Svenska Handelsbanken AB & Ors v ESMA (Board of Appeal) [2019] EUBOA BoA_2019_01 (27 February 2019)

[ Home ] [ Databases ] [ World Law ] [ Multidatabase Search ] [ Help ] [ Feedback ] [ DONATE ] Board Of Appeal Of The European Supervisory Authorities Decisions You are here: BAILII >> Databases >> Board Of Appeal Of The European Supervisory Authorities Decisions >> Svenska Handelsbanken AB & Ors v ESMA (Board of Appeal) [2019] EUBOA BoA_2019_01 (27 February 2019) URL: https://www.bailii.org/eu/cases/EUBOA/2019/BoA_2019_01.html Cite as: [2019] EUBOA BoA_2019_01html, [2019] EUBOA BoA_2019_1html [ New search ] [ Printable PDF version ] [ Help ] BoA D 2019 01 BoA D 2019 02 BoA D 2019 03 BoA D 2019 04 D E C I S I O N S given by the BOARD OF APPEAL OF THE EUROPEAN SUPERVISORY AUTHORITIES under Article 60.4 Regulation (EU) No 1095/2010 and the Board of Appeal’s Rules of Procedure (BOA 2012 002) in Appeals by Svenska Handelsbanken AB [Appellant] Skandinaviska Enskilda Banken AB [Appellant] Swedbank AB [Appellant] Nordea Bank Abp [Appellant] [together “the Appellants”] against decisions of The European Securities and Markets Authority (ESMA) [Respondent] Decisions Ref.: BoA-D-2019-01 BoA-D-2019-02 BoA-D-2019-03 BoA-D-2019-04 Board of Appeal William Blair (President) Juan Fernandez-Armesto (Vice-President) Beata Maria Mrozowska Katalin Mero Marco Lamandini Michele Siri Place of this decision: Frankfurt a. M. 27 February 2019 I n d e x Page I Introduction II The Board of Appeal III The Parties IV The right of Appeal V Background VI ESMA’s investigation VII The Decisions of the Board of Supervisors VIII The Appeals IX MiFID/MAR regimes applicable to investment recommendations / research X CRAR framework applicable to credit ratings XI The cases of the appellant banks XII The response of ESMA to the appeals XIII The applicable provisions XIV The Board of Appeal’s discussion and conclusions XV Summary of Board of Appeal’s decisions and relief granted XVI Decisions I Introduction 1. These are the decisions of the Board of Appeal of the European Securities and Markets Authority (ESMA) in respect of four appeals each of which raises issues which are the same or similar. On 9 October 2018, the Board of Appeal directed that the appeals were to be dealt with and heard at the same time. By agreement of all parties, the Board of Appeal’s decisions on all four appeals are dealt with in the same document. 2. The appellants, Svenska Handelsbanken AB, Skandinaviska Enskilda Banken AB, Swedbank AB, and Nordea Bank Abp, are credit institutions established in Sweden. They are referred to collectively in this decision as “the banks”. The question on the appeals is whether in issuing corporate research which included what have been called “shadow ratings”, the banks required to be registered under the Regulation (EC) No 1060/2009 on credit rating agencies (the “Credit Rating Agencies Regulation”, or “CRAR”). 3. The banks’ cases are that the activities fall within the “recommendations” or “investment research” exclusion in Article 3(2) of CRAR. By decisions issued on 11 July 2018, the Board of Supervisors of ESMA decided that the exclusion did not apply, and that the Regulation had been negligently infringed by each bank by failing to register. It adopted a supervisory measure in the form of a public notice, and imposed fines of EUR 495,000 on each bank. 4. These decisions were made public on 23 July 2018, and each is available on ESMA’s website. 5. The banks have appealed to the Board of Appeal under the provisions of Regulation (EU) No 1095/2010 (the “ESMA Regulation”), contending that the decisions of the Board of Supervisors are wrong, and that the cases should be remitted to the Board of Supervisors for amended decisions (a fifth bank, Danske Bank A/S, against which the same findings were made has not appealed). 6. ESMA contends that the decisions of the Board of Supervisors are correct, and that the appeals should be dismissed. 7. The facts and arguments of the four appellant banks are not in all respects the same. However, the main points for decision are common to all the banks and are in summary: (1) Whether the Board of Supervisors correctly interpreted the provisions of CRAR, and in particular whether the Board of Supervisors was correct in finding that Article 3(2) of CRAR, which provides that recommendations and investment research (as defined in the Market Abuse and MiFID Implementing Directives respectively) shall not be considered to be credit ratings, did not apply to exclude the activities of appellant banks from the requirements of CRAR, with the consequence that they had committed an infringement listed in Annex III of CRAR by issuing credit ratings without being registered by ESMA as credit agencies; (2) Whether in reaching its decisions the Board of Supervisors breached the principles of legality and legal certainty; (3) Whether the Board of Supervisors was correct to find that the appellant banks “negligently” committed the infringement, that being a condition of ESMA’s power to impose a fine; (4) Whether the Board of Supervisors correctly applied the sanctions provisions in CRAR, and whether the sanctions imposed were just and proportionate. 8. One of the banks (Skandinaviska Enskilda Banken) contends that the criteria of public disclosure or distribution by subscription in Article 2(1) CRAR have not been satisfied in its case. 9. Skandinaviska Enskilda Banken applied under the provisions of the ESMA Regulation to suspend the Board of Supervisors’ decision against it until the appeal proceedings were concluded. The Board of Appeal refused that application by decision of 30 November 2018. 10. The parties exercised their right under Article 60(4) of the ESMA Regulation to make oral representations, which were heard by the Board of Appeal in Frankfurt on 6 February 2019. II The Board of Appeal 11. ESMA’s Board of Appeal is part of the governance structure of ESMA (and the other European Supervisory Authorities of which it is a joint body under their founding regulations). The members are required to be independent in making their decisions, and undertake to act independently and in the public interest (Article 59 of the ESMA Regulation). 12. As an appeal body of ESMA, the Board of Appeal must decide whether the decisions of the Board of Supervisors were correct or not, and may confirm the decisions or remit the cases to the Board of Supervisors. Article 60(5) of the ESMA Regulation provides that: “5. The Board of Appeal may confirm the decision taken by the competent body of the Authority, or remit the case to the competent body of the Authority. That body shall be bound by the decision of the Board of Appeal and that body shall adopt an amended decision regarding the case concerned.” It is no part of the Board of Appeal’s function to decide policy, which is a matter for the Board of Supervisors (within the limits of the ESMA Regulation and its legal basis). 13. These are complex appeals, and all five parties have filed a considerable volume of written submissions, with supporting documents, and each has exercised its right under Article 60(4) of the ESMA Regulation to make oral representations. It is noteworthy that there is much more by way of legal submissions before the Board of Appeal than there was before the Board of Supervisors. The oral representations were particularly helpful in illuminating the issues. 14. The Board of Appeal acknowledges the high quality of these submissions. All the parties’ contentions have been taken into account, whether expressly referred to herein or not. The position of each appellant has been considered separately, and the decision in respect of each is set out at the end. In the event, the decisions of the Board of Appeal in respect of each appellant are the same, as they were before the Board of Supervisors. 15. Upon the hearing of the oral representations, the evidence was complete in relation to each of the appeals under Article 20 of Board of Appeal’s Rules of Procedure for the purposes of Article 60(2) of the ESMA Regulation. III The parties 16. The appellants are each legal persons established in the European Union (see Article 14(1) CRAR), namely: (1) Nordea Bank Abp (“Nordea”), a credit institution established in Sweden and represented in the appeals by Linklaters LLP. (2) Swedbank AB (“Swedbank”), a credit institution established in Sweden and represented in the appeals by Gernandt & Danielsson Advokatbyrå KB. (3) Skandinaviska Enskilda Banken AB (“SEB”), a credit institution established in Sweden and represented in the appeals by Allen & Overy LLP. (4) Svenska Handelsbanken AB (“Handelsbanken”), a credit institution established in Sweden and represented in the appeals by Roschier Advokatbyrå AB. 17. The respondent is the European Securities and Markets Authority (“ESMA”) established by Regulation (EU) No 1095/2010 and represented by Cleary Gottlieb Steen & Hamilton LLP. IV The right of appeal 18. The appeals are brought under the ESMA Regulation, which by Article 6 includes in its composition a Board of Supervisors and a Board of Appeal. 19. Article 6(5) of the ESMA Regulation provides for the Board of Appeal to exercise the tasks set out in Article 60. Article 60(1) gives a right of appeal against decisions of the Authority as follows: “Any natural or legal person, including competent authorities, may appeal against a decision of the Authority referred to in Articles 17, 18 and 19 and any other decision taken by the Authority in accordance with the Union acts referred to in Article 1(2) which is addressed to that person, or against a decision which, although in the form of a decision addressed to another person, is of direct and individual concern to that person.” 20. The Credit Rating Agencies Regulation is among the Union acts referred to in Article 1(2). It is not in contention that the appellant banks have a right of appeal to the Board of Appeal against the decisions of the Board of Supervisors under Article 60 of the ESMA Regulation. V The background 21. The factual background is largely uncontested. For many years, investment research and recommendations in respect of fixed income securities (such as corporate bonds etc) issued by corporates have been published by various Nordic banks including so-called “shadow ratings”. Though it has been said on behalf of the banks that the term is unfortunate, it is often used (including by one of the appellant banks in a credit research report), and the Board of Appeal will use it whilst making it clear that no legal conclusions can be drawn from its use. 22. These “shadow ratings” were not those of the “official” rating agencies (which registered under CRAR when it came into force), but composed (for example) by the bank’s credit analysts, and treated as part of the recommendation: as it was put in oral representations, the members of the bank’s credit research team would discuss among themselves, weigh up the material, and use their experience, in order to transcribe their views into a notation. As appears from the decisions of the Board of Supervisors, in the case of two of the banks, the ratings were based in whole or in part on the methodology of the “official” rating agencies. 23. Unlike “official” ratings published by the rating agencies (which the Board of Supervisors found are often, but not exclusively, paid for by the relevant issuer), these ratings were not paid for by the issuer. 24. An example of such a recommendation dated 19 December 2011 was produced in its oral representations by Swedbank in respect of a bond issue by a Swedish company. The recommendation is described at the top of the document as “overweight” (which in context is a recommendation to buy), and at the bottom of the document is stated, “We view […] as an investment grade issuer in the BBB+ range”. 25. This example was helpful with the caveat that the examples seen by the Board of Appeal (as part of the files of the appeals) differ widely in form, and this is the shortest. Another convenient example was provided by Swedbank which showed that in a 36 page credit research document there was (as Swedbank emphasised) a single line stating, “Issuer shadow rating: BBB-/Stable”. 26. Where there were relevant “public ratings” i.e. credit ratings on the same issuer or instrument issued by registered CRAs, the Board of Supervisors stated that such ratings were generally included in the research reports, where applicable, in close proximity to the bank’s shadow rating. (Swedbank is critical of this finding on the basis that it does not take account of the fact that only 10% of its research reports included reference to public ratings.) 27. As it was put in oral representations by the banks, the length, comprehensiveness and extent of the research reports and opinions on relative creditworthiness vary considerably. In some cases, the report itself is 30 or 40 pages long and the information and analysis on relative creditworthiness is part of the report. In other cases, the report is only two or three pages, in which cases all information in the report constitutes information on relative creditworthiness, presented with the bank’s view on relative creditworthiness. 28. However, whatever the precise form, the key point in the submissions as put on behalf of ESMA is the inclusion of an alphanumerical rating in the text. It is this which the Board of Supervisors considered brought the documents in question outside the investment research exclusion in CRAR. This is notwithstanding that it is common ground that these reports, or at least elements of the reports, can be characterised as MiFID investment research (or recommendations under the Market Abuse framework), which gives rise to the first and most important aspect of the dispute as to the true construction of the exclusion. 29. There is much material before the Board of Appeal to the effect that this practice of “shadow rating” was widely regarded as useful, particularly for small to medium sized enterprises (SMEs) for which the cost of obtaining an “official” rating could be prohibitive. This opinion appears for example in a proposal for amending the Credit Rating Agencies Regulation by Finans Norge (the industry organisation for the financial industry in Norway), which was also relied on by ESMA in its submissions (it advocates adding a further exemption to CRAR). It was also expressed by the European Commission Expert Group on Corporate Bonds in November 2017. The “shadow ratings” were also regarded as “public goods”, as illustrated in the market analysis in Economic Commentaries No. 7, 2014 on “ The development of the Swedish market for corporate bonds ” published by the Sveriges Riksbank (the Swedish Central Bank), which is referred to below. 30. As a matter of background, the credit ratings market in the EU is dominated by three agencies, S&P Global Ratings, Moody's Investor Services and Fitch Ratings. Despite measures to encourage the use of smaller credit rating agencies (see recital (11) of Regulation (EU) No 462/2013), ESMA’s November 2018 statistics show that between them these three agencies still have 93.4% of the EU market. 31. The reasons for the legal changes which have given rise to the present dispute have been explained by the Board of Appeal most recently in FinancialCraft Analytics Sp. z o.o. (formerly named Global Rating Sp. z o.o.) v ESMA , 3 July 2017, paragraph 33 et seq, and is well known. 32. In short, credit rating agencies largely fell outside the scope of financial regulation until the financial crisis of 2007-8. Credit ratings were (and are) perceived to help investors to understand the risks associated with a particular investment or financial instrument. However, the financial crisis and subsequent sovereign debt crisis of 2011-2012 brought to light manifest deficiencies in the operation of the agencies, the accuracy of their ratings, and their perceived lack of independence. 33. As it has been put by the European Commission, “In the period leading up to the financial crisis in 2008, credit rating agencies (CRAs) failed to properly appreciate the risks in more complex financial instruments. For instance, structured finance products backed by risky sub-prime mortgages were issued with incorrect ratings that were far too high. During the subsequent euro area debt crisis, certain countries were faced with abrupt bond sell-offs and higher borrowing costs following a downgrade of their credit rating”. Such deficiencies were (and are) seen as a threat to financial stability. 34. As stated in recital (10) of the Credit Rating Agencies Regulation (reference also being made to recitals 2 and 11): “Credit rating agencies are considered to have failed, first, to reflect early enough in their credit ratings the worsening market conditions, and second, to adjust their credit ratings in time following the deepening market crisis. The most appropriate manner in which to correct those failures is by measures relating to conflicts of interest, the quality of the credit ratings, the transparency and internal governance of the credit rating agencies, and the surveillance of the activities of the credit rating agencies...” 35. The EU response as summarised by the European Commission was as follows: “In response, the Commission made proposals to strengthen the regulatory and supervisory framework for CRAs in the EU, to restore market confidence and increase investor protection. The new EU rules were introduced in three consecutive steps. The first set of rules, which entered into force at the end of 2009, established a regulatory framework for CRAs and introduced a regulatory oversight regime, whereby CRAs had to be registered and were supervised by national competent authorities. In addition, CRAs were required to avoid conflicts of interest, and to have sound rating methodologies and transparent rating activities. In 2011, these rules were amended to take into account the creation of the European Securities and Markets Authority (ESMA), which supervised CRAs registered in the EU. A further amendment was made in 2013 to reinforce the rules and address weaknesses related to sovereign debt credit ratings.” 36. As is explained further below, the Credit Rating Agencies Regulation provided for the first time for credit rating agencies to register with their respective national financial supervisory authorities. CRAR was amended by Regulation (EU) No 513/2011 of 11 May 2011 which conferred responsibility for the registration and supervision of credit rating agencies on ESMA from 1 July 2011 (and, as appears from the above, amended again subsequently). VI ESMA’s investigation 37. Following a preliminary investigation, in December 2016 ESMA’s Supervision Department concluded that there were serious indications on the part of each of the appellant banks of the possible existence of facts liable to constitute one or more of the infringements listed in Annex III to CRAR, that is, the failure to apply for registration as a credit rating agency. 38. The banks took a different view, asserting that there was no registration requirement since the recommendations/investment research exclusion in Article 3(2) CRAR applied to the activities in question. 39. Nevertheless, the banks voluntarily ceased the practice of including what ESMA considered to be ratings within CRAR in their investment research and recommendations pending the outcome (Nordea, Swedbank and Handelsbanken in August 2016, and SEB in May 2018 following receipt of the Board’s initial Statement of Findings). There was therefore no need for ESMA to consider using its powers to prohibit or suspend the use of the ratings. 40. In January 2017, ESMA appointed an independent investigating officer (“IIO”) pursuant to Article 23e(1) of CRAR to investigate the matter. The IIO sent her initial statement of findings to each of the banks in June 2017. This contained a detailed examination of the law and facts in the case of each bank. Much of the analysis remains relevant on the appeals. 41. Over the next few months, the banks responded by way of written submissions. On 27 September 2017, the IIO submitted to the Board of Supervisors her file relating to the cases, which included an amended statement of findings. Though the facts were those applicable to the individual bank, the analysis and reasoning was the same for all the banks. She found that the banks had committed the infringement, but had not done so negligently. On that basis, by Article (1) of CRAR, there was no power to impose a fine. 42. The Board of Supervisors discussed the IIO’s findings and the cases at a number of meetings, and on 17 May 2018, ESMA sent the Statements of Findings to the banks on behalf of the Board of Supervisors. Over the next two months, the banks provided written submissions to ESMA in relation to the matter. 43. The Board of Supervisors discussed the cases further at its meeting on 11 July 2018, and on that day adopted the decisions that are challenged by the banks in these appeals. The Board of Supervisors agreed with the IIO’s conclusion as to the applicability of CRAR to the banks’ activities, but took a different view on the issue of negligence from that which she had taken. VII The Decisions of the Board of Supervisors 44. The four Decisions of the Board of Supervisors dated 11 July 2018 were in the same form for each of the banks. The Annex contained the Statement of Findings of the Board, and though the facts were those applicable to the individual bank, the central analysis and reasoning was the same for all the banks. 45. Some passages in the Decisions have been redacted on confidentiality grounds. The Board of Appeal has not been asked to look at the redacted passages. 46. The Decisions of 11 July 2018 include recitals which state that: “12. On the basis of the file containing the IIO’s findings and having considered the submissions made on behalf of [the bank], the Board finds that [the bank] negligently committed the infringement set out at point 54 of Section I of Annex III of Regulation (EC) No 1060/2009 [ that is, the failure to apply for registration as a credit rating agency ]. 13. Pursuant to Article 24 of Regulation (EC) No 1060/2009, the Board adopts a supervisory measure in the form of a public notice. 14. Pursuant to Article 36a of Regulation (EC) No 1060/2009, the Board also imposes a fine on [the bank] as calculated in the Annex to this Decision.” 