In the run-up to the financial crisis of 2007-2008, market participants relied heavily on the ratings that credit rating agencies assigned to financial instruments, including mortgage-backed securities, to determine creditworthy investment options. As mortgage holders began to default on their loans and many highly rated securities lost value, the poor quality of these ratings became apparent. Policy makers pondering financial regulatory changes to avoid future catastrophes should understand how regulatory actions facilitated a noncompetitive credit rating industry and propelled its members into the center of the bond information process, which in turn contributed to the financial crisis of 2007-2008.
A credit rating agency is a potential source of information for market participants who are trying to ascertain the creditworthiness of borrowers. Essentially, rating agencies offer judgments (they prefer the word "opinions"1) about the quality of bonds issued by corporations, governments (including U.S. state and local governments, as well as "sovereign" issuers abroad), and mortgage securitizers. These judgments come in the form of letter grades. The best-known scale is that used by Standard & Poor's (S&P) and some other rating agencies: AAA, AA, A, BBB, BB, etc., with pluses and minuses as well.2
John Moody published the first publicly available bond ratings (mostly concerning railroad bonds) in 1909. Moody's firm3 was followed by Poor's Publishing Company in 1916, the Standard Statistics Company in 1922,4 and the Fitch Publishing Company in 1924.5 These firms sold their bond ratings to bond investors in thick rating manuals. In the language of modern corporate strategy, their "business model" was one of "investor pays."
This relationship between the rating agencies and the U.S. bond markets changed in 1936 when the Office of the Comptroller of the Currency prohibited banks from investing in "speculative investment securities," as determined by "recognized rating manuals" (i.e., Moody's, Poor's, Standard, and Fitch). "Speculative" securities were bonds that were below "investment grade,"6 thereby forcing banks that invested in bonds to hold only those bonds that were rated highly (e.g., BBB or better on the S&P scale) by these four agencies. In effect, regulators had endowed third-party safety judgments with the force of law.
In the following decades, insurance regulators and then pension fund regulators followed with similar regulatory actions that forced their regulated financial institutions to heed the judgments of a handful of credit rating agencies.
NRSROS: BARRIER TO ENTRY
In 1975, the Securities and Exchange Commission (SEC) issued new rules that crystallized the centrality of the rating agencies. To make capital requirements sensitive to the riskiness of broker-dealers' bond portfolios, the SEC decided to use the ratings on those bonds as the indicators of risk.
However, the SEC worried that references to "recognized rating manuals" were too vague and that a "bogus" rating firm might arise that would promise "AAA" ratings to those companies that would suitably reward it and "DDD" ratings to those that would not. If a broker-dealer claimed that those ratings were "recognized," the SEC might have difficulties challenging this assertion.
To solve this problem, the SEC designated Moody's, S&P, and Fitch as "Nationally Recognized Statistical Rating Organizations" (NRSROs).7 In effect, the SEC endorsed the ratings of NRSROs for the determination of the broker-dealers' capital requirements. Other financial regulators soon followed suit and deemed the SEC-identified NRSROs as the relevant sources of the ratings required for evaluations of the bond portfolios of their regulated financial institutions.
Over the next 25 years, the SEC designated only four additional firms as NRSROs,8 but mergers among the entrants and with Fitch reduced the number of NRSROs to the original three by the end of 2000. NRSRO designation had become a significant barrier to entry into the bond-rating business because the SEC's support was quite important for potential entrants. Moreover, the SEC neither established criteria for a NRSRO designation nor provided any justification or explanation as to why it "anointed" some firms with the designation and refused to do so for others.
Also importantly, in place of the "investor pays" model established by John Moody in 1909, the agencies converted to an "issuer pays" model during the early 1970s whereby the entity that is issuing the bonds also pays the rating firm to rate the bonds. This change opened the door to potential conflicts of interest: A rating agency might shade its rating upward so as to keep the issuer happy and forestall the issuer's taking its business to a different rating agency.
