Responses by the Ministry of Finance of the Slovak Republic on the Public consultation on Credit Rating Agencies January 2011 Introduction The Slovak Republic in general welcomes and supports initiatives which aim to ameliorate regulation of financial market participants, including of credit rating agencies, and make it more effective. Below given remarks are build on our knowledge of present regulatory regime, our knowledge of the business processes, especially in the banking industry and our experience with regulation and supervision besides the consultation document itself. It is important to also add that the remarks should be read in the context of our perception of fundamental role assigned to the financial markets regulation namely protection of investors covered by any guarantee scheme and preserving sustainable stability of a financial system. General approach For any area of regulation of financial institutions and related businesses there are a few policy decisions that might be done before drafting any specific regulatory text. As soon as the overall policy decision is made, many specific issues might be resolved, decided and regulated just by application of the overall policy points to particular cases. In the case of credit rating agencies there the basic issues to be decided can be phrased as follows: 1. Who should/should not/can issue external ratings? 2. Who should/should not/can use external ratings? 3. Who is responsible for what? Current status The first point of rating issuance is already mostly satisfactorily solved by having in place the Regulation (EC) No 1060/2009 on credit rating agencies (CRA Regulation). This EC Regulation creates very different situation to that when the most of valid financial sector regulation using external assessments was approved (CRD, MiFiD, etc.). For the second point (usage of ratings) the status quo is largely given by a huge amount of financial sector regulation that includes many explicit and implicit references to existence of credit rating agencies and ratings issued by them. Furthermore, external ratings play important role in the internal processes of all market participants (not limited to regulated firms). The degree of reliance on external ratings is generally perceived as too extensive. For the third point it can be shortly said that all interested parties are responsible for their actions, however in detail their responsibilities differ significantly. Regulated institutions should comply with 1
regulation and are legally liable to their customers. As the EC paper sums 1 CRAs are currently largely not legally liable to investors in EU with exception of one Member State. Finally, the liability of regulators acting in good faith towards market participants is a very sensitive issue. There is a feeling that there should be some accountability also for regulators but to translate this liability into legal wording is extremely sensible political though fundamental issue. While the extent of external ratings use is perceived as too extensive by EC and some other interested parties, the fact that rating agencies started to be strictly regulated in the EU should mean inter alia diminished risk of hugely misplaced ratings. It does not matter whether we talk about degree of individual rating mistakes or number of wrong ratings. Moreover, taking the current status of regulatory framework as the starting point, it should be possible to change it over time in gradual steps, to one conforming to the following principles: A. Any market participant willing to assess the risk internally instead of relying on external rating should be allowed in an appropriate alternative to do so by regulation. B. The regulation should reward prudence and willingness to assess the risk, in other words the market participants should be appropriately motivated with relevant incentives to do own risk assessments whenever it is reasonable. The question of responsibility and in turn possible liability is more sensitive. The effort is to set up (or unify within EU) liability framework of a rating agency to investors in a case of losses caused by use of rating assigned in breach of CRA Regulation. While this try might be succeed, undoubtedly the question of analogous responsibility and in turn possible liability of regulators and supervisors will be raised in a case of probably wrong decision/requirement/prohibition by regulators or supervisors. The most important danger in wrongly set liability regulation would be that it kills the rating issuance market segment. 1. Overreliance on External Credit Ratings We perceive the matter of overreliance on external credit ratings to be an important issue which merits further discussion. From our point of view, however, some inconsistency can be perceived between the regulation of credit ratings currently in place and regulation that is being considered. On one hand, there is a general trend in the existing regulatory framework concerning credit rating agencies (CRAs) in response to the crisis to strengthen and tighten regulatory rules relating to them with the aim to enhance the credibility and creditworthiness of CRAs and their ratings on the financial market. On the other hand, there is, however, a countertrend aiming to eliminate references to and use of ratings in the EU legislation and thus de facto in the financial sector itself. We therefore consider crucial to strive for an appropriate balance between these two contradictory principles and approaches, as they are not following the same objectives. Main problems with many small CRAs as well as with national CRAs in our global trade world is the necessity of unification of procedures used by agencies, and simultaneously also necessity for investors to learn about agencies credibility and quality of theirs results. Credit rating agencies are extremely important actors on financial markets, and that is why they should be subject to an effective and robust regulation in the first place. 1 See remark 71 on page 24 of Consultation paper. 2
