Subject: ICAP Group s Response to European Commission s Public Consultation on Credit Rating Agencies

Transcription

1 2, El. Venizelou Ave, Kallithea, Greece Tel.: Fax: European Commission Directorate General Internal Market and Services Financial Services Policy and Financial Markets Securities Markets Athens, 5th January 2011 Subject: ICAP Group s Response to European Commission s Public Consultation on Credit Rating Agencies Dear Sirs, ICAP Group welcomes the opportunity to comment on European Commission s Public Consultation on Credit Rating Agencies (CRAs) regarding the strengthening of the EU regulatory framework for CRAs. In the following pages you can find ICAP s responses on the questions posed on the Consultation Paper. ICAP issues unsolicited credit ratings in compliance with the requirements of the EU Regulation 1060/2009. ICAP business model is based on a subscription basis (the investor requests from ICAP the assignment of a credit rating to an issuer). The issuers rated are Greek corporates as well as small and medium enterprises. The overwhelming majority of corporate debt in Greece has the form of banking loans and as a result, the credit ratings assigned to the obligors are primarily used by bank for decision making, portfolio credit risk measurement and regulatory capital requirements calculation. Moreover, a significant number of Greek corporates use ICAP Credit Risk Services in order to manage their credit risk generated by their business activities. In the light of the above, ICAP Group recognizes the importance of credit ratings in financial stability and has aligned all its procedures with the requirements of the External Credit Assessment Institutions (ECAI) introduced by the Directive 2006/48/EC and the requirements of Eurosystem Credit Assessment Framework (ECAF) introduced by the European Central Bank. ICAP Group supports European Commission s initiatives toward a comprehensive, efficient and viable regulatory framework that will ensure the long term stability of the financial markets. We are at your disposal for any further clarifications. Yours Sincerely, Leonidas Kotsaftis Director Credit Risk Services Division Panagiotis Avramidis Senior Manager Credit Risk Services Division

2 1. Overreliance on External Credit Ratings (Q1) Reliance over external ratings is partly attributed to lack of sufficient resources for implementing internal risk processes, an observation applicable mainly to smaller and less sophisticated financial institutions but also, in some cases, to medium and large ones. Nevertheless, enforcement of external ratings on a particular segment of the financial sector may have some serious side effects, primarily on competition. Financial firms eligible to use the standardized approach will be handed with the advantage of lower operational costs and they will be further put off from establishing internal risk processes and perform their own due diligence in the near future. In addition, the decline in demand for external ratings by larger financial institutions would shrinkage the presence of CRAs, especially those currently operating in the investors/subscriber pays business model, a development with negative effect on competition in the sector. (Q2) Internal ratings are assessments prepared by banks exclusively for obligors in their portfolios. This poses a restriction on the internal model s applicability to new market segments in which the bank has no presence. Moreover, internal models developed using solely the bank s portfolio may show biasness towards good-payer customers due to cherry picking applied in customer selection process. Furthermore, internal ratings tend to be more procyclical as they rely heavily on models processing financial information and market prices. Ratings procyclicality has been recognized as one of the causes that exaggerated the impact of the recent financial turmoil and therefore the proposition of exclusive dependence on internal ratings/models for the assessment of the financial firms portfolio credit risk contradicts the objectives of the new regulatory framework. Instead, using internal assessments along with external credit assessments, which will play the role of an independent external benchmark, could be more effective in promoting the stability of financial markets, maintaining the critical volume of demand that ensures CRA competition and ultimately safeguarding financial institutions long term solvency. The proposal to use both external credit assessments and internal ratings displays two advantages. First, direct comparison of the two independent credit risk assessments may reveal data inefficiencies, misinterpretation of risk factors and other model singularities such as high procyclicality or biasness. Hence, the final assessment should be free of such discrepancies, reducing the likelihood of risk underestimation. Secondly, the regulators will have an alternative benchmark measure of risk in their effort to assess the internal model s adequacy. The regulators will find rather useful to have a benchmark assessment, especially for small and medium businesses or other non-traded issuers characterized by scarcity of available information. (Q3) In principle, having two external ratings should reduce the reliance and improve stability. Nevertheless, external ratings in most of the cases show high correlation and they tend to move in similar directions. This inter-dependence reduces the effectiveness of the two external ratings measure. (Q4) Market data and prices of traded securities are forward looking in nature and compliment the usually backward looking nature of classic risk measures. Nonetheless, using market data to evaluate credit risk is likely to have a negative impact on market stability as procyclical effects will be exaggerated. 2