In the bond-information market, experience, brand-name reputation, and economies of scale are important features. The industry was never going to be a commodity business of thousands (or even hundreds) of small-scale producers. Nevertheless, regulators' actions surely contributed heavily to the dominance of the three major rating agencies. The SEC's belated efforts to allow wider entry into the NRSRO category during the current decade were too little and too late.9 The entrants could not quickly overcome the advantages of the "big three's" incumbency.
FUELING THE SUBPRIME DEBACLE
To a large extent, subprime lending fueled the U.S. housing boom that began in the late 1990s and ran through mid-2006.10 The securitization of the subprime mortgage loans, in collateralized debt obligations (CDOs) and other mortgage-related securities, encouraged subprime lending and led to the development of other financing structures, such as "structured investment vehicles" (SIVs), whereby a financial institution might sponsor the creation of an entity that bought tranches of the CDOs and financed its purchase by issuing short-term "asset-backed" commercial paper (ABCP). If rating agencies rated the CDO tranches in an SIV favorably, that favorable rating concomitantly meant high ABCP ratings (interest-rate risk and liquidity risk were apparently ignored in the ratings). Hence, the agencies' favorable ratings of mortgage-related securities were crucial for the securitization process.
Favorable ratings were important for at least two reasons: First, as has been discussed above, ratings had the force of law with respect to regulated financial institutions' abilities and incentives (via capital requirements) to invest in bonds. Furthermore, in addition to the prohibition on banks and savings institutions holding sub-BBB bonds, there was another major impact of ratings: Mortgage-backed securities—including CDOs—rated AA or better and issued by non-governmental entities qualified for the same reduced capital requirements (1.6 percent of asset value) as those issued by Fannie Mae and Freddie Mac, instead of the higher (4 percent) capital requirement that applied to mortgages and lower-rated mortgage securities. Higher ratings on larger fractions of the tranches that flowed from any given package of mortgage securities thus meant that regulated financial institutions could more readily buy these larger fractions. Second, the generally favorable reputations that the credit rating agencies had established in their corporate and government bond ratings meant that many bond purchasers—both regulated and non-regulated—were inclined to trust the agencies' ratings on the mortgage-related securities, even (or, perhaps, especially) if the market yields on those securities were higher than on comparably rated corporate bonds.
Meanwhile, the profits from gaining higher ratings on a larger percentage of tranches also motivated securitizers. These higher-rated tranches carried lower interest rates, which issuing firms would have to pay to the investors in those tranches, leaving a greater spread for the securitizers. Consequently, securitizers would be prepared to pressure the rating agencies, including threats to choose a different agency, to deliver those favorable ratings.
Currently, policy makers are favoring regulations that require ratings agencies to address conflicts of interest and to provide more transparency concerning their rating and their rating processes. However, such rules would increase the costs of being a rating agency and thus discourage entry, rigidify current arrangements and procedures, and discourage innovation. Ironically, they would likely make the incumbent rating agencies more important—and may well have little or no effect on the quality of ratings.
There is another, better route: Suppose that instead of granting credit rating agencies' judgments the force of law, regulators placed the burden of a safe bond portfolio directly on their financial institutions, but permitted them to make their own choices (subject to regulatory oversight) as to where and how they gained the information that would guide those safety judgments. Under this alternative policy route, banks (and insurance companies, etc.) would have a far wider selection as to where and from whom they could seek information and advice as to the safety of the bonds in their portfolios.
In this scenario, some institutions might choose to do the necessary research on bonds themselves or rely primarily on the information yielded by the credit default swap market. Others might turn to outside advisors that they consider reliable based on the advisors' track records, business models (including the possibilities of conflicts of interest), other activities that might pose conflicts, and anything else that the institutions considered relevant. Though they may still turn to credit rating agencies, institutions will also be able to seek the advice of investment banks, industry analysts, and upstart advisory firms that are currently unknown.