When assessing the framework for future regulatory changes to the current regime for CRAs, the principal aim to be followed should therefore be to establish a framework which would in the most efficient way enhance and strengthen quality of issued ratings and ensure that their issuance is credible and transparent, as well as increase role of regulators. Assessing framework which would aim to avoid the use of CRAs in practice should thus not be treated in isolation, but in the context of the above recommendation. As far as the use of own methods for (internal) credit risk assessment is concerned, there are 2x2 basic case with regards to ability and willingness to use internal estimates, together with the way how the situation can be treated in line with general principles: Institution...... can do reasonable internal credit risk assessment... can NOT do reasonable internal credit risk assessment (usually due to lack of enough data, either in terms of history length or count of observations)... is willing to do internal credit risk assessment no-problem: institution should be allowed to do internal credit risk assessment with reasonable safeguards institution should be allowed to use even internal credit risk assessments with appropriate floor, thereby retaining some minimal level or regulatory prudence with increasing amount of data available that will become no-problem situation above... is NOT willing to do internal credit risk assessment institution should be allowed to use external credit risk assessments with appropriate penalty, thereby motivation to move forward and find the will to use internal assessments no-problem: institution should be allowed to use external credit risk assessments with appropriate penalty, thereby motivation to move forward and find a way how to make internal assessments (in the first place) The main principle reflected in the above table is that if an entity is able but is not willing to move in wished direction its capacity of exposures creation shall be diminishing. One of many possibilities are e.g. capital adds-on as we perceive requirement on regulatory capital as requirements on limiting (credit) risk exposure. In the banking industry, capital requirements can be based on both external and internal credit risk assessments. Large exposure regime however relies only on the external rating with explanation that internal credit risk estimates are mutually incomparable between any two banks. However IRB banks have their internal ratings and PDs computed using models approved by competent authorities and as such they can be used as a reliable basis for setting up large exposures limits, e.g. by slotting PDs to regulatory bands. This approach would eliminate one of possibly unnecessary uses of external credit assessments. External ratings are implicitly not an objective measure of risk. But the argument that internal models are not an objective measure of risk 2 is wrong. Well designed and developed internal models are quite objective measures of risk, often with known degree of uncertainty (confidence) and with very limited subjectivity. 2 See paragraph (1) in 1.1. of consultative paper, p. 8 3
Given the principles above, idea of imposing any specific limits based on other variables than external ratings by regulators is wrong. It would just mean the change of one specific variable usable in the internal assessment by other one, with apparently no advantage (except doubtful leaving of already regulated CRAs). Flexibility in decision powers of the institutions should be retained and promoted to degree suitable to fulfil principles A and B mentioned above. For internal purposes, the use of internal credit risk assessment 3 should be appropriately encouraged. Specifically regarding question 11 on ability of financial institutions to produce reliable sovereign debt ratings for internal purposes: External credit rating agencies do have advanced knowledge compared to average even institutional market participant and thus external ratings do have added value compared to even the best internal credit risk assessments of sovereigns. 1.1. Reference to external ratings in regulatory capital frameworks for credit institutions, investment firms, insurance and reinsurance undertakings We in general welcome trends towards developing and using internal ratings and internal models for the calculation of capital requirements. There are sectors where this may be reasonable (for example banking, pensions fund sector). Alternatives to external ratings may include investors own assessment, strengthening the role of auditors, better quality of financial statement, information provided by issuers, etc. However, while responsibility of bankers for their own internal ratings should be enhanced (including responsibility of regulators to monitor internal risk monitoring systems of banks), it should be kept in mind that internal ratings are not a panacea and replacing external ratings completely by internal assessments is very low-probable. External ratings ensure comparability allowing for transparency for investors and their role in irreplaceable. We therefore also share the view that completely eliminating any reference to external ratings in capital requirement frameworks does not seem to be a realistic solution, as long as there are no other alternative measures of credit risk which could be used instead by all financial firms. Yet, the system of external ratings has been failing on more occasions, which was in particular revealed by the recent financial crisis. The efforts to improve it, with the aim to enhance the quality and transparency of issuing ratings and to increase role of regulators, must be