«Global TrusT and EThics in FinancE Innovative Ideas from the Robin Cosgrove Prize Carol Cosgrove-Sacks / Paul H. Dembinski Editors Trust and Ethics in ...»
246 Trust and Ethics in Finance Regulatory dependence of CRAs The final key issue to consider is the dependence of regulators on rating agencies. This extends the scope that stakeholders with vested interests should consider when addressing CRA-related regulatory issues and reflects the effects of poor ratings on financial stability. A case in point here is the dependence of Basel II’s capital adequacy standards for banks on ratings. This is particularly problematic in a scenario in which credit ratings do not adequately reflect credit risk. Here, “the banks’ capital structure might give the illusory impression that it constitutes a sufficient cushion against risk, which could threaten the safety and soundness of the banking system”.17The same is true for the US National Recognised Statistical Rating Organisation status, whereby a wide range of investors are required not to hold securities whose ratings are below “investment grade”, and ratings affect the risk weightings of banks’ assets in calculating capital adequacy ratios.18 Recent regulatory developments Appendix 1 shows the current regulatory framework for CRAs. As this is beyond the scope of this study, IOSCO, the main regulatory tool, as well as recent regulatory developments in the European Union and in the United States will be briefly highlighted.
IOSCO The International Organisation of Securities Commissions published its Code of Conduct Fundamentals for Credit Rating Agencies in 2004.
This code focuses on voluntary “corporate governance rules designed to (1) ensure quality and integrity of the rating process; (2) remain independent and avoid conflicts of interest; (3) assume their responsibility to market participants through greater methodology transparency and adeAmelie Champsaur, 2005.
Richard Portes, 2008.
Second Line of Defence 247 quate treatment of confidential information provided by issuers”.19 The key shortcoming of these provisions is that IOSCO does not address the issue of the enforcement of the code. Instead, it leaves enforcement to either national regulators or market mechanisms20 that, in light of the current crisis, have been ineffective.
European Union October this year saw the drafting of a law designed to prevent conflicts of interests between CRAs and their clients in the EU The legislation aims to forces CRAs to disclose their working methods and to encourage the emergence of new agencies. Under the draft proposal, CRAs will not be allowed to combine their work with consultancy and specific rules are given on how staff members are paid and how long they can work with clients. These are that (a) pay should be determined primarily by the quality, accuracy, thoroughness and integrity of their work; (b) for companies with more than 50 staff members, a four-year time limit would be placed on work with individual clients to prevent relationships from becoming too close and; (c) there would then have to be a two-year break before the individual worked with the same client.21 United States In the US, preliminary steps to address the issue of CRAs in light of their role in the current financial crisis were taken in the form of a hearing by the House Oversight and Government Policy Committee. Here, the current heads of S&P, Moody’s and Fitch were issued with firm statements after testifying with US legislators announcing clear intentions to hold CRAs accountable for their actions.22 This may be impetus for future EU-like regulatory procedures to be taken in the near future.
Stephen Castle, 29 October 2008. EU prepares tight rules on credit rating agencies. International Herald Tribune.
US Congress, 22 October 2008.
248 Trust and Ethics in Finance
Proposed regulatory options
With these key issues in mind, several regulatory options can and have been proposed. Apart from their role in the current financial crisis, reasons behind calls for regulation stem from studies that have refuted the claims that reputational concerns lead to reliable ratings by CRAs.
Here, it was previously argued that CRAs provide accurate information to market participants in order to safeguard their reputation of ensuring reliable ratings. However, it has been noted that “worrying about reputation is not the same as worrying about providing reliable ratings”.23 This is because CRAs can be too conformist, too conservative or too bold precisely because they worry about their reputation.24 It is for all the reasons mentioned here and above that various proposals for the regulation of CRAs have been put forward. A brief description and discussion of each option shall be given before their respective merits are analysed.
Status quo This will entail leaving the current status quo as it is, since market participants will lobby against anything stronger. Moreover, the French Autorité des Marchés Financiers believes that the voluntary code has been implemented in a “globally satisfactory manner”.25 The main problem with this proposal is that it does and cannot solve the incentives problems mentioned above.