47. These recitals are then restated and implemented in Articles 1, 2, and 3 of the Decisions. Article 3 states the fine imposed as EUR 495,000, the fine being in the same amount for each bank. The Public Notice is in the form of a draft appendix to the Decisions, and again is substantially in the same form. 48. In summary, the Board of Supervisors’ analysis and reasoning, based on the findings of the IIO, and the banks’ responses, was as follows: (1) Over the relevant period (between 2011 and 2016), the banks conducted credit research activities, which included the issuing of what were described as credit, creditworthiness, credit research, or investment research reports. These reports tended to relate either to the issuers of bonds or other debt instruments or to those instruments themselves. A number of these reports included opinions that were variously described as (for example): a. A ‘Corporate rating’, ‘Company Rating’, ‘Bond rating’ or ‘shadow rating’. ( Nordea ) b. A ‘shadow rating’, ‘Corporate rating’, ‘Swedbank estimated Issuer rating’, ‘Swedbank estimated Bond rating’ or ‘Swedbank issuer rating’. ( Swedbank ) c. A ‘Corporate rating’, ‘Stand-alone rating’ or ‘Credit rating’ ( SEB ) d. An ‘indicative rating’, ‘indicative corporate rating’, ‘indicative issue rating’ and ‘indicative issuer rating’ ( Handelsbanken ). These ratings are collectively referred to in the Decisions as “the Ratings”. (2) The Ratings were opinions on the creditworthiness of two of the types of entity, issuer, financial instrument or other asset specified in the definition of a credit rating, specifically debt instruments and the issuers of such instruments. (3) These opinions were issued using an established and defined system of rating categories. The Board of Supervisors considered that: a. “the Ratings considered by the Board all appear to use rating categories, involving as they do rating symbols representing differing levels of risk in relation to the entity, issuer, financial instrument or other asset specified in the definition of a credit rating being assessed”, e.g. ‘A-’, ‘BBB+’ and ‘BB-’” ( Nordea ) b. As “stated by Swedbank itself, “[b]asically our rating methodology is a similar approach as for the main rating agencies, [redacted due to confidentiality]”. The sample reports reviewed by the Board of Supervisors included rating letters representing differing levels of risk relating to the entity, issuer, financial instrument or other asset specified in the definition of a credit rating being assessed”. ( Swedbank ) c. The “Ratings considered by the Board all appear to use rating categories, involving as they do rating symbols representing differing levels of risk in relation to the entity, issuer, financial instrument or other asset specified in the definition of a credit rating being assessed e.g ‘A+’, ‘A-’ and ‘B+’”. ( SEB ) d. The “Handelsbanken has stated that its ‘core framework for analysing non-financial companies is a hybrid that is partly based on S&P’s methodology but which has been modified in key areas. The Board noted the rating categories employed by Handelsbanken that represented different levels of risk in the range from ‘AAA’ to ‘D’, where ‘[t]he scale measures the relative creditworthiness with AAA representing the highest indicative rating (and the lowest credit risk) and D (default) representing the lowest indicative rating’”. ( Handelsbanken ) (4) On that basis, the Board of Supervisors concluded that the Ratings appeared to meet the definition of a credit rating under CRAR. (5) The Board of Supervisors then considered the contention of the banks that their activities nevertheless fell within Article 3(2)(b) CRAR which provides that investment research as defined in MiFID and other forms of general recommendation relating to transactions in financial instruments or to financial obligations shall not be considered to be credit ratings. Its analysis was as follows: a. The legislation is not definitive as to whether ‘credit ratings’ and ‘investment research’ or ‘recommendations’ are mutually exclusive terms or if there is an overlap between them, or indeed if they are related in some other way. b. On the basis of the material before it, including the IIO’s view, and without expressing a firm or settled view, a credit rating is a distinct concept from recommendations and investment research in this context. c. If that is the case, it is also possible that a given document could contain both investment research or recommendation and a credit rating, depending on the character of the opinions put forward and the manner in which they are expressed. That is, an opinion, contained in a publicly-available document (or one distributed by subscription) that otherwise comprises investment research, which relates to the creditworthiness of an entity, issuer, financial instrument or other asset set out in the definition of a credit rating in Article 3(1)(a) CRAR and which is issued using an established and defined ranking system of rating categories, is likely to be considered a credit rating within the scope of that Regulation. d. This view, which was not settled, was reached taking into consideration the aims of CRAR. Producers of investment research and other forms of recommendation will be likely to be subject to regulation under MiFID and MAR (Market Abuse Regulation) in respect of its production. Nevertheless, CRAR establishes a separate regime with distinct objectives, for example that issued credit ratings are of adequate quality. If credit ratings (that is, opinions that meet the definition of a credit rating) could be included in investment research or other recommendations published by entities not registered as credit rating agencies, it is possible that these aims (such as credit ratings being of adequate quality) might be frustrated. e. It did not wholly accept the suggestion that investment research or recommendations could contain a rating scale and not be considered to be credit ratings, as its categorisation as a credit rating would seem to be more likely, in light of the aims of the Regulation. It follows that the legislator might indeed have intended to bring already-regulated entities within the scope of the CRAR if the activities of those entities extend to the substance of that legislation. f. It did not accept the assertion that investment research is exempted from the definition of a credit rating even where it ‘includes a view on creditworthiness of an issuer and/or a rating’. An opinion on creditworthiness issued using a rating category would be more likely to fall, depending on the particular facts, within the meaning of a credit rating. CRAR is concerned with the activities of entities, rather than with their existing state of regulatory supervision. (6) Upon that analysis, the Board of Supervisors formed the view that the Ratings are credit ratings within the meaning of CRAR. It reached this view on the basis of the facts in the case i.e. on the material in the IIO file. Considering the Ratings themselves, they fall most precisely within the CRAR definition of “credit ratings”. The Ratings did not fall within the definitions of either a recommendation or investment research. In particular, the Ratings (in the sense of the rating categories) do not in themselves appear to recommend or suggest an investment strategy, which would have been expected of investment research pursuant to its definition in Article 24 MiFID. (7) The banks’ occupation included the issuing of credit ratings on a professional basis. (8) The banks issued credit ratings that were disclosed publicly or distributed by subscription. (9) The banks had not applied for registration as a CRA. (10) As previously stated in decisions of the Board of Supervisors, negligence is established for a CRA where, as a professional firm in the financial services sector subject to stringent regulatory requirements, it is required to take special care in assessing the risks that its acts or omissions entail, and has failed to take that care. Further, as result of that failure, the CRA has not foreseen the consequences of its acts or omissions, including particularly its infringement of CRAR, in circumstances when a person in such a position who is normally informed and sufficiently attentive could not have failed to foresee those consequences. (11) The Court of Justice of the European Union (“CJEU”) jurisprudence suggests that the concept of a negligent infringement of CRAR is to be understood as denoting a lack of care on the part of a CRA in complying with CRAR. A high standard of care is to be expected of a CRA. (12) An entity must take special care to comply with CRAR. Steps to meet this obligation might have included, for example, an initial evaluation as to whether CRAR might apply to its issuing of the Ratings, the taking of legal advice on the scope and effect of CRAR and/or seeking advice from their National Competent Authority and/or ESMA on those issues. In addition to such an initial evaluation (that is, prior to CRAR’s implementation), the banks might have been expected to subject any initial opinion to periodic review. (13) There was a very limited amount of material on the file that might provide evidence of any special care that the banks had taken to comply with CRAR. The lack of documentary evidence did not assist in determining whether the banks did take special care at this time in assessing the risks of this activity. While this limited amount of material was not in itself evidence of negligence, the Board of Supervisors could take the fact of it into account when considering the test for negligence. (14) The Board of Supervisors had decided not to follow the manner in which the test for negligence was applied by the IIO. (15) If the banks had taken special care in assessing the risks of their conduct, they would not have failed to foresee that the issuing of the Ratings would amount to an infringement of the CRAR. A normally informed party in the position of the banks would have foreseen the consequences of their actions. That is, in not applying to be registered as a CRA, the banks were committing an infringement of CRAR. (16) The banks negligently committed the infringement, though it was not committed deliberately. (17) Article 24(1) of CRAR provides that where the Board of Supervisors finds that a CRA has committed one of the infringements listed in Annex III of the Regulation, the Board must adopt one or more of the supervisory measures listed in that Article 50. In accordance with Article 24(2) of CRAR, the Board of Supervisors considered that it was appropriate to issue public notices in respect of the Infringements, notwithstanding objections on behalf of the banks that compliance had already been achieved. (18) Having found that the banks negligently committed the infringement by not applying for registration for the purposes of Article 2(1) of CRAR, Article 36a(1) of CRAR required the Board of Supervisors to impose a fine for the infringement which was calculated in accordance with Annex IV of CRAR at EUR 495,000 in respect of each bank. VIII The appeals 49. The Notices of Appeal in this matter were filed by Nordea on 10 September 2018, by Swedbank on 7 September 2018, by SEB on 6 September 2018 and by Handelsbanken on 6 September 2018. 50. SEB also made an application for suspension of the decision so far as it concerns SEB pursuant to Article 60(3) of the ESMA Regulation. This raised issues which are separate and distinct from those raised generally in the appeals. 51. The Notices of Appeal were received by ESMA between 7 and 10 September 2018, and by the Board of Appeal shortly afterwards. 52. On 13 September 2018, ESMA made an application to the Board of Appeal seeking an extension of the three week period under which it was required to serve its Responses to the Notices of Appeal under the Board of Appeal’s Rules of Procedure, and asked to be permitted to do so within four months, that is, by 11 January 2019. This was on the basis that there were four separate appeals involving complex considerations and over 2,500 pages of documents. 53. A directions teleconference was held on 8 October 2018, agreement being reached that for convenience the four appeals were henceforth to be dealt with and heard at the same time, though remaining separate appeals, each appellant continuing to make its own representation. Though ESMA’s submissions as to complexity were justified, it was agreed that its Responses should be served within a shorter period than that sought, namely by 20 November 2018. 54. On 9 October 2018, an order was issued by the President of the Board of Appeal setting a timetable. 55. On 8 October 2018, ESMA served written submissions on SEB’s application for suspension. 56. On 9 October 2018, ESMA made a further application to the Board of Appeal seeking 10 weeks from 24 December 2018 in which to file a rejoinder and suggesting that the hearing of oral representations should take place at or after the end of March 2019. 57. On 18 October 2018, SEB served its response to ESMA’s written submissions on the application for suspension. 58. On 25 October 2018, the Board of Appeal ruled that any further rejoinder should be served by ESMA by 23 January 2019 and that the hearing of oral representations would be fixed for 7 (later brought forward to 6) February 2019. 59. On 29 October 2018, ESMA served a reply to SEB’s submissions on the application for suspension. 60. On 20 November 2018, ESMA’s Responses to the Notices of Appeal were served. 61. On 30 November 2018, the Board of Appeal issued a decision refusing SEB’s application for a suspension. The decision required an analysis of multiple issues raised by both parties. This decision is published at the same time as the present decisions. 62. The appellants’ Replies to the Responses were all served by 24 December 2018. 63. On 23 January 2019, ESMA served its rejoinder. 64. On 6 February 2019, pursuant to Article 60(4) of the ESMA Regulation, the parties’ oral representations were heard by the Board of Appeal in Frankfurt a. M. IX MiFID/MAR regimes applicable to investment recommendations/research 65. The appellants’ case is that Article 3(2) CRAR excludes MAR recommendations and MiFID investment research from the definition of a credit rating and therefore from the scope of CRAR. The following summary of aspects of the regulatory regime applicable to investment recommendations/research is based on the Responses to the Appeals filed on behalf of ESMA on 20 November 2018. 66. The provision of investment recommendations on financial instruments is regulated under both the framework established by MiFID (Directive No 2014/65/EU, the Markets in Financial Instruments Directive) and the framework established by MAR (Regulation (EU) No 596/2014, the Market Abuse Regulation). Both frameworks have been revised recently. 67. The purpose of the MAR framework is to ensure the integrity and transparency of markets by requiring persons that disseminate investment recommendations to the public (whether or not they are regulated) (i) to present such recommendations in a fair and objective manner, and (ii) to disclose any conflicts of interest. 68. The purpose of the MiFID framework is to ensure the protection of clients of investment service providers, by among other things requiring such investment service providers to avoid conflicts of interest when providing services to their clients. Among the implementing acts of MiFID is Commission Delegated Directive (EC) No 2006/73 (the “MiFID Implementing Directive”), adopted under the previous “MiFID 1” framework, and referenced in Article 3(2)(b) CRAR. (The current MiFID framework is known as “MiFID 2”.) 69. Under the MiFID framework: (i) The provision of personalised investment recommendations to clients either upon their request or at the initiative of the investment firm in respect of one or more transactions relating to financial instruments constitutes investment advice, which is a regulated investment service. (ii) The provision of general (non-personalised) investment recommendations to clients is an ancillary service. Investment service providers that provide general investment recommendations to their clients are subject to certain conduct rules aiming at ensuring fair presentation and the prevention of conflicts of interest. (iii) Investment research, which is defined in Article 24(1) of the MiFID Implementing Directive, is considered as a sub-category of investment recommendations (see Recital 28). (iv) The MiFID framework has introduced specific rules relating to research (including investment research). Specifically, it requires investment firms that provide investment advice or asset management services to pay other investment firms for any investment research received from them, so as to avoid being induced to refer clients for services on the basis of free research obtained rather than on the quality of services. 70. General (non-personalised) investment recommendations/research are an ancillary service and therefore not a reserved activity. They can be issued either by investment service providers (who must, where applicable, comply with the MiFID and MAR frameworks), or by any other unregulated market participants (who may not be subject to the MIFID framework but must in any event comply with the relevant provisions of the MAR framework). X CRAR framework applicable to credit ratings 71. The following summary of the CRAR regulatory regime is based on the Responses to the Appeals filed on behalf of ESMA on 20 November 2018. 72. As has been explained above, the framework applicable to credit ratings in CRAR and its implementing acts was introduced in 2009. It contains a broad set of regulatory requirements aimed at ensuring the integrity and transparency of the credit rating process and limiting mechanistic reliance on credit ratings by market participants, and includes registration requirements for issuers of credit ratings, governance, conduct and transparency rules. As stated in Recital 1: “Credit rating agencies play an important role in global securities and banking markets, as their credit ratings are used by investors, borrowers, issuers and governments as part of making informed investment and financing decisions. Credit institutions, investment firms, insurance undertakings, assurance undertakings, reinsurance undertakings, undertakings for collective investment in transferable securities (UCITS) and institutions for occupational retirement provision may use those credit ratings as the reference for the calculation of their capital requirements for solvency purposes or for calculating risks in their investment activity. Consequently, credit ratings have a significant impact on the operation of the markets and on the trust and confidence of investors and consumers. It is essential, therefore, that credit rating activities are conducted in accordance with the principles of integrity, transparency, responsibility and good governance in order to ensure that resulting credit ratings used in the Community are independent, objective and of adequate quality.” 73. The CRAR framework is more recent than the MAR and MIFID frameworks. While the MAR framework aims at protecting the integrity of the market in general and the MIFID framework aims at protecting clients of investment service providers, CRAR aims at regulating credit rating activities (see Article 1) consisting of the production and dissemination of a specific product, i.e. credit ratings. 74. The legislator chose to adopt a specific framework for the regulation of credit ratings activities due to their nature. Credit ratings synthesise a broad array of sources of information regarding market, issuer or instrument into a single rating notch (AAA; BB-, etc ) which constitutes a simple, readable and easily comparable metric. This has resulted in (i) a significant use by investors, borrowers, issuers and governments as part of making investment and financing decisions and (ii) their incorporation as a tool in various areas of financial regulation, e . g . to set capital requirements of banks and investment limits of asset managers. 75. However, while the use of credit rating is considered to have led to increased information efficiencies ( e.g ., reduced asymmetries of information between buy and sell side), it also led to a mechanistic reliance on credit ratings by market participants, detracting from internal qualitative and multi-sourced creditworthiness assessments, as well as producing pro-cyclical cliff and contagion effects. These negative effects were on full display during the 2007-2008 financial crisis and during the 2011-2012 sovereign debt crisis with abrupt downgrades spiralling into destabilising credit crises. These deficiencies are referred to above. 76. The risks related to mechanistic reliance on credit ratings by market participants led to calls by bodies such as IOSCO to examine the need to regulate CRAs as early as 2003. The EU Commission adopted a communication to this effect in 2006. These calls were reiterated with renewed urgency during the financial crisis, the Commission adopting a legislative proposal with respect to CRAs in November 2008, and the adoption of CRAR in 2009. CRAR was subsequently amended to take into account (i) the creation of ESMA as the single EU CRA supervisor and (ii) the specific nature of sovereign ratings following the sovereign debt crisis of 2011-2012. 77. The provisions of CRAR relevant to the present appeals are set out below. XI The cases of the appellant banks (1) Nordea 78. Nordea asks the Board of Appeal to overturn the Board of Supervisors’ findings and remit the case to ESMA in accordance with Article 60(5) ESA Regulation. 79. The Board of Supervisors erred as a matter of law in that it failed to reach a reasoned conclusion as to whether a document can simultaneously include both a credit rating and investment recommendations/investment research: (1) The central question is whether a document can constitute both an investment recommendations and investment research (“IR”) (within the meaning of the MAD/MiFID Implementing Directives) and a credit rating (as defined by the CRAR). Nordea’s position is that it cannot, meaning that a document falling within the definition of IR is to be excluded from the restrictions contained in the CRAR through the wording of Article 3(2). (2) The Board of Supervisors did not answer this question in clear terms, yet proceeded on the basis that a document can constitute both IR and a credit rating. Its reasoning on this central question is unclear and internally inconsistent. (3) Despite seeming to recognise that IRs and credit ratings were “ distinct concepts ”, the Board of Supervisors nevertheless concluded that “ a given document could contain both investment research or recommendation and a credit rating ” despite its earlier statement that the legislation on that point was “ not definitive ”. It lacked confidence in its own reasoning, concluding that its view was “ not settled ” but despite its uncertainty, it imposed a fine on the basis of that view of the legislation. (4) Nordea agrees with the Independent Investigating Officer (“IIO”), which the Board of Supervisors has not been able to conclude upon, that credit ratings and IRs are distinct and mutually exclusive concepts. It considers that the Board erred in its conclusion that such distinct concepts can co-exist within a single document or communication. (5) Credit ratings and IR were intended by European legislators to be regarded differently as is plain from the wording of Recital 10 of the MAD Implementing Directive which distinguishes them on the basis that the former are opinions on the creditworthiness of a particular issuer or financial instrument “as of a given date” , their purpose being to articulate the issuer’s creditworthiness, without encouraging (or otherwise) an investment in the particular issuer or instruments. (6) In contrast, the fundamental purpose of IR is to propose an investment strategy. IRs are forward looking and give a view as to the present/future value of the financial instrument/issuer. In the context of credit or fixed income research, that takes into account the likelihood that the issuer will meet its contractual obligations. It may include a view on a particular issuer’s creditworthiness, but this goes to the merits of purchasing that issuer’s securities. It is different from a credit rating published on a standalone basis. (7) The fundamental difference between a credit rating which offers an independent point in time view on creditworthiness, and MiFID research or a MAD recommendation where a view on creditworthiness is offered in support of and as an explanation for the recommended strategy, has been acknowledged on a number of occasions (including by the EC, IOSCO in 2008 and the MAD Implementing Directive recital 10) and is well understood in the market. (8) There is no legislative basis or regulatory guidance supporting the Board of Supervisors’ characterisation of a credit or investment research report not as a single document, but as a piece comprising a number of component elements, each of which should be individually analysed and characterised. (9) Credit ratings and IRs are published for different reasons, and a single firm cannot produce both types of material and present them side by side in the same document. Although the operative provisions of CRAR do not seem to expressly prohibit a MiFID investment firm registering as a credit rating agency, in practice the operational and organisational requirements with which credit ratings agencies are required to comply (e.g. the ban on undertaking advisory or consultancy services) means that it would not have been possible for Nordea to issue both credit ratings and IRs from a single legal entity. (10) For credit or fixed income investment research to be of value, it needs to express an opinion on the creditworthiness of the issuer which can readily be understood. It is not possible to give a view on credit quality in the absence of a reference to a scale or rating, or by comparison to something which is referenced to a scale. The Board of Supervisors’ decision has had the effect of causing market uncertainty as to the permissible limits of a firm expressing a view on credit quality. The “Ratings” were no more than the formal articulation or summary of the analyst’s view of creditworthiness, as expressed within the credit or investment research report. (11) Contrary to ESMA’s new point advanced after the decisions, buy, sell or hold options could never constitute “credit ratings”. (12) From a policy perspective, IR in the Nordic countries has been an important tool in enabling SMEs to gain access to funding, as acknowledged by the Swedish Securities Dealers Association. A key policy objective in the European Commission’s Action Plan Building a Capital Markets Union is to ensure that SMEs operating within CMU have access to financing. Poor information and lower liquidity are hurdles which may be addressed through the publication of high quality credit research. The decision of the Board of Supervisors compromises this, precluding banks from publishing credit research which could improve liquidity. (13) The Board of Supervisors’ suggestion that Nordea’s interpretation could frustrate the aims of CRAR is unsupported. Nordea accepts the need to take into account the particular aims of CRAR, e.g. that credit ratings are of an adequate quality. The Board of Supervisors’ conclusion that “ it is possible that these aims… might be frustrated ” is expressed hypothetically and without examples. (14) The Board of Supervisors makes reference to the “distinct objectives” of CRAR “ for example that issued credit ratings are of adequate quality ”, and to the essential aims and objectives as set out in Recital 1 that credit ratings are “independent, objective and of adequate quality” . These address conflicts of interest and internal governance arrangements; and the preparation processes and methodologies for generating credit ratings. The organisational, operational and disclosure requirements which MAD/MiFID impose on investment firms in the context of IR differ in wording and scope from those in CRAR, but there are significant similarities in the requirements, protections and safeguards included in the parallel legislative frameworks which Nordea, as an investment firm publishing credit or investment research reports, was required to comply with. (15) The differences which do exist, as between CRAR and the MiFID/MAD framework, provide no evidence therefore to support the Board of Supervisors’ conclusion that the aims of the CRAR would be frustrated, as Nordea is subject to substantially equivalent requirements. (16) The Board of Supervisors’ construction would deprive the Article 3(2)(b) CRAR exemption of any