Chapman Law Review
LIMITS OF DODD-FRANK’S RATING AGENCY REFORM
Copyright (c) 2011 Chapman Law Review; Claire A. Hill
The history of rating agency reform has not been inspiring. Until recently, it seemed stuck in an ever-repeating cycle of futility. A crisis would spur calls for reform, hearings would be conducted, the SEC would issue proposals and requests for comments, and ultimately, nothing would happen–until the next crisis, when the cycle would begin again. The Enron debacle, in which the rating agencies rated Enron’s debt investment grade until four days before Enron declared bankruptcy, did spur some action, including federal legislation and SEC regulations. Whatever else may be said about the Enron-spurred action, it failed to prevent rating agencies from rating low-quality securities as AAA. This misrating was an important cause of the recent financial crisis.
In response to the crisis, the Dodd-Frank Wall Street Reform and Consumer Protection Act of 2010 (“Dodd-Frank”) was passed. Dodd-Frank includes some reforms to the rating agency regulatory regime. In this article, I argue that while Dodd-Frank’s rating agency reforms are not bad, they also are not particularly good. They do not sufficiently address the core reasons why rating agencies gave such inflated ratings to subprime securities or why the agencies so grievously misrated other instruments, including Enron debt and debt involved in the Asian flu. I then explain the reasons for this conclusion, and make some suggestions for better rating agency reform.
I. Why Have the Rating Agencies Failed?
A. Two (Extreme?) Possibilities
Rating agencies have often been right in their ratings. Sometimes, however, they have been wrong, even spectacularly wrong. Why have they sometimes been so wrong? Consider the following two possibilities. The first is that the rating agencies are less than cutting-edge in their capabilities. Because of their privileged positions–thanks to government favoritism and market norms effectively requiring their use, something I will explain more in Part II–they have not had to fear large losses of market share. Because credit ratings agencies have not needed to compete vigorously on quality, they have not sought to excel; instead (in the more charitable version of this explanation) they have simply maintained some minimal level of competence. Rating agencies therefore have not needed to be very sophisticated in their financial analyses; they also have not needed to be particularly good at ferreting out fraud.
The second possibility as to why rating agencies have been so wrong at times is that they are corrupt. Because the agencies are paid by the issuers and the issuers can threaten to take their business elsewhere if they cannot get high ratings, the rating agencies have let themselves be bribed into giving high ratings even when such ratings are not warranted.
Many rating agency critics writing before the recent financial crisis seemed to believe the first explanation. Critics writing since the crisis seem to believe the second.
i. Before the Recent Financial Crisis
Notable rating agency fiascos before 2008 include: Enron (corporation’s extensive use of financial “techniques” that created a wholly false financial appearance), Orange County (ill-advised bets on interest rates using complex derivatives by county Treasurer Robert Citron, leading to bankruptcy), the Asian Flu (a cascade of recessions in the region leading to massive downgrading), Washington Public Power Supply System (“WPPSS,” pronounced “Whoops”) (cost overruns and other complicating factors led to default on bonds originally rated AAA), and National Century Financial Enterprises (bankruptcy after lying executives “deceive[d] investors and rating agencies about the financial health of NCFE and how investors’ money would be used”) .
Following are some criticisms of the rating agencies in these debacles attributing the bad rating agency performance to incompetence:
As to Enron: Joseph Lieberman, then-Chairman of the Committee on Governmental Affairs, said:
[T]he credit rating agencies were dismally lax in their coverage of Enron. They didn’t ask probing questions and generally accepted at face value whatever Enron officials chose to tell them. And while they claim to rely primarily on public filings with the SEC, analysts from Standard and Poor’s not only did not read Enron’s proxy statement, they didn’t even know what information they might contain.
Approximately one year after Enron, one commentator said: “[t]he rating agencies display the classic characteristics of an entrenched cartel . . . [t]hey’re lazy, unresponsive, and ultimately unhelpful. They tend to play catch-up, downgrading ratings only after financial weaknesses are revealed or ferreted out by more enterprising researchers.”
