Pages

Monday, August 3, 2009

A Practical Proposal for Rating Agency Reform

The premise behind most proposed regulatory regimes for rating agencies – managing conflicts of interests between issuers and investors – is misguided. The conflicts of interest faced by rating agencies are not qualitatively different than those faced by many other industries, such as the accounting or brokerage industries. The problem is actually more pedestrian – product quality versus profit. Regulating outside influences will not work if the source of the problem is internal.

The conflict stems directly from the concentration of power at the rating agencies. They create rating methodologies, assign ratings based on the same and then judge their own performance. There is no “separation of powers” in the credit rating world – the agencies act as judge, jury and executioner, and are not required to justify their actions to any regulator or other third party. This concentration of power, combined with regulatory demand for ratings, makes it too tempting to alter methodologies and procedures to maximize revenue.

Imagine, for example, if accountants could be the sole arbiters of what counts as income or expenses for a company’s financial statements. Imagine further that their word was final and not even the authorities could appeal the verdict. No one would be surprised by the resulting decline in accounting standards.

My solution is to reduce the concentration of power within the agencies. The proposed template would be based on the regulatory apparatus for other third party information providers in the capital markets – accountants and attorneys.

Like credit rating agencies, accountants and attorneys are private parties competing for business in the capital markets. While no one would argue that these parties are paragons of virtue, their ability to generate profits through manipulation of analyses and opinions is limited. One key safeguard is that neither party sets its own methodological standards. Accountants take their cue from national or international standards and the relevant regulators. Lawyers likewise only interpret and offer advice on legal standards set by authorities.

Standardized Methodology

Rating agencies should follow a standard set of methodologies as well. A body similar to the FASB could be empowered to set various standards for the rating agencies. These could initially include some credit basics such as the definition of ratings (e.g. what is the default probability of a AAA at 5 years), the frequency of surveillance updates and the types of data required to be presented for ratings.

The proposed central credit policy body would have two major effects in improving credit quality:

First, the central approach would stop the methodological “race to the bottom” that was at least partially responsible for today’s credit crisis. Mistakes will still be made, but rating agencies will no longer have to the incentive to play “one-upmanship” in order to maintain their market share.

Second, a standard methodology can be used to actually match the ratings with their regulatory usage. For example, the risk profile of a given rating should match the capital weight assigned to it – this is not the case today. Likewise, if the Federal Reserve requests that only securities rated AAA be allowed in a financing program, the quantitative (e.g. probability of default) measures will have a consistent meaning across the agencies. This is not the case today.

The rating agencies will argue that standardizing methodologies will lead to a loss of independent opinion. This is simply not true; despite the differences in methodologies, the bankers made sure that the ratings issued in structured finance were nearly identical. Moreover, the supposedly different methodologies used by the rating agencies prior to the crisis did nothing to prevent it. If a rating agency had actually proffered a different methodology, especially a more grounded, more conservative analysis, it would have been run out of business.

In fact, the rating agencies do not have to give up their current methodologies and approaches. What they can be asked to provide is a new “regulatory rating” which fits the needs of regulators. This would be no different from having different accounting standards for different tasks. For example, GAAP accounting and tax accounting. Each can produce different results, but is tailored to the needs of the requisite purpose. Lawyers, too, deal with a multitude of jurisdictions and conflicting laws. Ratings should be no different.

Central Ethics Body

Another function that is currently in the sole control of the rating agencies is the adjudication of ethics and professional codes. The International Organization of Securities Commissions has promulgated a code of conduct for the ratings industry. While it is a very good code, the enforcement of its principles is left entirely to the rating agency itself. There are no appeals or penalties for non-enforcement. This is another example of concentration of power within the rating industry. Other professionals in the capital markets (stockbrokers, accountants and lawyers) are subject to ethical standards which can be and are enforced beyond the employer itself. In my personal experience the lack of enforceability leads to precisely the result one would expect – loose or non-existent enforcement.

A central ethics body needs to be established with the authority to adjudicate violations of the IOSCO or any other code of ethics for the ratings industry. The proceedings of this body need not be public, but they should be known to the regulatory bodies with proper jurisdiction such as the SEC.

In order to empower this body to fairly enforce its rulings, a fine-tuned remedy is required. I propose that rating agency analysts be licensed individually to provide “regulatory ratings” only. A licensing requirement would put rating analysts on par with their colleagues in the securities industry. Removal of an individual’s ability to provide ratings is a sufficient punishment for most violations of conduct within the ratings agencies. The knowledge that one may lose her livelihood will increase the incentives for analysts to behave ethically.

One additional benefit of licensing would be increased understanding of regulatory requirements within the rating agency. I have been surprised and frustrated by analysts’ lack of knowledge of applicable regulation or even securities laws. Other regulatory licensing standards (e.g. Registered Representative) require the candidate to have a basic understanding of the legal environment surrounding her profession.

Conclusion

I believe that the two proposals above, the creation of standardized methodologies and the use of ethics bodies, would significantly improve the quality of the work performed by rating agencies. They would also allow regulators to more efficiently discharge their responsibilities. Additionally, these proposals would also be simple to implement and would cause a minimal amount of capital market dislocation.


- Former rating agency analyst

3 comments:

Current Sane Person said...

The key to unraveling this entire article is that lawyers are there to interpret, you know, the law. Ratings analysts are there to offer an opinion on something's credit quality. Bankers spent the last decade whining about how ratings methodologies weren't transparent and the agencies listened intently to their customers. Once the methodologies became transparent, bankers happily gamed the now-transparent system for ridiculous ratings. Now they want to cry that the ratings didn't make sense?

Fact is, if the industry doesn't like someone's ratings, that rating agency will stop getting business. If the industry actually cared about correct ratings, then they would put their money where their mouth is rather than shouting from the rooftops about regulation, a set of even more stringent and transparent methodological standards, which would further allow them to rip off their unsuspecting investors. That's assuming, of course, that any of their investors survive.

Tjemme said...

You suggest that rating agencies are their own jury. That is not quite the case. Professional investors also buy services from rating agencies - and they have made them so large. Professional investors feel that the ratings provide value added (as a service for their end-investor clients, or as a check / supplement on their own research).

Having said that, I agree with you that (credit) risk measurement requires more standard setting. I personally think that professional risk management organisations (like GARP or PRMIA) can play a role here. Accountants force companies to release accounting information. Risk managers should force companies to release (standardised) risk information. This is what 'Basel' attempts to achieve for banks (in a longer-term process) and 'Solvency' for insurance companies. Regular publishing of regulatory stress tests may also help (once the details have been finalised).

In terms of a ethics body, the insurance industry has the concept of a 'certifying actuary'. This person plays the role of a 'risk accountant'. He has the duty to report centrally to regulators if he feels things may go wrong.

GP said...

@ Tjemme, thanks you make good points.

I can't respond on our guest contributor's behalf, but may I ask what you mean when you say that "professional investors also buy services from rating agencies - and they have made them so large."

I think it's fair to say that the rating business -- a monopoly -- is where they have become "large," as you say. I'm not sure what particulr services you're referring to, but data, analytics and research are almost negligible, peripheral sources of revenue for the rating agencies versus the rating themselves. Or might you disagree?

Similarly, the evaluations teams they have are separated from the rating agencies.