effect. The whole purpose of this provision is to take recommendations which have the features of credit ratings outside of the scope of the regulation. So if we are not in that territory, it is simply irrelevant. (17) The question whether a rating can simultaneously constitute IR and a credit rating is fundamental. The Board of Supervisors seemed to conclude that the distinction between credit ratings and IR was based not on whether the document included a view on creditworthiness but on whether or not an established and defined ranking system of rating categories was used. This would mean only IRs which do not express an opinion “using an established and defined ranking system of rating categories” would fall within the scope of the Article 3(2) exemption. But the article is intended as a carve-out – such documents could never constitute “credit ratings” within Article 3(1)(a) CRAR as this requires both an opinion on creditworthiness and use of an established rating scale. Such documents would therefore not require the benefit of any exemption. (18) Whilst Nordea does not dispute the general principle that exemptions from Community law should be construed narrowly, context is important and the legislators must have had an aim in including the exemption. (19) There is no question of Nordea’s interpretation leading to a disapplication of the entire framework introduced by CRAR. 80. The Board of Supervisors imposed supervisory measures “without expressing a firm or settled view” on the scope of the exemption contained in Article 3(2) CRAR which is a breach of due process. (1) Nordea’s interpretation of the Article 3(2) exemption has been commonly adopted in the Nordic markets. No concerns about the practice have been raised by ESMA or local regulators and Nordea itself received no enquiries prior to ESMA’s inquiry in April 2015. (2) The Board of Supervisors has put forward an interpretation of the Article 3(2) CRAR exemption that is inconsistent with the understanding in the market, and in circumstances where it has been unable to come to a definitive conclusion on issues that are critical to determining the scope of the exemption. Where ESMA has been unable to reach a “firm and settled view” , the appropriate course would have been for it first to determine its preferred construction of CRAR and then to issue clarificatory guidance. (3) It is a breach of due process to proceed with an enforcement investigation and issue supervisory measures in circumstances where (i) ESMA’s interpretation of the scope of the Article 3(2) exemption is inconsistent with that adopted in the market supported (or not disputed) by local regulator without communicating its views to firms; and (ii) ESMA’s own views do not seem to be definitive. Nordea had a legitimate expectation that a material revision to the interpretation of CRAR such as articulated in the decision would be clearly communicated, particularly as ESMA has sought to clarify aspects of the CRAR through publication of Q&A on areas of uncertainty. (4) If the Board of Supervisors now takes the view that the application of the Article 3(2) CRAR exemption is uncertain, the taking of enforcement proceedings and issuance of supervisory measures, without providing an opportunity to assess and remediate market practice, is unjust, particularly since Nordea took prompt remedial action and ceased publishing credit research reports consequent to ESMA’s appointment of an IIO and feedback through the Supervisory team. 81. The Board of Supervisors’ construction of the Article 3(2) CRAR exemption undermines the fundamental principle of legal certainty under Community law. (1) Legal certainty is a fundamental principle of Community law which requires that rules should be clear and precise, so that legal persons may be able to ascertain unequivocally what their rights and obligations are and take steps appropriately. (2) The wording of Article 3(2) CRAR and Recital 20 CRAR is clear and unambiguous. (3) The Board of Supervisors indicates on multiple occasions that the regulations which it has analysed to reach its view are fraught with interpretive difficulty, such that their meaning is far from unequivocal. On the Board of Supervisors’ own case, the principle of legal certainty has been undermined. (4) Further, if the Board of Supervisors is unable to definitively set out and conclude upon the correct legal interpretation there can be no legally certain basis on which to prove a breach of the legislation. 82. Even if Nordea did commit the infringement alleged (which it denies), it did not do so negligently. (1) Nordea accepts that it had responsibility to assess the applicability of the regulation to its contracts. (2) However, the Board of Supervisors was wrong to find Nordea acted negligently in adopting an interpretation of the CRAR which is consistent with a logical and good faith reading of the legislation. Nordea cannot be regarded as having acted negligently where ESMA has concluded that the exemption gives rise to interpretative difficulty and on balance should be interpreted differently. Nordea’s interpretation does not fall below the standard of care to be expected by a reasonable market participant; multiple third parties reached the same conclusion, the other bank participants, and the Ministry of Finance in Finland. (3) Although the Board of Supervisors states that Nordea should have regard to the entirety of the EU in considering its conduct, the issues concerning interpretation of Article 3(2) CRAR are likely to have been most acute only in those markets where shadow ratings are published. Apart from the Nordic market, Nordea is only aware of one other market - the Schuldschein market in Germany – which continues to operate using shadow ratings. (4) Whilst the Board of Supervisors correctly notes that Nordea did not review the legal analysis which underpinned its reliance on the Article 3(2) CRAR exemption on an ongoing basis, this is not indicative of negligence. The legislation did not substantively change, no new regulatory guidance on the matters was published and no enquiries or concerns were raised with Nordea by ESMA or its local regulator. (5) With respect to penalty, Nordea took prompt steps to cease publishing shadow ratings once it became clear that ESMA’s view was that the exemption does not apply to investment research that includes a shadow rating (notwithstanding its logical, good faith reading to the contrary). (2) Swedbank 83. Swedbank submits that the central question of the appeal case is whether activities of an entity that is already regulated by certain Union acts (investment research) should also be regulated by additional Union acts (credit ratings) if these activities fulfil the criteria for both activities. Swedbank submits that this question should be answered in the negative in this case, since the latter Union act explicitly sets forth that activities under the former Union act are excluded from its scope of application. 84. The Appeal relates to whether Swedbank committed an infringement of CRAR and, if so, whether such infringement was committed negligently. The conduct of Swedbank under scrutiny is the issuing of credit research reports that included written opinions on relative creditworthiness of corporate bonds and the issuers thereof. 85. The central question is not whether credit research reports fulfilled the criteria for being credit ratings. The central question is whether they fulfilled the criteria for being investment research in the sense set forth in Article 3(2)(b) CRAR. The reason that this is the central question is that investment research in the sense set forth in Article 3(2)(b) shall not be considered credit ratings and thus is excluded from the scope of the CRAR. 86. Consequently, in order to determine whether Swedbank committed the alleged infringement, one first has to determine whether the research reports and the views expressed therein constituted investment research, in the widened sense referred to in Article 3(2)(b) of the CRAR. If so, they were excluded from constituting credit ratings and Swedbank was, accordingly, out of scope of regulation under CRAR. 87. The scope of Article 3(2)(b) and its relationship to the definition of a credit rating is also essential in determining the issue of negligence. If the legislation is unclear and objectively open for different interpretations, it would be a stretched and overstated conclusion to hold that Swedbank acted negligently in making its assessment that the issued reports were exempted from regulation under CRAR. 88. This central question of the proper interpretation of Article 3(2)(b) and its relationship to the definition of a credit rating has not been answered by ESMA in the Board of Supervisors’ Decision. Instead, ESMA acknowledges the lack of definitive legislation and thus refrains from expressing a “settled view” on the issue. 89. Swedbank submits that it did not commit any infringement, since the research reports and the views expressed therein constituted investment research in the sense referred to in Article 3(2)(b) of CRAR. Thus, they were excluded from constituting credit ratings and Swedbank was, accordingly, out of scope of regulation under CRAR. In any event, Swedbank submits, in particular having regard to ESMA’s own admission that the legislation is not definitive on this issue, that Swedbank’s assessment cannot have been negligent. 90. More particularly, the appealed Decision states that Swedbank has committed the infringement set out in point 54 of Section I of Annex III of the CRAR, i.e., that Swedbank was under an obligation to apply for registration under Article 2(1) of CRAR and has failed to do so. 91. This finding is based on the fact that Swedbank conducted credit research activities between 1 June 2011 and 31 August 2016, which included the issuing of reports that included written views/opinions of relative creditworthiness of corporate bonds and the issuers thereof. 92. The term “opinions” is used here to refer to opinions regarding the creditworthiness of a debt instrument or the issuer thereof published by Swedbank in the research reports, expressed as an assessment that includes both the rating category and the accompanying line of reasoning by Swedbank. The length, comprehensiveness and extent of the research reports and the opinions varies from one individual example of a report to another. In some reports, the accompanying information and analysis presented with the rating category itself is extensive and in such cases the opinion comprises several pages of analysis and reasoning. In other examples, the research report is only one page long, in which case the entire document is the opinion, as the entirety of the document comprises the analysis and conclusion presented by Swedbank in relation to the creditworthiness of the relevant bond or issuer. 93. The Board of Supervisors’ decision is founded upon the notion that the opinions constituted credit ratings as defined in Article 3(1)(a) CRAR and that it was part of Swedbank’s occupation during the relevant time period to issue the opinions on a professional basis. Hence, ESMA formed the opinion that Swedbank was a credit rating agency under an obligation to be registered as such pursuant to Article 2(1) CRAR. ESMA also found that Swedbank committed the alleged infringement negligently. Swedbank submits that these findings, upon which the decision was based, are incorrect. 94. In its Responses and oral representations, ESMA has defined the term “ratings” extremely narrowly. ESMA's definition of ratings does not cover the reports as such, nor the entire opinions on creditworthiness, but only the expression or symbol, for example BBB-, stable. 95. Swedbank submits that the opinions did not constitute credit ratings within the meaning of Article 3(1)(a) CRAR, since the opinions were excluded from the applicability of CRAR by virtue of Article 3(2)(b). Thus, Swedbank was under no obligation to be registered as a credit rating agency under Article 2(1) CRAR. If this objection is accepted by the Board of Appeal, the appeal must be allowed. 96. If expressions of a view of the creditworthiness of an entity or its instruments made using a rating category were to always and unconditionally be subject to regulation under the CRAR as argued by ESMA, Article 3(2) CRAR would have no operative effect. In effect, ESMA argues that Article 3(2)(b) would only exclude investment research that does not fulfil the criteria for being a credit rating. Hence, Article 3(2)(b) would have no meaning, since investment research that does not fulfil the criteria for being a credit rating in the first place is out of the scope of CRAR. 97. In any event, Swedbank should not bear the risk of unclear legislation. If the Board of Appeal was to find that Article 3(2)(b) does not exclude investment research from the scope of CRAR, which is a far-fetched interpretation of the provision, the risk of this ambiguity must rest with the legislator. 98. In reality, what the Board of Supervisors’ decision stipulated was that there was no settled view as to whether it was possible for a particular statement to qualify under Article 3(1)(a) CRAR and Article 3(2) CRAR, or whether these provisions concerned distinct concepts. This is not, as ESMA contends, merely a question of whether there is an overlap in the applicable regulatory frameworks, but a fundamental question to determine whether the opinions qualify under Article 3(2)(b) and thereby are exempted from its scope of applicability. In this important aspect, the Board of Supervisors’ decision was based on a non-settled view, as explicitly explained in the decision itself. 99. If the Board of Appeal finds that the opinions did constitute credit ratings within the meaning of Article 3(1)(a) CRAR, and that Swedbank was consequently obligated to be registered as a credit rating agency pursuant to Article 2(1) CRAR, Swedbank’s requests for relief should nonetheless be granted if the Board of Appeal agrees that Swedbank’s omission to be registered as a credit rating agency was not committed negligently. Swedbank made an internal assessment of the applicability of the CRAR to its activities, concluded that the CRAR did not apply, and thereafter had no reason to question its initial assessment. Since Article 1 of the Board of Supervisors’ decision stipulates that Swedbank negligently committed the infringement set out in point 54 of Section I of Annex III of the CRAR, the decision cannot be upheld if the Board of Appeal finds that Swedbank’s failure was not negligent. 100. Moreover, if the Board of Appeal finds that Swedbank’s omission to be registered as a credit rating agency was not negligent, Article 2 of the Board of Supervisors’ decision cannot be upheld since it establishes the issuing of a public notice that refers to the negligent infringement set out in Article 1 of the Board of Supervisors decision. 101. Finally, if the Board of Appeal finds that Swedbank’s omission to be registered as a credit rating agency was not negligent, Article 3 of the Board of Supervisors’ decision cannot be upheld since it establishes the imposing of a fine that, pursuant to Article 36(a)(1) of the ESMA Regulation, can only be imposed when a credit rating agency has negligently committed an infringement listed in Annex III of the CRAR. 102. Conclusively, if the Board of Appeal finds that Swedbank did not commit the infringement set out in point 54 of Section I of Annex III of the CRAR, or did not do so negligently, Swedbank’s requests below shall be granted. 103. When reaching its findings and adopting the Board of Supervisors’ decision, ESMA has failed to properly review the facts of the case. Specifically, the Board of Supervisors appears not to have reviewed the research reports containing the opinions, instead basing its decision solely, or at least to a great extent, on the description of the opinions included in the Independent Investigating Officer’s Statement of Findings. These shortcomings have led ESMA to reach an incorrect conclusion in its assessment of the opinions in relation to Article 3(2)(b) CRAR, and has consequently had a determining effect on the outcome of the decision. 104. Based on these submissions, Swedbank requests that the Board of Appeal: (1) declares that Swedbank did not negligently commit the infringement set out in point 54 of Section I of Annex III of the CRAR as set out in Article 1 of the Board of Supervisors’ decision, and remit the case to ESMA to adopt an amended decision in accordance therewith; (2) declares that a supervisory measure in the form of a public notice as set out in Article 2 of the Board of Supervisors’ decision shall not be issued, and remit the case to ESMA to adopt an amended decision in accordance therewith; and (3) declares that the fine as set out in Article 3 of the Board of Supervisors’ decision shall be repaid to Swedbank, and remit the case to ESMA to adopt an amended decision in accordance therewith; (4) or, if the Board of Appeal finds that Swedbank did negligently commit the infringement set out in point 54 of Section I of Annex III of the CRAR, declare that ESMA made procedural errors that affected the outcome of the Board of Supervisors’ decision, and remit the case to ESMA to adopt an amended decision following the rectification of the procedural errors. (3) SEB 105. SEB’s case in summary is that it is a licensed credit institution authorised by the Swedish FSA to provide investment services, including investment research and recommendations, principally under the MiFID and MAD framework. Such an authorisation is designed to protect investors and the stability of the financial system. It requires SEB to comply with various conduct of business obligations when providing investment services to its clients, and it is highly regulated. 106. During the relevant period, SEB produced credit research reports in relation to companies active in the Nordic corporate bond market, in particular small and medium sized enterprises which cannot afford to have public ratings and need to access the Nordic bond market to finance their business and development. 107. These credit research reports constituted investment recommendations and research under the MiFID and MAD, and they included an assessment of the creditworthiness of bond issuers, which is obviously an integral and vital part of the investment research as they enable the selected investors to make informed decision in the light of the credit risk taken. 108. The investment research and recommendations “ for distribution channels ” included assessments on the creditworthiness of bond issuers, in compliance with the EU applicable legal and regulatory framework, and in particular with the MiFID and the MAD requirements. 109. These assessments of the creditworthiness of the issuers covered by SEB’s Credit Research Department have always been provided by SEB as part of investment research and recommendations governed by the MiFID and MAD framework, exclusively to a small group of existing clients vetted by SEB’s Coverage organisation and selected by its Credits Sales’ team on a discretionary basis. These clients were selected on a discretionary basis. They could not subscribe. 110. On the facts, the criteria of public disclosure or distribution by subscription in Article 2(1) CRAR has not been satisfied. ESMA is wrong in its response to claim that this is a new point raised by SEB, and it has always been a central part of SEB’s case. 111. As an already highly regulated credit institution, SEB has full regard for the regulatory framework, and does not seek to circumvent CRAR. A comparison of the stringency of the MiFID/MAR and CRAR regimes is out of place. SEB is not bound by the disclosure rules in CRAR such as apply to methodologies. 112. Investment research and recommendations are already thoroughly regulated. In drawing the distinction, the definition of a credit rating in Article 3(1) CRAR must be precisely determined first. 113. SEB’s credit research activities do not and have never fallen within the scope of CRAR, in particular since the creditworthiness assessments included by SEB in its credit research reports during the Relevant Period: (1) are an integral part of a communication which qualifies as an investment recommendation or research under the MIFID and MAD framework and, as such, cannot be considered as credit ratings pursuant to Article 3(2) of CRAR, and (2) were neither issued “ for public disclosure ”, nor distributed “ by subscription ” as required by Article 2(1) of CRAR (see in this regard Art 2(2)(a)): the reports were not distributed to the public at large. 114. The credit reports cannot qualify as both a credit rating under the CRAR and investment research or recommendation under the MAD and MIFID requirements: (1) The exemption provided in Article 3(2) CRAR exists for a reason and cannot be treated as superfluous, and (2) Most, if not all Nordic banks, have always considered that they fell within the exemption. There is no material suggesting that a supervised credit institution expressing an opinion on the creditworthiness of an issuer has to register as a CRA, nor has any bank done so. 115. The principles enshrined in the EU Charter of Fundamental Rights, in particular the principles of legality and proportionality of offences and penalties, apply to the current proceedings, given the nature and potential consequences of these proceedings. 116. Pursuant to the right to a fair trial and the presumption of innocence which applies to these proceedings, it is up to ESMA to evidence that the credit research reports issued by SEB constituted credit ratings under the meaning of the regulation, and it has not done so. 117. The European Court of Human Rights and the Court of Justice consistently consider that the principle of legality applies to administrative sanction proceedings, particularly fines, and that “a penalty, even of a non-criminal nature, cannot be imposed unless it rests on a clear and unambiguous legal basis”. 118. The Board of Supervisors’ decision is based on an interpretation of CRAR which breaches the principles of legality of offences and penalties as developed in the case law, in particular the Board of Supervisors’ interpretation of: (1) The definition of a credit rating in Article 3(1) CRAR, which is merely speculative and contributes to making the scope of CRAR ambiguous and unclear; the Board’s reasoning is expressed in the conditional tense and is not definitive. (2) The exemption of investment research and recommendations in Article 3(2) of CRAR, which matches neither the ratio legis nor the objectives of CRAR, and which is unprecedented and unforeseeable for already regulated credit institutions; (3) The credit ratings “ disclosed publicly or distributed by subscription ” under Article 2 CRAR, which has no legal grounding and lacks a sufficient factual demonstration; (4) The concept of negligence in Articles 24 and 36 CRAR, which is equally baseless from both a legal and factual perspective. 119. The products are clearly different from those from the rating agencies, and there is no risk of confusion. 120. There is no factual or legal justification for the Board of Supervisor’s finding of negligence. The practical examples cited by ESMA in its 2012 Consultation Paper are directed at credit rating agencies and do not address the issues in the present matter. The obligations which it considered SEB should have complied with are speculative. 121. The term “shadow rating” is an unfortunate one in this regard, and so far as SEB is concerned, its reports fell within Article 3(2), and there was no negligence of any kind. 122. Any fine imposed on SEB would be disproportionate and therefore illegal, in particular since: (1) It would not be appropriate and necessary for attaining the objectives pursued by CRAR, and (2) The credit research reports under consideration have not caused any harm to the internal market or any individual or group of investors or issuers. SEB’s credit research activities enable investors to make informed investment decisions in light of credit risk, and help to preserve an important source of financing for small and medium sized enterprises on the Nordic bond market. They thereby contribute to the protection of investors and the stability of the internal market. 123. Accordingly, SEB asks the Board of Appeal to: (1) Quash the Decision, (2) Decide that SEB’s credit research activities are excluded from the scope of CRAR pursuant to Article 3(2) of the Regulation; and (3) Decide that SEB has not committed the infringement set out in Point 54 of Section 1 of the Annex III of the Regulation. 124. Alternatively, if the Board of Appeal considers that SEB committed the infringement, it asks the Board of Appeal to: (1) Decide that SEB did not commit the infringement negligently, (2) Rule that no fine can be imposed on it. (4) Handelsbanken 125. The Board of Supervisors’ decision is the final step in ESMA's investigation that commenced in April 2015 into the so called "shadow ratings" contained in investment research reports on debt instruments, labelled credit research reports, by Handelsbanken and other market participants, prepared by the capital markets division of Handelsbanken and issued to investors. The debt instruments concerned were corporate bonds on the Nordic market. 126. The Board of Supervisors’ decision concerns the interpretation of Article 3(2) CRAR and the specific exemption afforded to investment research. Handelsbanken has maintained throughout ESMA's investigation that the exemption regarding investment research in Article 3(2) is clear and that it applies to Handelsbanken's credit research reports on debt instruments. 