It could have been a scene out of a Frank Capra movie (“Mr. Smith Goes to Washington”). Individual senators, with no background in finance or the securities industry and no relevant committee memberships, took to the Senate floor last week and began to offer amendments to the “Restoring American Financial Stability Act of 2010” (the financial reform legislation that Senator Christopher Dodd, D-Conn., has been carefully orchestrating for the last year). Normally, such amendments from junior senators (think of Jimmy Stewart in his role as “ Mr. Smith”) would have been quickly rejected for their naiveté, and the Senate would have moved on. But instead, last week, the atmosphere was very different, and these amendments passed by overwhelming margins, despite the opposition of Senator Dodd.

What explains this? The two amendments—one by Senator Al Franken, D.-Minn., and the other by Senator George S. LeMieux, R-Fla.,—both addressed the credit rating agencies (everyone’s favorite villain), and both proposed bold and easily understood answers to a complex problem. After weeks of hearings and enforcement actions directed at Goldman Sachs and other investment banks, Congress had heard the grassroots reaction, and their constituents clearly wanted change (and maybe some punishment too). They voted accordingly.

The problem, however, is that the two amendments adopted by the Senate last week follow divergent and ultimately inconsistent paths. Although few, if any, defend the performance of the credit rating agencies, reformers fall into two opposed camps when it comes to prescriptions. One group (probably the more numerous) wants to address the conflicts of interest inherent in their “issuer pays” business model, on the premise that if the rating agency were not beholding to the issuer for its fee, it could perform as a legitimate gatekeeper.1 The alternative position is that there has been excessive reliance placed on the credit rating agencies and their role should be reduced.2 These latter critics argue that the rating agencies have been improperly delegated a regulatory licensing authority under which many categories of institutional investors cannot buy debt securities unless they have been rated “investment grade” by one of the rating agencies recognized by the SEC as “Nationally Recognized Statistical Rating Organizations” (or “NRSROs”).

Agreeing with the first critique that conflicts of interest were the key problem, Senator Franken proposed an amendment that would create a “Credit Rating Agency board” as a self-regulatory organization under the SEC. This board would select the credit rating agency for the issuer in the case of all initial ratings on structured finance products. Thus, the issuer would still pay the rating agency’s fee, but would not select the rating agency.

Senator LeMieux’s amendment seeks instead to end the quasi-governmental power accorded NRSRO rating agencies by amending those federal statutes that compel institutional investors largely to limit their purchases of debt securities to those that are rated “investment grade” by NRSROs. Under his amendment, these federal agencies—the FDIC, the Comptroller of the Currency, and the SEC—would instead be instructed to develop their own standards of credit-worthiness.

Endearing as this spectacle is of senators striving to outdo each other in their zeal to crack down on those responsible for the 2008 meltdown, these two amendments work at cross purposes. One reform is attempting to eliminate conflicts so that rating agencies can be trustworthy gatekeepers, while the other is seeking to downsize the overall role they play and force investors to rely on other sources of information. Worse yet, once NRSROs are stripped of the special benefit accorded their ratings, the major rating agencies may simply surrender their NRSRO designations and thereby outflank pending legislation that largely applies only to NRSROs.

Of these two paths to regulatory reform, which is more feasible? This column will attempt an assessment after first reviewing the latest empirical evidence on the performance of the credit rating agencies. Then, it will examine one other amendment that will be voted on next week. A growing number of senators, including some Republicans, are sponsoring legislation to reverse Gustafson v. Alloyd Co.,3 the Supreme Court’s 1995 decision that held §12(a)(2) of the Securities Act of 1933 inapplicable to private placements. Because §12(a)(2) provides for negligence-based liability and shifts the burden to the defendant to demonstrate its due diligence, the rationale for reversing Gustafson is that its negligence standard would force investment banks selling CDO offerings to provide fuller disclosure and conduct greater factual investigation. Although its likely fate or that of even more controversial attempts to enact a broad fiduciary duty standard for broker-dealers cannot be easily predicted as this column goes to press, all are indirect responses to the Goldman Sachs congressional hearings and the SEC’s complaint.

A. The Latest Evidence

Were the credit rating agencies misled by investment banks eager to market Collateralized Debt Obligations (“CDOs”), as New York Attorney General Andrew M. Cuomo has recently alleged? Or did they willingly acquiesce, manipulating their models to produce inflated AAA ratings? Recent studies find that CDOs were responsible for the greater part of the damages to the banking sector,4 and most CDO notes (at least those issued prior to mid-2007) were rated AAA.5 Given that the underlying collateral for most CDOs was subprime mortgages, it may seem hard to understand how the majority of the notes issued to investors could be rated AAA, even with the subordination of the bottom tranches to the AAA tranche.6 Was it sound financial engineering or hocus pocus?