As to Orange County, Betsy Dotson, an assistant director with the Government Finance Officers Association said: “[m]any of our members have called us wondering where the rating agencies were. It’s a logical question . . . . Some of these losses may have been avoided if the rating agencies had spotted problems earlier.” Christopher Cox, then a U.S. Representative (and later SEC Commissioner), said: “[h]ow is it that the rating agencies, the SEC, the entire market, was unable to learn about what [Robert ] Citron was doing?”
As to the Asian flu, one common critique was that the rating agencies were caught unaware (or, as one article memorably said, “with their pants down.”) In the same vein is this critique of the rating agencies made in the aftermath of the WPPSS debacle: Moody’s Investors Service “simply did not recognize that credit analysis in this new situation required looking upstream to the impact on the venture’s sponsors.”
ii. Since the Financial Crisis Began
A stark contrast exists between the foregoing picture of rating agencies as hapless dunces and the picture painted of their behavior rating subprime mortgage securities, the securities whose decline in value precipitated the financial crisis. A critical fact is that most rating agencies, including Moody’s, Standard & Poor’s, and Fitch Ratings (sometimes colloquially known as the “Big Three”), are paid by the issuers whose securities they rate, creating a conflict of interest. A search on Google.com for “rating agency conflict of interest” on April 11, 2011 yielded approximately 688,000 hits. Paul Krugman apparently subscribes to this view. He wrote a column entitled “Berating The Raters” in which he says, “[c]learly the rating agencies skewed their assessments to please their clients.” Joseph Stiglitz also emphasizes rating agency conflicts. Stiglitz noted that
[t]he incentive structure of the rating agencies also proved perverse. Agencies such as Moody’s and Standard & Poor’s are paid by the very people they are supposed to grade. As a result, they’ve had every reason to give companies high ratings, in a financial version of what college professors know as grade inflation.
Former employees of rating agencies, especially Moody’s, have given powerful testimony blaming the misratings involved in the financial crisis on this conflict of interest. One Moody’s employee, Eric Kolchinsky, testified that
[i]n my opinion, the cause of the financial crisis lies primarily with the misaligned incentives in the financial system. Individuals across the financial food chain, from the mortgage broker to the CDO banker, were compensated based on quantity rather than quality. The situation was no different at the rating agencies.
B. The (Obvious) Third Possibility
The foregoing indicates an interesting, and largely unnoticed, shift in the reasons given by commentators as to why rating agencies have sometimes spectacularly misrated. Are rating agencies dunces, corrupt, or both? I have argued elsewhere that the agencies are neither, although both their quality and independence need significant improvement. As to the former, I argued that the agencies were not cutting-edge, given that they didn’t have to compete vigorously for business. I have disputed, though, the argument that their ratings provide no information whatsoever–that the agencies are complete dunces. Government favoritism and established market practices allowed rating agencies to be slackers (and/or dunces), but only up to a point. Past that point, markets, and probably government, would have intervened.
I have also argued that the agencies were generally not wholly self-consciously corrupt. Rather, they engaged in considerable self-deception and reckless, if not willful, blindness, similar to the self-deception practiced at many investment banks and law firms. Moreover, the agencies are clearly capable of not being corrupted when they are being paid by issuers; their ratings for plain-vanilla debt, for which they are also paid by the issuers, are not generally considered to be compromised. But there are some clear indications of corruption during the recent financial crisis. A few former rating agency employees have described the single-minded pursuit of market share at the expense of credit quality. One notorious incident discussed below concerns Moody’s’ discovery that it had mistakenly applied its rating model so as to give too high a rating to a particular type of instrument. Moody’s’ response was to change the model so that the instruments would continue to have a high rating.
What accounts for the increasing perception and, to some extent, reality, of greater corruption in rating agencies post-Enron? Very simply, that there was a much higher payoff to being corrupt (or self-deceiving, or recklessly or willfully blind) than had previously been the case, and a higher cost to not being so. This is because rating agencies began to compete with one another for business. Previously, the government favoritism and market norms had created what was sometimes called a “partner monopoly” with Moody’s and Standard & Poor’s as the partners. Another agency, Fitch, gradually became an acceptable alternative. Thus, an issuer’s threat to take its business elsewhere became a credible one.