127. The Board of Supervisors’ decision nevertheless concluded that the credit research reports of Handelsbanken fell within the scope of CRAR and that Handelsbanken has been negligent in omitting to register as a credit rating agency. 128. Due to the actions of ESMA and its interpretation of CRAR in clear contradiction of its wording, a central question in this matter is whether the alleged lack of clarity of Article 3(2) CRAR can result in any burden and responsibility for Handelsbanken. 129. Handelsbanken has all along noted that if the provision is to be held to mean anything else than its clear wording, this should be done in legislation as the interpretation suggested in the Board of Supervisors’ decision would entail a fundamental change to CRAR and a considerable broadening of its scope. 130. ESMA has not been able to satisfactorily explain how the alleged lack of clarity should in practice have been apparent to Handelsbanken before the investigation by ESMA commenced. The "doubt" regarding the scope of CRAR has in fact been introduced and created by ESMA a number of years after CRAR first entered into force. Hence, there are important and fundamental principles of foreseeability, proportionality and legal certainty which go to the heart of this case. 131. Should the Board of Supervisors of ESMA ultimately be confirmed in its interpretation of Article 3(2) of CRAR, in clear contradiction of its wording, such a result has not been foreseeable to Handelsbanken. There has been no legal certainty for it on which to base its actions. Therefore, Handelsbanken cannot be considered to have infringed CRAR or to have been negligent in its actions. For the same reasons – the lack of legal certainty and foreseeability – it was not correct to impose a fine on Handelsbanken. 132. The Board of Supervisors’ decision breaches CRAR and fundamental principles of EU law and must therefore under Article 60(5) of the ESAs Regulation be remitted. 133. The grounds for appeal are in summary form: (1) Incorrect determination of the scope of CRAR: the relevant investment research on debt instruments, referred to as credit research reports, issued by Handelsbanken, does not fall within the scope of CRAR because it falls within the exemption in its Article 3(2)(b) afforded to investment research. (2) The objective of CRAR supports this conclusion, which the Board of Supervisors’ decision fails to acknowledge. The objective was to fill a regulatory gap, not impose double regulation on regulated financial institutions such as Handelsbanken that issue investment research related to debt instruments. This activity is regulated by other Union legislation. (3) The practical impact of the Board of Supervisors’ decision is that the exemption in Article 3(2) is rendered virtually meaningless. The Decision ignores Court of Justice case law on the interpretation of exemptions. (4) The Board of Supervisors’ decision breaches the principle of legal certainty since the doubt regarding the scope of CRAR has in fact been introduced by ESMA a number of years after CRAR entered into force and in contradiction to its wording. This was not foreseeable to Handelsbanken, and there has not been sufficient legal certainty for Handelsbanken to base its actions on. (5) Incorrect determination of negligence: in the circumstances, the standard of "special care" is not appropriate because a market participant that, taking into account all relevant circumstances, acts on the basis of a plausible interpretation of a legislative act cannot reasonably be considered negligent. Further, an alleged lack of clarity in the relevant legislative act should not prejudice a market participant which has arranged its operations in accordance with a reasonable interpretation of the language of that act. (6) In the circumstances, Handelsbanken cannot be held to have acted negligently. A duty of care cannot be extended beyond reason, taking practical and financial considerations into account. The test suggested and steps proposed are far too onerous and unrealistic. They would paralyse the operations of market actors and supervisory authorities and would not be practically meaningful. Furthermore, Handelsbanken could simply not have been able to foresee that such steps were required. (7) The sanctions are imposed in breach of fundamental principles of EU law. There was no uncertainty as to Handelsbanken's compliance. Thus, as a starting point, there were no grounds for ESMA to sanction Handelsbanken. In addition, the requirement of legal certainty must be observed all the more strictly in the case of rules liable to entail adverse financial consequences of private subjects. In this case, neither the alleged infringement nor the proposed sanction has been foreseeable. 134. Reliance is placed on the background to the appeal. (1) Handelsbanken became aware of ESMA's investigation when it received the first request from ESMA in April 2015 in which it requested information on the so called "shadow ratings" contained in "credit research reports" regarding corporate bonds prepared by the bank’s capital markets division. The relevant investment research concerned corporate bonds in the Nordic market. (2) The broader underlying purpose of the reports was to assist the growing market for corporate bonds – especially for SMEs – in the Nordic region by providing analyses to support well-founded investment decisions. (3) The reports served as an important tool for both investors and bond issuers. The former often had no other realistic means of evaluating the strength and value of the Nordic corporate bonds in a local market context. In addition, the issuers could often not finance their operations with a bond issue without any such underlying assessment of their creditworthiness. (4) In order for the reports to fulfil this purpose and constitute a useful tool for investment decisions, they needed to include an assessment of the creditworthiness of the bond issuer otherwise the reports would be virtually meaningless to market participants. (5) The creditworthiness of a company, in particular for investment purposes, needs to be assessed on some scale that can relate this fact to the creditworthiness of other issuers and other alternative investments. That was the case for the indicative ratings (also called "shadow" ratings) included in Handelsbanken's investment research reports. (6) The investment research on debt instruments included a credit analysis of a company and thus constituted an integrated product with many contributing elements. The alleged "credit rating", i.e. the indicative ratings, was therefore not a naked or specific or separate product, but formed an integral and — on the part of investors — expected part of the overall assessment of the creditworthiness of the company. (7) The investment research on debt instruments was never intended to commercially compete with the credit rating agencies and had a distinct role on the local securities market in which ratings from rating agencies were not commercially available for relevant issuers and investors. (8) All the major Swedish banks acting in the Nordic market issued similar reports during the relevant time period. It could therefore be considered market practice in the Nordic region. (9) The practice did not provoke any action from the Swedish regulator, Finansinspektionen, that supervises all the Swedish banks for both prudential and market-conduct purposes. (10) Handelsbanken emphasises the timeline. It originally and still considers that CRAR is clear with respect to the investment research exemption in Article 3(2)(b). It originally considered and still maintains that its credit research reports constitute such investment research and are within the exemption and excluded from the scope of CRAR. Handelsbanken held and still holds that this is a natural and reasonable reading of CRAR, doing justice to the policy concerns underpinning the Regulation. (11) The practice and understanding of Handelsbanken also corresponded to market practice in the Nordic region. There appears to have been a consensus — independently arrived at by various unconnected and unaffiliated market participants (including issuers, investors and intermediaries) active in the Nordic market — that reports of this kind were exempted from the scope of CRAR by virtue of Article 3(2)(b). (12) Thus, from June 2011 — when Handelsbanken commenced publication of the relevant reports — until 28 April 2015 when ESMA commenced its investigation, there was no indication from any external source in the form of guidance from domestic or supranational regulators or other means that could reasonably be taken to have indicated to Handelsbanken that it could potentially fall within the scope of CRAR, despite the clear wording of the Regulation to the contrary. (13) When it became apparent during ESMA's investigation that there was a potential risk that its credit research reports could be considered to fall under CRAR, Handelsbanken first amended its practice and then, after a meeting with ESMA officials, ceased it altogether. Handelsbanken therefore acted prudently and with caution as a responsible market participant as soon as the potential infringement had been brought to its attention. (14) In the Decision, the Board of Supervisors (repeatedly) maintains that it has not formed a "settled view" on the central issue of how to interpret the exemption in Article 3(2)(b). Nevertheless, it adopted an interpretation against the wording of CRAR that considerably broadens its scope, and imposed sanctions on Handelsbanken for alleged breaches of CRAR despite an uncertainty of its scope that the Board of Supervisors itself admits to. (15) This was done without any substantiated support for that interpretation other than the policy aims of the Board of Supervisors and its conjecture as to the intentions of the legislator with respect to CRAR. (16) In its reply to ESMA in the investigation process, Handelsbanken noted that an amendment of CRAR would be necessary for its reports to fall within the scope of CRAR. Handelsbanken has maintained this position throughout the investigation. It is not for ESMA to change the scope of CRAR – that remains the prerogative of the Union legislator. (17) Furthermore, in a case such as the one at hand that revolves around an alleged "unclarity" in CRAR of which the relevant market actors were not aware, and that goes against its wording, ESMA attempts to impose a "special duty of care" on market participants. This is unreasonable in the circumstances and would in essence be a retroactive duty, from the date of the decision, unascertainable to market participants. (18) The Board of Supervisors of ESMA elaborates a list of far-reaching steps that market participants should take in order to demonstrate that they have not acted negligently. Both the special duty of care and the steps that ESMA attempts to impose in the decision are without foundation. (19) Notwithstanding the matter being "unclear", ESMA has failed to provide guidance or information to the market, or otherwise indicated to market participants that it would interpret CRAR as it now does. 135. Handelsbanken requests Board of Appeal to remit the case to the Board of Supervisors for the adoption of an amended decision that it had no duty to apply for registration under CRAR and did not infringe point 54 of Section I of Annex III CRAR. If the Board of Appeal finds an infringement, it should find that Handelsbanken has not been negligent, and if the Board of Appeal finds negligence, it should find that the sanction violates fundamental principles of EU law. XII The response of ESMA to the appeals 136. In response to the appeals, it is submitted on behalf of ESMA in summary that: (1) ESMA has not erred in law in the interpretation of the definition of “credit rating” under Article 3(1) CRAR and in the interpretation of Article 3(2) CRAR. ESMA’s interpretation of these provisions is fully in line with the text, context and intent of the relevant provisions. Conversely, the appellants’ interpretation has no legal basis and deprives CRAR of any practical application. (2) ESMA has not breached the principles of due process or legal certainty or legality of offences and penalties or the appellants’ legitimate expectations. It has not infringed the appellants’ fundamental rights, or the principle of proportionality. (3) ESMA’s position on the matter relevant to the case at hand is clear and settled and was entirely foreseeable. In addition, ESMA provided the appellants no assurances giving rise to legitimate expectation that their conduct would not constitute an infringement of CRAR requirements. (4) ESMA fully assessed the factual circumstances surrounding the distribution of the ratings to conclude that the conditions existed for the ratings to be deemed to have been disclosed publicly or distributed by subscription. (5) The appellants infringed CRAR by acting negligently. The appellants are significant credit institutions required to comply with a high standard of care. The appellants failed to take appropriate steps to satisfy themselves that their conduct did not infringe CRAR. (6) The appellants’ claim that ESMA breached the principle of proportionality and did not correctly apply the mitigating factors is unfounded. ESMA applied CRAR faithfully and had no discretion to take into account the mitigating factors invoked by the appellants. (7) ESMA requests the Board of Appeal to (i) dismiss the appeals as entirely unfounded and (ii) confirm the Decisions in their entirety. 137. It is not in dispute that the appellants have issued ratings, which are opinions on the creditworthiness of financial instruments and issuers, expressed through a standardised rating scale. Except in the case of SEB, it is also not in dispute that the ratings have been disclosed publicly and/or distributed by subscription, and do not fall within any of the exceptions listed in Article 2(2) CRAR. The ratings are therefore subject to CRAR, and the appellants have infringed Article 14(1) CRAR by issuing the ratings without having applied for registration. 138. Article 3(2) CRAR does not affect this conclusion. This excludes content that constitutes investment recommendations/research for purposes of MAR/MIFID as well as other opinions about the value of a financial instrument or a financial obligation. It does not exclude credit ratings that are included in documents that also include investment recommendations/research or other opinions. 139. Neither Article 3(2) CRAR nor any other provision of CRAR excludes credit ratings based on the type of document or materials in which they are included or on the type of support through which they are disseminated. 140. So long as the credit rating is expressed in the form of a standardised rating symbol and is widely available to market participants, it can have the detrimental effects on investor confidence and market stability that CRAR seeks to prevent. 141. Conversely, ratings that do not have those characteristics, e.g. are private ratings that are not widely disseminated, and therefore do not present risks for market stability, are excluded from the application of CRAR under Article 2(2). 142. If the legislator had thought that credit ratings included in investment recommendations/research reports present less of a risk to financial stability as a result of such inclusion, then it would have included them explicitly among the exemptions listed in Article 2(2) CRAR, which it did not do. 143. The notion that credit ratings included in investment recommendations/ research reports or any other document containing the elements described in Article 3(2) should be excluded from CRAR entails an absurd result. It would mean that market participants (including registered CRAs) would be able to circumvent the application of CRAR simply by including credit ratings in documents containing investment recommendation/research, or even merely opinions on the value of a financial instrument as per Article 3(2)(c). Such an interpretation would not only frustrate the purpose of CRAR but lead to a disapplication of the entire framework introduced by CRAR. It is inconceivable that this was the interpretation intended by the legislator. 144. The contention that the existence of two separate regulatory frameworks entails that CRAR does not apply to credit ratings included in investment recommendation/ research reports is misguided because the purposes of different. CRAR is not a “double regulation”. 145. The contentions based on Article 3(2)(b) CRAR that a communication cannot be both a credit rating and investment research within MiFID, that credit ratings should not be considered as such when they are included in investment recommendations/ research reports, or cannot be separated from such reports, or viewed exclusively as investment recommendations/ research, or that a document cannot qualify both as investment research and as a credit rating, or that the two concepts cannot coexist within a single document, or that the rating must be considered as an integral part of the investment recommendations/ research reports, have no legal basis. CRAR regulates credit ratings regardless of the documents or material in which they are included. A communication can be both a credit rating and investment research. The fact that the credit rating may be a “minuscule” part of the reports is irrelevant. 146. The practical impossibility for a bank that conducts investment research to simultaneously issue credit ratings does not make the issuing of credit ratings by the bank legal or permissible. It is wrong to contend that MiFID/MAR requirements are substantially equivalent to those under CRAR, which are more stringent and precise. 147. The inclusion of credit ratings in reports will trigger the application of CRAR, since no provision of CRAR or of MIFID/MAR exempts credit ratings from CRAR on the basis of the documents in which they are included. 148. The appellants’ claim that credit investment research needs to contain an opinion expressed through a scale in order to have any practical meaning to investors is not supported by any evidence. 149. Claims that ESMA’s interpretation of Article 3(2)(b) CRAR renders this provision meaningless are wrong. Its purpose and effect is to state that investment recommendations/reports do not constitute credit ratings and do not need to be regulated under CRAR. Specifically, Article 3(2)(b) makes clear that “buy”, “hold” or “sell” recommendations, although they are technically standardised, are not credit ratings for purposes of CRAR. 150. ESMA’s position is settled on all matters relevant to the case at hand. ESMA’s interpretation of CRAR was entirely foreseeable (having been stated in a consultation paper), and it did not breach the principle of legitimate expectations. 151. The Decisions are based on Article 3(1) CRAR and specifically on the fact that the ratings issued by the appellants are credit ratings as defined in Article 3(1)(a), issued by an entity that qualifies as a CRA as defined in Article 3(1)(b). ESMA’s interpretation of Article 3(2) is in line with the wording of the article which excludes from CRAR investment recommendations/ research, not credit ratings that may be included in such reports. 152. References in the Decisions to matters being “not settled” are referring to the broader considerations relating to possible regulatory overlaps between the CRAR and MIFID frameworks, i.e. whether credit rating activities of CRAs can also be subject to MIFID/MAR, in which case the CRAR and MIFID/MAR frameworks would overlap (a question that the Board of Supervisors notes is “not settled”). However, these general considerations do not affect the Board’s clear finding according to which, in the case at hand, the ratings issued by the appellant are credit ratings for purposes of CRAR and not excluded under Article 3(2). 153. By stating that the ratings serve the same purpose as credit ratings issued by CRA’s in a region where issuers cannot afford public ratings from registered CRA’s, the appellants merely confirm that the ratings serve the same main purpose of credit ratings under CRAR. The fact that there appears to have been consensus in the Nordic Region that ratings included in investment recommendations/research reports are outside CRAR is irrelevant and does not justify the conduct of unauthorised activities. 154. SEB is wrong to claim that the Board of Supervisors only relied on three credit reports to make its assessments that its communications to its clients were “publicly disclosed or distributed by subscription”. SEB itself provided ESMA with figures regarding the percentage of its credit investment research that included an assessment of the creditworthiness of an issuer or debt instrument to the effect that from 2011 to 2016 approximately 100% of its credit investment research included such creditworthiness assessment. 155. The Board of Supervisors did state both the factual and legal basis of their reasoning, by referring to (i) the relevant criteria, as set out in Article 2(1) CRAR and (ii) the factual basis upon which it found that the Ratings met such criteria. ESMA has stated in its Guidelines and recommendations on the scope of the CRA Regulation that it is sufficient, for a rating to be considered as “distributed by subscription”, that the rating be provided to different persons belonging to a list of subscribers, without any specific number. Indeed, “[a] rating which is provided to different persons belonging to a list of subscribers does not fall within the definition of “private rating” in Article 2(2)(a) of the CRA Regulation”. 156. According to Article 1(7) of Commission Directive 2003/125/EC, distributions channel is defined as “a channel through which information is or is likely to become publicly available” and “likely to become publicly available information” is defined as information to which a large number of persons have access. According to the figures provided by SEB, the reports were distributed to 1147 clients. Also on the basis of this element, the Board of Supervisors rightly considered that the ratings could not qualify as “private ratings” under Article 2(2)(a) CRAR. 157. The CJEU case-law holds that negligence should be understood as “entailing an unintentional act or omission by which the person responsible breaches his duty of care which he should have and could have complied with in view of its attributes, knowledge, abilities and individual situation”. Under a “normal care standard”, negligence requires that the author of the infringement “has not foreseen the consequences of his action in circumstances where a person who is normally informed and sufficiently attentive could not have failed to foresee them.” 158. The CJEU has also consistently held that “persons carrying out a professional activity, who are used to having to proceed with a high degree of caution when pursuing their occupation” “can on this account be expected to take special care in assessing the risks that such an activity entails”. Under the “high standard of care”, negligence is established where the author of the infringement fails to take “special care” in assessing the risks that its acts or omissions entail. Under that standard, a more active approach is required from such entities which must take certain specific actions and/or followed certain steps, including, for example, seeking “appropriate legal advice” or “expert advice to assess, to a degree that is reasonable in the circumstances, the consequences which a given activity may entail”. 159. According to the CJEU, the “special care” standard applies to “persons carrying on a professional activity, who are used to having to proceed with a high degree of caution when pursuing their occupation.” Financial service providers and CRAs play an important role in the economy of the EU, as well as the financial stability and integrity of the financial markets. ESMA is of the view that a high standard of care is to be expected of a CRA: see Recital 1 CRAR. 160. Given their importance, financial services providers are required to have in place robust internal governance systems to ensure compliance with applicable regulatory requirements. As to credit institutions, see Directive (EU) No 2013/36 (“CRD IV”) and the guidelines adopted by the European Banking Authority (“EBA”) for its application. It is also true of CRAs, which are required under CRAR to set up a compliance function and to “ establish adequate policies and procedures to ensure compliance with its obligations under [CRAR]. ” 161. Given the importance of the requirements applicable to credit ratings, and the appellants’ role as some of the main providers of financial services of the Nordic region, ESMA believes that a high standard of care was to be expected of the appellant when they decided to issue credit ratings. 162. In the Decision, the Board of Supervisors provided examples of concrete actions or steps that diligent professional persons might have taken in similar circumstances, e.g. carefully evaluated whether CRAR might apply to the Ratings, sought legal advice on the scope and effect of CRAR and/or advice from their National Competent Authority and/or ESMA on those issues; in addition, the Board of Supervisors indicated that the appellants might have been expected to submit any initial legal position to periodic review. 