I participated in a very interesting debate in the NY Times regarding how to fix the rating agencies.
I end my contribution to the debate as follows:
How do you fix this (if it’s not obvious already)?
Eliminate perverse incentives. Whoever wants to buy an asset should have to pay to have it rated. Credit agencies shouldn’t be paid by the same entities they might have to chastise.
It would also help if agencies could no longer hide behind the excuse that their rating was only an opinion, rather than empirical research they must stand behind. There's no need to do a reliable job if you face no credible legal liability, and the government essentially limits the competition you face.
For six years, I have run circles around the three major agencies with timely and accurate predictions of where regional banks, commercial/investment banks (Bear Stearns collapse, Lehman Brothers), insurers, commercial real estate, residential real estate, US home builders (Lennar), and the pan-European sovereign debt crisis participants were heading. If I can do it, the agencies can too.
One thing many commenters seem to be confused about is the ability for investors to pay for ratings. You don't get anything for free. Never does something emanate from nothing. Any credible advice HAS to be paid for, period! S&P actually sells equity research to the end user, yet gives away fixed income research. Which do you think is the most credible? Most fized income investors are institutions, who are more than capable of paying for advice, and regularly do so anyway.
We can fix the problems we have with rating agencies as end users, but you have to realize that the agencies themselves are not broken. It appears as if the agencies are broken only if you don't understand their business model...
This clip is an excerpt from the VPRO documentary on rating agencies, a worthwhile view. In the meantime, let's revisit my historical viewpoints on the topic: