Senator Al Franken’s plan to regulate the rating agencies via a new “Credit Agency Rating Board,” is such a hare-brained idea that I’m surprised it hasn’t drawn more criticism.
Have you heard about the Franken Amendment? It’s part of the Senate's version of the financial reform bill. The amendment aims to eliminate the “issuer-pays” conflict of interest that’s said to have undermined the agencies’ objectivity during housing boom (when the agencies routinely gave AAA ratings to subprime CDOs that quickly defaulted), which in turn contributed to the subsequent bust. Under Franken’s plan, the issuer would still pay for the rating, but it wouldn’t be able to choose which agency provides it—and so wouldn’t be able to play one agency off against another. Instead, the government would do the hiring, via the aforesaid Credit Rating Agency Board, that would rotate assignments.
As I say, a bad idea. Here’s why:
1. The “issuer-pays” conflict isn’t why the rating agencies screwed up.
You’ll get no argument from me that the rating agencies blew it in a very big way when they stamped all those AAAs on the mountain of subprime CDOs issued during the housing bubble. But, notably, structured-products was the only product line where the agencies fell down. Their ratings of, say, corporates and municipals held up just fine. If “issuer pays” is the pervasive conflict that Franken and the rest seem to think, then corporate and municipal issuers would have had the same clout over the agencies that CDO issuers did, and would have been able to browbeat them into issuing undeservedly high ratings, right? But that didn’t happen. In the agencies’ more traditional businesses, the issuer-pays model worked just fine, as it has for decades. (Oh, sure, rating agencies will occasionally let an Enron or two slip through. Those can cause unexpected losses that can be expensive for the debtholder, and are entirely regrettable. But they are one-offs that don’t threaten the global financial system the way the subprime CDO crackup did. The rating agencies’ failures rating mortgage CDOs were of an entirely different order.)
Rather, the problem at the root of the subprime CDO fiasco was the complex structure of the instruments themselves, and the models the agencies used to assess their risk. The bonds were just too darn complicated for anyone to independently analyze, while the models (as is typical) failed to anticipate the likelihood of extremely negative outcomes. Investors who wanted to buy the paper had no choice but to rely on the agencies’ work which, we now know, was nowhere near up to the task.
2. Whether Sen. Franken knows it or not, the problem of overblown credit ratings has already solved itself.
Let’s do a thought experiment. Suppose your local securities salesman calls you and pitches on a new issue he’s selling that consists a hodgepodge of other issues—residential mortgages, junk bonds, you name it. Not to worry about the credit quality of the underlying portfolio, he adds. The whole package is rated AAA by S&P and Moody’s. Would you buy it? Actually, you might. But your decision wouldn’t be based on the imprimatur of the two agencies. The deal would have to be simple enough that you’d be able to do your own work and come to a decision. That’s not the way things worked at the cycle’s peak. Then, the deals were so complex that buyers couldn’t do their own analyses and had to depend on the rating (and the wrapper). Now, offerings like that are unsellable. In structured products, credit ratings are simply no longer credible—and for good reason
3. A Credit Rating Agency Board would give debt buyers a false sense of security.
The agency’s involvement in the rating process might cause a buyer to give more credence to a rating than it deserves. Why? Because the buyer would believe, wrongly, that the agency had been especially rigorous in determining the rating, since it was no longer operating under that issuer-pays conflict. But, as noted, the problem wasn’t issuer-pays in the first place. If the risk models are just as bad as they were before, the rating will still be suspect—regardless of who hired the actual agency that provided it and which “conflicts” the agency is or isn’t operating under. The intervention of a government rating-agency-assigner that would implicitly bless the agency’s work only complicates things.
Franken’s amendment passed 64-35 in the Senate and is said to have a fighting chance to survive in the conference version of the bill. So what is to be done? Well, there’s another amendment in the Senate bill. It’s sponsored by Georgia LeMieux and Maria Cantwell (Republican of Florida and Democrat of Washington, respectively), and it might solve the rating-agency problem altogether. The Cantwell-LeMieux amendment essentially cuts the agencies out of the process by rescinding a 1930s-era requirement that banks, pensions, and insurers only invest in debt instruments rated a certain minimum investment grade. If Cantwell-LeMieux stays in the final bill, institutional fixed-income investors will no longer have to rely on credit ratings, and will have to do their own analysis, instead. Which, given that they are professional investors, makes a certain amount of sense.
The rating agencies bear more than a little of the blame for the inflating of the housing bubble and its subsequent collapse. One way or another, they should be reformed, or they’ll play some similar role in another bubble and collapse sometime in the future. The Cantwell-LeMieux approach fixes the problem. The Franken approach does not.