163. The appellants consider that CRAR provides for a clear exemption for credit ratings included in investment recommendations/research, thus suggesting that these indicative steps were not necessary. 164. However, the Board of Supervisors was not imposing additional steps not contemplated by the law but merely indicating examples, consistent with the above CJEU case-law, of steps that a diligent bank “ might have ” taken, with the understanding that other measures could also have been taken, to comply with the high standard of care. Had a proper legal assessment have been carried out by the appellants as to the scope and purpose of CRAR, that would have at least raised questions as to the legality of their credit rating activities which could have been answered through the steps mentioned. 165. In deciding whether the appellants met the high standard expected of them, the Board of Supervisors had to consider those steps that the appellants did or did not take in the particular circumstances. Whether any of these steps were taken is an objective criterion relevant to the question of whether the appellant in question met that high standard of care. 166. The fact that the authorities did not raise any concern or that there were no material changes in the provisions of CRAR is irrelevant. This does not relieve an entity of its duty of care or of its obligation to ensure that it complies with applicable regulations. 167. It cannot be said in good faith that there was no possibility that the ratings, which are expressed in the form of standardised rating notches, could potentially be subject to CRAR. Only an inadequately short assessment made without appropriate due care and seriousness could have led to such a conclusion. 168. The banks’ interpretation of Article 3(2) CRAR is not supported by the text, is manifestly contrary to the regulatory purpose of CRAR, and would deprive the CRAR of any practical application. 169. Reliance cannot be placed on the argument that other market participants were issuing ratings without complying with CRAR: each regulated entity is responsible for conducting its own assessment of its obligations under the law and comply with it and the fact that no one in the Nordic region was compliant is not an excuse for not having complied with the duty of care. 170. The national authorities are not in charge of interpreting and enforcing CRAR, and the fact that they did not raise any issues is not relevant. The absence of concerns raised by ESMA is also irrelevant as it is settled case-law that the fact that specific conduct has not yet been examined by the authorities does not exonerate an undertaking. The appellants should have foreseen that their credit ratings activities were likely to be subject to CRAR. 171. The appellants’ behaviour demonstrates that, rather than adopting conduct that one would expect from skilled and informed entities (such as carefully analysing the application of regulations, which should have led at least to questions being raised as to whether the ratings could potentially be subject to CRAR), they relied on an incorrect reading of CRAR to continue credit rating business activities and clearly failed to comply with their duty of care. 172. The appellants’ defence suggesting that due to the complexity of CRAR and/or the uncertainties in the interpretation of CRAR they cannot be held negligent, is unfounded. The broader considerations that the Board of Supervisors considered “not settled” are irrelevant. 173. The appellants cannot reasonably claim an uncertainty or complexity raised after the fact by another party (the Board of Supervisors) that they have never previously raised as grounds to retroactively evade the duty of care that they were required to comply with at the time. 174. Even if the appellants had claimed that they had difficulties in interpreting CRAR when they decided to issue credit ratings, the case law of the EU courts shows that such difficulties do not relieve the entities of their duty of care. They had to “seek all possible clarifications to ensure it did not infringe [the relevant provisions]” (CJEU in Firma Söhl & Söhlke ). 175. ESMA cannot take into account a hypothetical outcome that external counsel and the national authority if consulted would have reached the same conclusion as the appellants did. 176. The IIO’s conclusions to the effect that there was no negligence were not binding on the Board of Supervisors, insofar as the decision-making power in the matter of supervisory measures and sanction is solely vested in the Board of Supervisors. 177. As to Nordea: (1) It appears to have taken no concrete actions or steps, which confirms ESMA’s assessment that it failed to comply with the required standard of care. (2) The fact that the National Competent Authorities allegedly interpreted CRAR as it did consist only of very general statements made by the Ministry of Finance of Finland, which show that said authorities expressed no opinion on the specific question at hand. (3) Despite issuing what itself describes as “Corporate ratings”, “Company Ratings”, “Bond ratings” or “shadow ratings”, the bank did not appear to have taken any active step to assess whether it was required to comply with CRAR, and was unable to produce any documentary evidence of such assessment. 178. As to Swedbank: (1) Though the appellant stated it had carried out an initial assessment of its position in relation to CRAR, it did not produce documentary evidence of the assessment and appeared not to have taken any steps externally, such as with its National Competent Authority or ESMA, to confirm its assessment nor seemed to have reviewed its position periodically. (2) The appellant did not meet the high standard of care required of it, and a properly informed person in its position, sufficiently attentive and having taken special care in assessing the risks involved in issuing ratings, would not have failed to foresee that the ratings were likely to be considered to be credit ratings subject to CRAR. (3) No actions or steps other than its alleged initial assessment appear to have been taken by the appellant. (4) Despite issuing what the appellant itself describes as “shadow ratings”, “Corporate ratings”, “Swedbank estimated Issuer ratings”, “Swedbank estimated Bond ratings” or “Swedbank issuer ratings”, it did not appear to have taken any active step beyond its alleged initial assessment to confirm whether it was required to comply with CRAR. (5) While claiming that the assessment process was extensive and rigorous and carried out with highly competent in-house legal experts, the appellant was not able to provide any record of its alleged initial assessment of its position in respect of CRAR. 179. As to SEB: (1) Though the appellant stated that it had carried out an initial assessment of its position in relation to CRAR and concluded it was not applicable to it given the scope of the exemption provided by Article 3(2), it has not produced documentary evidence of the assessment and no record was kept of the conclusion that the legislation was inapplicable. (2) The appellant appears not to have taken any steps externally, such as with its National Competent Authority or ESMA, to confirm its assessment nor seemed to have reviewed its position periodically. (3) The appellant did not meet the high standard of care required of it, and a properly informed person in its position sufficiently attentive and having taken special care in assessing the risks involved in issuing ratings, would not have failed to foresee that the ratings were likely to be considered to be credit ratings subject to CRAR. (4) No actions or steps other than its alleged initial assessment appear to have been taken by the appellant. (5) Despite issuing what it describes as “Corporate ratings”, “Stand-alone ratings” or “Credit ratings”, the appellant did not take any active step beyond its alleged initial assessment to confirm whether it was required to comply with CRAR. (6) While claiming that the assessment process was extensive and rigorous and carried out with highly competent in-house legal experts, the appellant was not able to provide any record of its alleged initial assessment of its position in respect of CRAR. 180. As to Handelsbanken: (1) Although the appellant appears to have carried out an initial assessment of its position in relation to CRAR, it has not produced documentary evidence of the assessment, and appears not to have taken any steps externally, such as with its National Competent Authority or ESMA, to confirm its assessment, nor “seems to have reviewed its position periodically. (2) The appellant did not meet the high standard of care required of it and a properly informed person in the position of the appellant, sufficiently attentive and having taken special care in assessing the risks involved in issuing ratings, would not have failed to foresee that the ratings were likely to be considered to be credit ratings subject to CRAR. (3) No actions or steps other than its alleged initial assessment appear to have been taken by the appellant. (4) Despite issuing what itself describes as “Indicative ratings”, “Indicative corporate ratings”, “Indicative issue ratings”, or an “Indicative issuer ratings”, the appellant did not take any active step beyond its alleged initial assessment to confirm whether it was required to comply with CRAR. (5) While claiming that the assessment process was extensive and rigorous and carried out with highly competent in-house legal experts, the appellant was not able to provide any record of its alleged initial assessment of its position in respect of CRAR. XIII The applicable provisions 181. The following are the provisions principally applicable to the issues arising on the appeals. There are relatively few of them. 182. Article 1 of CRAR sets out the Regulation’s ambit and purpose: “This Regulation introduces a common regulatory approach in order to enhance the integrity, transparency, responsibility, good governance and independence of credit rating activities, contributing to the quality of credit ratings issued in the Union and to the smooth functioning of the internal market, while achieving a high level of consumer and investor protection. It lays down conditions for the issuing of credit ratings and rules on the organisation and conduct of credit rating agencies, including their shareholders and members, to promote credit rating agencies’ independence, avoid conflicts of interest, and the enhancement of consumer and investor protection.” 183. The limits of the applicability of CRAR are addressed in a number of recitals, of which Recital 20 relating to investment research and investment recommendations is most relevant to the appeals. “Investment research, investment recommendations and other opinions about a value or a price for a financial instrument or a financial obligation should not be considered to be credit ratings.” 184. The definition of a “credit rating” is stated in Article 3(1)(a) as follows. “‘credit rating’ means an opinion regarding the creditworthiness of an entity, a debt or financial obligation, debt security, preferred share or other financial instrument, or of an issuer of such a debt or financial obligation, debt security, preferred share or other financial instrument, issued using an established and defined ranking system of rating categories;” 185. The definition of “rating category” in terms of a “rating symbol” follows in Article 3(1)(h): “‘rating category’ means a rating symbol, such as a letter or numerical symbol which might be accompanied by appending identifying characters, used in a credit rating to provide a relative measure of risk to distinguish the different risk characteristics of the types of rated entities, issuers and financial instruments or other assets;” 186. As foreshadowed in Recital 20, the exclusions are stated in Article 3(2) which provides that: “2. For the purposes of paragraph 1(a), the following shall not be considered to be credit ratings: (a) recommendations within the meaning of Article 1(3) of Commission Directive 2003/125/EC; (b) investment research as defined in Article 24(1) of Directive 2006/73/EC and other forms of general recommendation, such as ‘buy’, ‘sell’ or ‘hold’, relating to transactions in financial instruments or to financial obligations; or (c) opinions about the value of a financial instrument or a financial obligation.” Commission Directive 2003/125/EC is the Implementing Directive for the Market Abuse Directive (MAD). Directive 2006/73/EC is the MiFID Implementing Directive: see above for a description of these regimes. 187. The definition of a “recommendation” referred to in Article 3(2)(a) is found in Article 1(3) of the Implementing Directive for the Market Abuse Directive: “Article 1(3) “recommendation” means research or other information recommending or suggesting an investment strategy, explicitly or implicitly, concerning one or several financial instruments or the issuers of financial instruments, including any opinion as to the present or future value or price of such instruments, intended for distribution channels or for the public.” 188. Recital 10 of the Implementing Directive for the Market Abuse Directive is also relied on by the appellant banks to rebut dual characterisation: “Recital (10) Credit rating agencies issue opinions on the creditworthiness of a particular issuer or financial instrument as of a given date. As such, these opinions do not constitute a recommendation within the meaning of this Directive. However, credit rating agencies should consider adopting internal policies and procedures designed to ensure that credit ratings published by them are fairly presented and that they appropriately disclose any significant interests or conflicts of interest concerning the financial instruments or the issuers to which their credit ratings relate.” 189. The definition of “investment research” referred to in Article 3(2)(b) is found in Article 24(1) of the MiFID Implementing Directive: “Article 24(1) For the purposes of Article 25, ‘investment research’ means research or other information recommending or suggesting an investment strategy, explicitly or implicitly, concerning one or several financial instruments or the issuers of financial instruments, including any opinion as to the present or future value or price of such instruments, intended for distribution channels or for the public, and in relation to which the following conditions are met: (a) it is labelled or described as investment research or in similar terms, or is otherwise presented as an objective or independent explanation of the matters contained in the recommendation; (b) if the recommendation in question were made by an investment firm to a client, it would not constitute the provision of investment advice for the purposes of Directive 2004/39/EC.” By Recital 28, investment research is treated as a sub-category of the type of information defined as a recommendation in the Implementing Directive for the Market Abuse Directive. 190. ESMA’s Board of Supervisors’ power to impose a fine where a credit rating agency has, intentionally or negligently infringed CRAR is found in Article 36a (added by amendment in 2011) as follows: “Fines 1. Where, in accordance with Article 23e(5), ESMA's Board of Supervisors finds that a credit rating agency has, intentionally or negligently, committed one of the infringements listed in Annex III, it shall adopt a decision imposing a fine in accordance with paragraph 2. An infringement by a credit rating agency shall be considered to have been committed intentionally if ESMA finds objective factors which demonstrate that the credit rating agency or its senior management acted deliberately to commit the infringement.” Article 36a goes on to specify how fines shall be calculated. XIV The Board of Appeal’s discussion and conclusions (1) Introduction 191. These appeals raise important questions as to the relationship between the regulation of credit rating activities, which until relatively recently fell largely outside the regulatory perimeter in the EU, and the wider activities of investment research and recommendations by banks and others which, as these appeals show, may find the use of symbols such as AAA uniquely powerful in conveying opinions of a particular creditworthiness. The case is also unusual in that it concerns a particular market, the Nordic debt market, rather than an individual institution. 192. The resolution of the central question in the case – whether the credit ratings included within the research reports issued by the banks bring the reports within the scope of the Credit Rating Agencies Regulation, or whether the banks are correct to say that the reports are covered by the recommendations/ investment research exclusion in the Regulation – depends on the true construction of a relatively few legal provisions. Though both appellants and respondent argued on the appeals that the effect of the provisions is clear, the Board of Appeal has not found them to be straightforward to interpret. The history of the provision shows that the current wording changed in the drafting process, but does not provide reasons for the change. In any case, the line is not an easy one to draw. 193. Important questions are also raised as to the correct approach to the making of a finding of negligence on the part of a financial institution which triggers regulatory sanctions in the form of fines. The banks maintain that it is wrong in principle to find negligence where the meaning of the rules said to be breached are uncertain and unsettled. The response on behalf of ESMA is that the rules are clear, and that in any case it was for the banks to ensure that they complied with them, and that the functioning of the regulatory system is dependent on institutions observing a high standard of care in this regard. 194. As noted above, the governance structure of ESMA is designed so as to give an avenue of review within ESMA itself. The Board of Supervisors’ decisions – there being four decisions here – are subject to a right of appeal to ESMA’s Board of Appeal. As stated at the outset of this decision, the Board of Appeal must decide whether the decisions of the Board of Supervisors were correct or not, and may confirm the decisions or remit the cases to the Board of Supervisors under Article 60(5) of the ESMA Regulation. It is no part of the Board of Appeal’s function to decide policy, which is a matter for the Board of Supervisors (within the limits of the ESMA Regulation and its legal basis) . (2) The construction issue The relevant provisions 195. The relevant provisions are set out in full above. In summary: (1) By Article 2(1), CRAR applies to credit ratings issued by credit rating agencies registered in the EU and which are disclosed publicly or distributed by subscription. (2) By Article 3(1)(b), “credit rating agency” means a legal person whose occupation includes the issuing of credit ratings on a professional basis. (3) By Article 3(1)(a), “credit rating” means an opinion regarding the creditworthiness of a financial obligation or of the issuer of a financial obligation, “issued using an established and defined ranking system of rating categories”. (4) By Article 3(1)(h), “rating category” means “a rating symbol, such as a letter or numerical symbol which might be accompanied by appending identifying characters, used in a credit rating to provide a relative measure of risk to distinguish the different risk characteristics of the types of rated entities, issuers and financial instruments or other assets”. (5) By Article 3(2), recommendations within the meaning of the Market Abuse Directive and investment research within the meaning of MiFID shall not be considered to be credit ratings. The issues on the appeals 196. For an entity to be found to have committed an infringement of Article 14(1) CRAR, each of the following elements must be satisfied: (i) the relevant entity is a legal person established in the Union; (ii) the legal person has issued credit ratings as defined by Article 3(1)(a) and (h); (iii) the occupation of the legal person included the issuing of such credit ratings on a professional basis (in which case the legal person will fall within the definition of “credit rating agency” in Article 3(1)(b)); (iv) the credit rating agency issued such credit ratings that were disclosed publicly or distributed by subscription; (v) the credit rating agency has not applied for registration for the purposes of Article 2(1); (vi) no exemption or exclusion applies. 197. It is not in dispute that the banks are registered in the Union carrying on their activities on a professional basis, and none of them have applied for registration under CRAR. None of the exemptions in Article 2(2) apply. SEB does however dispute that its research reports were disclosed publicly or distributed by subscription, and so fall outside CRAR on that ground, and this is separately considered below. 198. As to (ii), under Article 3(1)(a) and (h), a credit rating is an opinion regarding the creditworthiness of an entity or financial instrument issued using an established and defined ranking system of rating categories: a rating category is a rating symbol such as a letter or numerical symbol used in a credit rating to provide a relative measure of risk. In oral representations, the term “symbol” as used in CRAR was paraphrased as an alphanumerical symbol regarding creditworthiness. 199. It is not in dispute that communications in the form of recommendations and credit research reports were issued by the appellant banks over the relevant period, and included symbols such as a letter or numerical symbol – e.g. A+, BBB+, or such like – regarding creditworthiness, and that an established and defined ranking system of rating categories was used. (The position as found by the Board of Supervisors as regards each bank is set out above.) These are what have been referred to as “shadow ratings”. 200. The first question on the appeals, which is a question of construction or interpretation of the relevant provisions, is whether the Board of Supervisors was correct to find that the ratings fell within the CRAR definition of “credit ratings” and did not fall within the definitions of either recommendations or investment research so as to be excluded under Article 3(2). 201. It is submitted on behalf of the appellant banks that the Board of Supervisors was wrong, and that a communication cannot be both a credit rating and recommendations/investment research, so that there was no obligation on them to register for the purposes of CRAR. It is further submitted that the decisions contravene the principle of legal certainty, in that the Board of Supervisors acknowledged that their view of the effect of the provisions was unsettled, but nevertheless proceeded to impose sanctions on the banks based on their unsupported view as to the effect. 202. It is submitted on behalf of ESMA that the Board of Supervisors was correct, and that recommendations and investment research on the one hand, and credit ratings on the other, are in principle mutually exclusive concepts. Even if one accepts that they may overlap, the banks’ ratings fell into the credit rating category, and were not exempt under Article 3(2). It submits that the decisions of the Board of Supervisors were settled in the relevant respects, and that the principle of legal certainty was not breached. 203. The appellants have put forward a number of contentions in support of their positions which the Board of Appeal will now consider under separate headings. The “double regulation” contention 204. A contention made by each of the appellant banks relates to the existing regulatory framework to which they are subject as credit institutions. As it is put, as credit institutions, the banks are already highly regulated. Further, investment research and recommendations are also thoroughly regulated, and there is no case for subjecting such material to another layer of regulation. 205. As is submitted, the organisational, operational and disclosure requirements which MAR/MiFID impose on investment firms in the context of investment research/ recommendations differ in wording and scope from those in CRAR, but there are significant similarities in the requirements, protections and safeguards included in the parallel legislative frameworks which the banks, as investment firms publishing credit or investment research reports, are required to comply with. 206. The objective of CRAR is to fill a regulatory gap, not impose double regulation on regulated financial institutions such as the appellant banks that issue investment research related to debt instruments. This activity is regulated by other EU legislation. Investment research on debt instruments was never intended to commercially compete with the credit rating agencies. The banks’ products are clearly different from those of the rating agencies, and there is no risk of confusion. Further, it is evident from the provisions of CRAR (e.g. recital 22) that the focus was on the particular position of institutions the business of which was credit rating. 