Moreover, the amount of business at issue is near-infinite. The ratings business originally involved rating debt securities for companies that wanted to raise money and found it worthwhile to do so. A “natural” and fairly low limit of such issuances thus exists. By contrast, structured finance securities are made up of other instruments; structuring the securities is itself a business entered into for profit. There is thus incentive and ability to structure more such instruments. The raw materials–the underlying instruments–originally were needed. But as structuring technology and investor appetites got more sophisticated, the need for raw materials ceased to be much of a constraint. Securitization securities could be crafted out of other securitization securities (“collateralized debt obligations cubed,” or CDO3s) and they could be “synthetics,” comprised of bets tracking the performance of “real” instruments. Thus, the only “natural” limit to the structure of securities was investors’ appetites, and investors proved voracious indeed.
II. Some Institutional Background
Some institutional background is in order. Who are these rating agencies that benefitted from government favoritism, and that are favored by markets? Why are there so few of them? What form did the government favoritism take? What effect did the reforms adopted in response to the Enron debacle have? Below, I briefly discuss some answers to these questions.
The ratings agencies at issue, the Big Three, are Standard & Poor’s, Moody’s Investors Service, and, to a lesser extent, Fitch Ratings. Standard & Poor’s and Moody’s both date back to the early 1900s. They have had the vast bulk of the rating agency market for a very long time. Fitch has existed for a considerable length of time as well, and is an amalgam of several smaller rating agencies. Over the years, Fitch has increased its prominence, especially in the area of structured finance securities, the type of securities involved in the financial crisis. Markets have expected issuers to get ratings for their securities from both Standard & Poor’s and Moody’s (or, in the case of some structured finance securities, one rating from Moody’s or Standard & Poor’s and the other from Fitch); this expectation apparently may translate into lower buyer valuations for securities issued without their ratings.
Government favoritism is shorthand for two things that, especially in tandem, have contributed to the continuing dominance of Moody’s and Standard & Poor’s and more recently, Fitch. One is that since 1931, the government has required or encouraged certain types of investors to prefer financial instruments that rating agencies rate highly. The other is that in 1975, the SEC began designating particular rating agencies as “Nationally Recognized Statistical Rating Organizations,” or “NRSROs.” The government’s requirement or encouragement that some investors buy highly rated instruments became a requirement or encouragement that such investors buy instruments rated highly by an NRSRO. There were no set procedures to become an NRSRO. Very few agencies were accepted; the SEC noted that not many applied. Prominent among those accepted were Moody’s, Standard & Poor’s, and Fitch.
Certainly, the process to obtain NRSRO designation was opaque at best. One applicant, LACE Financial, noted that it was only contacted twice by the SEC: once to say its application had been received, and again, eight years later, to say the application had been denied. Another disappointed applicant, Egan-Jones, reported being told by the SEC that the SEC did not want to reveal the criteria for becoming an NRSRO, lest the applicant find a way to meet them.
Why do so few rating agencies have the vast bulk of the market? Many blame the government and the NRSRO designation process for this state of affairs. I have argued elsewhere that the dynamic is more complicated. It is not as though the only agencies receiving the NRSRO designation were Moody’s, Standard & Poor’s, and Fitch. Issuers could have used other NRSROs, but they typically did not. Moreover, even when regulations only required the use of one agency, issuers might use two. A newspaper article written in 2004 noted that “[i]nvestors expect ratings from Moody’s and S&P, each of which controls about 40 percent of the market.” The article quoted Dessa Bokides, a former Wall Street banker as saying, “[y]ou basically have to go to Moody’s and S&P . . . . The market doesn’t accept it if you don’t go to both of them.” Moody’s and Standard & Poor’s were the market anointed raters; Fitch made some inroads, especially in structured finance. Nobody had both the will and the way to change this state of affairs. But why two, and why these two? Why was Fitch (and not one or more other agencies) able to make inroads when it did? Nobody has a good answer to these questions, but it is clear that the norms at issue are sticky indeed.
