Credit rating agencies play a key role in domestic and international markets for fixed-income securities, and the use of their ratings is increasingly prominent in regulatory requirements. Only two firms, Standard & Poor's and Moody's, with Fitch as a distant third, "overwhelmingly dominate the business," as the Washington Post recently observed. A central factor in this dominance is that the Securities and Exchange Commission limits new entry and competition. Thus the government both mandates demand for rating agency services and severely restricts supply. The result is lack of customer choice and (exactly as theory would predict) exceptional profitability for what may be fairly termed a government-sponsored cartel. Several steps should be taken immediately to increase competition in the rating industry, which would increase customer choice, price competition, innovation, and the variety of analysis available to investors.
A Notable Franchise
To be designated a "nationally recognized statistical rating organization" (NRSRO) by the SEC is an extremely valuable franchise. If granted, it allows entry into a cartel with only three U.S. members, which represent about 95 percent of sector revenues. Two of these agencies represent about 80 percent of the revenue. Since it is common for securities issuances to have two ratings, they need not compete much against one another. Obviously, it is essential to the economics of the business that Standard & Poor's and Moody's, the dominant two, and Fitch, the third, have been designated NRSROs by the SEC; it is equally essential that other potential competitors have not been. The only other current NRSRO is a Canadian firm, Dominion.
Doubtless it was never the intent of the SEC to create a cartel or the related cartel profits, but this has been the result of its actions (and inaction) over the last three decades since introducing the NRSRO label in 1975. This designation is not a statutory requirement, but a regulatory initiative, originally a very narrow one: to specify which ratings could be used to calculate broker-dealers' required capital. The idea was subsequently picked up by various other regulators, which linked their regulations to the SEC designation. It is now embedded in large numbers of investment and capital regulations for securities firms, mutual funds, banks, thrifts, insurance companies, and government-sponsored enterprises.
In December 2002, a Shadow Financial Regulatory Committee statement pointed out that "the use of NRSRO ratings in federal and state securities laws and regulations has expanded dramatically." It continued: "However, the term 'NRSRO' remains undefined in SEC regulations . . . as has the process for obtaining NRSRO designation from the SEC." This remarkable situation remains true, in spite of a lengthy report on the matter submitted by the SEC to Congress in January 2003, a subsequent SEC "Concept Release," and congressional hearings and concerns. At subcommittee hearings of the House Committee on Financial Services in April 2003, for example, Chairman Michael Oxley's statement and questions included the following:
There are also concerns regarding . . . whether anticompetitive barriers to entry exist for ratings firms seeking recognition by the SEC.
I am concerned that the commission may have allowed an oligopoly to exist.
What would prevent any of the agencies from exercising monopoly power or pricing for their services?
Our goal is to try to get more competition, more entries into the market [as] you have obviously heard from Chairman Baker, Ranking Member Kanjorski, and myself.
So far none of this has brought any change.
Professor Lawrence J. White of New York University, having examined the history of this issue at length, concluded that the results of the NRSRO regulation are that
Incumbent bond rating firms are protected, potential entrants are excluded, and new ideas and technologies for assessing the riskiness of debt (and therefore the allocation of capital) may well be stifled. This entry regulation is a perfect example of good intentions gone awry.
Needless to say, competitive markets typically lead to innovation, economic growth, efficiency, and customer choice. They also lead to rates of profitability that tend to reflect the market cost of equity capital. If it is true that the rating agency sector is a government-sponsored cartel, we ought to see evidence of that in high rates of profitability at the dominant firms. Although the financial results of Standard & Poor's and Fitch listings are not publicly available in any informative way, since they are both divisions of larger companies (respectively, McGraw Hill and Fimalac, a French company), Moody's is a publicly traded corporation with standard SEC disclosures. Its exceptional profitability, detailed below, is fully consistent with our expectation. It is a fair assumption that the returns of Standard & Poor's are similar.
Moody's profit performance is outstanding by any measure. (See Table 1.) In 2003, it had operating income of $663 million on revenues of $1.2 billion for an operating profit margin of 53 percent. Its net after-tax profit of $364 million compared to a year-end equity account of negative $32 million (negative $327 million the year before). This makes the return on equity ratio (ROE) hard to calculate! But whatever you think it is, it is impressive.