Nationalising the agencies The first proposal to be considered is the nationalisation of the CRAs. This comes from the classification of ratings as public goods which should therefore have public funding as they are essential in ensuring that the global financial plumbing operates smoothly. Some have Beatriz Mariano, 12 July 2008, Do reputational concerns lead to reliable ratings? www. voxeu.com Ibid Ibid Second Line of Defence 249 taken this proposal even further by calling on governments to set up a public rating agency (PRA).26 The key feature of such a PRA is that it would not aim to make profit and would not have an interest in providing overly generous rating thus avoiding the incentives and conflict of interest problems mentioned above. It is proposed that such a PRA be funded with taxpayers’ money, that it should be organised at the European level and combined with the European Financial Services Authority (EFSA) in the EU The main problem with this proposal is overcoming the lack of political will combined with the powerful lobby that is sure to be presented by the CRAs.
Abolish official recognition of CRAs’ ratings This proposal would involve eliminating the regulatory license of CRAs by removing the NRSRO designation and merely requiring agencies to register with the regulators. Intuitively, it can be seen that awarding CRA ratings with official recognitions places the ratings above suspicion by investors who then fail to conduct their own risk assessment before making investment decisions. No official recognition will force investors to see ratings as one of many tools that can be used to inform investment decisions, and thus allow them to gauge risk more accurately. For CRAs the barriers to entry and almost “god-like” status will be removed. The main problems with this proposal are that such a move will increase the regulatory burden on the regulators and would suppress the role of CRAs in Basle II, which, after all the effort put into crafting it, is likely to encounter great opposition.27 Reviving subscriptions Prior to the 1970s, the CRAs revenue came from a levy on the investors who used their ratings. At this time, the CRAs were not “tied up in Roel Beetsma, October 27, 2008. The ase for public sector credit rating agencies. www.rgemonitor.com Richard Portes, 2008.
250 Trust and Ethics in Finance the system” as they are today.28 This would reduce the incentives and free rider problems mentioned above. The main problem with this proposal is that it is linked to the public good aspect as described by Portes who notes that there will be free rider problems and payment by the user of the information will be either suboptimal or unenforceable, which is the reason why the CRAs stopped this practice in the 1970s.
Providing additional information apart from ratings This would address the accuracy problems highlighted above. This should include details such as assessment of the liquidity characteristics of the instruments and of likely volatility of its market price. In addition to this, it has been suggested that agencies provide a range of risk for each instrument rather than a point estimate29 and develop a distinct rating scale for structured finance products.30 The main problem with this proposal is that the agencies do not seem well equipped to do that as these kinds of risk are almost impossible to measure quantitatively. The CRAs are likely to lobby strongly against such an expansion of the depth of their work.
Introducing explicit legal liability for negligence and malfeasance At the moment, CRAs successfully maintain legal immunity from malfeasance claims on the ground that they are only financial journalists publishing their opinions, which are protected by free speech, although, as pointed out by Portes, Moody’s is much more profitable than the Financial Times, suggesting that they are earning some rents (their “reputational capital” and the regulatory license conferred by their role in the financial system.31. Apart from the evident shortcomings of the free speech argument, being able to hold them more legally accountable will Tara Perkings, 2007.
Richard Portes, 2008.
Ferguson et al, 2007.
Richard Portes, 2008.
Second Line of Defence 251 be a way to ensure that they take full responsibility for their ratings, and as such, be more thorough and meticulous in the provision of these ratings. However, the main problem with this proposal is that quite simply, CRAs will be sued out of business should such legal liability be in place as every investor who makes losses on a AAA rated product will demand redress.
Separating rating from consultancy and advisory functions This will need to go beyond Chinese walls32 and will entail ensuring the CRAs do not engage in consultancy work and that their structured finance business is stopped. This will eliminate the conflict of interest problems mentioned above. The main problem with this proposal is that the CRAs are likely to put up great resistance to giving up a highly remunerative line of work and to accept a dramatic decrease in their revenue stream.
Decrease barriers of entry The rationale for this is that it would allow more CRAs into the market and create competition that will improve the performance of CRAs.
However, the main problem with this proposals is the natural monopoly tendencies of the industry alluded to above as well as the possibility that more players will exacerbate the incentives problems and there will be more competition to gain issuers business, which may be done by providing the highest rankings.