One reason why the rating agency industry is as concentrated as it is may be the difficulty with business strategies that would-be competitors might use. Competing on price is not apt to lure issuers; after all, the ones making the decisions are not bearing the costs themselves. Any benefits issuers might enjoy from saving their companies money would be significantly outweighed by the reputational and financial costs their companies would bear if an issuance were not well-accepted by the markets. Competing on laxity is the more obvious strategy, but it is self-defeating if done too nakedly; the new agency cannot be seen by markets to be easily bribed. Thus, the rating agency needs some credible reason for its apparent laxity. The obvious reason that springs to mind is greater “expertise.” This is by some accounts precisely what Fitch did to become the third ranked agency, especially in structured finance. However, Fitch still had to fight perceptions that it was more lax than Moody’s and Standard & Poor’s, and it developed and carried out a strategy to do so.
So, it is perhaps not surprising that the rating agency industry has been concentrated, with very little competition. The concentration had its costs, but it may also have had an important benefit. As an article on WPPSS notes:
[t]hese alumni [of the rating agencies who claim that there were insiders who wanted to downgrade WPPSS] are quick to allege a reason for the rating agencies’ restraint: they are rating their clients . . . . Whoops, for example, has paid the two agencies more than $400,000 over the past ten years. But since marketing a big bond issue without agency ratings would be unthinkable, it’s not clear that paying even substantial fees can buy issuers the upper hand.
Indeed, one important issue flagged in the Enron debacle was the extent to which the accounting firm Arthur Andersen felt it had to do its client’s bidding to keep the client, and was willing to do things that were arguably dishonest as a means to that end. When rating agencies misrated Enron’s securities, it was not because they were trying to keep Enron’s business by helping Enron create a false financial appearance. Rather, the picture painted at the time was that the rating agencies were not canny chicanerers- they were clueless. As a rating agency veteran was quoted in The Economist in 2003 as saying, “[w]e may be incompetent . . . but we’re not dishonest.”
That was then. What about now? The Credit Rating Agency Reform Act of 2006 requires the SEC to award NRSRO designations for applicants who meet specified criteria. The Act also requires increased disclosure and oversight. The Act did not prevent rating agencies from disastrously misrating subprime mortgage securities. It has led to the designation of more NRSROs, including LACE Financial and Egan-Jones. By the time the Act was passed, Fitch was becoming a more plausible second (that is, additional) rater than it previously had been, especially in the structured finance arena. Issuers began to play agencies against each other, now being able to threaten credibly (unlike when there had been a partner monopoly consisting of Moody’s and Standard & Poor’s) that they might take their business elsewhere. Might the two developments be linked? It seems possible, although it also seems possible that Fitch, having marketed itself as a particular expert in structured finance simply succeeded in doing what it had set out to do. Whatever the cause, the effect is clear and disastrous. Former Moody’s employee Mark Froeba described the shift in culture at Moody’s. According to Froeba, a culture of integrity was replaced by a culture that single-mindedly pursued market share. Inaccurate ratings were a foreseeable result. In testimony before the Financial Crisis Inquiry Commission, Froeba stated:
I have tried to show that Moody’s managers deliberately engineered a change to its culture intended to ensure that rating analysis never jeopardized market share and revenue. They accomplished this both by rewarding those who collaborated and punishing those who resisted. In addition to intimidating analysts who did not embrace the new values, they also emboldened bankers to resist Moody’s analysts if doing so was good for Moody’s business. Finally, I have tried to provide you with an example of the extent to which the new culture corrupted the rating process. The adjusted European CLO Rating Factor Table appears to have been adopted for the sole purpose of preserving Moody’s European CLO market share despite the fact that it might have resulted in Moody’s assigning ratings that were wrong by as much as one and a half to two notches.