stepped-up. For example, as far as banking is concerned, there are certain areas where internal credit risk assessments are desirable and well placed. This is in particular in cases when the bank holds more appropriate and reliable information on particular assets than CRAs. In this context, internal assessments should be strengthened and promoted, and using market based risk indicators might be promoted in this context. However, for some special products, in particular various investment and securitized products, internal assessment might not be appropriate, as it would not capture the whole reality relating to complex financial instruments, and external ratings thus seem to be a better solution. When assessing to what extent internal credit risk assessment is needed, the following should be furthermore taken into account: There is a need for a sectoral approach, as it is not possible to implement one solution to different sectors of financial market. 3 In fact any other significant risk can be substituted here. 4
Costs of internal assessments, and the possible problem of the lack of skilled personnel to undertake these assessments, should be taken into account, in particular in case of smalland medium-sized financial firms. One of the proposals to use at least two external ratings in order to be able to compare them would also substantially increase the costs of rating. 1.2. Use of external ratings for internal risk management purposes We in general welcome proposals on how to decrease use of external ratings for internal risk management purposes. We share the view that financial firms should not mechanically rely on ratings, but that they should be obliged to carry out their own due diligence, as it is the primordial purpose of risk management they are required to have in place. We in particular welcome imposing disclosure requirements to issuers of structured finance instruments, according to which credit rating agencies would have the possibility to access the information necessary for issuing unsolicited ratings of these instruments. We believe that this requirement would substantially increase protection of investors, transparency and competition on the rating market, as it would increase the number of ratings per instrument so that users of ratings would be able to rely on more than one rating for the same instrument. It would also mitigate risks and conflicts of interest relating to issuer-pays principle. 1.3. Use of external ratings in the mandates and investment policies of investment managers As already highlighted above, we in general welcome reducing the use of external ratings in the mandates and investment policies of investment managers. We also welcome implementing obligation to introduce measures on investment managers to ensure that the proportion of portfolios that is solely reliant on external ratings is limited. As regards flexibility clause, its implementation in investment mandates and policies may prevent the cliff effects and volatility on the financial markets. However, we would also like to highlight that introduction of flexibility clause might be accompanied with possible risks. One of these may be that it would allow deviating from original characteristics and profile of the funds, in which the investor decided to invest in, and would thus expose him to various risks without his prior notice. It is therefore important that time limits for deviation from external rating thresholds are flexible and do not trigger fire sales. 2. Sovereign Debt Ratings We believe that sovereign debt ratings provide useful information to investors and thus should be maintained. Imposing more transparency and disclosure regarding sovereign debts, if effectively designed, would not be harmful and we therefore generally support it. We also take into account that there are important specificities of sovereign debt ratings. However, we believe that it is important to maintain level playing field between private and sovereign debt sector. In this context, we do not consider concrete recommendations proposed in the documents to be the right approach, as their potential negative consequences might outweigh their positive contribution to the stability and well-functionning of the credit rating markets. 5
We do not support to extend periods for informing the country to which CRAs are in the process of issuing the rating from 12 hours to 3 working days. Prolonging the time period, although meant to provide sufficient space for correction of any possible factual errors in ratings, would unduly widen the window for any of the imaginable abuses of internal information and might in extreme cases open the door for a lobbying or interventions in the process of assessment and evaluation. Three working days represent a substantially long time period during which disclosing information on ratings might cause important damages to financial markets or might encourage speculative transactions relating to the rating announced. Furthermore, rules of market competition would encourage the CRAs to issue as credible ratings as possible which eliminates the risk of possible factual errors. In case these would still occur, we believe that the 12-hour period for their correction, taking into account a number of possible negative consequences longer period might imply, is sufficient and appropriate. Regarding sovereign debt ratings we therefore believe that maintaining level playing field between private and sovereign debt sector is important. The focus should be placed on developing a framework which would ensure to the most possible extent the quality and credibility of ratings. One way to achieve this is for example implementing civil liability of credit ratings for breaking rules, which would also eliminate risks of possible errors in the ratings. As far as proposals to ensure transparency relating to