One of the major factors leading to the subprime crisis was the failure of credit rating agencies’ assessments of structured finance products. What are these ratings agencies, how did they go wrong, and what can regulators do to address the problems?
The failure of the credit rating agencies: Evidence
Credit rating agencies have been part of the financial system since the first ratings were issued by Moody’s in 1909. In 1975, the Securities and Exchange Commission formalised the role of the ratings agencies (and created a regulatory barrier to entry) by creating and granting the designation of Nationally Recognised Statistical Rating Organisation (White 2002).1 Despite complaints of agencies’ irresponsiveness during the East Asian Financial Crisis (1997) and the failures of Enron (2001) and Worldcom (2002), few foresaw the magnitude of the problems today.
The failure has come with the advent of structured finance products. Subprime mortgage originations increased tenfold from 1996 to 2006 (Chomsisengphet and Pennington-Cross 2006; Ashcraft and Schuermann 2008), and from 2002 to 2006 Moody’s profits tripled, with an average profit margin around 50% (Lowenstein 2008).2 Therefore it is important to understand how rating structured product is different from rating corporate bonds. A key difference is the complexity of the products. Gorton’s (2008) broad exploration of the engineering of structured products demonstrates the degree of complexity of these assets and the difficulty in identifying their underlying cash flows. This complexity made the agencies’ models less reliable. To make things worse, the rating agencies were dependent on issuer due diligence to document the expected cash flows.
As “gate keepers” of the fixed income markets, rating agencies had a key role to play in reducing the asymmetric information between the issuer of a security and investors. Yet three characteristics of rating agencies are troublesome. First, since the beginning of the 1970s, rating agencies have been paid by the issuers. In an investigation by the Securities and Exchange Commission in 2008, it was found that senior analytical managers and supervisors participated in fee discussions with issuers and that the analytical staff also discussed ratings decisions and methodology in the context of fees and market share. At the same time, rating agencies provided related consulting services to issuers. Second, rating agencies are not liable for the quality of their ratings as their responsibility in information provision to the public has been found in courts to be closer to that of a journalist than to that of an auditor. Finally, the possibility for issuers to approach several rating agencies and “shop” for the best possible rating implies that the market only sees the most optimistic ratings that have been purchased by the issuer.
Market regulation has acknowledged the important role of gate keepers, not only in the US through the creation of the Nationally Recognised Statistical Rating Organisation designation, but also under Basel II banking regulation, which acknowledges a critical role for rating agencies. Have they fulfilled the role that regulation bestows upon them?
A number of recent papers have analysed the credit rating industry. They incorporate two common elements. The first element is that the key change in the ratings industry has been an increase in complexity of the products rated (i.e. structured finance versus corporate bonds). The second, related, element is that some investors in the market for these products are naïve. The main focus of these papers is to identify agencies’ conflicts of interest and examine the perverse effect of shopping.
Conflicts of interest
Bolton et al. (2008) model the agencies’ conflict of understating credit risk to attract more business and find the intuitive results that rating agencies are more prone to inflate ratings when there is a larger fraction of naive investors in the market or when expected reputational costs are lower. More interestingly, to the extent that in booms the fraction of naive investors is higher, and the risk of getting caught understating credit risk is lower, their analysis predicts that rating agencies are more likely to understate credit risk in booms than in recessions. Furthermore, they find that the ability of issuers to restructure financial products (tranching) can reduce welfare, as the only benefit of tranching is to make rating agencies comfortable with issuing a good rating. Otherwise, tranching does not add any risk-diversification or incentive benefits in their model.
Conflicts of interest are also examined by Mathis et al. (2008) in a dynamic model where reputation is endogenous. Their analysis provides an additional insight – when the fraction of income that comes from rating complex products becomes too large, rating agencies tend to inflate ratings.
Skreta and Veldkamp (2008) argue that when financial products are simple, the correlation among ratings is high and there is not much gain from shopping. When products become complex, however, an increase in the variance of the ratings provides incentives to shop.
Both Skreta and Veldkamp and Bolton et al. demonstrate that increasing the number of rating agencies worsens the shopping problem and reduces welfare. This indicates that fostering entry into the ratings business may not make it more efficient. In empirical work, Becker and Milbourn (2008) find that in the market for corporate bonds, increased market share by Fitch decreased the accuracy of ratings and made them pander more to issuers.
