The financial crisis has provided an unexpected crash course on credit rating agencies, such as Moody’s, Fitch and Standard & Poor’s, which stamped triple-A ratings on a broad spectrum of subprime mortgage securities — implying that they were nearly risk free — then back-pedaled when the debt collapsed, taking with it the global economy.
In the United States, the resulting clamor for ratings agency reform led to a U.S. Securities and Exchange Commission (SEC) proposal that sought to mitigate conflicts of interest and enhance disclosures, require ratings firms to differentiate ratings for structured products, and nearly eliminate the role of ratings in SEC regulations. But the final rules the SEC adopted last December were far less stringent than the ones it had proposed six months earlier, in June.
Europe, on the other hand, became the world’s most stringent regulator of ratings agencies when the European Union and the European parliament approved a package of ratings agency rules on April 23. The question now is whether the U.S. will follow Europe’s lead and build on the new regulations, or go a different route.
“After proposing some radical and promising reforms of the regulation of the ratings industry, the SEC settled in December on the two least consequential and least controversial proposals,” says Wharton finance professor Richard J. Herring. The new SEC rules separated consulting from rating — a measure Herring believes is unenforceable — and barred ratings agencies from accepting gifts over $25 from the companies whose debt they rate.
“The EU responded with a potentially much more consequential set of reforms,” Herring notes. “In part, this was probably motivated by a lack of confidence that the SEC reforms would make much difference, and in part it may have been motivated by concern that Europe is the only major area without an active, home-grown ratings industry…. In principle, the SEC had oversight of the ratings agencies, but one cannot help but share the European view that they bungled it badly.”
The new EU regulations require credit rating agencies operating in Europe to register with the Committee of European Securities Regulators (CESR) and comply with rigorous rules. The rules include these, which state that the agencies:
- May not provide advisory services.
- Must disclose the models, methodologies and key assumptions upon which ratings are based.
- Must differentiate the ratings of complex products with a specific symbol.
- Must publish an annual transparency report.
- Must have at least two directors on their boards whose salary does not depend on the ratings agency’s business performance.
- Must create an internal function to review the quality of their ratings.
“It’s not entirely clear how this [set of rules] will work in practice, but it makes it possible for European standards to be quite different from standards in the rest of the world,” says Herring, who worries that a gap in regulations between countries could “fragment world capital markets. If different countries in the EU come up with different regulations from each other, or from us or from Japan, it will be more difficult to compare securities across countries, and thus will become a barrier to the integration of world capital markets.”
According to Wharton finance professor Franklin Allen, a long-term disparity of regulations between the U.S. and Europe could potentially affect the world’s bond underwriting business, which is heavily dependent on the agencies’ ratings. “If we do nothing, would underwriting business come here?” Allen asks. “Coca-Cola can issue bonds anywhere in the world. If it can do it [more easily] here, maybe they’ll come here…. So if we do nothing, maybe it would help our underwriters.” On the other hand, he adds, “It could go the other way, too. If people don’t trust the ratings here, Coca-Cola could go to Europe to get its ratings. It could go both ways.”
But will the U.S. follow the EU in beefing up credit rating agency regulations?] On April 15, just days ahead of the European vote, the SEC held a roundtable discussion of ratings agency reform. But afterward, many issues remained unresolved. Key among them: how to eliminate conflicts of interest; how much information ratings agencies should be required to disclose; whether ratings agencies should develop a special rating for structured products, as the EU now requires; whether ratings agencies should be held liable for their ratings; and how much of a role ratings should play in government regulations.
“I think that the EU reform is a logical and important step in the right direction,” says Wharton insurance and risk management professor Kent Smetters. “But the EU reforms alone don’t go far enough for the U.S. A key problem in the U.S. is that we have an oligopoly for the provision of ratings. A lot of people don’t realize it, but U.S.-based ratings agencies are actually sanctioned by the government through its regulations on acceptable commercial paper for financial institutions, as well as other rules. We need to encourage more competition in the provision of this type of information.”
In some ways, banks in the U.S. have little choice but to pay attention to ratings agencies because U.S. regulations essentially require them to do so. When John Moody published the first publicly available bond ratings in 1909, investors paid to get the information in thick ratings manuals, and the business was driven by supply and demand. But the relationship between the bond market and ratings agencies changed drastically in the 1930s, notes New York University economics professor Lawrence J. White in public comments to the SEC on April 15.
