What Should Be Done about the Credit Rating Agencies?

| | Comments (0) | TrackBacks (0)

by Edward I. Altman, Sabri Oncu, Anjolein Schmeits, and Lawrence J. White

The three large global credit rating agencies - Moody's, Standard & Poor's, and Fitch - were central players in the subprime residential mortgage debacle of 2007-2008. Their initial overly optimistic ratings on mortgage-related securities encouraged the housing boom and bubble of 1998-2006. When house prices ceased rising and began to fall, mortgage default rates rose sharply, and the prices of the mortgage bonds cratered (as did their ratings), wreaking havoc throughout the U.S. financial system.

The Securities and Exchange Commission has already expanded its regulation of the rating agencies, and congressional legislation may well insist on more. But to understand the proper route forward, it's important to understand how we got to where we are today.

Rating agencies provide judgments - typically in the form of a letter grade - about the creditworthiness of bonds that are issued by corporations, governments, and packagers of asset-backed securities.  Although the three large agencies dominate the field, they are not the only source of such information.  There are smaller rating firms; some large financial institutions do their own research; all of the large securities firms have "fixed income analysts" that offer similar information to their clients; and specialized risk consultants offer credit models and default estimates.  It should also be remembered that the main investors in the bond markets are financial institutions.

Until the 1930s, the use of ratings by investors was entirely voluntary.  In 1936, however, bank regulators first began requiring banks to heed the ratings of a select group of raters.  In essence, these ratings acquired the force of law!  In subsequent decades, regulators of other categories of financial institution - insurance companies, pension funds, securities firms, and money market funds - similarly required their firms to heed these ratings.

In 1975 the SEC crystallized these requirements by creating a new designation for rating agencies - Nationally Recognized Statistical Rating Organization - and ushered the three large rating agencies into this category.  The SEC subsequently became an opaque barrier to entry, allowing only four more raters to obtain the NRSRO designation during the next 25 years.  But mergers among these four and with Fitch meant that only the original three had the crucial NRSRO designation at year-end 2000 - which meant that only these three rating agencies could "bless" the expanding volume of mortgage bonds with their all-important ratings - especially their top "AAA" and "AA" ratings - for institutional investors.

One other important piece of history:  Although the raters' original business model involved selling thick rating manuals to investors, in the early 1970s the industry changed to its current model:  The bond issuers pay the raters.  Although the potential conflicts of interest of this newer model are clear, the raters didn't succumb to those conflicts while they were rating "plain vanilla" corporate and government bonds from the 1970s through the 1990s.  There were so many issuers that a rater could safely ignore any single issuer's entreaty (or threat) for a better rating, and issuers were sufficiently transparent that errors would be quickly spotted, with adverse consequences for a rater's reputation.  The raters' initial ratings were generally accurate, and the raters' reputations were generally good, although they were often slow to react to new information.

The widespread issuance of mortgage bonds in the decade of the 2000s heightened the conflicts:  There were fewer issuers, each with large amounts of business for a rater; the profit margins were larger; and the bonds were more complex, so that errors would be harder to spot.  The rating agencies succumbed, with disastrous consequences.

So, how should public policy with respect to the rating agencies proceed?  We see three important objectives:

- Eliminate regulatory reliance on ratings.  The disastrous consequences of the three large raters' errors were greatly magnified because seven decades of financial regulation had thrust these three into the center of the bond markets and had forced the major players in those markets to heed their ratings.  Without such forced reliance, bond market investors - primarily financial institutions - would be able to access a wider array of opinions as to bonds' creditworthiness.

Since the goal of having safe bonds in banks and other regulated institutions should remain, regulators should replace their current delegations of safety judgments to the NRSRO rating agencies with a more direct system of scrutinizing the bond portfolios of their institutions.  Also, the boards and senior managements of financial institutions should not be able to abrogate their fiduciary responsibilities by uncritically accepting information from any source.