In a recent study of 916 CDOs issued between January 1997 to December 2007, two finance professors find that the rating agencies deviated from their quantitative models and adjusted their ratings upward, making discretionary adjustments to create a larger AAA tranche.7 The result of these discretionary adjustments was to increase the AAA rated tranche of the CDO by on average 12.1 percent over the size implied by the rating agency’s own model. More surprisingly, they find that “only 1.4 percent of AAA CDOs closed between January 1997 and March 2007 met the rating agency’s reported AAA default standard.”8 In the absence of adjustments, they estimate “that the AAA tranches would have been rated BBB on average.”9 The aggregate result of these adjustments was to inflate the value of these 916 CDOs by some $86.22 billion. Although the rating process may necessarily involve a subjective element, it raises eyebrows to learn that they find that 84 percent of all adjustments were positive.10

This aggregate data dovetails with the testimony in April of former rating agency personnel before the Senate Permanent Subcommittee on Investigations. Eric Kolchinsky, Moody’s former managing director in charge of rating subprime mortgage-backed CDOs, testified that managers at Moody’s were expected to build, or at least maintain, market share and that “[i]t was an unspoken understanding that loss of market share would cause a manager to lose his or her job.”11 Adjustments then were a necessity for the manager intent on survival in a market that became significantly more competitive with the rise of Fitch Ratings after 2000. Moreover, as the structured finance market became the principal profit center for the rating agencies, they also confronted for the first time a very concentrated market. Precisely because a limited number of underwriters dominated subprime mortgage-backed securities, the rating agencies were not in their traditional posture of dealing with thousands of corporate issuers who came only intermittently to the bond market. Instead, they were confronting a very concentrated market in which a few investment banks could credibly threaten to move their business if they did not receive the inflated ratings that they wanted.

This evidence suggests that conflicts of interest are the primary problem. Although the proponents of the “regulatory license” school view the ratings agencies as possessing de facto governmental power, such a quasi-governmental agency actually has less incentive to inflate its ratings, because it has a protected market. Only an agency facing loss of market share has strong economic incentives to adjust its ratings upward. Moreover, the same rating inflation occurred in Europe where the NRSRO concept is unknown. On this basis, the critical priority should be addressing the conflict of interest problem.

B. The Franken Amendment

The Franken proposal amends §15E of the Securities Exchange Act of 1934 to create a “Credit Rating Agency Board” (the “board”) as a self-regulatory organization under the SEC. Its jurisdiction is limited to ratings on “structured finance products,” which term it authorizes the SEC to define, but clearly the term includes all asset-backed securities. Hence, ratings on corporate debt would not be affected. NRSRO agencies would apply to the board to become eligible to provide initial ratings on structured finance products. An issuer seeking an initial credit rating on a structured finance product would be required to submit a request to the board, which would select the rating agency for the initial rating (but the issuer would then be free to hire any additional rating agency it wished). The board is expressly authorized to use a lottery or a rotating assignment system, but it is prohibited from considering the issuer’s preference. The rating agency so chosen may specify its fee, but is limited to charging a “reasonable fee,” as determined by the SEC. Presumably, this both protects the issuer from excessive fees and the investing public from implicit bribery.

Potentially, the board would be administering an enormous patronage system. The fact that the rating agency so chosen could charge any fee so long as it does not exceed a “reasonable” ceiling (as determined by the SEC) means that there would be little, if any, competition in this market.

A key problem with this system is that the rating agencies participating in it will have little incentive to compete in terms of quality or their investment in research. So long as they are as competent as their rivals, they will keep their position in a shared monopoly.

What would have been a superior structure? A competitive “subscriber pays” rating system might have been created, if instead of regulating the issuer, Congress had mandated that institutional investors could not purchase debt securities unless they first obtained a rating from a rating agency not paid by the issuer. This would have created a “subscriber pays” market that would have been competitive. But it would have required Congress to regulate the users of financial information and require them to pay, and this may have been less politically feasible.

Unfortunately, a truly competitive and unconflicted marketplace in which rating agencies compete based on the quality of their research will not develop either on its own or under the Franken Amendment. Financial information is a public good, and it is virtually impossible to keep non-subscribers from learning the ratings assigned by a subscriber-paid ratings agency. Thus, because free riders can learn and rely on financial information that is costly to develop without paying, credit rating agencies cannot prosper under an unregulated “subscriber pays” system. In this light, the Franken Amendment may create a second-best substitute, and it likely will incentivize some new entrants to enter the business.

