Rethinking Credit Ratings

If you buy dinner from a three-star Michelin restaurant you can be confident of getting a good meal, but today if you buy a bond from an AAA-rated borrower you can be rather less sure of getting your money back.

Michelin maintains the value of its brand by restricting its prestigious awards to just a handful of the very best establishments. Unfortunately the rating agencies have been rather freer with their favors, bestowing their best AAA rating on huge quantities of what we now know to be distinctly subprime investments.

Today's wave of credit losses and ratings downgrades is not unusual , the corporate credit bubble, which just a few years ago brought down Enron and WorldCom, followed a very similar pattern. Failing to spot one credit bubble is unfortunate, but missing two of the biggest credit bubbles in history within a single decade suggests the problems are more than bad luck.

Policy makers have already tried improving the performance of the rating agencies with the Credit Rating Agency Reform Act of 2006. The aim of the act was to encourage greater competition within the industry in the hope that this would lead to an improvement in the quality of work. Recent events suggest the problems have not been fixed.

First, the obvious problem: Rating agencies make their living by selling ratings to borrowers. Understandably these borrowers are reluctant to pay good money for bad ratings, which puts a fundamental conflict of interest at the very heart of the industry. The natural way to resolve this conflict would be to reorganize the whole business so that investors are obliged to pay for the rationings that they use.

At first sight having investors pay for ratings sounds like a good idea, but it is difficult to see how it would work in practice. Ratings would have to be confidential to prevent investors from free-riding on each other. Then there is the problem of pricing; a one-price-fits-all policy would disadvantage smaller investors, while pricing the service based on assets under management would be unwieldy. Given these problems, it is understandable that the 2006 act left the pricing model unchanged, opting instead for more competition. As yet, the extra competition has not materialized, but even if it were to do so the result could actually make matters worse by driving raters to compete for market share with even more lenient ratings.

Now the more subtle problems: Calculating a credit rating involves estimating the borrower's future cash flows and assessing the value of any collateral being offered. Credit analysts tend to work "bottom up," meaning that they focus on analyzing an individual borrower's circumstance based on the assumption that the overall macroeconomic climate will remain unchanged.