The failure of credit ratings agencies to do their job – warn investors of the true risks entailed by the subprime mortgage securities they rated – was at the heart of the financial crisis. Policy makers since have wrestled with how to “fix” the ratings process going forward. Although the Securities and Exchange Commission has required the agencies to disclose more of their methodology, the ratings process is still less than transparent. The issuer-pay rating agency business model has been criticized as a central cause and new agencies designated by the SEC after 2008 moved away from this model, though they have since moved back. Various additional ideas to fix the system have been put forward but none has been adopted: randomizing the choice of ratings agency, or replacing private ratings with those of a public agency, such as the Securities and Exchange Commission.
Faulting the issuer-pay model for the Crisis, which has been in continuous use for more than 40 years cannot explain the sudden explosion and subsequent collapse of the securitization market, which occurred over a much shorter period. We offer a different approach here: by showing how the absence of a single, numerical, public structured credit scale to serve as a yardstick of structured credit quality in the U.S. debt capital markets provides a more plausible explanation for the problems in structured finance in particular. Transparent, numerical benchmarks of credit risk relating to structured credits should not only fix structured finance going forward, and ideally help resuscitate the market but in a more sensible fashion. In addition, we will argue that such benchmarks also are a necessary component to a prudent system of capital regulation and for accurately informing investors of true credit risk, just as speed limits are a necessary component of vehicular traffic regulation.
Structured credits refers to securities backed by pools of receivables, such as mortgages or loans, frequently with different “tranches” with different claims on the cash flows generated by the financial instruments backing the securities. As such, structured securities are a lot more complex than the standard corporate bond, which promises the same interest and principal payments per bond to all bondholders.
This paper elaborates several themes relating to the credit ratings process in general, and to the ratings of structured financial products, such as collateralized debt obligations (CDOs), that were common before the crisis:
- Ratings should be forward looking, but have been backward looking in practice
- Risks of default in structured securities change over time in ways that are not true for corporate or sovereign debt, and which are not reflected in the methodology for rating structured products
- Ratings upgrades and downgrades lag reality by significant margins and are biased by the desire of ratings agencies to show smooth progressions in ratings
- Rating should reflect expected losses, but this is not the case for structured products
- Investors in structured products relied too blindly on ratings, allowing more knowledgeable parties to take advantage of them. This information asymmetry led to the creation of too many securitized instruments.
- The foregoing problems were especially manifest in CDOs, where the ratings of the underlying ecurities were the key to the ratings of the CDOs themselves, compounding the problems
Central to fixing structured finance, in our view, is changing the way structured securities are rated. For more than a century, investors have been accustomed to alphanumeric ratings of corporate bonds (AAA, AA, A, BBB and so on). The different letters on the corporate ratings scale represent the enterprise’s distance to default. For lenders, who cannot know precisely the source or quantity of cash available to the firm when future debt comes due, a corporate rating is a useful gauge of future payment ability based on what is known today.
But this is not the case for structured products, whose credit quality is completely determined by the amount of cash collected on a dedicated pool (or pools) of receivables and distributed through to investors determined by security’s structure, that cash must be counted. The analysis of cash flow adequacy requires numbers, or a cardinal rather than an ordinal (lettered) scale.
Securities that are rated only in an ordinal fashion – that is, ranked in order of likelihood of default – can be misleading, allowing sophisticated parties in the know to take advantage of naïve investors. Ordinal rankings can also lead to a sense of false comfort for investors and policy makers, which contributed to the global Financial Crisis.
To complicate matters, no two credit rating agencies use the same benchmarks, which, in structured finance in particular, is equivalent to saying no two credit rating agencies count cash the same way. This is not only violates the “law of one price,” but the confusion for investors creates an opportunity for rating agencies to tamper with their own input scales undetected, advantaging preferred clients with cheaper funding and competing invisibly for market share.
Having a public performance benchmark scale for rating structured credits would narrow the sophistication gap between arrangers and investors, empowering the latter to conduct their own value analysis and “see through” false ratings. Increased investor vigilance should motivate more competitive pricing for high quality debt and undermine the perverse incentives currently at work in the market. The existence of a public scale would put a foundation under rating agency oversight, which many in Congress have urged, while lessening the market’s uncritical dependency on ratings.
Given the market’s propensity to exploit information asymmetries, it is not surprising that a public benchmark scale has not yet materialized spontaneously to heal the broken market. Nor is it likely to. Rating agencies for their part are no more likely now than before to volunteer to work together voluntarily to accommodate greater public scrutiny or diminish their own power.
As a classic public good, the structured credit scale needs public support and may need to be developed by one or more federal regulatory institutions that support the use of the structured credit scale in the regulatory landscape. Logically, such an initiative could come from the SEC, either on its own or by suggestion from the Financial Stability Oversight Council (FSOC), or from the Federal Reserve. Each regulatory choice addresses a dimension of the problem. In the final section of the paper, we analyze the trade-offs of each arrangement.
Disclosure: Ann Rutledge is a founding principal of R&R Consulting, which is an independent credit rating agency that applied in 2011 to be a Nationally Recognized Statistical Rating Organization. Brookings learned about this application subsequent to the article’s publication and regrets not disclosing this information at the time of publication.