Options analysis Table 1 looks at the extent to which the policy proposal addresses the problem category as highlighted above. Here, “yes”, “no” and “maybe” are used to indicate whether or not the proposal addresses the problem.
Ibid 252 Trust and Ethics in Finance
z^ D z z^ z^ EK ^ z^ z^ EK D z EK EK EK D z z^ z^ d t This exercise shows that no one solution can address all the problems relating to CRAs. This means that a policy package comprising key suggestions from each proposal will be needed.
The elements to be included in such a policy package should depend on the feasibility level of the implementation of that policy. Feasibility here refers to the political, lobbying and other practical obstacles that can potentially be leveled against the proposal by key market participants. Table 2 shows the relative feasibility options of each proposal on a scale of 1 (high feasibility) to 5 (low feasibility).
Here, it can be seen that introducing legal liability and nationalisation will most probably encounter the greatest political and lobbying obstacles as it threatens the very existence of CRAs. This is followed by no official recognition, reviving subscriptions and separating rating from consultancy and advisory functions. This category threatens the revenue stream of CRAs. Compared to these options, asking CRAs to provide information is less threatening and allowing more CRAs into the market even less so as the major CRAs have already established their market dominance.
Second Line of Defence 253
Recommendations In light of the above analysis, several recommendations can be made.
These shall be based on the identification of the core problem with the current role of CRAs in the global financial system seems to be related to dealing with innovation, in terms of the valuation of new and complex instruments. Here, the models used by the CRAs, which fail to factor in tail events and the limited amount of information provided by CRAs, are particularly problematic. Linked to this is the fact that the god-like stature of CRAs blind investors to the inherent uncertainty associated with ratings which are treated as indicators of guaranteed returns on investments instead of probabilities. This is particularly true in the case of rating new and complex instruments as was seen in the current financial crisis.
With this in mind, I would propose nationalising the ratings function (ratings as public goods) of CRA through the setting up of a public sector rating agency as proposed by Beetsma. Here, I would propose giving CRAs two options (1) Delegate their rating function to the public sector rating agency whilst continuing with the structured finance and consultancy arm of their business or (2) Cease their structured finance and consultancy activities and focus only on providing ratings.
254 Trust and Ethics in Finance Should the CRA choose the second option, I would propose that ratings be paid for by subscription by investors, because, although the conflict of interest problem will be addressed, the question of incentives will not. With the growth of capital markets around the world, and the need investors have of ratings to inform their decision, the problem of insufficient revenue faced by CRAs in the 1970s should not be an issue today. The only change will be that instead of making obscenely large profits, CRAs’ profits will fall within the normal range of other financial services providers.
The second caveat that will come with choosing option 2 is that the CRAs’ performance will be monitored by regulators to ensure the quality of ratings. Should CRAs be seen to fall below par based on a predetermined list of key performance indicators, they will be either nationalised or have their official recognition revoked.
In addition to this, it must be ensured that ratings be accompanied by additional information beyond point-estimates as highlighted above and that CRAs’ models be shown to take into account events beyond normal distributions with different instruments having different rating scales.
Basle 2, and other regulatory structures that rely on ratings, must then insist on the use of an aggregated rating of each of the main accredited agencies to increase accuracy, minimise potential errors. Here, it must be noted that the weighting of different factors by different CRAs may vary according to a set of different conditions and variables thus making certain aspects of the each CRA’s rating process essentially subjective.33 Moreover, aggregation will also potentially mitigate the effects to the high correlation and interdependence of all the lines of defense.
Ideally, each has to be independent and strong. However, strengthening the second pillar in the current environment will invariably lead to the strengthening of the third pillar as well.
Amelie Champsaur. 2005.
Second Line of Defence 255 Finally, to prevent this particular failure in the future, whenever innovation occurs, there must be a freezing of ratings for new instruments for six months from the time the product is introduced onto the market.
Here, it should be rated U as in unknown, and all known information about the product should be provided to investors. This will force investors to make their own investment decision based on their risk sensitivity levels.
Beetsma, R., 2007. “The Case for Public Sector Credit Rating Agencies”:
Castle, S., 2008. “EU prepares tight rules on credit rating agencies”, New York Times, October 29 2008.
Champsaur, A., 2005. “The regulation of credit rating agencies in the US, and in