3 Indeed, a drop in the asset price, i.e. an increase in the yield, will lead to a credit rating downgrade which in turn will result in lower asset price, creating a vicious cycle leading to higher instability. Smoothing techniques can be employed in order to reduce the impact, but it is doubtful whether it will yield stable regulatory capital requirements. In addition, the following two factors should be considered: i. market based risk measures are not available for all issuers / issues and ii. market prices are susceptible to manipulation by market participants. Hence, we see this proposition as a contradiction to the financial stability objective of the new regulation. (Q5) The recent crisis that started from a particular segment of mortgage loan market was diffused across markets and economies mainly through securitization and structured products. The increased number of structured products seen in the last decade is attributed to the objective of higher growth rates which, due to the regulatory requirements, could be achieved either by raising more capital or selling assets. This has transformed the business model of banks from loan providers to loan originators and distributers. The securitization was also propelled by the external credit assessment institutions highly optimistic credit rating assignments that ignored the long term economic fluctuations. A solution that could improve the credit risk inherited by securitized investments is to impose minimum credit quality requirements to all or the majority of securitized underlying exposures and limit the inclusion of high risk obligations to the asset pool. The securitization process has provided the opportunity for creating low risk products from a pool of high risk debt obligations through the process of payment prioritization. Although the idea of tranches improves the inherited risk, especially for senior investors, the correlation among the pooled assets and the high risk asset base could ultimately lead even the lower risk tranches into losses. Securing that individual credit quality remains above a minimum limit and that correlation estimates objectively reflect the risk interdependencies, the securitization products will be less sensitive to economic changes. (Q7) Regulatory advancements in the last two decades have advocated the development of internal risk management processes. Especially, the introduction of foundation and advanced IRB approaches in Basel II Accord capital calculations were an explicit embracement of internal ratings by regulatory authorities. Nonetheless, the recent crisis occurred after Basel II implementation, showing to all participants that one-dimensional risk management process, either internal or external, has serious weaknesses. Instead, tighter cooperation among lenders, ratings agencies and regulators could be an effective way to promote best practice risk management. Financial firms should be explicitly obliged to carry out their own due diligence but this task could be complemented with external risk assessments in an effort to identify discrepancies and implement adequate adjustments. If external ratings are used as an independent variable (input) of IRBs models, the weight assigned to them should be limited and the internal rating process should be mainly based on internal information regarding the issuer. 3

4 (Q8) There are two categories of information that should be disclosed to supervisors in order to enable them to monitor internal risk management processes: information on the rating model and information on the rating implementation according to credit policy. The first set of information can be used by supervisors in order to evaluate: i. the contribution of external credit ratings to the outcome of the internal process, ii. the internal rating process adequacy and the underlying assumptions of key parameters (including external rating) and economic conditions and iii. the sensitivity of risk components (PD, Internal Rating) on the changes of the external credit rating. The second set of information should aim to reveal material policy overrides that ultimately make internal rating processes redundant. This is an area that goes beyond the purpose of the external rating impact assessment but it is regarded as a shortcoming that requires further improvement as exhibited by the recent failures of credit policy discipline and inefficient monitoring processes of policy implementation. (Q9) Recent financial turmoil showed that the Boards of Directors or other governing bodies of financial firms were not involved to the required extent in the review of the use of credit ratings in risk management processes and investment mandates. The establishment of stateof-the-art risk management processes does not guarantee a balanced portfolio, if policies and business decisions ignore these risk assessments. Top management that consistently ignores the risk warnings in pursuit of higher short term growth is putting the organization s future at stake. It is understood that restricting risk usually has a short term negative effect on return growth, a consequence hardly acceptable by the Boards of the financial firm. The issue mainly arises because defaults happen so rarely due to long periods of economic and financial market stability that it is hard to convince the Board to consider such an event. Hence, an effective application of risk management processes requires the full commitment and higher involvement of management in order to bridge the gap between the risk management division and the growth-oriented business management. 4