The operative question for the “Big Three” rating agencies is whether to apply to the board and receive their share of initial ratings—or instead focus on being paid by the issuer to provide a second rating. My sense is that Moody’s and S&P will prefer to solicit issuers and will not seek a share of the initial rating business. Issuers have long secured two ratings, but they may be less willing to pay for a third. Thus, if one rating is assigned by the board and another is purchased, both Moody’s and Standard & Poor’s face some likely loss of market share, as issuers may no longer hire both of them.

One last thought on this approach: if it is successful, one wonders how long it will take before institutional investors begin to suggest that auditors could also be hired in a similar fashion.

C. The LeMieux Amendment

The LeMieux Amendment mandates that the FDIC, the Comptroller of the Currency and the SEC no longer rely on investment grade credit ratings, but instead must develop their own “standard of credit-worthiness.” This is no small task, and it may require regulations of telephone book-length.

More importantly, a hidden crisis has been unintentionally precipitated by the LeMieux Amendment. Because, under it, the NRSRO designation no longer confers much benefit or regulatory power on a credit rating agency, the major credit rating agencies could easily decide to surrender their NRSRO license and thereby escape most of the controls and restrictions in the pending legislation, which largely apply only to NRSROs. The only real cost would be that they would then not be eligible for selection by the Credit Rating Agency Board as an initial rater. But if Moody’s and S&P both ceased to be NRSROs, issuers would likely still hire them because the new rating agencies chosen by the board would generally be unknown. The result is a potential embarrassment for Congress, as a major legislative effort might be easily outflanked. One answer is, however, available under the House legislation, which denies existing NRSROs the right to deregister so long as their revenues from ratings exceed $250 million annually. Obviously, the conference process in finalizing this legislation will be critical.

The more ideological proponents of downsizing the role of credit rating agencies might still approve of an outcome under which Moody’s and S&P simply deregistered. They have suggested that “investment grade” credit ratings should not be necessary, because investors might better rely on credit spreads. This is a very academic argument that ignores both how little sophistication some smaller institutions have and how inefficient, thin, and subject to manipulation some debt markets are. Moreover, there is a major conflict of interest problem lurking here too. Smaller pension funds, union funds, school boards, charitable bodies, hospitals and colleges—all tend to lack experienced debt specialists on their financial staffs and cannot easily afford to hire them. In the future, they may also face requests to buy the debt securities of small and affiliated entities that are not truly creditworthy. These occasions did not arise in the past because these institutions had to buy “investment grade” debt, and, for all their sins, the credit rating agencies did not give investment grade ratings to “fly by night” companies. But those “fly by night companies” may now become able to peddle their debt offerings to smaller institutions, and conflicts of interest lurking in the background may explain the purchase decision.

In this light, a superior compromise would have been to continue to allow “investment grade” ratings to supply the governing standard for the purchase of corporate bonds and notes, but delete the references to investment grade only in the case of “structured finance products.”

Ultimately, sensible regulation of the credit rating agencies should recognize that rating agencies do have two critical advantages over alternative approaches:

First, specialization is efficient. It is more cost-effective for a rating agency to hire or train specialized experts than for each large institutional investor to do this on its own. Inefficient duplication results if all institutional investors hire their own debt specialists.

Second, there are economies of scale in the production of information. Asking each institutional investor to do its own credit analysis will produce a sub-optimal level of research and inferior credit models.

For these reasons, public policy should focus on curbing conflicts of interest and not seek to downsize or eliminate credit rating agencies.

D. The ‘Gustafson’ Repeal

Senator Carl Levin, D-Mich., has offered an amendment co-sponsored by senators Tom Coburn, R-Okla., Jack Reed, R-R.I., and Ted Kaufman, D-Del., to “restore Congressional intent that the statutory term ‘prospectus’ apply to not only public, but also private offerings of securities.” It thus seeks to reverse Gustafson v. Alloyd Co. Inc.,12 which found §12(a)(2) of the Securities Act of 1933 to be inapplicable to private placements because that section proscribes material misstatements or omissions “by means of a prospectus.” Ignoring the broad statutory definition of prospectus in §2(a)(10), the five-justice majority looked instead to §10 of the act, which specifies the information which must be set forth in a “prospectus” used in a public offering. Thus, they found that private placements were exempt from §12(a)(2). This was a bizarre rationale that united Justices Scalia, Thomas, Ginsburg and Breyer in separate dissents. (In truth, the contract of sale at issue in Gustafson could also have been found not to constitute a prospectus even under §2(a)(10), but this narrower interpretation would have had little impact).

