I propose a better (?) system for rating bonds
The housing bubble could never have happened if the three ratings agencies — Moody’s, Standard & Poor’s, and Fitch — hadn’t slapped “AAA” seals-of-approval on thousands of bundles of junk subprime mortgages. The ratings agencies similarly failed to detect many fraudulent firms, including Enron. The problem is not an inability to assess risk but a greed-driven allergy to issuing honest ratings:
Internal e-mails from the various firms show that warning bells were repeatedly ignored. A 2006 email from an S&P employee cited in the report said the rating agencies “have all developed a kind of Stockholm syndrome,” held captive by banks. Another e-mail cites an employee openly fretting about the agencies facing their “Nixon moment,” when the housing bubble finally burst.
In the bond rating business, market competition fails spectacularly because market competition drives ratings agencies to inflate ratings to please customers:
Levin’s report includes testimony from an endless stream of former employees who say they felt pressure to grant top ratings to new bond issues, lest they lose deals to competitors.
…[After Enron], lawmakers blamed the virtual duopoly that Moody’s and S&P held on the ratings market, and the law made it easier for competitors to join the industry.
That didn’t work. Instead, newcomers simply accelerated the “race to the bottom,” and made a nefarious strategy called “ratings shopping” even easier. If a bank wanted to issue a mortgage-backed security, it could shop it to various agencies and pick whichever firm offered the highest rating. The temptation for competitive forces to overwhelm good accounting was enormous.
Government efforts to regulate the firms haven’t worked well either:
[R]egulating the credit agencies has proven to be a near-impossible task. Not only are they intertwined in almost every corner of the market, they are intertwined with government agencies, too. Most state pension funds, for example, require that their managers only invest in funds with high ratings, which acts as a de facto endorsement of the agencies' work.
And the agencies have such a strong role in the markets that they can bully regulators into backing off.
In July 2010, with the increased liability provision of Dodd-Frank set to kick in, ratings agencies scored a victory by telling bond issuers they could no longer include the ratings on marketing materials. Because SEC rules require credit ratings to appear, the ratings agencies effectively created a Catch-22, threatening to shut down the entire asset-backed bond market. Regulators, faced with that potential calamity, backed off, and said issuers could temporarily issue bonds without the ratings. In December, the SEC made the change permanent, “effectively exempting companies from part of the U.S. Dodd-Frank Financial Reform Act,” according to a Bloomberg News report at the time.
Here’s a possible solution. We must separate the search for ratings business from the process of assigning a rating. I propose setting a standard price scale for those seeking bond ratings. The government could require (or coerce) every eligible ratings agency to either get out of the business entirely or commit to rating every bond and receiving 1/N of the rating fee bond issuers pay (where N is the number of ratings agencies; currently, N = 3).
Each rating agency would be assured an equal cut of every bond rating fee and would not benefit from manipulating ratings. Because firms' reputations would be built on the accuracy of their ratings, each firm would have an incentive to be accurate. And this incentive could be greatly enhanced if a periodic evaluation of existing ratings agencies were made. New ratings agencies could be granted eligibility, and agencies that had performed very poorly (on some objective, public metric) could be excluded. This would likely increase the quantity and quality of ratings each issuance receives, benefiting all who rely on those ratings.
Posted by James on Tuesday, August 09, 2011