BANQUEFINANCE ASSOCIES Conseil en activités financières

Transcription

1

2 Comments on the draft proposals prepared by the European Commission for introducing formal regulation of credit rating agencies (CRA) 1. General observations 1.1. Credit rating agencies role on European capital markets has considerably expanded during the last 15 years, and CRA provide today a critical service for credit risk evaluation. The importance of their role, and also certain shortcomings occurred at the beginning of the decade (Enron, Worldcom and a few others on the US market, Parmalat in Europe ) has already translated into a substantial increase of regulatory oversight and disclosure requirements for rating agencies, formalized in reference to the principles for a code of conduct defined by IOSCO and implemented, on a self-regulatory basis, by CRA under close monitoring of supervisory authorities (SEC in the US, CESR in the EU). While the approach taken of comply or explain may seem relatively benign, the different reviews made on the compliance by CRAs with IOSCO principles has so far exhibited very few breaches. Nevertheless, the current financial crisis has placed CRAs on the limelight again, especially regarding the structured finance and monolines sectors, and the massive losses encountered by investors despite the high level ratings initially granted have accredited the opinion of a failure by CRAs for performing their role of assessing risk accurately. As a consequence, restoring confidence on CRA s role, an essential task for enabling credit markets to return to normal conditions, probably necessitates strong signals at the regulatory level, and to substitute an explicit and formal regulatory approval to a now less credible self-regulatory framework. The initiative of the Commission for implementing a directive in this regard appears therefore justified The objectives of this new framework can be manifold, and need to be clearly articulated. The first and most important- objective is to help providing credit markets with more accurate and meaningful ratings, clearly understood and disseminated. The second objective is to ensure that the quality of credit ratings, their relevance and their meaning can be sufficiently homogeneous and comparable, among credit rating agencies themselves, but also for the investing community across various jurisdictions: considering the global nature of capital markets, any material deviation in the regulatory framework between for instance Europe and the United States the two most important financial areas- that would create inconsistency in methodologies and operating processes would be extremely damaging. The Commission should pay particular attention to this imperative: provisions to be implemented in Europe can differ from other jurisdictions (document- Introductory remarks 3), but must remain compatible. The third objective is to maintain an acceptable level playing field for rating

3 agencies, and especially not to reinforce the oligopolistic nature of the rating business by imposing regulatory obligations which only a very limited number of large players of similar sizes, implementing the same methodology with the same operating processes could comply with: in an opinion-based service, it is extremely important for new players to be able to enter the business, provide new perspectives and methodologies, gain in credibility and challenge established players. The understanding of the Commission s proposal is that the new regulation would only concern rating agencies used by market participants for complying with legislative requirements, i.e. provisions referring to credit ratings in European legislation. This approach is valid, but it should be clear that the ability for entities outside the present regulatory scope will be maintained, in accordance of course with European financial legislation (i.e. MIFID or Transparency directive for instance). Nevertheless, an excessively prescriptive stance by supervisors on methodologies and technical requirements for producing ratings would have a negative impact in that sense, namely imposing a correct and unique perspective for producing those ratings, reducing innovation and making very difficult for new players to establish themselves over time. An excessive focus from regulators towards prescribing methodological guidelines would also be dangerous, making them de facto validate the accuracy of the ratings conducted in respect to such prescriptions, a task that supervisors are neither legitimate nor technically capable to perform As a provisional conclusion, a regulatory framework should focus on disclosure of CRA s methodologies and ratings and prevention of conflicts of interest in respect to IOSCO s stated principles, as also strengthening investors education for understanding credit ratings better. It should aim to supporting innovation in methodologies and plurality of comparable opinions across rating institutions and internationally. It ultimately should enhance the quality and accuracy of credit ratings. For achieving this goal, it is crucial for supervisors NOT to interfere, directly or indirectly, in the analytical process. However, regulators may be able to assess the accuracy of ratings as a critical element for granting or maintaining authorizations to operate, relying on the performance statistics (defaults and transition matrixes) that CRA would be requested to publish on a regular basis (cf. p of the document). A substantial deterioration of ratings performance may be for supervisors a warning signal questioning the relevance of the accreditation of the relevant CRA The last important point is of course the designation of the relevant authority. The complexity of the European regulatory framework, notwithstanding the merits of the Lamfallussy process, may make the CRA supervisory framework cumbersome, costly and hardly effective. Separating this process between an Authority in charge of granting authorizations and local capital market regulators in charge of enforcing compliance is likely to be a very complex and ineffective approach. In a perfect world, establishing a unique supervisor in charge of the whole value chain of CRA regulation would have the big advantage of clarity and simplicity. This task could be performed by CESR, working in an extended capacity and with ad hoc resources, closely coordinating with domestic regulators. The second best solution would be to designate a local supervisor, acting on behalf of the other member states, in charge of covering all aspects of a

4 specific CRA regulation. This regulator could be designated according to the criteria described by the Commission in its proposal (cf. p). Such a solution could then be, all things being equal, comparable to the process implemented for ECAI recognition, which has proven effective. 2. Factual comments in the document 2.1. I-Introductory remarks, p.1 end of 3 rd paragraph: when different, provisions introduced in Europe should remain compatible with those in other jurisdictions 2.2. I-Introductory remarks p st paragraph; p. 11 Draft Directive, Title I Article 1-Proposal for a directive (4) p.7: it must be clear that ratings opinions can continue to be freely produced and disseminated by not eligible for complying with European legislation requirements on ratings 2.3. II- Options related to the authorisation process and the supervisory architecture pp. 3-5: establishing a single authority (CESR?) both for authorization and supervision would be the most effective framework. Otherwise, option 1 (devolution to a unique national supervisor with CESR support) would be preferable to separating authorization and supervision between different bodies Proposal p. 8 (6) and Directive p. 16 art. 9 item 5: the interactive nature of structured finance ratings (various iterations between the arrangers and the agency) will make the distinction between recommendations regarding the design of a structured finance instrument (prohibited) and a proper rating opinion (authorized) very difficult to assess Proposal p.8 (9): regulation should refrain setting up stringent metrics regarding resources allocated 2.6. Proposal p.8 (10): distinction should be made between analysts directly in charge of preparing the rating reports, where standards for rotation are useful, from the members of the rating committees. It is not realistic to ask a rotation in rating committee membership, with the risk rating committees losing rapidly in expertise and relevance Proposal p. 9 (14): while convergence in standards is welcome, regulators should not be in the position to impose it, which would force them to define standards and enter the analytical field, which may differ across rating agencies for valid reasons (for instance default definition or probability of default vs expected loss) 2.8. Proposal p. 9 (16): what is the reason for a 4 year threshold? 2.9. Directive p. 14 Chapter II, article 7: the rationale for prohibiting non executive members of the supervisory/administrative board to be renewed is not obvious, since they may bring continuity and experience in their supervisory role Directive p. 15 art.9- item 3 a): the 5% level is probably excessive and may prove inapplicable; it should be managed by an obligation of disclosure. Clarification whether this % should be understood at the consolidated or regional levels (e.g. worldwide/ Europe) is needed cf. 2.6 above Directive p.17 art. 12 item 5: the article should only establish the principle of informing the market on any substantial change in methodologies. Details should be defined by the CRA, according to the importance of the changes Directive art. 17 item 1, second paragraph: considering that the withdrawal of the authorization to operate of an accredited CRA may have lethal implications

5 for its business, the obligation to retain 3 years of records may be difficult to comply with Annex II section B (pp.29-30) item 3 and 5 seem redundant Annex II section B p. 30 item 3: if the CRA considers that the quality of the information provided is not satisfactory, it should refrain from rating the transaction