We are encouraged that the SEC has already taken some tentative steps in this direction.  The bill passed by the House of Representatives in December (H.R. 4173) goes further, calling for all financial regulators to cease such delegations.  The senate bill recently introduced by Banking Committee Chairman Christopher Dodd is unclear on this topic.  We hope that the Congress's final legislation follows the lead of the House on this matter.

However, neither the SEC nor the two bills recognize that once regulatory reliance on ratings has been eliminated, the need for the NRSRO designation has also been eliminated.

- Credit information should be more nuanced than just a simple letter grade.  A single letter grade - or even a more precise numerical estimate of the probability of default during a specified time period - doesn't convey an adequate amount of information for sensible investment decisions.  For example, the degree of certainty that the rater places on the estimate, or a set of alternative estimates for a set of "what if" alternative economic scenarios, would be useful and is quite feasible with today's data availability and analytical technologies.  In retrospect, investors might have been more cautious about investing in "AAA" securities if they had been told how a 25% decline in housing prices would affect those securities.

If the bond information market is opened to new competition, through our first suggestion, it's far more likely that some entrants could provide such information -- and the competitive pressures might even cause some of the incumbents to do the same.  The two bills would require that rating agencies provide such information.  However, it's unclear to what extent regulation is needed to spur this process.  A major risk is that regulation could rigidify the process - e.g., by mandating certain kinds of information and prohibiting others.

- Reduce the potential conflicts of interest.  That the potential conflicts of the issuer-pays business model can become real is patently clear.  How best to address them, however, is less clear.

A major advantage to the issuer-pays model is that the information is disseminated instantaneously to the market; the same cannot be said of an investor-pays model.  Also, the investor-pays model has its own conflicts:  Before the bonds are issued, investors would prefer pessimistic ratings; after investors have the bonds in their portfolios, they don't want downgrades (or at least not until after they have sold the bonds).

With greater entry and more choice for institutional investors, these investors may shun information providers that are perceived to have excessive conflicts.  Again, it is unclear whether regulation can be a net help:  Mandated efforts to address conflicts may raise costs, which could discourage entry; and they risk locking in existing structures and procedures.  Ironically, by discouraging entry, regulation could make the incumbents more important!

One possible direction (on which the House bill asks for a study) that would preserve the information advantages of the issuer-pays model while eliminating the conflicts of interest would be the establishment of a "clearinghouse" - perhaps by the SEC - of qualified rating firms.  When an issuer requests a rating, the clearinghouse could randomly assign a rating firm from among the qualified group, which would eliminate the ability of the issuer to "shop around".  But the success of this approach would depend crucially on the quality control of the clearinghouse.

The House bill has an extensive array of mandates for NRSROs with respect to addressing conflicts of interest and providing transparency of methodologies, as well as rating histories.  Since any final legislation is likely to insist on such measures, we hope that they have minimal consequences in discouraging new entry.

0 TrackBacks

Listed below are links to blogs that reference this entry: What Should Be Done about the Credit Rating Agencies?.

TrackBack URL for this entry: http://w4.stern.nyu.edu/mt/mt-tb.cgi/132

Leave a comment

About RegulatingWallStreet.com

The Dodd-Frank Act, signed into law in July 2010, represented the most significant and controversial overhaul of the U.S. financial regulatory system since the Great Depression. Forty NYU Stern faculty, including editors Viral V. Acharya, Thomas F. Cooley, Matthew P. Richardson, and Ingo Walter, provide a definitive analysis of the Act, expose key flaws and propose solutions to inform the rules’ adoption by regulators, in a new book, Regulating Wall Street: The Dodd-Frank Act and the New Architecture of Global Finance (Wiley, November 2010).

About Restoring Financial Stability

Previously, many of these faculty developed 18 independent policy papers offering market-focused solutions to the financial crisis, which were published in a book, Restoring Financial Stability: How to Repair a Failed System (Wiley, March 2009).

About the Authors