207. The Board of Appeal’s view is as follows. It accepts that there is already a substantial layer of regulations on the banks, both as credit institutions and investment service providers, and in so far as they issue investment research and recommendations. It also accepts the wider point that a strong objective of CRAR was to fill a regulatory gap and bring rating agencies into a formal system of regulation which had not been there before, regulating credit ratings issued by CRAs registered in the EU which are disclosed publicly or distributed by subscription. There is nothing in the material that the Board of Appeal has seen to suggest that it was foreseen that the Nordic banks would find themselves within the provisions of CRAR, or that their activities in any way posed a threat to financial stability, or that their existing practice would have to change. On the contrary, as stated above, this was (and is) seen as beneficial to the Nordic debt market, particularly in the case of SMEs. 208. All these points may carry some weight in seeking to interpret the legislative provisions. However, in the Board of Appeal’s view, each of them is of limited weight. As was submitted on behalf of ESMA, credit rating activities are subject to a specific regulatory framework, which is distinct from the MAR/MIFID frameworks. (1) As opposed to investment recommendations/ research governed by the MAR/MIFID frameworks, credit rating activities are reserved activities, i . e . they may be conducted only by entities that are registered in accordance with Article 14(1) CRAR (or otherwise recognised under CRAR). (2) Credit ratings are subject to specific requirements that do not have any equivalent in the MAR/MIFID frameworks and aim at addressing specific risks that these ratings may present to the financial system, including provisions (i) limiting the use of credit ratings by market participants (Article 4 CRAR), (ii) imposing certain requirements on the credit rating process and methodology (Article 8 CRAR) and (iii) imposing requirements in terms of presentation and disclosure of credit ratings including discontinuance of ratings and unsolicited ratings (Article 10 CRAR and Section D of Annex I CRAR). (3) Whereas investment recommendations/research activities remain supervised principally by national market or banking authorities, credit rating activities are supervised exclusively by a single EU authority (ESMA), and are outside the scope of competence of national market or banking authorities. 209. There is a further point made by the banks in this respect, namely that because of the nature of the CRAR requirements, it would not have been possible for them to issue both credit ratings and investment research and recommendations from a single legal entity. This point is accepted on behalf of ESMA. 210. Again, this may have some limited weight in the sense that it is consistent with the banks’ key contention that a communication cannot be both a credit rating and recommendations/investment research. However, it does not resolve the issue whether as a matter of construction of the provisions, the banks are exempted from the ambit of CRAR under the exclusion in Article 3(2) so as to enable the activities lawfully to be carried out by the same entity. 211. Finally under this heading, it is convenient to deal with a further contention made on behalf of the banks, namely that the creditworthiness of a company, in particular for investment purposes, needs to be assessed on some scale, and given that investment research and recommendations are expressly carved out of CRAR under Article 3(2), there is no reason why this should not be done through alphanumerical symbols. 212. Again, even if this proposition is factually correct, it seems to the Board of Appeal that this contention has limited weight. ESMA accepts that there is no provision in CRAR or in the MiFID framework that prohibits the inclusion of a rating scale in investment research report, subject to the requirement of registration if what has been used amounts to a credit rating within CRAR. In principle, there is no reason why banks using credit ratings within CRAR should not be subject to the rules that govern the issuance of credit ratings while other market participants should be subject to these rules. In the view of the Board of Appeal, the issue remains as stated above, namely whether as a matter of interpretation the banks were exempt from the requirement of registration under Article 3(2). Practice in the Nordic markets 213. The banks’ case is that their practice and understanding corresponded to market practice in the Nordic region, and there appears to have been a consensus that their ratings fell within the Article 3(2) b) CRAR exclusion (a point which has not been disputed). The banks also point to the absence of action from the Swedish and other Nordic regulators, even when they had direct responsibility for CRAR. 214. The banks submit that there is a good reason for this, because a key policy objective in the European Commission’s Action Plan Building a Capital Markets Union is to ensure that SMEs operating within CMU have access to financing, which the banks have facilitated. In that regard, following ESMA’s action against the banks, a new rating agency called Nordic Credit Rating AS was set up and registered with ESMA in 2018. It assigns credit ratings to financial institutions and corporate entities based primarily in Denmark, Finland, Iceland, Norway, and Sweden. Three of the five banks fined by ESMA in 2018 are shareholders. 215. The Board of Appeal’s view is that policy issues are certainly relevant to the construction issue, and that the banks can legitimately claim that their activities in this respect have been beneficial in a wider sense. Practice in the Nordic markets, including the approach of the supervisory authorities, on the other hand, is important on the negligence issue but has limited weight as regards the question of the correct construction or interpretation of the scope of the Article 3(2) CRAR exclusion, since all participants may have been proceeding on an erroneous view of the exclusion. The lack of legal certainty and lack of due process contentions 216. The appellants contend that ESMA proceeded without regard to the principle of legal certainty and with a lack of due process. The banks submit that the lack of due process arises from the imposition of sanctions in the absence of legal certainty, and the issues will be dealt with together. 217. SEB (supported by the other banks) relies on the principles enshrined in the EU Charter of Fundamental Rights, in particular the principles of legality and proportionality of offences and penalties as applicable to the current proceedings. The European Court of Human Rights and the Court of Justice consistently consider that the principle of legality applies to administrative sanction proceedings, particularly fines, and that a penalty, even of a non-criminal nature, cannot be imposed unless it rests on a clear and unambiguous legal basis (see e.g. Case 117/83, Könecke v BALM , § 11). 218. Much authority has been cited particularly by Handelsbanken and SEB, and the principle is not in doubt. The Court of Justice has held that the principle of legality “form[s] part of the legal order of the Community” (Case T-472/13, Judgment of 8 September 2016, § 761; Case C-352/09 P, Judgment of 29 March 2011, § 81; Case C-345/06 Judgment of 10 March 2009, Heinrich , § 44) and is “a fundamental principle of Community law” (Case C-94/05, Judgment of 16 March 2006, Emsland-Stärke GmbH v Landwirtschaftskammer Hannover , § 43). 219. The principle “requires that EU rules enable those concerned to know precisely the extent of the obligations which are imposed on them, and that those persons must be able to ascertain unequivocally what their rights and obligations are and take steps accordingly”, and the “imperative of legal certainty must be observed all the more strictly in the case of rules liable to have financial consequences” (Case C- 158/06, Judgment of 21 June 2007, Stichting ROM-projecten v Staatssecretaris van Economische Zaken , § 25). 220. The Board of Appeal accepts that these principles are applicable to the present administrative sanction proceedings brought against the banks for breach of the provisions of CRAR, and this is not understood to be in dispute. 221. The challenges to the decisions of the Board of Supervisors in this regard largely arise from the way in which the Board dealt with the contention advanced to it that a communication cannot be both a credit rating and investment research, and that it is not possible for only part of the communication to be classified as investment research allowing another part to be characterised as credit ratings. The banks regard this as central to their appeals. 222. The banks submit that despite seeming to recognise that recommendations/ investment research and credit ratings are “distinct concepts”, the Board of Supervisors nevertheless concluded that “a given document could contain both investment research or recommendations and a credit rating” despite its earlier statement that the legislation on that point was “not definitive”. It is said that the Board of Supervisors lacked confidence in its own reasoning, concluding that its view was “not settled”, but then despite its uncertainty, imposing a fine on the basis of that view of the legislation. 223. In the Board of Appeal’s view, the correct way of looking at the decisions is as follows. In considering the distinction between credit ratings and investment research and recommendations, the IIO had expressed the view that these are mutually exclusive concepts. This point was taken up by the banks, and the Board of Supervisors proceeded to deal with it. The Board of Supervisors considered that the legislation was not definitive in this regard. On the basis of the material before it, but without expressing a firm or settled view, it appeared to the Board of Supervisors that a credit rating is a distinct concept from recommendations and investment research in this context. The Board of Supervisors considered that a given document could contain both investment research or recommendations and a credit rating. The Board of Supervisors did not wholly accept the suggestion that investment research or recommendations could contain a rating scale and not be considered credit ratings, and said that its categorisation as a credit rating would seem to be more likely in the light of the aims of CRAR. The Board of Supervisors’ view as to whether these concepts were mutually exclusive or whether there could be an overlap was, as the Board of Supervisors said, not settled. 224. Having thus considered the matter, the Board of Supervisors concluded that: “ … the Ratings are credit ratings within the meaning of the CRA Regulation. [The Board] has reached this view on the basis of the facts in the case i.e. on the material in the IIO file. Considering the Ratings themselves, they would appear to the Board to fall most precisely within the CRA Regulation definition of ‘credit ratings’. This view follows from the analysis summarised … above. The Ratings do not appear to fall within the definitions of either a recommendation or investment research. In particular, the Board has noted that the Ratings (in the sense of the rating categories) do not in themselves appear to recommend or suggest an investment strategy, which would have been expected of investment research pursuant to its definition.” 225. It is correct that this part of the Board of Supervisors’ decision identifies certain points that had been raised with it, without taking a definitive view. It seems to the Board of Appeal that this is unobjectionable, particularly since reconciling the provisions is not easy. The approach of the Board of Supervisors was to decide the matters it had to decide, whilst leaving wider points open for further consideration against the factual context in which such points may arise. In the Board of Appeal’s view, this was a sensible approach. 226. This approach could be a basis to impugn the decisions on grounds of legal certainty if a conclusive view on the matters it regarded as unsettled was necessary for the Board of Supervisors to reach a decision. But (in the Board of Appeal’s view) it was not necessary to decide every aspect of the arguments put to it. As put to the Board of Appeal, the central question on the appeals is whether a document can constitute both investment research and a credit rating. It is not apparent to the Board of Appeal that the issue was put to the Board of Supervisors with quite such clarity. However, the banks accept that the Board of Supervisors concluded that a given document can constitute both investment research or recommendation and a credit rating, and proceeded on this basis (e.g. § 5.1.3 and 5.1.3(iv) of Nordea’s submissions). 227. The criticism seems to be of the reasoning in support of that conclusion, but that does not in itself raise issues of legal uncertainty. The fact that a given provision of financial regulation is open to different interpretations does not, in the Board of Appeal’s view, necessarily invoke the principle of legal certainty in respect of sanctions. As stated below, on the authority of the Denkavit case, a lack of clarity and precision of the rule breached may, depending on the circumstances, be a factor relevant to the question of negligence leading to a fine. But it does not follow for example that, as in the present case, the issuance of a public notice is in any way impugned. This appears from the provisions of Article 24(1) CRAR (as amended) which give power to issue a public notice, and also give power to require the infringement to be brought to an end, a power which did not have to be exercised in the present case because the banks voluntarily desisted. 228. The most that can be said by way of criticism – and the Board of Appeal accepts that this is a valid criticism made by the banks – is that the Board of Supervisors did not expressly engage with the banks’ contention now advanced with great clarity that if the inclusion of rating symbols within investment research is sufficient in itself to amount to “credit ratings” so as to bring the investment research within CRAR, then the exclusion in Article 3(2) is redundant in that it could never have any application. This is a contention that arises as a matter of interpretation and is considered as such below. 229. For these reasons, the Board of Appeal does not accept the appellants’ submission that to uphold the Board of Supervisors decisions would in essence be to impose a retroactive duty, from the date of the decisions, unascertainable to market participants. 230. Legitimate expectation arguments were raised by the appellants, in the sense that the banks contend that they had a legitimate expectation that the position taken by the Board of Supervisors would be made clear in advance by guidance or otherwise. However, the Board of Appeal does not consider that these arguments add anything substantive to the case as to legal certainty and due process. 231. In the context of due process, several of the banks have contended that ESMA failed properly to review the facts of the case. That is not accepted by the Board of Appeal. The procedure by which ESMA reached its decisions is set out above, and the Board of Appeal is satisfied that all relevant factual matters were properly taken into account. 232. It is convenient to deal here with a contention advanced on behalf of ESMA in this context to the effect that it had in fact previously issued guidance on the issue in question, and that ESMA’s position was consequently foreseeable, and the appellants were in a position to know which acts or omissions would make them liable. 233. The position is that since 2011, ESMA has published various guidelines dealing with various aspects of CRAR, but none of them deal with the construction issue that arises in these appeals. 234. Prior to the guidelines, ESMA had issued a Consultation Paper (ESMA/2012/841) in 2012 dealing with various issues, but not the construction issue that arises in these appeals. It is however contended that in §34 under the heading “Practical Examples” ESMA provided clear guidance. 235. The Board of Appeal rejects this contention. The example is focused on methodology, and in context does not relate to the issue on these appeals. There is no reason why the appellant banks should have read it as doing so. The legislative history of the provisions dealing with investment research and credit ratings 236. In her reports, the IIO was rightly concerned to ascertain the legislative history of the provisions dealing with investment research and credit ratings with a view to assisting to establish the correct construction of Article 3(2). The Board of Appeal largely supports her analysis. 237. The Commission’s original proposal of 12 November 2008 (COM (208) 704 final) defined “credit rating” for the purpose of delineating the scope of the CRAR essentially, but for few minor precisions, in the same way as in the adopted CRA Regulation. 238. Article 3(2) of the Commission’s original proposal, on the other hand, was different and had the opposite purpose of the current text of Article 3(2) in the adopted CRAR. In the Commission’s proposal, paragraph 2 was as follows: “For the purposes of point (a) of paragraph 1, credit ratings shall not be considered recommendations within the meaning of Article 1(3) of Commission Directive 2003/125/EC”. 239. The intended purpose was to specify that a credit rating, albeit being an opinion regarding the creditworthiness of an issuer or of a debt security, was not to be confused with an investment recommendation within the market abuse framework. In other words, paragraph 2 was clearly intended, in the Commission’s original proposal, as a clarification for the benefit of credit rating agencies. The original proposal did not address the question whether the provision of (regulated or unregulated) investment services, in the form specifically of investment recommendations, investment research and other opinions by entities which were not credit rating agencies, could be considered a credit rating activity. 240. The ECON Committee Report of 1 April 2009, Rapporteur MEP Jean-Paul Gauzès (A6-0191/2009) did not table any amendments to the original Commission’s proposal in this respect. 241. Also the ECB, in its Opinion of 21 April 2009, did not raise any specific issue in this regard. The ECB noted instead, as to the scope of the proposed Regulation that: a) “[The proposal] alternates between, on the one hand, the objective of introducing a ‘common approach to ensuring the high quality of credit ratings to be used in the Community’ and ‘all ratings used by financial institutions governed by Community legislation are of high quality and issued by credit rating agencies subject to stringent requirements’ and, on the other hand, the more limited objective of requiring registration only for the credit rating agencies established in the Community and seeking to ensure that their credit ratings are used for regulatory purposes by financial institutions in the EU”. In footnote 10, the ECB quoted specifically Recital 28, second sentence, referring “to the requirement that competent authorities should have the necessary means to ensure that ‘ratings for use within the Community are issued in compliance’ with the proposed regulation. b) “The proposed regulation does not clarify the rules applicable to securities for which a prospectus has been published under Directive 2003/71 (…) and which are rated under the proposed regulation”. This intersection between CRAR and this different piece of European legislation pertaining to capital markets regulation was then addressed by Article 4(1), second sentence of the adopted CRAR. 242. On 23 April 2009, the European Parliament adopted its “position on first reading for the adoption of the proposed Regulation”. This text contains – for the first time, so as far as the Board of Appeal can verify from publicly available Parliamentary documents the Board of Appeal was able to identify – both recital (20) and Article 3(2) as they currently stand in the adopted CRAR. 243. The Board of Appeal could not find any Parliamentary preparatory documents illustrating the reasons why the new recital (20) and the new text of Article 3(2) were inserted in the EP position on first reading. It may be that the original provision was considered inappropriate in so far as it could suggest that ratings, albeit not recommendations under MAR, may nonetheless be close to, or have elements in common with, an investment recommendation (an issue which in some aspects had come up in U. S. disputes against rating agencies). It may be that the sentence “For the purposes of point (a) of paragraph 1, credit ratings shall not be considered recommendations (…)” was considered inaccurate by the co-legislators, because it is difficult to see how the fact that a credit rating is not a recommendation could be “for the purposes of point (a) of paragraph 1”, as far as point (a) just defined what, for the purpose of CRAR, credit ratings are. (To this effect, the sentence should have rather been that “Credit ratings for the purposes of [or: as defined under] point (a) of paragraph 1 shall not be considered recommendations”). 244. There may have been other reasons of which the Board of Appeal is not aware, for instance suggestions from the industry. Be that as it may, what is known is that negotiations between the co-legislators led to the result that the original Article 3(2) of the Commission’s proposal was fundamentally transformed. A provision originally directed at credit rating agencies became a provision directed at entities issuing recommendations under MAD or engaging in investment research and other forms of general recommendations under MiFID or otherwise providing “opinions about a value of a financial instrument or a financial obligation”. 245. Yet, the incipit “For the purposes of paragraph (1)(a)…” remained, just in a slightly different form from the original “for the purposes of point (a) of paragraph 1”, but in a completely different context: one in which, as noted, the original purpose of paragraph (2) was reversed, it not being specified that the rating that was not a recommendation for the purposes of MAD, but that recommendations under MAD and MIFID were not ratings, to the effect that those providing such recommendations should be deemed outside the scope of CRAR. 246. The Board of Appeal considers it quite likely that the intention of the co-legislators when adopting the new Article 3(2) was to broadly clarify that all the three types of “opinions” listed in the amended paragraph (2) should be considered opinions different from those which qualify as ratings irrespective of the fact that they included a ranking system of rating categories (as specified in the last part of Article 3(1)(a). In the Board’s view, however, it is also quite possible that, through the reference to paragraph (1)(a) – which qualifies ratings not only as opinions but as opinions expressed using a ranking system of rating categories – the co-legislators intended to be specific so as to address specifically recommendations, investment research and other opinions which include a rating (as the reference to the entire letter a) of paragraph 1, textually taken, would imply). 247. The Board of Appeal notes, in this respect, that the circumstance that Article 3(2), as it stands now, was not entirely drafted from scratch in its current version but was amended during the legislative process, and that its incipit (which is posing the interpretative dilemma in the instant case) was already there in the Commission’s proposal with the purpose of defining what credit rating is, casts some doubts on the real intention of the co-legislators. Considering that legislators cannot be assumed to be infallible in drafting rules, in this specific case it cannot even be excluded, in the Board’s view, that the reference to paragraph (1)(a) remained with the same function it had originally (albeit with a quite unfortunate wording from the beginning), and thus to specify what a credit rating is, without any intention to restrict the class of the recommendations and research investments covered by Article 3(2) only to the sub-class of those among them which in fact included also a ranking system of rating categories. 248. In other words, in the Board of Appeal’s view, it cannot be excluded by reference to the legislative history that Article 3(2) was just intended to make into a rule the principle stated in recital (20), which simply reads that “investment research, investment recommendations and other opinions about a value or a price of a financial instrument of a financial obligation should not be considered to be credit ratings” (it must be recalled in this regard that the articles of a Regulation must be interpreted in the light of the relevant recitals), without any further qualification concerning the required inclusion in the recommendations or investment research of opinions on creditworthiness using a rating scale. 249. The Board of Appeal acknowledges therefore that there is a significant ambiguity in the wording of Article 3(2) and in the combined reading of Article 3(1)(a) and Article 3(2) and that this ambiguity cannot be resolved with certainty by looking at the legislative history of the provision. The Board of Appeal acknowledges also that a strong argument can be made out that, taken on its literal meaning, the wording of the two provisions and their combined reading, could also be interpreted as advocated by the appellants. That would consider the reference to the recommendations, investment research and other opinions mentioned in Article 3(2) as being further qualified by the fact that they also include a rating. But the legislative history does not support the view that the literal meaning of the provision is clear and unambiguous, and the Board of Appeal agrees with the conclusion of the IIO that it is also necessary to consider whether the literal interpretation as advocated by the appellants would make CRAR devoid of at least some of its purposes and would, to some extent, contradict the scope of the Regulation, as defined in Article 2, opening an unreasonable loophole in the system (subject to Article 4, Use of credit ratings). The correct construction of the Article 3(2) exclusion 250. The starting point of the Article 3(2) exclusion in terms of construction is Recital 20 of CRAR which provides that, “Investment research, investment recommendations and other opinions about a value or a price for a financial instrument or a financial obligation should not be considered to be credit ratings.” 