“Eager to encourage banks to invest only in safe bonds, bank regulators issued a set of regulations that culminated in a 1936 decree that prohibited banks from investing in ‘speculative investment securities’ as determined by ‘recognized rating manuals,'” White wrote. “Banks were restricted to holding only bonds that were ‘investment grade’ — e.g., bonds that were rated BBB or better on the S&P scale….”
Consequently, “banks were no longer free to act on information about bonds from any source that they deemed reliable,” White wrote. “They were instead forced to use the judgments of the publishers of the ‘recognized rating manuals’ (i.e., Moody’s, Poors, Standard [prior to their merger] and Fitch).”
The SEC “crystallized the rating agencies’ centrality in 1975” when it created a new category of ratings agency called an NRSRO — or “nationally recognized statistical rating organization” — and grand-fathered the three big ratings agencies into the new special status, White says. Around that same time, the three large ratings agencies changed from the traditional “investor pays” model to an “issuer pays” model — meaning that the company or entity issuing the bonds also pays the ratings agency for the rating. The dominance of the “issuer-pays” model continues to this day.
The conflicts of interest inherent in the “insurer-pays” model have provided much fodder for debate.
“There were a lot of discussions on who should pay for ratings” during the April 15 roundtable, says Wharton professor of insurance and risk management Anastasia V. Kartasheva. “On one side, the economics are very simple. If the issuer pays for the ratings, obviously the ratings agency has an incentive to inflate the ratings. But on the other side, if you look at the market participants — for example, investment banks, insurance companies, other financial intermediaries — they are on both sides of the market. So they are both issuers and investors.”
Kartasheva says if ratings agencies switched back to an investor-pays model, there might be less disclosure of ratings, because investors would not want to share something they had paid for. And a large investor, like a hedge fund, could influence a ratings agency. “So you just create a different type of conflict,” Kartasheva says.
James C. Allen of the CFA Institute, a Charlottesville, Va.-based association of investment analysts and financial advisors, agrees. “The implication is that if only investors were paying for it, there wouldn’t be conflicts of interest. That’s just not the case,” says Allen (no relation to Franklin), who is director of capital markets policy at the CFA Institute’s Centre for Financial Market Integrity. For example, if a large investment firm didn’t want a ratings agency to downgrade U.S. Treasuries, it could threaten to cancel its subscription to one ratings agency and go to another, Allen says.
“If you start pushing it out to the investors, you could get a situation where you have one investor who makes up a fairly significant piece of [an agency’s] revenues,” Allen notes. “Essentially, there are always going to be conflicts of interest. So the best way to deal with them is to disclose them [and be] as forthright as possible.”
Government regulations will not necessarily eliminate all conflicts of interest, experts say. “If governments become too deeply involved in the rating process, there would be considerable concerns about bias,” Herring says. “For example, imagine a French registered subsidiary of S&P rating a French state-owned corporation, or even a sovereign issue by France.”
Wharton’s Franklin Allen points out that governments, too, issue bonds that ratings agencies must evaluate. “One of the interesting issues that doesn’t come up in the discussion: Ratings firms also rate sovereign debt,” says Allen. That fact was driven home recently when Standard & Poor’s warned Britain that it may lose its triple-A rating, sending tremors through global bond markets. As countries around the world have taken on more debt to shore up faltering economies, several have seen their ratings fall: Ireland and Spain recently lost their triple-A rating, and Iceland has seen its rating drop from single-A plus to triple-B plus. Allen wonders if highly regulated ratings agencies could find themselves under pressure to keep ratings of government debt stable.
“Suppose Moody’s downgraded the U.S. from an AAA to an AA,” Allen says. “Moody’s might worry that it would somehow be punished for that [through stricter government regulations]. That’s a worrying issue, [and] another conflict of interest.”
Kartasheva suggests that “there are going to be conflicts of interest whenever people exchange money. That’s part of the game. You cannot eliminate it completely. Even if you nationalize ratings agencies, how can you make sure that [the government] would not force [them] not to downgrade companies such as GM, for political reasons — to keep employees in place [and so forth]. There is no way that you can eliminate conflict of interest. So the only way to go is just to have different sources of information.”
Expanding the amount of information in the market is tricky. The SEC has anointed just a handful of NRSROs, and the three incumbent ratings firms (Moody’s, Fitch and S&P) still dominate the market. There have been calls in the U.S. to force ratings agencies to disclose a greater amount of information about their ratings, in part to make it easier for new players to enter the market.
