On December 3 the Securities and Exchange Commission approved tighter regulations on the credit rating agencies, hoping curbs on conflicts of interest will prevent the kind of ratings-grade inflation that played such a key role in the credit crisis. The SEC voted to require greater disclosure and to ban agencies from rating securities they have helped issuers create.

While the ratings firms applauded the moves, reform advocates were disappointed. The changes fell far short of remedies initially proposed by the SEC in June and supported in a December 1 statement by the Financial Economists Roundtable, a 15-year-old group of top economists from around the world that meets every year to tackle economic issues. By omitting two critical elements of the June proposal, the SEC pulled the teeth that would have made the regulations effective, according to the FER.

“It’s a quarter of a loaf at best. The one thing they did was the one thing that we thought was unenforceable,” says Wharton finance professor Richard J. Herring, the roundtable’s executive director, referring to a ban that would prevent agencies from rating securities for which they have provided paid advice to the issuer.

The FER’s December 1 statement urged stronger public disclosure of data used in the agencies’ computer models, less use of ratings in governmental regulations and a requirement that ratings on riskier types of securities state a margin for error. Indeed, most of these ideas had been tentatively proposed by the SEC staff in June, but were dropped, diluted or delayed in December. FER members met July 12-14 in Glen Cove, N.Y., and refined their statement through the summer and fall.

Wharton finance professor Marshall E. Blume, one of several Wharton faculty members of the roundtable, describes the SEC’s June proposal as “very bold,” but says the “the final document is very limited.”

In addition to the ban on rating securities an agency helped create, the SEC barred raters from accepting gifts worth more than $25 from clients. The agencies also will have to disclose statistics on their upgrades and downgrades for each type of asset they rate, and they must describe the steps they take to verify information used in ratings. Additional disclosure requirements are meant to shed light on potential conflicts of interest. The three main ratings firms — Moody’s Investors Service, Standard & Poor’s and Fitch Ratings — supported the new rules, which will take effect over a number of months.

Tougher Rules Opposed on Wall Street

Wall Street and the ratings firms had opposed many of the stronger provisions as too inhibiting. SEC Chairman Christopher Cox said the agency would continue to study ideas like special ratings treatment for volatile securities and reducing use of ratings in regulations. A June 2007 law gave the SEC authority to regulate the agencies, and it then did a 10-month study which found “significant weaknesses in ratings practices.”

Debate over regulating the ratings industry, dominated by the three largest firms, is far from over. Top Democratic members of Congress and President-elect Barack Obama have talked of the need for improving practices on Wall Street.

Regulators, lawmakers, economists and watchdog groups — and the ratings industry itself — are looking for ways to prevent the kind of misjudgments that led the ratings agencies to issue top-quality grades to mortgage-backed securities that later collapsed as the housing bubble burst.

One problem with the current system, according to Blume: Investors have become overly reliant on ratings, using them as a substitute for their own analysis. “Let’s say you have two triple-A bonds. One has a lot higher yield than the other. The investors or their board or the people responsible should have asked why it has a much higher yield, and those questions were not being asked.”

One factor was lack of experience with securities backed by new forms of loans, such as subprime and other types of exotic mortgages, which became popular early in the decade. Not only were the loans untested, they were issued to borrowers who could not have qualified under the more stringent guidelines of the past, making it hard to judge how the loans would perform in the kind of economic downturn that ensued.

In the easy-lending climate that followed the dot-com stock collapse at the start of the decade, other factors were at work as well. One was the rise of private-label securitization, the bundling of mortgages into bonds sold to investors. Though securitization had long been done by government-created companies like Fannie Mae and Freddie Mac, it was now being done by companies operating with looser practices and little or no government regulation. Investors were attracted by the high yields paid by these securities, and they relied on the ratings agencies to assess the risk that homeowners would stop making monthly payments and drive down the securities’ values.

The ratings firms gave many of the new mortgage securities highly desired AAA ratings, making them look as safe as other investment-grade securities such as corporate bonds, and making the mortgage securities attractive to investors limited by law or policy to owning only top-rated securities. That made lenders even more eager to issue exotic mortgages. When the house-price bubble burst, the economy began to sink and rate resets on adjustable-rate loans raised borrowers’ monthly payments, homeowner defaults soared and many of the newer mortgage securities fell to a fraction of their original value.

Ratings Inflation

Today, most experts agree that far too many mortgage securities received investment-grade ratings they did not deserve. The ratings industry primarily blames the underlying loans’ short track record, which led to flaws in its computer models of risk, compounded by the unprecedented nationwide housing downturn. But others say common business practices prodded the agencies to act with bias and blinders.

“We viewed the role of the ratings agencies as having been significant in the problem of subprime-related debt, and even more so in making it a systemic problem,” Herring says of the FER analysis. Errors in rating mortgage securities undermined the agencies’ credibility with other types of ratings, making the credit crunch worse, argues Herring, who is also co-director of the Wharton Financial Institutions Center.