Nine Months 2004 Annualized
N/A (negative equity)
Net Operating Cash Flow
Moody's total investment as of December 31, 2003, defined as non-current assets plus net working capital, was $509 million. The company generated a notable after-tax return on investment (ROI) of 74 percent. This is by definition an unleveraged return. Warren Buffett, who knows a good thing when he sees it, has Berkshire Hathaway owning 16 percent of Moody's common stock (as of March 2004).
The 2004 profit performance through September 30 is equally worthy of note. Operating profit margin was 55 percent. Net after-tax profit for the nine months was $302 million or $403 million annualized. The equity account became positive, averaging about $58 million, which leads to an annualized rate of return on equity of about 690 percent. The annualized after-tax ROI was 71 percent, while net operating cash flow was $378 million ($504 million annualized) and the company had accumulated cash of $451 million by the end of the third quarter.
There is no reason to think that the profitability of the other dominant rating agency is not also highly attractive, but there is good reason to doubt that such results could be sustained in a competitive market.
What Does "NRSRO" Status Confer?
To make the rating agency sector more competitive, the role of the NRSRO franchise must be addressed. In the first place, there is the term itself: "nationally recognized statistical rating organization." This implies some sort of market test, suggesting that the rating agency has achieved acceptance and credibility with a large number of investors and other financial actors by the quality of its ratings and their performance over time. At the time of its introduction, NRSRO did in fact reflect such a market test, but it no longer does. Now it really means only one thing: SEC-approved rating agency. So the term is fundamentally misleading.
As noted above, there have been very few approvals granted by the SEC, the criteria are not well defined, and the approval process, judging by the experience of potential competitors who have attempted to gain approval, is purely ad hoc.
If you think back to the position of the SEC at the time of its first NRSRO regulation in 1975, it is easy to see why as a first designation of acceptable sources for ratings, it would be attractive to ask which rating agencies had met the market test and were widely accepted or already "nationally recognized." At that point, the concept was an endorsement of the market evolution of the past. But once any new competitor had as a prerequisite, before it could gain national acceptance by the market, to gain the approval of the SEC--once, in other words, all market evolution was heavily constrained by the regulatory restriction and granting of franchises--the logic of the original idea was turned upside down.
Simply consider that when John Moody published his first ratings in 1909, or when Poor's Publishing Company published its first ratings in 1916, or the Fitch Publishing Company issued its first ratings in 1924, they were not yet "nationally recognized." They all had to fill a market need successfully and compete their way into becoming widely used by financial market participants. But this competitive market opportunity and test is no longer available to a potential John Moody of today.
Today the NRSRO designation is embedded in the rules of multiple financial regulators, not only the SEC, as an essential restriction on investment and financial market decisions and an important factor in calculating capital requirements. No matter what the market might think, there is only one way to get to be an NRSRO: to be approved as one by the SEC. Many commentators have pointed out that this involves an insuperable logical problem or Catch 22.
The SEC's 2003 Concept Release on rating agencies states that to be approved as an NRSRO:
The single most important criterion is that the rating agency is widely accepted in the U.S. as an issuer of credible and reliable ratings by the predominant users of securities ratings.
Nothing is more obvious than that under the current regulatory regime, it is not possible to be "widely accepted in the United States by the predominant users of ratings" unless you are already designated an NRSRO. But you cannot become an NRSRO unless you are already widely accepted by the predominant users! Here's a pretty problem in bureaucratic reasoning.
The same SEC Concept Release observes:
Some commentators believe that the NRSRO designation acts as a barrier to entry into the credit rating business.
There is no doubt that these commentators, which include the Department of Justice, are correct.
In March 1998, the Department of Justice submitted the following statement of position to the SEC:
The Department opposes . . . a "recognition" requirement, as currently formulated. According to this requirement, a rating organization would have to be recognized as an issuer of credible and reliable ratings by the predominant users of securities ratings in the U.S. in order to receive NRSRO status. The adoption of such a criterion is likely to create a nearly insurmountable barrier to de novo entry into the market for NRSRO services. For this reason, the recognition requirement is likely to be anticompetitive and could lead to higher prices for securities ratings than would otherwise occur.
The Department of Justice commented further that "the recognition requirement might preclude the procompetitive benefits of de novo entry by smaller firms," and that "the commission's recognition requirement creates a problematic barrier to entry in an industry that is already highly concentrated."
This position was certainly unambiguous and, it would seem, convincing. Yet five years later, the SEC Concept Release repeated the same anticompetitive, Catch-22 language to which the U.S. Department of Justice objected.