But how much information about the agencies’ models and methodologies should be disclosed? The new European regulations require credit rating agencies to “disclose to the public the methodologies, models and key rating assumptions used in the rating process.” In contrast, the SEC rules passed in December require only that ratings firms publicly disclose a random sample of 10% of their ratings within six months of issue. “Definitely, just the wording [of the European regulations] is much stronger” than that of the American rules, says Kartasheva. “Here we are discussing whether they should disclose models and what they should discuss. But [in Europe], they are saying they must disclose models.
“And that’s not necessarily a good thing, again, because if everyone is going to disclose models, then the models will tend to look like each other. So they are going to converge,” Kartasheva adds. “Ideally, you would like to create an industry where people have incentives to look for the best models to describe what is going on. But if everything becomes transparent, it will be easier to replicate. First, you have less incentive to invest in innovation. Secondly, it means that everyone will have the tendency to look the same…. And that’s one of the reasons why it’s still being discussed here, because people raised this issue.”
Along similar lines, a new European rule that requires ratings agencies to distinguish structured products with a special symbol is bogged down by controversy in the U.S.
According to Kartasheva, one reason for the pushback on such a change is that a new scale or special designations for structured products might cause confusion among investors, and could make it difficult to compare structured products with other types of securities. It could also be laborious for pension funds and wealth managers who may have ratings written into their investment guidelines. For example, a pension fund may be required to keep a certain amount of securities with an AA rating: How would a new rank for a structured security compare on that scale?
“Having a different scale is not a bad idea, but the transition and the matching between the two scales becomes less obvious,” she says. “Each scale becomes more informative, but the comparison between scales becomes more difficult.”
Allen of the CFA Institute says some banks balked at the idea, not only because of costs and confusion, but also because of the stigma such a symbol might carry. “They were worried that if they brought these things to the CFOs, they would say, ‘What’s that? Oh, structured finance? We’re not going to have any of that stuff!'”
Nervousness surrounding the accuracy of ratings has led some to suggest that ratings agencies should be held liable if their calls fall too far off the mark. “That’s another interesting point,” says Kartasheva. “If you implement this kind of rule, the best reaction that the agency can have is to just issue a rating that is very vague — for example, just one letter grade … and then no one can sue.” So an attempt to improve the quality of the ratings by holding ratings firms liable could end up making the ratings less informative and less accurate, she adds.
Many of the proposals that call for holding ratings agencies responsible for inaccurate evaluations have come from pension fund or money managers, Kartasheva points out. “Do they just want to put the blame on ratings agencies so they don’t have to take that much care or due diligence on the portfolios that they manage?”
Kartasheva says the real question people should be asking in the U.S. is not how to regulate the ratings agencies, but what role they should play in the American financial system. “Why do we all of a sudden want to regulate [so] much? And I think the reason for that, in part, is that these ratings are used a lot in regulation. That’s another big issue that definitely has to be addressed down the road: To what extent do we want to have these ratings used in regulation?”
Ratings have become less of an evaluation and more of a license since their use in banking regulations and contracts is so widespread, Kartasheva says. Regulations often call for an investment grade rating. If a bank has an investment grade rating, it will be required to hold less capital. But what exactly does an “investment grade” rating mean? It does not actually specify any risk measure or nail down an exact probability of default, she notes. And yet, certain ratings allow for more lenient regulatory treatment. “So on one side, the use of ratings provides a simple framework … but on the other side, I think that it might also create some moral hazard issues in the banking sector, because now you potentially have a rating to back you up. So it provides you with a license.” Ratings agencies should “be a part of the process but not the cornerstone of the process,” Kartasheva suggests.
Although the SEC originally proposed scaling back the use of ratings in banking regulations, the final rules adopted in December still allow their regulatory use.
“The final proposal did allow use of ratings in regulation, which I think is a bad thing,” said Wharton finance professor Marshall E. Blume. “You’re basing a regulation on a commercial organization’s view of the world. And I don’t think that’s a good way to do it…. It removes the incentive to do your own security analysis…. In the recent crisis, companies did not question why they earned more from subprime securities, because they had been stamped with a ‘safe’ rating.” He says such companies should have asked: “‘Well I’ve got a half a percent more — why?’ Most people didn’t do that. They relied totally on the regulations and ratings agencies.”
Herring agrees. “I continue to believe that the original SEC proposals were a superior way to go. The merit in these proposals is that it keeps the ratings business out of politics.”