To reverse Gustafson, the proposed amendment inserts the parenthetical phrase “(as defined in §2(a)(10) of this title)” after the word “prospectus” in §12(a)(2) and then modifies the definition of “prospectus” in §2(a)(10) to include securities and offerings that are exempt “pursuant to the provisions of sections 3 or 4″ of the act.

This language may be broader than necessary because it will apparently pick up secondary market transactions, including those made pursuant to the “Section 4(11/2)” exemption. Nonetheless, it will clearly cover Rule 144A sales by a dealer or distributor. Procedurally, the advantage to a plaintiff from using §12(a)(2) is that it need not allege or prove scienter and thus avoids the pleading rules of §21(d)(b) of the Securities Exchange Act. Defendants are given to a due diligence defense by §12(a)(2), but this requires them to “sustain the burden of proof” of showing that they “did not know and in the exercise of care could not have known of such untruth or omission.”

Although the fate of this amendment cannot not be safely predicted, it is clearly congressional payback, aimed at investment banks that packaged toxic deal after toxic deal over the last decade. Both the Franken and LeMieux amendments could be modified at the conference stage when the House and Senate bills must be reconciled. But because both passed with over 60 -vote majorities, it is unlikely that they will disappear. Their combined impact, however, is to give the major rating agencies a strong incentive to deregister as NRSROs (if they can). In addition, an even greater danger for the rating agencies is that the House bill, passed in December 2009, authorized a new private cause of action against credit rating agencies based on a gross negligence standard. If it were to pass (which still seems unlikely), it is doubtful that the major credit rating agencies would continue to rate structured finance products. Indeed, if this provision were in fact to pass, the proponents of downsizing the credit rating agencies would ironically win an unintended and inadvertent victory, as rating agencies fled the field of structured finance. While the future outcome is still murky, the law of unintended consequences seems to be applying once again with its usual perverse vengeance.

John C. Coffee, Jr. is the Adolf A. Berle Professor of Law at Columbia University Law School and director of its Center on Corporate Governance.

Endnotes:

1. For a discussion of the “reputational capital” model, which assumes that financial gatekeepers, such as auditors, investment banks, and credit rating agencies, pledge their reputational capital as in effect collateral for their professional opinions, see John C. Coffee, Jr., GATEKEEPERS: The Professions and Corporate Governance (2006). Suffice it to say that a variety of special factors—the oligopoly shared by the three major rating agencies, their virtual immunity from liability, and the high costs of entry into the field—may make this model less applicable to credit rating agencies.

2. Professor Frank Partnoy has been the leading advocate of this position. See Frank Partnoy, “The Siskel & Ebert of Financial Markets?: Two Thumbs Down for the Ratings Agencies,” 77 Wash. U. L. Q. 619 (1999); see also Partnoy, “Overdependence on Credit Ratings Was a Primary Cause of the Crisis,” (available at http://ssrn.com/abstract=1430653).

3. 513 U.S. 561 (1995).

4. See Markus Brunnermeier, “Deciphering the Liquidity and Credit Crunch 2007-2008,” 23 J. of Econ. Persps. 77 (2009) (emphasizing role of CDOs in magnifying the impact of the credit crisis).

5. See John M. Griffin and Dragon Yongjun Tang, “Did Subjectivity Play a Role in CDO Credit Ratings?” (2009) (available at http://ssrn.com/abstract=1364933).

6. For this view from a financial economist, see John Hull, “Credit Ratings and the Securitization of Subprime Mortgages,” Paper presented at the Financial Reserve Bank of Atlanta 2010 Financial Markets Conference, “Up From the Ashes: The Financial System After the Crisis,” May 11, 2010.

7. See Griffin and Tang, supra note 5, at 1.

8. Id. at 4.

9. Id. at 5.

10. Id. at 4.

11. See Statement of Eric Kolchinsky before the Senate Permanent Subcommittee on Investigations, Hearing on “Wall Street and the Financial Crisis: The Role of Credit Rating Agencies.” April 23, 2010 at p. 1. He noted that in mid-October 2007, when Moody’s discovered that its market share in subprime bond ratings had “dropped from 98% plus to 94% in the third quarter,” it set off an internal crisis and his superiors “demanded an accounting of the missing deals.” Id. at 3.

12. 513 U.S. 561 (1995).