With this competitive approach, the bond-information market will open up to new ideas about ratings business models, methodologies, and technologies, as well as new entrants, in ways that have not actually been possible since the 1930s. The answer to whether the "issuer pays" business model could survive under this alternative policy scenario rests on whether bond buyers can ascertain which advisors provide reliable advice. If they can (which seems reasonable, since the major transactors in the bond markets are financial institutions), then they would be willing to pay higher prices (and thus accept lower interest yields) on the bonds of any given underlying quality that is rated by these reliable advisors. The competitive process would determine the outcome of the "issuer pays" business model.
A series of regulatory actions brought the judgments of a few rating agencies to prominence. Regulators and investors came to rely significantly on securities ratings by these agencies. A careful look at the evolution of the interaction between the rating agencies and regulatory decision making reveals that financial regulators created a non-competitive industry and propelled it into the center of the bond information process. The three largest firms in that industry became complacent and harbored potential conflict-of-interest issues.
Going forward, policy makers should withdraw their delegations of safety judgments to the credit rating agencies. This would encourage competition, entry, and innovation in the bond information business, while financial institutions would (as they should) bear the burden of justifying the safety of their bond portfolios to their regulators.
1. The rating agencies favor that term because it allows them to claim that they are "publishers" and thus enjoy the protections of the First Amendment of the U.S. Constitution (e.g., when the agencies are sued by investors and issuers who claim that they have been injured by the actions of the agencies).
2. For short-term obligations, such as commercial paper, a separate set of ratings are used.
3. Dun & Bradstreet bought Moody's firm in 1962; subsequently, in 2000, Dun & Bradstreet spun off Moody's as a free-standing corporation.
4. Poor's and Standard merged in 1941 to form S&P; S&P was absorbed by McGraw-Hill in 1966.
5. Fitch merged with IBCA (a British firm that was owned by French business services conglomerate FIMILAC) in 1997.
6. United States Comptroller of the Currency, Purchase of Investment Securities, and Further Defining the Term "Investment Securities" as Used in Section 5136 of the Revised Statutes as Amended by the "Banking Act of 1935," Section II (February 15, 1936). This rule still applies to banks today. This rule did not apply to savings institutions until 1989. Its application to savings institutions in 1989 forced them to sell substantial holdings of "junk bonds" (i.e., below investment grade) at the time, causing a major slump in the junk-bond market.
7. "Commodity and Securities Exchanges," Title 17, Code of Federal Regulations, section 240.15c3-1 (1998). Also, in the early 1990s, the SEC again made use of the NRSROs' ratings when it established safety requirements for the short-term bonds (e.g., commercial paper) that are held by money-market mutual funds.
8. The SEC bestowed the NRSRO designation on Duff & Phelps in 1982; on McCarthy, Cristanti & Maffei in 1983; on IBCA in 1991; and on Thomson Bank Watch in 1992.
9. The Sarbanes-Oxley Act of 2002 included a provision that required the SEC to send a report to Congress on the credit rating industry and the NRSRO system. The report simply raised questions rather than addressing the issues. In early 2003, the SEC designated a fourth NRSRO (Dominion Bond Rating Services, a Canadian firm), and in early 2005 the SEC designated a fifth NRSRO (A.M. Best, an insurance-company rating specialist). Congress later passed the Credit Rating Agency Reform Act (CRARA) in 2006. The act instructed the SEC to cease being a barrier to entry and specified the criteria for designating new NRSROs. In response to the legislation, the SEC designated three new NRSROs in 2007 (Japan Credit Rating Agency; Rating and Information, Inc. [of Japan]; and Egan-Jones) and another two NRSROs in 2008 (Lace Financial and Realpoint). There are currently ten NRSROs.
10. The debacle is discussed extensively in Gary Gorton, Slapped in the Face by the Invisible Hand: Banking and the Panic of 2007 (National Bureau of Economic Research, 2008); Viral Acharya and Matthew Richardson, "Causes of the Financial Crisis" Critical Review: A Journal of Politics and Society 21, nos. 2-3 (2009); Joshua Coval, et al., "The Pricing of Investment Grade Credit Risk during the Financial Crisis" (preliminary paper, USC Business School, 2009); and Mayer, et al., "The Rise in Mortgage Defaults: Facts and Myths" Journal of Economic Perspectives 23, no. 1 (2009).