Summary
As subprime mortgage losses cascade throughout the global financial system, attention has turned to the structure and performance of the bond rating industry. Faulty ratings on securities backed by subprime mortgages are believed responsible for billions of dollars in losses. This White Paper argues that any such faulty ratings are due in part to conflicts inherent in the issuer-pays rating agency business model. Such conflicts, when combined with existing structured finance practices, have led to widespread rating shopping. Efforts to increase competition among rating agencies may exacerbate the problem unless fundamental changes occur in the structured finance area, particularly in attitudes toward unsolicited ratings.
Analysis
Background
Today’s credit rating industry traces its roots to the 19th century commercial credit bureaus, whose simple grading systems helped facilitate trading among merchants. At the same time, the birth of the US corporate bond market was underway as financing for the bulging nation’s railroads was desperately needed. Large sums were required and the expected horizon of borrowing was many years. Because banks alone could not meet these needs, bonds became the preferred financing vehicle for the many, initially quite separate, railroad companies dotting the landscape of the late 1800s. Seeing a need for information to assist potential buyers of railroad bonds, several enterprises began marketing “manuals” of financial data and other statistics. Poor’s manuals were soon joined by John Moody’s railroad and industrial manual. This was a difficult business because the barriers to entry were fairly low and high volumes were necessary to cover printing and distribution costs. Moreover, demand was highly dependent upon market conditions, which tended to be quite volatile around the start of the 20th century. After losing his manual business following the panic of 1907, John Moody decided to implement an idea suggested to him by an associate. He created a system that graded bonds according to their investment quality. The 1909 publication of Moody’s Analyses of Railroad Investments thus marked the beginning of the bond rating industry. In the following decades, Standard Statistics (later merged with Poor’s) and subsequently, Fitch, joined the ratings fray. By convention, bond ratings are opinions of relative credit quality.1 These are expressed using comparable, simple rating systems. Most rating agencies rely on a rating system expressed, from highest to lowest, as AAA, AA, A, BBB, BB, B, CCC, CC, C and D, while Moody’s has continued to use its Aaa, Aa, A, Baa, Ba, B, Caa, Ca and C system. As the acceptance of ratings grew, so did their application. The large rating agencies today assign credit ratings to corporate bonds, commercial paper, preferred stock, syndicated bank loans, sovereign nations, municipal obligations, infrastructure projects, structured finance transactions, bank deposits and mutual funds. Managing Conflicts Prior to 1970, rating agencies did not accept payment from rated bond issuers.2 Instead, they financed their rating operations through manual sales and investment advisory services. Rating agencies were well aware of the conflicts of interest posed by the “issuer-pays” business model. By accepting payment from an issuer, a rating agency sacrifices its independence. It has a vested interest in the success of a bond offering and in the welfare of the issuer. Despite this conflict, the issuer-pays model now dominates the industry. Market features and business practices have evolved to help offset this conflict of interest. These safeguards, however, do not eliminate the conflict. Reputation Risk First and foremost, the credibility (and therefore the value) of a rating presumably derives from the reputation of the issuing agency. Any agency suspected of selling high ratings would, in a free market, see its business deteriorate as such ratings would not influence bond pricing decisions. The market would discount or ignore ratings of agencies whose reputation is tarnished. It is argued that building a stellar reputation requires a long-term horizon and view. Yet managers of publicly owned rating agencies are subject to intense short-term pressure to demonstrate earnings growth. It takes tremendous discipline to turn away business, particularly when competitors are building market share. Recent rating mistakes, while undoubtedly harming reputations, have not materially hurt the rating agencies.3 On the contrary, rating mistakes have in many cases been accompanied by an increase in the demand for rating services. One could conclude that reputation risk is not an important deterrent to poor ratings. Separating Analysis from Business Pressures Independent, non-conflicted ratings do not take into account revenue implications for the rating agency. A popular practice that helps meet this objective is the rating committee. More specifically, a rating committee where those with business objectives have little, or at best, equal voting rights, can help resist some of the pressure exerted by an issuer. Even better would be a rating committee in which such individuals play no role whatsoever. Also, a larger committee may (though not necessarily) have a smaller stake in the rating outcome, further improving rating independence.