251. This language is reflected in Article 3(2) itself, which adds definition to the terms “recommendations” and “investment research” by reference to the Market Abuse and MiFID Implementing Directives respectively. 252. It is not in issue that the banks’ various communications fell within the ambit of MiFID and MAR. As noted above, the question of construction or interpretation is whether the Board of Supervisors was correct to find that the ratings within those communications fell within the CRAR definition of “credit ratings” and did not fall within the definitions of either recommendations or investment research so as to be excluded under Article 3(2). 253. As set out above, the appellant banks have a number of key contentions in this respect (the below is taken from Nordea’s submissions as the first presenter at the oral representations hearing, but equally any of the banks’ submissions making these points could be taken). These are restated here for convenience. The points dealt with above are not repeated: (1) A document cannot constitute both investment recommendations and investment research and a credit rating. A document falling within the definition of recommendations and investment research is excluded from the restrictions contained in CRAR through the wording of Article 3(2). (2) Credit ratings and recommendations/investment research are distinct and mutually exclusive concepts. The Board of Supervisors was wrong to conclude that such distinct concepts can co-exist within a single document or communication. (3) Credit ratings and recommendations/investment research were intended by to be regarded differently: Recital 10 of the MAD Implementing Directive distinguishes them on the basis that the former are opinions on the creditworthiness of a particular issuer or financial instrument “as of a given date” , their purpose being to articulate the issuer’s creditworthiness, without encouraging (or otherwise) an investment in the particular issuer or instruments. (4) In contrast, the fundamental purpose of recommendations/ investment research is to propose an investment strategy. These are forward looking and give a view as to the present/future value of the financial instrument/issuer. It may include a view on a particular issuer’s creditworthiness, but this goes to the merits of purchasing that issuer’s securities. (5) There is no legislative basis or regulatory guidance supporting the Board of Supervisors’ characterisation of a credit or investment research report not as a single document, but as a piece comprising a number of component elements, each of which should be individually analysed and characterised. (6) The Board of Supervisors’ construction would deprive the Article 3(2) CRAR exemption of any effect. The whole purpose of this provision is to take recommendations which have the features of credit ratings outside of the scope of the regulation. So if we are not in that territory, it is simply irrelevant. (7) The question whether a rating can simultaneously constitute recommendations/ investment research and a credit rating is fundamental. The Board of Supervisors seemed to conclude that the distinction was based not on whether the document included a view on creditworthiness but on whether or not an established and defined ranking system of rating categories was used. This would mean only that recommendations/investment research which do not express an opinion “using an established and defined ranking system of rating categories” would fall within the scope of Article 3(2). But the article is intended as a carve-out – such documents could never constitute “credit ratings” within Article 3(1)(a) CRAR as this requires both an opinion on creditworthiness and use of an established rating scale. Such documents would therefore not require the benefit of any exemption. (8) Whilst the general principle that exemptions from Community law should be construed narrowly is not disputed, context is important (C-340/94 E. J. M. de Jaeck v Staatsecretaris von Financien , para 17) and the legislators must have had an aim in including the exemption. 254. ESMA appeared to accept that recommendations/investment research on the one hand, and credit ratings, on the other hand are distinct concepts, and that is the view of the Board of Appeal. The point is well made by reference to the legislation referred to in Article 3(2). Recital 10 of the Implementing Directive for the Market Abuse Directive states that credit rating agencies issue opinions on the creditworthiness of a particular issuer or financial instrument as of a given date, and as such, these opinions do not constitute a recommendation within the meaning of the Directive. Article 24(1) of the MiFID Implementing Directive defines “investment research” as research or other information recommending or suggesting an investment strategy. Credit ratings may inform an investment strategy, but they do not recommend it. 255. The question is what follows from that. The banks draw a bright line, submitting that a document cannot constitute both recommendations/ investment research and a credit rating, and that once it is ascertained that the document falls into the former category, the Article 3(2) exclusion applies. As it was put on behalf of Handelsbanken, the reports were an integrated whole, and the “credit rating” was not a naked or separate product but formed an integral part of the overall assessment of the creditworthiness of the company concerned. 256. The banks’ response to ESMA’s contention that the inclusion of credit ratings within such reports will trigger the application of CRAR is that this is a circular contention, because Article 3(2) excludes such reports from the definition of credit rating in Article 3(1)(a), and the Board of Supervisors was wrong to conclude that such distinct concepts can co-exist within a single document or communication. 257. The Board of Appeal’s view is as follows. It has already acknowledged that a strong argument can be made out by the appellants on the literal meaning of the wording of Article 3(2), by which recommendations within the meaning of the Market Abuse Directive and investment research within the meaning of MiFID shall not be considered to be credit ratings. The argument on the literal interpretation is supported by the wording of Recital 20. 258. The Board of Appeal also considers that a strong argument can be made out by the appellants that the Board of Supervisors’ construction would deprive the Article 3(2)(b) exclusion of any effect (which is not permissible: Case C-8/01 Taksatorringen v Skatteministeriet ). ESMA had difficulty in identifying any situation in which (on its construction) the exclusion would operate. The suggestion that an example was found in Article 3(2)(b) itself in that “buy”, “hold”, “sell” recommendations could arguably fall under the definition of “rating category” was not convincing given the definition of “rating category” in Article 3(1)(h). In oral representations, it was frankly accepted on behalf of ESMA that it did not have examples of products that fall into both categories, but maintained that it is possible that these products could exist, and that there could in the future be a product which both gives an opinion on creditworthiness and gives a recommendation as to investment strategy, and which is expressed in the form of a scale. 259. Further, though as ESMA points out, Article 3(2) is not in the exemptions section of Article 2(2), the historical analysis above explains why that is so. The Board of Appeal’s view is that it has the same effect as an exempting provision, and in any case it cannot be interpreted in such a way as to leave it without content. 260. However, a strong argument is not necessarily a correct argument, and there are important points that go in the contrary direction. As explained above, Article 3(2) in the Commission’s original proposal was changed in the Parliamentary process, and a provision originally directed at credit rating agencies became a provision directed at entities issuing recommendations or engaging in investment research and other forms of general recommendations under MAD/MiFID. Whilst the banks have referred to the acknowledged benefits of their products in the Nordic debt market particularly as regards SMEs, they have not suggested that the change to Article 3(2) was in any way a response to this consideration. 261. In this regard, the principle is that where the wording of an EU law provision is clear and precise, its contextual or teleological interpretation may not call into question the literal meaning of that provision, as this would run counter to the principle of legal certainty (Koen Lenaerts, José A. Gutiérrez-Fons, To say what the law of the EU is: methods of interpretation and the European Court of Justice , p. 7). The Board of Appeal’s opinion is that the legislative history does not support the view that the provision in question should be considered as clear and unambiguous in such a way as to be decisive on the question of interpretation. 262. There is certainly no indication that it was intended to open a major exception to the operation of CRAR. Though the banks understandably played the practical consequences of this down, the effect of their interpretation would be that market participants (including those not subject to the MiFID framework because general investment recommendations/research is not a reserved activity, and also because Article 3(2)(c) also includes a potentially very wide class of “opinions about the value of a financial instrument or a financial obligation” which are not recommendations or investment research as defined in letters (a) and (b)), and even (potentially) registered CRAs, would be able to avoid the application of CRAR simply by including credit ratings in documents containing recommendations or investment research or even “opinions about the value of a financial instrument”. In other words, subject to the market abuse framework, anyone could at least in theory issue credit ratings so long as the ratings were included in a document that fell within the Article 3(2) definitions. That this would be so was not in dispute. These ratings could not have the regulatory use set out in Article 4 (this Article expressly requiring that for regulatory purposes credit ratings can be used only if they are official and issued by registered credit rating agencies), but would nonetheless be (and present themselves as) credit ratings. 263. This was a consideration also relied upon by the Board of Supervisors in reaching their decision as to the ambit of Article 3(2), though understandably in more guarded language. The Board of Appeal accepts that it is very unlikely that this was the interpretation intended by the legislator. Given all the circumstances, it regards this as a decisive factor in reaching a true interpretation of the provisions. 264. This does not necessarily deprive Article 3(2) of effect, the purpose of which may simply be one of clarifying the general position of investment recommendation or investment research which is consistent with its treatment in earlier legislation. Alternatively, there is some force in ESMA’s submission that a product may be developed which does fall into both categories, which is not an impossible outcome given the propensity of the financial markets to change over time. The fact that it is unable presently to give examples is of limited significance. 265. Further, as the Board of Supervisors noted, the banks’ ratings (in the sense of the rating categories) do not in themselves appear to recommend or suggest an investment strategy, which would have been expected of recommendations or investment research under the applicable definitions. 266. In oral representations, ESMA referred to the example produced by Swedbank of credit research in respect of a bond issue by a Swedish company. This is described above. The recommendation is described at the top of the document as “overweight” (which in context is a recommendation to buy), and at the bottom of the document is stated, “We view […] as an investment grade issuer in the BBB+ range”. 267. The Board of Appeal found it helpful to put the issue into the context of a specific example like this. (To deal here with another point raised by the banks, as a matter of analysis, it makes no difference that the BBB+ is only a small part of the document overall – this is a feature of any alphanumerical symbol, including those used by the rating agencies.) 268. ESMA’s reading of this document is that the BBB+ content is a credit rating, whereas the content "overweight" is an investment recommendation. This is consistent with the Board of Supervisors’ decisions, and accepts that the same communication may be treated as having different components for regulatory purposes. It may also be added that if the recommendation “overweight” was accompanied by an official credit rating issued by a registered rating agency (as it appears to be sometimes, albeit less often, the case in investment research produced by the appellant banks), the presence in the same document of such a credit rating would not make the whole document a credit rating (Article 3(2) clarifies that such a transmutation is prevented); both the recommendation content and the rating content would remain what they are, but in such a case the coexistence of these two components in the same document prepared by the bank would be fully lawful, because the official credit rating is not issued by the bank but by a registered rating agency. 269. The Board of Appeal has come to the conclusion that the interpretation adopted by the Board of Supervisors is the correct interpretation. It considers that the same communication may have content which consists of recommendations or investment research within the Article 3(2) exclusion, and content which consists of “credit ratings” within the meaning of the definitions of “credit rating” in Article 3(1)(a) and (h) of CRAR. If so, and if the other requirements of CRAR are satisfied, the communication, for its rating component, will fall within CRAR, and the issuer will require to be registered. It follows that the banks’ appeals on the construction issue must fail. SEB’s case that its research reports were not disclosed publicly or distributed by subscription 270. SEB claims that its investment research reports (including the shadow ratings) were only distributed to a ‘selected’ group of investors and were therefore not “disclosed to the public” or “distributed by subscription” for the purposes of Article 2(1) CRAR. The IIO statement of findings shows in paragraph 152 that the investment research reports were accessible via the SEB portal and that some clients were further emailed such reports. The number of recipients of such reports was, as of 20 April 2016, 1,147. 271. Article 2(1) CRAR provides that it applies to credit ratings “which are disclosed publicly or distributed by subscription”. In turn, Article 2(2)(a) provides that CRAR does not apply to “private credit ratings produced pursuant to an individual order and provided exclusively to the person who placed the order and which are not intended for public disclosure or distribution by subscription”. The same principle is also stated in recital (19) CRAR. 272. The Board of Appeal considers that shadow ratings distributed to a large number of bank’s clients must be considered “disclosed to the public” or “distributed by subscription” even assuming that the clients do not pay any specific fee for such shadow ratings or for being included in the list of the addressees of such investment researches or recommendations (including the shadow ratings) and even if these potential investors are only a selected fraction of the overall clients of the bank. 273. In the Board of Appeal’s view, as a matter of principle, the notion of ‘public’ in respect of any reserve of regulated financial activity to duly authorised entities complying with the requirements set out in the relevant regulation is broad, and this is necessary to prevent circumvention of such registration requirements. Any ‘private’ exception (like the one referred to in Article 2(2)(a) CRAR) must be subject, therefore, to strict interpretation. 274. This is also confirmed, in the Board of Appeal’s view and as already noted by the IIO in her statement of findings, by the fact that investment recommendations under MiFID (more specifically, Article 24(1) of the 2006 Commission MiFID I Directive) are intended for distribution channels or for the public, and a distribution channel is defined as a channel through which information is, or is likely to become, publicly available (compare paragraph 154 IIO’s SEB statement of findings). Consistently, in the MAR context, as also noted by ESMA in its submissions, ESMA’s Final Report Draft technical standards on MAR of 28 September 2015 (ESMA/2015/1455, § 340) clarifies that “ESMA holds the view that an investment recommendation is intended for distribution channels or for the public not only when it is intended or expected to be made available to the public in general, but also when it is intended or expected to be distributed to clients or to a specific segment of clients, whatever their number, as a non-personal recommendation”. The Board of Appeal notes, in this respect, that the appellant acknowledges that its investment recommendations and researches are either investment recommendations under MAR and/or investment researches under MiFID and cannot be considered therefore strictly ‘private’. 275. The Board of Appeal considers further that, as noted by ESMA in its submission, a list of persons who are granted access to a service via a portal or through delivery to their email addresses (which SEB does not contest) must be considered a list of subscribers under Article 2(1) CRAR, and this is the same manner in which registered CRAs provide access to their ratings to their subscribers. 276. The Board of Appeal finally holds that the fact that the addressees represent a selected number of the bank’s clients becomes irrelevant once it is acknowledged that these addressees are in the hundreds (more than a thousand in the instant case). It is also irrelevant whether the addressees pay a specific subscription fee or are offered the service as part of the services they receive from the bank. (3) The negligence issue The issue 277. The issue is whether the Board of Supervisors was correct to find that the appellant banks “negligently” committed the infringement listed in Annex III of CRAR by issuing credit ratings without being authorised by ESMA to do so, that being a condition of the ESMA’s power to impose a fine in Article 36a CRAR. 278. Also relevant in this regard is Recital 18 of Regulation (EU) No 513/2011 amending CRAR stating that “ESMA should also be able to impose fines on credit rating agencies, where it finds that they have committed, intentionally or negligently, an infringement of Regulation (EC) No 1060/2009.” No definition of negligence is provided in CRAR, but as appears below, the test has been stated by the Board of Supervisors in a number of decisions, as well as in the decisions in hand. 279. It is not contended that the infringement was intentional, that being the other condition under Article 36a. The applicable law 280. In its decisions, the Board of Supervisors noted that it had previously set out its views in relation to the negligent commission of an infringement. It considered that the test for negligence established by the case law of the Court of Justice (citing the Opinion of Advocate-General Mayras in Case 26/75 General Motors Continental NV v Commission , and Case C-308/06 Queen on the application of International Association of Independent Tanker Owners (Intertanko) v Secretary of State for Transport , para 77 (3 June 2008)) is as follows: “Negligence is established for a CRA where, as a professional firm in the financial services sector subject to stringent regulatory requirements, it is required to take special care in assessing the risks that its acts or omissions entail, and has failed to take that care. Further, as result of that failure, the CRA has not foreseen the consequences of its acts or omissions, including particularly its infringement of the CRA Regulation, in circumstances when a person in such a position who is normally informed and sufficiently attentive could not have failed to foresee those consequences.” 281. In a careful analysis of the law, the Board of Supervisors continued as follows: “Negligence is an Union law concept in the context of the CRA Regulation, albeit one which is familiar to, and an inherent part of, the 28 Member States’ legal systems, and must be given an autonomous, uniform interpretation. It would appear, from the provisions of Articles 24 and 36a of the CRA Regulation, that the term ‘negligence’ in the context of that Regulation requires more than a determination that an infringement has been committed. It is clear from the second subparagraph of Article 36a(1) of the CRA Regulation that a negligent infringement is not one that was committed deliberately or intentionally. This position is further supported by caselaw in which the CJEU has said that negligence may be understood as entailing an unintentional act or omission (see the Intertanko case above). The CJEU jurisprudence suggests that the concept of a negligent infringement of the CRA Regulation is to be understood as denoting a lack of care on the part of a CRA in complying with the CRA Regulation. The Board notes the position taken by the General Court in the Telefonica case, where the General Court spoke of persons “carrying on a professional activity, who are used to having to proceed with a high degree of caution. They can on that account be expected to take special care in assessing the risks that such activity entails” (Case T-336/07 Telefónica, SA and Telefónica de España, SA v European Commission , para. 323). Similarly the Board considers that in circumstances where, operating within the framework of a regulated industry, an entity which holds itself out as a professional entity and carries out regulated activities should be expected to exercise special care in assessing the risks that its acts and omissions may entail. The Board is of the view that a high standard of care is to be expected of a CRA. The nature and extent of the requirements imposed on CRAs by Annex I of the CRA Regulation, and of the corresponding infringement provisions under its Annex III, appear to reflect the weight given to these considerations by the legislator. The Board considers that in order to ensure a high standard of care by CRAs, the acts and omissions of a CRA should be judged with these considerations in mind.” 282. The parties made a number of points by way of commentary on or supplement to the legal test. In particular: (1) SEB pointed out that the General Motors and Intertanko cases were in the competition and maritime fields. But whilst credit rating is obviously a very different activity, the Board of Appeal does not think that this renders the principles inapplicable, nor was this suggested. (2) Nordea cited Joined Cases C–283/94, C-291/94, and C–292/94 Denkavit International BV v Bundesamt für Finanzen in submitting that context is important when considering negligence, and as showing that the fact that multiple parties formed the same view of legislation is relevant as regards whether the required standard of care is satisfied. (3) The Board of Appeal notes that in Denkavit the question was whether an action lay against a state in the context of taxation for a sufficiently serious breach of Community law. This has to do with the question whether the law was breached intentionally, rather than negligently, as is the issue in the present case. Nevertheless, the Board of Appeal considers that the case may be of relevance in the context of negligence also, and notes that the Court said that “One of the factors that may be taken into consideration in this regard is the clarity and precision of the rule breached” (§ 50) (4) On behalf of ESMA, reliance was placed in oral presentations on the Marine Harvest Case (T-704/140, 26 October 2017) in which the Court found that the applicant, a large European company, behaved negligently in construing the relevant Regulation, and in so doing acted at its own risk and could not legitimately rely on the allegedly “reasonable” nature of its interpretation (§ 259). (5) The appellants pointed out that the Marine Harvest Case arose in the competition field, but the Board of Appeal does not think that this renders the principles inapplicable, and some of them are relevant to the instant appeals. Though fact sensitive, as all negligence cases are, it is helpful in showing how the Court went about the analysis of the issue. (6) In the context of the high standard of care expected from professionals, ESMA relied on Case C-48/98, Firma Söhl & Söhlke v Hauptzollamt Bremen , November 11, 1999, paras 56-59, in which the Court said that, “As regards the care taken by the trader, it must be noted that, where doubts exist as to the exact application of the provisions, non-compliance with which may result in a [sanction decision being issued], the onus is on the trader to make inquiries and seek all possible clarification to ensure that it does not infringe these provisions”. 