5 3. Enhancing Competition in the Credit Rating Industry (Q23) Competition in the CRAs sector could be strengthened by encouraging either new players entering the sector or existing agencies expanding their rating processes in other market segments. According to the records recently published by CESR, 23 CRAs filed for registration under the new Regulation 1060/2009. Despite the considerably large number, global competition across market segments, including sovereign and structured products credit ratings still is rather limited, since most of these CRAs currently operate on a local base and/or in a rather narrow market segment (e.g. rating exclusively small & medium enterprises and domestic corporates). The reasons that explain the limited competition in the sector are: High start-up costs Reputational issues Regulatory uncertainty Setting up a new CRA or rating division, requires a significant investment in human resources and infrastructure. However, the complexity of the rating business and its significant role in financial markets justify these high costs. It can also be viewed as a fence that deters the entrance of speculators that could severely impair the market. Reputational issue is an additional concern for new players and it refers to the dominance of established global CRAs with large exposure in media in contrast to smaller CRAs or start ups that are less publicly exposed. The new regulation and the publication of the list of registered CRAs could potentially help to bridge this reputational gap between the global rating agencies and the small and medium CRAs. Regulatory changes create an additional source of uncertainty. Recent regulatory requirements, although necessary, have increased the regulatory overhead (e.g. establishment of new functions, credit rating reporting to central depository CEREP) and consequently the operational cost of CRAs. Moreover, frequent changes in regulation create an additional burden for compliance monitoring and adjustment. In practice, the additional overhead has higher impact on small and medium sized CRAs than the global market players because of price flexibility as well as availability of resources. In the light of the above, the following propositions could reduce the overhead and promote competition within the rating sector: Revisit certain requirements of Regulation 1060/2009 such as the establishment of new functions (the Internal Review Function and the Internal Audit have a number of overlapping responsibilities that do not justify the operation of two independent functions, especially for small and medium sized CRAs). Avoid explicit references to particular global CRAs that may give the impression of endorsement. For instance, the authorities can establish an independent central rating scale where all CRA rating scales will need to be mapped, instead of using the rating scale of particular CRAs as it stands now. 5

6 Maintain a stable regulatory framework for CRAs. Frequent and overlapping recognition processes create an additional overhead that obstructs new entrances. (Q24-26) The proposition that ECB or National Central Banks could provide in house credit ratings entails two flaws: If ECB or the National Central Banks decided to provide ratings, competition in CRAs sector could be severely distorted. No financial institution or investor would choose to use ratings from an alternative rating provider if they could make use of the regulators own assessments since they would view these ratings as having the authorities approval. This situation would ultimately lead to a single dominating player. It is unclear who would be the supervisor of such a function. The new supervisory authority (ESMA) could be playing this role but certain functions of rating agencies are still supervised by ECB (Liquidity funding through Eurosystem ECAF) or by National Central Banks (Standardized Approach of Basel ECAI), creating obvious conflicts of interest. On the other hand, national level players could be an alternative to ECB or National Central Banks. The advantage in this case would be a more effective supervision. However, the establishment of national players is likely to reduce competition for the same reasons as explained above in the cases of ECB and National Central Banks. Moreover, the national segregation raises questions regarding the operation of these players and the impact of national credit assessments on global markets. It should be mentioned that in all three options the assignment of sovereign credit ratings may not be possible due to the apparent existing conflicts of interest. (Q27) A European Credit Rating Agency could promote competition, especially in the fields of credit assessments of sovereigns or structured products, which are dominated by a small number of CRAs, but its effectiveness will depend on the market s view of its independence. Therefore, the creation of a European Credit Rating Agency could in principle be a useful alternative, in the aforementioned areas, if it is free of any direct or implied support from authorities. (Q28) Regarding the entry barriers, a number of proposals that could lower the operational as well as the compliance costs of CRAs have already been provided in (Q23). In addition, already registered CRAs could be encouraged to expand their business either geographically or in new market segments by simplifying the cross border supervisory process through a higher coordination among member states. (Q29) Publicly traded debt of multinational corporations or sovereigns as well as structured products are only a part of the portfolio of obligors that are required by the Basel II Accord to be rated. Particularly, the majority of corporate debt in Europe is in the form of banks loans, in contrast to the bond-driven debt market in the US. Consequently, European Authorities should support small and medium sized CRAs that provide credit ratings for this significant, not traded, debt market. As a niche market, it requires a good knowledge of the local market, in conjunction with an effective and transparent business model. 6

