structure and content of the relevant Database(s); and

(ii) regular updates to the Databases to enable the relevant Purchaser(s) to operate on the market immediately and compete effectively (such updates to be provided with the same regularity as the updates provided to Thomson or Reuters customers respectively to the extent desired by the Purchaser(s).

3. The Divestment Assets shall not include:

Customer contracts other than those specifically included, accounts or pricing information (other than in any customer contracts to be assigned); redistribution contracts other than those specifically included; existing operations and operational assets; hardware or software acquired by the Parties on the open market; trademarks, service marks or brands or any licenses thereto (including without limitation any rights to use the names "Thomson", "Reuters", "Worldscope", "Multex" or "EcoWin" alone or in connection with any of the Divestment Assets); any IP rights or customer contracts relating to other product offerings, platforms or delivery mechanisms of the Parties (such as Thomson One, Datastream, and Thomson Quantitative Analytics product families, Reuters 3000 Xtra, Reuters Knowledge or EcoWin Pro product families, in each case, including associated Excel add-ins, the Parties' API's and RMDS); network technology; land and buildings; goodwill; and employees other than Key Personnel and Personnel; any obligation to support or maintain any software or other IP transferred to the Purchaser except as set forth herein or in any transitional arrangements.

SCHEDULE 2 - LIST OF KEY PERSONNEL

[Confidential - APPROXIMATELY 30 KEY PERSONNEL IN TOTAL FOR FUNDAMENTALS, RESEARCH AND ESTIMATES AND ECONOMIC DATA]

SCHEDULE 3 - ADDITIONAL PERSONNEL

[Confidential - APPROXIMATELY 600 ADDITIONAL PERSONNEL FOR FUNDAMENTALS AND ESTIMATES]