The first ratings agencies, or statistical rating organizations (SROs), were formed more than 100 years ago to provide an objective, credible source of information on the risks of bonds issued by railroads and other U.S. companies. They made money by selling their findings to investors, and they did well if issuers’ default rates matched those predicted by the ratings. “Over time, the accumulation of reputational capital by successful SROs made entry difficult for new SROs,” the FER says in its statement. “The result is that two or three SROs have dominated the market for credit ratings, and did so long before the SEC began to designate particular SROs as Nationally Recognized Statistical Ratings Organizations in the 1970s.”

A turning point came in the 1930s, when federal and state regulators began to use ratings in rules for banks and other financial institutions. Some institutions, for example, are permitted to own only highly rated securities, and many capital requirements are keyed to ratings on the institution’s holdings. “The existence of such regulatory consequences was bound to intensify pressure on SROs to inflate the grades of lower-rated securities, because regulated clients routinely explore and develop ways of reducing their regulatory burdens,” according to the FER.

Other changes came more recently. “The spread of photocopying technology facilitated unauthorized reproduction of SRO rating manuals, which undermined the traditional user-pays revenue model,” the FER says. “SROs responded by shifting to a business plan in which the issuer pays for their services. This plan intensified SRO conflicts of interest with issuers. Issuers and underwriters actively shopped for ratings and were unwilling to pay for ratings they deemed too low.”

A mushrooming of elaborate forms of securities in recent years gave issuers ever greater amounts of leverage over the ratings agencies. “A further weakness in issuer-pays arrangements is that they undercut incentives to monitor and downgrade securities in the post-issuance market,” the statement says. “The re-rating of securities is usually paid for by a maintenance fee that is collected in advance from each issuer. Few issuers are eager to be monitored closely, especially when monitoring is apt to result in downgrades, and so it is not surprising that ratings are seldom downgraded until long after public information has signaled an obvious deterioration in an issuer’s probability of default.”

Herring notes that ratings downgrades for subprime residential mortgage-backed securities have dramatically exceeded those of corporate bonds carrying the same initial ratings. One survey showed that well over 50% of the subprime-backed securities that started with A ratings were subsequently downgraded, compared to well under 10% of the corporate bonds that started with the same ratings. Nearly all the corporates were downgraded just one notch to the next lowest category — A to BBB — while more than four fifths of the subprime securities dropped by three categories. “When they get downgraded they get downgraded by huge amounts rather than by very modest amounts,” Herring says.

To tackle the problem, the SEC in June proposed three remedies, all supported by the FER.

The first focused on public disclosures. It would, for example, have required the SROs to state how well their ratings had predicted defaults in various securities categories over the previous one, three and 10 years. It would have required release of data used in ratings models, and a discussion of how ratings are reviewed and modified and whether changes in models or procedures are applied to existing ratings retroactively.

This proposal also would have barred SROs from rating securities created with the SRO’s paid assistance. The FER agrees the dual roles create a conflict of interest but concluded there is so much back and forth in the normal rating process it would be “impossible for the courts to distinguish ratings services from advisory services in a definitive way.” This was one of the provisions adopted by the SEC on December 3.

The FER suggests that the SEC or Congress might also consider new disclosure requirements for issuers, to make public any data that could affect the health of each security it had issued.

The SEC’s second proposal would have imposed extra requirements on ratings for complex securities, such as those based on mortgages, to recognize their special risks. While an ordinary corporate bond is rated according to views about how well a single company will be able to shoulder its payments to bond holders, a mortgage security may depend on the abilities of thousands of heavily indebted homeowners to make monthly payments. Moreover, mortgage pools are typically sliced into tranches of varying risks. Prices of mortgage securities and other types of securitized debt can therefore be much more volatile than prices of ordinary bonds, subjecting ratings to greater margins for error.

“Introducing a differentiated [rating] scale is one way to alert investors that downside margins for error are much larger for securitized claims than for ordinary debt,” the FER said.

The SEC did not adopt this proposal.

Nor did the SEC adopt the third proposal, to reduce use of ratings in SEC regulations. Blume argues this was the most valuable proposal, because it would force securities investors to do more research on their own. Hence, an individual security might be scrutinized by multiple organizations rather than just the handful of ratings agencies. The FER said the move would “remove a major source of pressure for ratings inflation.” Incorporating ratings into regulations “enshrines those ratings agencies into law and gives the ratings agencies very strong power,” Blume said. “The ratings agencies have conflicts of interest. They are not always right. They make systemic errors, and that should not excuse the managers of, say, a mutual fund, from doing their own evaluation.”

Blume, Herring and the FER statement all express concerns that use of ratings in regulations is on the rise in many countries. The European Union, for instance, uses ratings in capital requirements, while ratings are incorporated into the Basel II international banking guidelines. “It really does bias the system toward ratings inflation,” Herring said, “and it’s getting more and more widespread and more and more pernicious.”

Though the SEC’s December 3 vote disappointed advocates of tougher rules, Herring thinks the debate is far from over, noting that President-elect Barack Obama’s transition team recently asked the FER for its views on ratings reform. “It does mean they are paying attention to these things.”