Transparency
Publicly available rating methodologies, providing sufficient detail to guide a layman towards plausible rating outcomes, are one of the most important tools to counteract the issuer-pays conflict. A transparent methodology makes it difficult to justify a higher-than-warranted rating outcome. Transparency in the financial situation of the rated issuer or obligation is also important in managing the conflict. For one thing, an issuer with publicly available financial data is open to scrutiny by a wide range of market participants. It is easier for investors to apply rating criteria and compare with published ratings when there is financial transparency. Moreover, as a defense against “rating shopping,” any rating agency can (in principle) assign a rating to an issuer with transparent financial reports.4 As discussed below, many structured finance transactions fail this transparency test.
Objectivity
A related characteristic, objectivity, can provide a defense against conflicted ratings. What is typically meant by objectivity is that a rating methodology is based on non-subjective, observable, criteria. Objective ratings are not subject to the whims of any particular analyst or rating committee. Complications can arise, however, when trying to balance an objective methodology against the desire to be “forward looking,” or able to include new or unanticipated factors into a rating. Analysts will often argue that they need to be flexible in applying a methodology. So long as the arguments for deviating from a published methodology in a given situation are clear (and publicly available), this complication can be managed.
Scale
A large rating agency is less likely to suffer financially by assigning low ratings to a given issuer. Since no single issuer can materially affect the revenue of the rating agency, the temptation to sell a high rating is more easily offset by reputation concerns. Consequently, a larger rating agency is more likely to have the financial resources, and therefore discipline, to stand up to any individual rated issuer.
Governance
In addition to these defenses, various corporate governance safeguards must also be in place. It is generally agreed that a large, diversified corporate parent should not own a rating agency. Corporate pressures may cause the agency to “lowball” ratings for competitors of its sister companies. If the rating agency is publicly owned, the board of directors (often with their own corporate affiliations) must not participate in rating decisions. One could argue that the need to meet financial targets of any kind places undue pressure on the quality of ratings under the issuer-pays framework.
Competition and Ratings
Because the rating industry tends to be dominated by a few large firms, many observers assume that greater competition can improve the quality of ratings. One of the goals of the Credit Rating Agency Reform Act of 2006 is to open the NRSRO recognition process to a wider array of firms.5 Unfortunately, increased competition can instead lead to rating shopping and a race to the bottom, in terms of ratings quality. It is useful to examine briefly what is meant by ratings quality. Rating quality is difficult to quantify. Most market observers equate rating quality with rating accuracy. Although there are no official measures in place for determining rating accuracy, the basic idea is that the more accurate a rating system, the better it discriminates ex ante between those issuers (or obligations) that default and those that do not.6 Because defaults tend to be somewhat rare, establishing a rating system’s accuracy can be difficult, particularly if one focuses on a single industry or region. Many users of ratings are focused not on the accuracy of ratings, but rather on subjective features, such as speed of execution, responsiveness to inquiries, or other aspects of service.7 In the absence of clearly articulated and observable rating system objectives, competition among rating agencies often occurs along these dimensions. While commendable as goals, these have nothing to do with protecting investors. The target market for bond ratings most accurately falls under the label “institutional buy-side.” That is, today’s rating agencies are organized to cater to large fund managers and other investor-agents. These well-funded participants hold tremendous sway with banks, broker dealers and bond-issuing companies. Many asset managers are themselves governed by ratings-based investment guidelines. These guidelines, in turn, lead many of these professionals to “game” ratings, rather than view them as helpful investment signals.8 Consider the rating needs of a typical bond fund manager. When deciding whether or not to buy a particular bond, the manager wants an accurate, independent opinion of the bond’s credit risk. Upon purchasing the bond, however, the manager’s interest in an accurate rating deteriorates. In particular, the manager does not want to see the bond’s rating downgraded. In addition to causing a possible decline in price and subsequent portfolio losses, a downgrade may actually force the manager to sell the bond (due to the aforementioned guidelines), even if he or she is disposed to keep it. In other words, a rating agency focused on pleasing fund managers will not necessarily provide a product that protects investors.
Network Effect
It is widely accepted that competition and market forces offer benefits under most circumstances, in terms of resource allocation and efficiency. Where there is a market failure, however, the competitive solution may not be optimal. One type of market