7 As mentioned earlier, there is currently a significant number of European small and medium sized CRAs operating at national level and assigning ratings mainly to corporates. These participants could form a Network with Pan-European geographical coverage with the following objectives: To improve the members existing methodologies and procedures regarding the sector already operating, i.e. corporate ratings. Through the exchange of local views and practices, the Network could increase the quality of the assigned credit ratings. To develop jointly new methodologies and procedures regarding sectors that the CRAs do not have presence, i.e. sovereign and structured finance ratings. The result could be a common framework that will be further adjusted by each CRA, according to its business needs and local market particularities. (Q30) According to our view, the effective business model requires a balance of human intervention and more automated procedures. Nonetheless, the new CRA regulation inclines towards the opposite direction and hence it, not only fails to promote competition and reduce entry barriers, but effectively benefits large and globally established agencies. It is necessary that the CRAs Regulation strikes the right balance among objectivity, accuracy, independence and consistency of credit ratings and the increase of small and medium sized CRAs market share. 7

8 4. Civil Liability of Credit Rating Agencies (Q31) On 15/12/2010, an amendment of CRAs Regulation 1060/2009 was approved by the European Parliament. This amendment deals with the establishment of a common supervisory authority for all CRAs established in EU (ESMA), and the sanities, either pecuniary or not, that should be applied in case that a CRA circumvents, intentionally or negligently, the provisions of the Regulation. As a consequence, a common EU-level law regime has already been introduced and all CRAs are going to be treated equally in case of violations of the Regulation, ensuring a level playing field and consistent application among them. (Q32) After the approval of the amendment of Regulation 1060/2009, sanities related with breaches of the provisions of the Regulation have been set up. CRAs should only be penalized in cases of Regulation breaches and not in cases where financial institutions and investors have wrongly made investment decisions based solely on credit ratings and overlooked the probability of default which still exists for investment grade credit ratings. Credit ratings are merely opinions regarding the creditworthiness of a counterparty or financial product and not investment proposals since the latter requires additional information regarding the investor s portfolio diversification and risk-return profile. Furthermore, investment losses liability which is not related to regulation infringement could lead CRAs to refrain from rating non investment grade companies and issuers, causing a distortion to the market. Consequently, we believe that CRAs should be liable only for Regulation breaches. (Q33) The CRA Regulation and its provisions were established in such a way that both solicited and unsolicited ratings are equally treated. According to the Regulation and its amendment, solicited and unsolicited ratings stand against a common liability regime and face the same penalties and pecuniary sanctions. Hence, there is no reason of segregation for liabilities related to regulation breaches. 8

9 5. Potential Conflicts of Interest due to the Issuer-Pays Model (Q34) Although the CRA Regulation has already taken into consideration a lot of measures in order to minimize conflicts of interest between CRA and issuers, there is still the possibility of rating distortion due to the issuer-pays model. There is an inherent conflict of interest that stems from the fact that the issuer pays for the assigned rating and can choose the rating agent. In general, this combination is flawed as it implies a motivation for revenue growth that opposes the principles of independence and objectiveness. (Q35) The most feasible alternative of the issuer-pays model is the subscriber/investor pays model that could lead to more accurate investment decisions. However, the cost incurred to the investor paying for the unsolicited rating, has so far been a hurdle to the development of this business model. This business model could be further promoted by the following initiatives: The information required for the assignment of a solicited rating should become publicly disclosed. As a result, all interested third-parties (investors, financial institutions, other CRAs) will be able to perform their own diligence and assign their own ratings (unsolicited). Public access to issuer s information will reduce the hurdles (cost for data collection, data unavailability) and it will promote transparency and ratings comparability. The latter will provide an additional incentive to CRAs to provide accurate solicited credit ratings as they will know that the issuer/issue may be unsolicitedly re-assessed by third parties. In addition, the CRAs that provide unsolicited ratings through the subscriber/investor pays model should be independent of the investors as the opposite (section 5.1 cases b and c) would create serious conflicts of interest. Indeed, a CRA dependent on investors could be used to manipulate the market and promote the investments of the controlling investors. 9