The quality of disclosed information
Pagano and Volpin (2008) address the puzzle of coarse information transmission by an agency to the market. The standard economic view is that more information increases market efficiency, yet information on the underlying loans and the tools to analyse them were not made easily accessible by the issuers or the ratings agencies. The point they make is that in a market with both sophisticated and naïve investors, only the first will be able to interpret and process additional complex information, and this will create a “winner’s curse” problem for naïve investors. The implication is that limiting information may help develop the primary market, as it will allow unsophisticated agents to access it at a more level playing field. Nevertheless, they point out that the cost of this information coarseness is that the secondary market will be less liquid.
Academic commentators seem to have reached a consensus on two issues. First, the ratings industry must be regulated to address agencies’ fundamental conflicts of interests. Second, shopping for ratings should be banned to reduce the conflict of interest of issuers.
One solution to the ratings inflation problem that is often proposed is to return to the “investor pays” system that was in place until the beginning of the 1970s. This system vanished because of free riding (coinciding with the introduction of affordable photocopying)., This could be implemented through the taxation of investors to eliminate the free riding, but that may be costly for the government to implement and oversee.
Another method of restoring the investor pays model would be to let exchanges pay for the ratings. Thus, Mathis et al. (2008) suggest creating a platform that separates the rating agencies from issuers. The platform would take payments from issuers and assign products to one or more rating agencies to rate. If the platform can balance the interests of issuers and rating agencies, it can eliminate conflicts of interest and shopping. It remains unclear though whether this would not simply amount to replacing one form of conflict of interest by another. Why wouldn’t exchanges also pressure rating agencies for good ratings in order to increase their own business, the trading of more securities on their exchange?
An agreement between New York State Attorney General Andrew Cuomo and the three main rating agencies in the summer of 2008 requires that issuers pay rating agencies upfront for their rating. Bolton et al. argue that upfront payments to rating agencies eliminate conflicts of interest for rating agencies (to the extent that they can be enforced) but cannot prevent shopping by issuers. They argue that augmenting the “Cuomo plan” by requiring that all rating agencies automatically disclose their ratings could help mitigate shopping. Oversight on analytical standards may be necessary as well, since upfront fees might create a moral hazard problem: rating agencies may not engage in costly efforts to maintain the quality of their analysis.
1 The NRSRO designation was granted to 7 firms, which through mergers and acquisitions, consolidated into the 3 firms Moody’s, Standard & Poor’s, and Fitch. The NRSRO designation was not granted to any new firms (despite applications) between 1992 and 2006.
2 Moody's profits are the easiest of the rating agencies to measure since they are a public stand-alone company.
Ashcraft, A. and T. Schuermann (2008), "Understanding the Securitization of Subprime Mortgage Credit ," mimeo, Federal Reserve Bank of New York.
Becker, B., and Milbourn, T. (2008), "Reputation and competition: evidence from the credit rating industry," mimeo, Havard Business School.
Bolton, P., Freixas, X., and J. Shapiro (2009), “The Credit Ratings Game”, mimeo, Columbia University and Universitat Pompeu Fabra.
Chomsisengphet, S. and A. Pennington-Cross (2006), "The Evolution of the Subprime Mortgage Market," Federal Reserve Bank of St. Louis Review, January/February, 31-56.
Gorton, G. (2008), "The Panic of 2007", mimeo, Yale University.
Lowenstein, Roger (2008) "Triple-A-Failure," New York Times Magazine, April 27
Mathis, J., Mc Andrews, J. and J.C. Rochet (2008), "Rating the Raters," mimeo, TSE and New York Fed.
Pagano, M. and Volpin, P. (2008), "Securitization, Transparency, and Liquidity," mimeo, Università di Napoli Federico II and London Business School.
Skreta V. and L. Veldkamp (2008) “Rating Shopping and Asset Complexity: a Theory of Ratings Inflation”, mimeo, NYU.
White, Lawrence, 2002, "The Credit Rating Industry: An Industrial Organization Analysis", in Levich, Richard M., Giovanni Majnoni and Carmen Reinhart, eds. Ratings, Rating Agencies and the Global Financial System, Kluwer Academic Publishers, Boston.