Because he really misses it and thinks he could bring a lot to the table. To help make his case, he sat down today for a little chit-chat with the Observer about what he would be doing to the ratings agencies right now had that whole...you know what it is, we don't need to say it...not happened:
All you need is a common law fraud concept that people—and you go back to the emails, just as we did in the analyst case—and again, I’m not saying “let’s relive the past” This is a more theoretical matter. Go through the emails, and you would’ve seen—“this isn’t a triple-A, but they’re a good client, and we’re gonna…”—that tension between what ratings were put on a product, and one’s belief or recognition that they may not deserve it. There are many theories about what would be there, but you have to get the evidence, to state the obvious. I don’t want to say “gee, they should’ve been prosecuted.” But there should’ve been greater scrutiny over the years, and the structure has always been problematic. It was next on our hit parade, if I had been there for that.
But he wasn't just going to prosecute the bad apples - he would have cleaned up the whole system. Again, if there'd only been time:
The best answer—and I think the marketplace is moving to this—is to essentially tell the ratings agencies: “You’ve got to earn your credibility.” Let’s remove from them the position they had for many years, which was the government saying “You are designated as agencies to which we ascribe a certain elevated position. And you have been given this power by the government without having earned it.“ There’s no reason for that. And now I’m going to sound like a freemarketeer. ...
Well, one type of intervention might just be: Take away all the government imprimatur that the rating agencies have and say ‘These guys are no better than what you have made of them in the past. They really haven’t spotted anything important. And you’ve got to do your own work.”
Leaving aside Spitzer's desire not to sound like a "freemarketeer," this is something a lot of people want. Like Congress, who mandated it in the Dodd-Frank Act. Or the agencies themselves.
It's quite appealing. Some of the blame for the 2008 financial crisis probably does fall on investors who blindly bought AAA structured credit just because they assumed that the AAA rating made it safe - and then saw it go horribly wrong. We might all have been better off had those investors been forced to do their own credit work.
The problem is that calling it a "free market" solution, or presenting it as reducing government involvement, isn't really right. Investors don't use ratings as a factor in investment decisions (if and when they do) because the government tells them to, and removing references to the agencies in the rules won't do much to change what investors think of them.
But investors do want certainty in figuring out how they'll be regulated - and the ratings agencies give them more certainty. The regulations that rely on ratings from S&P and Moody's - that give the ratings agencies their government imprimatur - cover things like bank and broker capital requirements and money market fund investment requirements. Banks and brokers want to know what capital requirements a particular security will bring, and the ratings are a pretty handy way of figuring that out: you buy a security with this rating, you get that risk-weighting. The proposals for removing ratings from those rules have gotten a lot of negative comments not from the agencies but from bank and broker groups, because they would change bright-line simple rules into fuzzy rate-your-own-portfolio-on-a-multifactor-test guidelines. Which would create a lot more work for banks, and also expose them to more liability if they get anything wrong.
That doesn't necessarily make it a bad idea, and since it's in Dodd-Frank it more or less has to happen eventually. But it's misleading to think that making regulatory compliance more complicated is a free-market solution - or that market participants will want it. The solutions that are more market-driven - and Spitzer endorses some of these as well - include increased competition among agencies and changing how the agencies are paid so that they don't have an incentive to inflate ratings to win business. The latter idea, in particular, should get more attention if the government's current investigation of S&P digs up emails along the lines of "this deal looks like it was structured by two pounds of ground sirloin, but we have to rate it AAA to keep the issuer's business" (wait, haven't we seen those already?).
Because of course, even when ratings aren't government mandated, they're still open to corruption if the company being rated is the client. No regulation forces you to buy or not buy stocks based on equity analysts' ratings, and actually it's hard to think of investors who would admit that they care about those ratings at all. But as Spitzer points out, the equity research scandal - in which analysts would call a stock "such a piece of crap" in email and then give it a Buy rating to help win business - appears to be the template for the current S&P investigation. Which Spitzer would be happy to help out with.