283. Subject to these points, there was no challenge by the appellants to the analysis of the law stated by the Board of Supervisors, and the Board of Appeal adopts it as correct. The facts 284. The facts are as set out above, and where necessary in the findings below. The facts were dealt with in detail in the case of each appellant by the IIO and the Board of Supervisors. As has been seen, there are differences between the various banks which are relevant to the negligence issue, and which the Board of Appeal has taken into account. It notes however that the fines imposed were the same in each case, and it has not been submitted on behalf of ESMA in the appeals that a different outcome may apply as between the banks because of any factual differences between them. The Board of Appeal will adopt the same approach. Discussion of the issues 285. In its submissions, ESMA rightly emphasises that financial service providers and CRAs play an important role in the economy of the EU, as well as in the financial stability and integrity of the financial markets. A high standard of care is to be expected of such persons, and that includes the appellants in the Nordic region. This is reflected in financial regulation. Such persons are required to have in place internal governance arrangements to ensure compliance with applicable regulatory requirements (as to credit institutions, see Directive (EU) No 2013/36 (“CRD IV”) and the guidelines adopted by the European Banking Authority (“EBA”), as to CRAs, see CRAR Annex I, Section A, para 3). 286. In its decisions, the Board of Supervisors sets out the steps that the banks might have taken to satisfy that high standard of care (this was stated in the same terms as regards each bank): “ … such steps might have included, for example, an initial evaluation as to whether the CRA Regulation might apply to its production of the Ratings, the taking of legal advice on the scope and effect of the CRA Regulation and/or seeking advice from their National Competent Authority and/or ESMA on those issues. In addition to such an initial evaluation (that is, prior to the CRA Regulation’s implementation), the Board considers that [the bank] might have been expected to subject its initial conclusion to periodic review.” 287. Judged against those steps, which it has been emphasised on the appeals are not additional steps not contemplated by the law but merely examples of steps that a diligent bank “might have” taken with the understanding that other measures could also have been taken to comply with the high standard of care, the Board of Supervisors went on to consider the position as regards each bank. 288. In each case, it found that the high standard had not been met, and noted in particular the absence of any documentary material to support the banks’ contention that they had reviewed the position, a point also remarked on by the IIO. The banks do not challenge the absence of material, and in several cases have expressed regret (though each takes the position that this is not indicative of any negligence on their part). 289. In the circumstances, the Board of Supervisors noted the finding of the IIO to the effect that there was no negligence, but decided not to follow her findings in that respect. 290. In its submissions, ESMA says that it is settled case-law that the fact that specific conduct has not yet been examined by the authorities does not exonerate an undertaking from compliance. The Board of Appeal agrees with this submission. 291. The Board of Appeal also agrees that an initial evaluation of CRAR might normally have been expected, and where appropriate, that evaluation periodically reviewed, and that failure to do this may be an indicator of negligence. In fact, the Board of Supervisors accepted that Swedbank, SEB and Handelsbanken had conducted such an initial review, though they had not produced any documentary evidence of it. (Nordea did not doany detailed analysis of the applicability of the Regulation, but any differences between them were not treated as a differentiating factor as between banks.) None of the banks had periodically reviewed their initial evaluation. The question however is whether the failure to do so indicates negligence with much force in the present case. One reason for the IIO’s finding that there was no negligence was her finding that the practice of “shadow ratings” was common in the Nordic countries and had taken place for many years without any concerns being raised by the regulators or others, and the Board of Appeal agrees that this is significant. 292. As to seeking advice from their National Competent Authority, the banks (all Swedish banks) maintain that their national supervisor (Finansinspektionen) would not normally give advice in a situation like this. In any event, as ESMA itself has pointed out, national authorities are not in charge of interpreting and enforcing CRAR (their responsibilities only subsisted for a short period before being transferred to ESMA). 293. However, ESMA is on stronger ground in its submission that the banks could have sought advice from ESMA itself (c.f. the EC in competition cases – the Marine Harvest case, ibid, at § 256), although it should also be acknowledged that ESMA conceded at the hearing of the oral representations that such guidance would have been given through informal advice or the use of Q&As so that its advice, if sought, to discontinue such a long established practice in the market, if not adopted in the form of a binding decision, would not have offered to the banks the same possibility to resort to judicial protection that they enjoy in the instant case in respect of the appealed decisions. 294. As regards taking legal advice, the position is perhaps not so clear. The banks point out that they have substantial legal resources in-house. Further, as stated by the Court in the Marine Harvest case, ibid, at § 286, “an undertaking may not escape imposition of a fine where the infringement of the competition rules has resulted from that undertaking erring as to the lawfulness of its conduct on account of the terms of legal advice given by a lawyer”. In fact, Nordea did take outside legal advice, but this was in 2017, and by then it had ceased using shadow ratings in the light of ESMA’s investigation. 295. ESMA submits that had a proper legal assessment have been carried out by the appellants as to the scope and purpose of CRAR, that would have at least raised questions as to the legality of their credit rating activities which could have been answered through the steps identified by the Board of Supervisors. The banks respond that a legal assessment would have shown that their activities were lawful as falling in the Article 3(2) exclusion. ESMA responds that it cannot be said in good faith that there was no possibility that the ratings, which are expressed in the form of standardised rating notches, could potentially be subject to CRAR. Only an inadequately short assessment made without appropriate due care and seriousness could have led to such a conclusion. 296. The Board of Appeal notes that the submission that it cannot be said in good faith that there was no possibility that the banks’ ratings could potentially be subject to CRAR is a very low threshold, and that ESMA’s case in relation to the legal certainty issue is that the provisions are clear and settled. The Board of Supervisors found that if the banks had taken special care in assessing the risks of their conduct, the banks would not have failed to foresee that the issuing of the Ratings would amount to an infringement of the CRAR. 297. In this respect, the unusual feature of the present case is that the practice was prevalent in a particular market, namely the Nordic market, and the consensus was that it fell within the Article 3(2) exclusion. This diminishes the force of the foreseeability contention, and also the force of ESMA’s point that each regulated entity is responsible for conducting its own assessment of its obligations under the law regardless of what others in the Nordic region were doing – in itself, that is correct, but what others do in a market may be relevant in the assessment (c.f. Denkavit , ibid). 298. Again, whilst it is correct, as submitted on behalf of ESMA, that the fact that the authorities did not raise any concerns is irrelevant because it does not relieve an entity of its duty of care or of its obligation to ensure that it complies with applicable regulations, that submission has to be seen against the unusual feature of this case. The Board of Appeal’s conclusion on the negligence issue 299. The main points of ESMA’s case may be summarised as follows. When CRAR came into force, it was incumbent on the Nordic banks to check whether their practice, however long-standing, of including ratings in their recommendations and investment research complied with the new regulatory regime. This is brought into particular focus in the case of the two banks (Swedbank and Handelsbanken) whose ratings used or were partially based upon the methodologies of the “official” rating agencies, but it applies to all of them. The banks claim to have made an initial appraisal, but were unable to produce any documentation in support. Had they done a proper appraisal, or kept the matter under review as they should have done, they would have ascertained that it was (as ESMA has contended on the appeals) clear that they fell outside the recommendations/investment research exclusion in Article 3(2) of CRAR. Alternatively, they could have made enquiries of ESMA, and any doubts would have been resolved. 300. When one adds to that the admittedly high standard of care imposed by law upon the banks, it is clear to the Board of Appeal that a strong case of negligence on the part of the banks can be made out, and that case has been strongly argued on behalf of ESMA on the appeals. 301. These points notwithstanding, it is also plain there are factors that support the contrary result. 302. The first relates to practice in the Nordic market. As noted above, from the material before the Board of Appeal, a clear picture has emerged which was largely accepted by the IIO and not challenged on the appeals. For many years, there was a practice in the Nordic debt market by which banks included in their credit reports or the like ratings of an alphanumerical nature. The evidence is that this worked well, and was particularly useful to SMEs for whom the “official” rating agencies were too expensive. 303. The material cited by the banks to support the contention that it was not appreciated by the authorities that CRAR impacted on this practice includes: (1) A statement by Finansinspektionen (the Swedish financial supervisory authority) of 10 February 2009 referring to the limited scope of CRAR to activities in Sweden and making no reference to the use of ratings by banks, and the statement by the Swedish Better Regulation Council (Sw. Regelrådet), a specific decision-making body within the Swedish Agency for Economic and Regional Growth (Sw. Tillväxtverket), a national Swedish authority, that CRAR would “as far as can be foreseen have limited effects on undertakings.” (2) Similar statements by the Ministry of Finance in Finland in 2010 and 2011 (Finnish Governmental Proposal (RP 42/2010 rd) and Finnish Governmental Communication (U 76/2011 rd)) to the effect that there were no credit rating agencies in Finland that would be subject to CRAR. (3) Economic Commentaries No. 7, 2014 on “ The development of the Swedish market for corporate bonds ” published by the Sveriges Riksbank (the Swedish Central Bank) in which the categorisation of official ratings, internal ratings, and credit assessments by banks is used. The commentary describes the role of the banks in the Swedish debt market, referring specifically to the term “investment grade” as referring to “bonds with a credit rating higher than BB + (S&P and Fitch), Ba1 (Moody’s) or with an equivalent credit assessment from the banks”. There is no suggestion that the banks’ practices contravened regulations in any way. The commentary refers to Handelsbanken, Nordea, SEB, Swedbank and Danske Bank as the “main representatives of the issuers on the Swedish market” (these being the banks subject to the regulatory sanctions in the instant cases). (4) A note issued by the Swedish Securities Dealers Association, Corporate Bond Market: the reference to rating scale in investment research under supervisory scrutiny , 12 August 2016, drawing attention to the view indicated by ESMA that shadow rating could be considered as credit rating and as such reserved for registered credit rating agencies, and saying that in, “Article 3(2) of CRAR there are specific exemptions for investment recommendations, investment research and/or opinions. Therefore Nordic market participants have continued to reference to rating scales in investment research after the regulation came into force”. Although (as was pointed out on behalf of ESMA) this was published after the investigation began, it nevertheless has value as a source in the Board of Appeal’s view. 304. When the Credit Rating Agencies Regulation came into force, it appears that the banks performed an initial appraisal of the effect, albeit undocumented. They proceeded on the basis that their activities fell within the recommendations/investment research exclusion in CRAR. It has not been disputed in these appeals that this represented the consensus in the Nordic market. This is not therefore a case in which an individual institution has taken a particular view of a regulatory provision, but a case in which a particular view prevailed in a market. It is evident that the authorities, including the financial regulator and the central bank in Sweden where each of the appellant banks is established, did not appreciate that CRAR impacted on the banks’ practice either. 305. It is true, as has been submitted on behalf of ESMA, that this is not conclusive since it is up to individual institutions to determine for themselves whether or not they comply with applicable regulations. Nevertheless, in the view of the Board of Appeal, it is a significant point on the issue of negligence. In particular, it contradicts the contention that the appellants should have foreseen that their credit ratings activities were likely to be subject to CRAR. 306. Although it has been argued in these appeals that the effect of the exclusion is clear and that the banks fell outside it, the Board of Appeal for reasons set out above, does not agree as to the asserted clarity. The Board of Appeal, like the Board of Supervisors, has decided (and is firmly of the view) that the banks’ argument is nevertheless wrong. The banks’ interpretation would, as explained, have seriously undermined the operation of CRAR. However, the case of Denkavit (supra) supports the view that in context of negligence, one of the factors that may be taken into consideration in this regard is the clarity and precision of the rule breached. In the present case, whilst in no way negating the doctrine of legal certainty, this is a factor against a finding of negligence. 307. This conclusion dose not contradict the submission made on behalf of ESMA that the mere fact that a person’s understanding of a regulatory provision is reasonable does not imply an absence of negligence in compliance with the regulation – the Board of Appeal agrees with the submission. However, the present case is very different. 308. It is evident from the terms of the Board of Supervisors’ decisions that it itself regarded matters that have been raised as central points in the banks’ arguments as unsettled, declining to express a concluded view on them. Contrary to the banks’ submissions, this gives rise to no question of breach of due process. The Board of Supervisors’ handling of the matter was right in the Board of Appeal’s view, and its conclusion on the main point was correct. 309. However, the fact remains that the Board of Supervisors had doubts which it rightly expressed, and that is a factor which in itself tends to show that the banks were not themselves negligent in the view they took. Any doubts are now resolved, subject to any further ruling by the Court of Justice which is the ultimate arbiter of all questions of EU law. 310. Finally, once it became clear that ESMA was challenging their practice, the banks desisted. They were right to do so, and not to continue the practice pending a final decision by the Board of Supervisors, which in the event was unfavourable to them. 311. It is plain for all the above reasons that this is a very unusual case. In the circumstances, the Board of Appeal respectfully differs from the Board of Supervisors, preferring the view of the IIO that there was no negligence on the part of the banks in this case. They will, however, be unable to resume the practice following the ruling on the correct interpretation of the exclusion without fulfilling the conditions in CRAR. (4) The issues arising on the sanctions provisions in CRAR 312. Though the appellants have suggested that there was in the circumstances no need for a public notice on the basis that they had ceased their activities already, this is not correct. By Article 24(1) CRAR, where “ESMA's Board of Supervisors finds that a credit rating agency has committed one of the infringements listed in Annex III, it shall take one or more of the following decisions: … (e) issue public notices”. In view of the infringements by the banks of Annex III of CRAR by issuing credit ratings without being authorised by ESMA to do so, the Board of Supervisors was obliged to take a supervisory measure, and since the activities had ceased, was fully entitled to take the view that a public notice was appropriate. 313. Since the infringements were not committed by the appellants intentionally or negligently, by Article 36a CRAR the Board of Supervisors’ could not adopt a decision imposing a fine. 314. It follows that the questions as to whether the fines imposed by the Board of Supervisors were just and proportionate and other issues on the calculations of the fines no longer arise, and the Board of Appeal will not comment on the parties’ contentions in that regard. XV Summary of Board of Appeal’s decisions and relief granted 315. The Board of Appeal has rejected the contention of the appellant banks that their practice of including “shadow ratings” in their credit research reports and similar reports fell within the exclusion in CRAR which covers investment recommendations and investment research. 316. The Board of Appeal has rejected the contention of the appellant banks that the Board of Supervisors’ decisions to this effect are vitiated under the doctrines of legal certainty and lack of due process. 317. The Board of Appeal agrees, therefore, with the Board of Supervisors on the central point arising on these appeals, namely that the activities of the appellant banks fell within the provisions CRAR with the consequence that to carry them on the banks required to be registered under CRAR. 318. Though the appellant banks voluntarily desisted during the course of ESMA’s investigation, the Board of Supervisors was right to find that infringements of CRAR had taken place by reason of the non-registration. 319. In those circumstances, the Board of Supervisors was obliged to take supervisory measures, and since the activities had ceased, was fully entitled to take the view that public notices were appropriate. 320. The Board of Supervisors correctly analysed the legal requirements for the establishment of negligence, which is a precondition for imposing a fine. 321. The Board of Appeal, however, respectfully differs from the Board of Supervisors in its assessment (reflected in the public notices) of whether the appellant banks acted negligently. 322. In the very unusual circumstances of this case, in which the banks’ practice was one which had been carried on in the Nordic debt markets for many years, it not being appreciated that CRAR impacted on this practice, and applying the high standard care required of banks, the Board of Appeal has concluded that the infringements by the banks were not committed negligently. 323. That being so, and it not being suggested that the infringements were committed intentionally, the Board of Supervisors’ could not adopt decisions imposing a fine. 324. Accordingly, under Article 60(5) of the ESMA Regulation, the Board of Appeal remits the cases to the Board of Supervisors to adopt amended decisions regarding the cases. The Board of Appeal does not propose to specify or give detailed instructions as to the amendment/s, and leaves it to the Board of Supervisors to adopt such decisions based on the findings of the Board of Appeal. Costs 325. All parties confirmed at the hearing of the oral representations that they wished to apply for legal costs. The Board of Appeal’s view in the light of its decisions is that there should be no order as to costs. Although the banks were unsuccessful in challenging the decisions of the Board of Supervisors as to the scope of the investment recommendations and investment research exclusion, they were successful in relation to the findings of negligence. XVI Decisions BoA D 2019 01: Svenska Handelsbanken AB 1. For the reasons given above, and pursuant to the provisions of Article 60(5) of Regulation (EU) No 1095/2010, the Board of Appeal unanimously decides as follows. 2. The decision taken by the Board of Supervisors (ESMA 41-137-1147 of 11 July 2018) is not confirmed. The case is remitted to the Board of Supervisors to adopt an amended decision based on the findings of the Board of Appeal. 3. The Secretariat is instructed to forthwith send a certified copy of this Decision to the parties, informing them of the right of appeal under Article 61 of the ESMA Regulation, and to file the original in the Secretariat’s records. 4. The original of this decision is signed by the Members of the Board of Appeal in electronic format, as authorised by Article 22.2 of the Rules of Procedure, and countersigned by hand by the Secretariat. BoA D 2019 02: Skandinaviska Enskilda Banken AB 1. For the reasons given above, and pursuant to the provisions of Article 60(5) of Regulation (EU) No 1095/2010, the Board of Appeal unanimously decides as follows. 2. The decision taken by the Board of Supervisors (ESMA 41-137-1153 of 11 July 2018) is not confirmed. The case is remitted to the Board of Supervisors to adopt an amended decision based on the findings of the Board of Appeal. 3. The Secretariat is instructed to forthwith send a certified copy of this Decision to the parties, informing them of the right of appeal under Article 61 of the ESMA Regulation, and to file the original in the Secretariat’s records. 4. The original of this decision is signed by the Members of the Board of Appeal in electronic format, as authorised by Article 22.2 of the Rules of Procedure, and countersigned by hand by the Secretariat. BoA D 2019 03: Swedbank AB 1. For the reasons given above, and pursuant to the provisions of Article 60(5) of Regulation (EU) No 1095/2010, the Board of Appeal unanimously decides as follows. 2. The decision taken by the Board of Supervisors (ESMA 41-137-1152 of 11 July 2018) is not confirmed. The case is remitted to the Board of Supervisors to adopt an amended decision based on the findings of the Board of Appeal. 3. The Secretariat is instructed to forthwith send a certified copy of this Decision to the parties, informing them of the right of appeal under Article 61 of the ESMA Regulation, and to file the original in the Secretariat’s records. 4. The original of this decision is signed by the Members of the Board of Appeal in electronic format, as authorised by Article 22.2 of the Rules of Procedure, and countersigned by hand by the Secretariat. BoA D 2019 04: Nordea Bank AB 1. For the reasons given above, and pursuant to the provisions of Article 60(5) of Regulation (EU) No 1095/2010, the Board of Appeal unanimously decides as follows. 2. The decision taken by the Board of Supervisors (ESMA 41-137-1149 of 11 July 2018) is not confirmed. The case is remitted to the Board of Supervisors to adopt an amended decision based on the findings of the Board of Appeal. 3. The Secretariat is instructed to forthwith send a certified copy of this Decision to the parties, informing them of the right of appeal under Article 61 of the ESMA Regulation, and to file the original in the Secretariat’s records. 4. The original of this decision is signed by the Members of the Board of Appeal in electronic format, as authorised by Article 22.2 of the Rules of Procedure, and countersigned by hand by the Secretariat. By order of the Board of Appeal, this decision is published with clerical mistakes corrected pursuant to Article 23 of the Board of Appeal’s Rules of Procedure. William Blair (President) (SIGNED) Juan Fernandez Armesto (Vice-President) (SIGNED) Marco Lamandini (SIGNED) Beata Maria Mrozowska (SIGNED) Katalin Mero (SIGNED) Michele Siri (SIGNED) On behalf of the Secretariat Kai Kosik (SIGNED) A signed copy of the decision is held by the Secretariat BAILII: Copyright Policy | Disclaimers | Privacy Policy | Feedback | Donate to BAILII URL: https://www.bailii.org/eu/cases/EUBOA/2019/BoA_2019_01.html