III. Dodd-Frank and the Goals of Rating Agency Reform: Returning to the Stories of Dunces and Corruption
Dodd-Frank’s rating agency reform was enacted in response to the rating agencies’ disastrous misrating of subprime securities. The reform has two important goals. One is to decrease reliance on ratings. Notwithstanding what is often said, the problem is not simply that the rating agencies’ ratings were too high. Rather, it is that people invested on the strength of those ratings. We would not need to care what rating agencies did (or how badly they rated) if nobody listened to them. The other goal is to improve the quality of ratings. The drafters of Dodd-Frank presumably recognized that people will continue to be influenced by the agencies, at least in the short- to moderate-term, no matter what the government does. If rating agencies will be influential for some time to come, it behooves government to make them better if at all possible.
How does Dodd-Frank try to achieve these aims? As to the first aim, elimination of reliance on the agencies, it mandates the elimination of references in statutes and regulations to NRSROs. Something else is to replace these references. The rationale behind the new law is simple and unassailable. The government has many reasons to care about the quality of financial instruments and firms that hold them; it therefore needs a measure of quality. It has used ratings, but they proved spectacularly unreliable. The problem is that there is no ready alternative. Moreover, the market norms of using ratings from rating agencies–indeed, particular rating agencies–will not disappear even if the statutory and regulatory references are removed. Market practices are sticky, and market actors have strong incentives to abide by them. Even now, after Moody’s, Standard & Poor’s, and Fitch have done so badly, and when other rating agencies are NRSROs, the Big Three are still highly influential. As the title of an article published when the agencies’ misratings should have been fresh in investors’ minds indicates, In Rating Agencies, Investors Still Trust. All this suggests that at least in the short term and perhaps for the moderate term as well, reliance on NRSRO ratings and in particular, the Big Three’s ratings, will continue.
As to the second aim, improving the quality of ratings, Dodd-Frank has a multifaceted approach.
For rating agencies, it provides for:
Will Dodd-Frank succeed in improving rating agency quality? To appraise its prospects for doing so, let us consider again why rating agencies have grossly misrated. I argued above that while some corruption existed, there was also a client-oriented perspective that many have (in my view, mistakenly) characterized as corruption. That perspective was borne of a broader lack of a competitive edge. A true competitive edge would have kept agencies more receptive to receiving disconfirming evidence–evidence that many instruments for which their clients sought high ratings were really junk.
Insofar as we think rating agencies are corrupt, or somewhat so, most of the Dodd-Frank reforms listed above may actually do a better job perfecting concealment than improving quality. Dodd-Frank does not ask rating agencies to guarantee results, only processes. Many skilled and well-meaning professionals are in the business of helping companies document that they are using appropriate processes. It may be harder to do so if the companies are in fact corrupt than if they are not, but it is presumably not impossible. If a corrupt company is trying to hide its corruption, it may be able to do so even from prying eyes of bureaucrats and litigants. And if a company has convinced itself that it is behaving appropriately, there is even less to hide. Moreover, while the corrupt agency will be better at hiding what it does, the more honest yet more incompetent agency may fool bureaucrats too; bureaucrats’ monitoring is scarcely cutting-edge, and may catch only egregious errors. The truly incompetent agency may be caught: it may also be incompetent in hiding its non-compliance with regulatory requirements. But by the time a lawsuit is initiated, significant losses may already have been suffered.
But what about the expanded procedures to deal with conflicts of interest and the increased role of independent board members? Here, again, there are several reasons to be wary. These types of approaches have been used before for corporations, without significant positive effect. Certainly, increased independence on corporate boards has scarcely been a panacea; the type of independence that is needed is independent-mindedness, not independence as it is formally defined. Whether a rating agency is (mostly or somewhat) corrupt, less than cutting-edge, or both, it will certainly be able and inclined to use processes that pass muster given the level of scrutiny even a highly technically sophisticated board member is apt to use. It seems unlikely that a highly corrupt agency will be unable to conceal its improprieties. It also seems unlikely that an agency simply engaging in self-serving self-deception would not also manage to fool its director-monitors. Procedures to guard against conflicts have the same types of shortcomings. The thoroughly corrupt agency simply lies, and the self-deceiving agency simply convinces itself all is well. The essential conflict, that the issuer is the client, remains, no matter what formal separations of duties and outside monitoring is mandated.
What is needed in the moderate term is vigorous competition. But in the short term, competition has proven a disaster. Why? Because the Big Three have “issuer pays” as their business model. Their incentives have been to direct their efforts (and for those who were actually corrupt, their guile) at “accommodating” their clients, working with their clients to achieve the desired ratings. If the agencies were simply and straightforwardly corrupt, they might be dissuaded from this course of action by sufficient monitoring and accountability. In the likelier scenario, in which self-deception plays a significant part, the agencies are going through the motions, guided by their highly-paid counselors, and do not really know they are sanitizing ultimately suspect and self-serving conclusions.
The solution is to get away from an “issuer pays” model, in which those paying for ratings are the securities’ sellers, and return to “subscriber pays,” in which ratings are paid for by people buying research as to securities’ quality But how? When an issuer is issuing securities, how can a “subscriber pays” rating be arranged? When the legislation that would become the Dodd-Frank Act was being considered in the Senate, Al Franken proposed an amendment that would have achieved some of the benefits of “subscriber pays” consistent with an “issuer pays” model in the context of ratings that have been particularly catastrophic, those of structured finance securities such as subprime securities. The amendment was adopted by the Senate, but dropped in the committee reconciliations of Dodd-Frank. Franken’s amendment would have required the creation of a board that would select the rating agency to be used to rate a structured finance issuance. The issuer would not be able to “fire” the agency and select another. It could use a second agency as well, but it would have to contend with the first agency’s rating.
One important reason agencies are not presently competing vigorously on quality is that market norms dictating the use of Moody’s, Standard & Poor’s, and Fitch are sticky. One way to achieve more vigorous competition is by giving Moody’s, Standard & Poor’s, and Fitch viable competitors–competitors the markets will accept. The government had a real chance in this area. Through the Franken amendment’s board, it could have given other agencies more visibility and acceptance in the marketplace by requiring issuers to use them. This probably would not have led to the most vigorous possible competition on quality, however, as there is no reason to suppose the board would have been expert at, or even inclined to, seek the “best” agency. But the board could and would have set minimum standards of competence. The increment of quality beyond that minimum increment is less important than attenuating the tie between issuers and rating agencies, as the bill would have done. The board contemplated by the Franken amendment may yet come into existence, giving us a chance to test its usefulness: Dodd-Frank requires a study into the feasibility of such a board or a like mechanism.
There are other optimistic scenarios, some more likely than others. Investors might be chastened by their experiences in the present financial crisis and become more wary of rating agency ratings, a prospect I consider quite unlikely given that even egregious misratings have not stopped market participants from listening to the rating agencies (recall the New York Times article discussed above, In Rating Agencies We Still Trust). The references in the statutes and regulations to rating agency ratings might be replaced with references to far more accurate measures of creditworthiness (also very unlikely-after all, if such measures could readily be developed, they would have been developed and in use already, certainly by market participants if not in regulations). One scenario, perhaps more likely than these, is that other agencies, such as Egan-Jones and now Kroll (the agency that bought one of the post-Enron designated rating agencies, LACE Financial, which is run by a well-known “corporate sleuth”) might, through aggressive positioning and perhaps good track records, manage to erode the market norm of using Moody’s, Standard & Poor’s and Fitch. Egan-Jones and, for some of its ratings, Kroll, are using a different business model from that used by the Big Three, where the subscribers pay rather than the issuers. And for Kroll, even where the issuer is paying, Kroll is effectively touting its critical-mindedness and ability to ferret out fraud–something Moody’s, Standard & Poor’s, and Fitch stated was not their strength when trying to explain and excuse their ratings of Enron. These are among the optimistic scenarios. The pessimistic scenario is that history, in the form of short investor memories and sticky market norms, repeats itself yet again. Dodd-Frank, as enacted, does not do enough to prevent this from occurring.