As a matter of corporate strategic analysis, the rating agency business has significant inherent barriers to entry in any case, including the need to establish reputation, reliability, and integrity; the prestige factor involved in the purchase of opinions and judgments; and natural conservatism in institutional risk management policies. To add to this a "distortionary entry restriction regime" (to use Professor White's phrase) insures a noncompetitive outcome.
Recommendations for Action
Accuracy in Labeling. The term NRSRO, with its implication that it represents a test of market acceptance--which it did thirty years ago but does not now--should be dropped altogether. It is clear that what NRSRO really means today is simply "SEC-approved rating agency." If the SEC continues to require its own approval of rating agencies for regulatory purposes, the designation should express the fact, viz.: "SEC-approved rating agency." This terminology would be accurate; it might also increase the incentive to adopt clear criteria for approval and a standard approval process.
Eliminate the Catch 22. If the SEC continues to require its approval of rating agencies, the anticompetitive criterion of having to be "widely accepted in the U.S. by the predominant users of securities ratings" in advance of having the chance to gain wide market acceptance should be eliminated. This was the recommendation so clearly made by the Department of Justice.
Independent Action by Other Regulators. Other regulators do not have to apply the SEC's rating agency limitations on competition. They could develop their own approvals--for example, "FDIC-Approved Rating Agency," or likewise for any other regulator, mutatis mutandis. This would add a healthy element of multiple regulatory perspectives, as well as encourage market competition.
Encourage New Entry by Specialized Competitors. In the consideration of rating agencies by any regulator, the approval of firms providing ratings for specialized purposes (as the SEC has occasionally done in the past) should be encouraged. Such specialization could usefully be an industry--for example, financial institutions; a country--for example, Japan; an instrument type--for example, residential mortgage securities; or any other logical domain defined by competence and knowledge.
This approach would allow new entrants to add competition where they are best able, to serve defined market segments with potential quality improvements or innovations, to have the opportunity to demonstrate their value to the market, and to grow organically if they succeed in gaining acceptance.
Financial Reporting by Rating Agencies. If the SEC continues to require its approval of rating agencies, it should require them to disclose regular financial statements on their ratings business. This would entail no change for Moody's, which files statements with the SEC already, but the addition of all approved ratings firms would allow the market to test the theory proposed by this paper: that an exceptional level of profit is induced by the government's structuring of the sector as a cartel. As new entry occurs, an equally interesting test of the effects of competition would be possible.
The Best Case. In the truly procompetitive and best case, not only would the term "NRSRO" be dropped, but the regulatory requirement of designation of approved rating agencies itself would be eliminated. That requirement has produced unintended effects never imagined when it was introduced in 1975, and it is time for it to retire.
In its place, the responsibility to choose among rating agencies and their services would belong to investors, financial firms, securities issuers, creditors, and other users of ratings--in short, to the market. Imagine that! Every firm for which it is relevant, especially those that purvey securities or deposits to the public, should have among its management and financial policies definitions of how it will use credit ratings and whose ratings can be used for what. These policies should be appropriately disclosed and would be reviewed by auditors and regulators, as appropriate.
Under these desirable circumstances, a competitive market test will determine which rating agencies turn out to be "widely accepted by the predominant users of securities ratings," and competition will provide its normal benefits of better prices, innovation, customer choice, and efficiency.
1. "Borrowers Find System Open to Conflicts Manipulation," Washington Post (November 22, 2004).
3. "SEC Standards for Designating Nationally Recognized Credit Rating Organizations," Statement of the Shadow Financial Regulatory Committee, No. 183 (December 9, 2002).
4. House Committee on Financial Services, Subcommittee on Capital Markets, Insurance, and Government-Sponsored Enterprises, "Rating the Rating Agencies," hearing on April 2, 2003.
5. "The SEC's Other Problem," Regulation (Winter 2002-03).
6. See the statements of Barron Putnam and Sean J. Egan to the House Subcommittee on Capital Markets, Insurance, and Government-Sponsored Enterprises, "The Ratings Game," hearing on September 14, 2004.
7. Securities and Exchange Commission, "Concept Release: Rating Agencies and the Use of Credit Ratings under the Federal Securities Law," 2003.
8. "Comments of the U.S. Department of Justice before the Securities and Exchange Commission," March 1998. Emphasis added.
Alex J. Pollock is a resident fellow at AEI.