sovereign debt ratings are concerned, we generally support them. However, they should be appropriately designed so that possible negative consequences of additional transparency requirements are avoided. For example, requirement relating to CRAs to disclose free of charge their full research reports on sovereign debt ratings might to some extent hinder their competition, as this would de facto means disclosing their knowhow on which their ratings are based. CRA that does not genuinely wish market participants to share its methodology in detail can formally fulfil all such requirements yet successfully obscure the important methodology details. Then only the market forces, namely equivalent of Basel II market discipline, can help to distinguish between open and effectively obscuring CRA. Rolling averages of bond prices can be used at least as warning signal complementary to rating; CDS spreads can also have some warning role in evaluation of credit risk. The idea to use ECB as a rating provider should be considered carefully as ratings from ECB may become one of leading indicators at least for sovereign and financial sector and the processing will require cooperation with National central banks. Current review periodicity of (sovereign) ratings requested by CRA Regulation is one year. While the internal repayment ability of a country would not change very quickly, the relevant environment can change. Therefore we do not recommend shortening the review period but rather rating agencies should set up a set of triggers for rating review even inside the legal period. The postponement of sovereign rating publication after close of European markets is irrelevant in today's globalized world. The business runs effectively 24 hours every day via organized markets all over the world or other channels. Online dealers are just one example. Finally, commitment of EU Member States not to pay for ratings is useless; the conflict of interest in the Issuer-Pays model should be resolved systemically. Here, the CRA Regulation again seems to be the key as mentioned in the last sub-section of the document. 6
3. Enhancing Competition in the Credit Rating Industry We in general support enhancing and promoting competition in the credit rating industry. However, we do not consider necessary and appropriate that a public sector should be incorporated in the structure of CRAs. We believe that credit rating industry is not an area of particular social or national interest, where public intervention is desirable, and it should therefore be market-driven and subject to rules of competition. We therefore see a limited space for state intervention without threatening independence of CRAs. However, State could find ways to promote competition through indirect economic incentives (for example tax holidays, etc). Among the suggestions of this section two stand out worth of commenting. Imposing the rating assignment process to relevant central banks seems to be an interesting idea meriting further elaboration. The preferred format of this task includes the following steps: Task 1. Publicly available data gathering 2. Data processing 3. Publication of basic rating report 4. Additional data gathering in cooperation with rated entity 5. Data processing 6. Publication of advanced rating report Paid by Central bank internal resources Customer requesting advanced rating procedure It should be however born in mind that having central banks as rating agencies creates potential conflict of interests. On one hand many central banks of EU Member States are expected to supervise or will be expected to help with supervision of rating agencies by ESMA in the near future according to expected update of CRA Regulation. On the other hand some of these central banks would produce ratings and therefore would be competing with some supervised agencies. In our view it is not necessary for any official structures such as EC on European level, some ministries on national level to step into creation of European rating agency in any form. Rather it would be more in the role of a government body to set up appropriate legal and effective framework so that public information is easily accessible by any interesting person. 4. Civil Liability of Credit Rating Agencies We perceive the proposal to implement civil liability of CRAs to be a sensitive issue, going beyond the framework of regulation on CRAs, as existing national regimes on civil liability are not harmonized. However, we believe that implementation of civil liability of CRAs in national regimes may be a very effective measure how to ensure viable and credible credit rating industry. It should also be carefully and duly assessed what concretely should be subject to possible civil liability. For example, we do not support implementation of civil liability of CRAs for unsolicited ratings, as this could not only exacerbate moral hazard problems, but also lead to elimination of issuance of unsolicited ratings. 7
5. Potential Conflicts of Interest due to the Issuer-Pays Model We support initiatives which aim to reduce potential conflicts of interest due to the issuer-pays principle which is a predominantly used model of remuneration in the credit rating industry. We believe that further analysis is needed with respect to possible pros and cons relating to various alternative solutions to the issuer-pays principle. We do agree that there could be a distorting influence of a fee-paying issuer over the determination of a credit rating. In order to reduce conflict of interest due to the issuer-pays model, option 5.1. Subscriber/Investor-Pays model seems to be appropriate, while (c.), where group of investors jointly hire rating agency, appears to be the best of the proposed solutions. The hired agency will be involved in investors decision through investor s ability to profit and contribute to fee. 8