The Senate's Disappointing Rating Agency Reform Proposal
So far, I've been pretty positive about the Senate's version of systemic risk regulation (.pdf). In a proposal that otherwise calls for sweeping changes to the current regulatory framework, I was extremely disappointed to see the meager reforms the plan has in mind for rating agencies. It essentially seeks to have the ratings industry work as before, but with some minor changes. That's kind of like giving a new coat of paint and oil change to a rusty 1973 Ford Pinto with an engine that once exploded: even if you get it running once more, it might just blow up again under certain conditions.
For starters, the Senate wants better controls in place during the rating process. Although that makes a lot of sense on paper, here's the thing: the process is already pretty well-controlled.
In my experience, there are several analysts working on a deal who specialize in the type of security a bank wants rated. Those analysts then decide the assumptions that should be in place, given the specific characteristics of the bond. They then bring their assumptions and decisions to a sort of broader credit committee at the rating agency who needs to sign off on those final ratings.
As you can see, this isn't a situation where a rogue rating analyst can just slap AAA ratings on whatever he feels like. Controls weren't the problem.
Another idea that sounds great would be if rating agencies established methodologies around the ratings they give. Here's the problem: they already do this too. For example, click here (.pdf -- yes, you have to register, but it's free) for Moody's Approach to Rating U.S. Residential Mortgage-Backed Securities. You can find such methodologies for everything the agencies rate.
The Senate also hopes to make these methodologies more transparent. Well, isn't the fact that you can search and access any of these methodologies on their websites for free pretty good transparency already? Again, establishing transparent methodologies wasn't the problem.
This is perhaps my favorite part. The Senate wants to ensure each rating analyst:
(1) meets standards of training, experience, and competence necessary to produce accurate ratings; and
(2) is tested for knowledge of the credit rating process.
Let's try to imagine a test question that could have shown up on a periodic exam that might have helped to show a mortgage-backed securities (MBS) rating analyst was ill-equip to handle his job prior to the mortgage market's collapse:
Question #23: Can real estate prices decrease?
A. They did once, but that won't happen again.
B. Maybe, but the bankers who send us MBS deals swear they won't.
C. Yes, but we don't think they will, especially not dramatically.
D. Yes, and consequently, we should rarely provide AAA ratings to subprime MBS without truly dramatic credit enhancement.
The correct answer, of course, should be "D." Exactly every rating agency analyst would have gotten that one incorrect. But frankly, I'm unconvinced that whatever regulator wrote this question would even have gotten the answer right in 2005. A testing standard like this would not have prevented the problem.
Right To Deregister
Another hilarious part of the Senate's proposal would give the Securities and Exchange Commission the right to deregister a rating agency if it performs poorly. I wonder, then, what would have happened this past year if this provision were already in place? All three performed quite poorly -- so would they all now be deregistered, leaving us in a world with no rating agencies? Dare to dream. Instead, however, this would never happen. Having an oligopoly of agencies causes their ratings to converge. As a result, there would never be a lone outlier to deregister.
Here's an interesting idea, however. The Senate's plan appears to provide investors with the opportunity to sue rating agencies -- something that currently isn't often accomplished due to the rating agencies' opinions traditionally being granted First Amendment protection. Investors can sue when agencies have failed:
(i) to conduct a reasonable investigation of the rated security with respect to the factual elements relied upon by its own methodology for evaluating credit risk; or
(ii) to obtain reasonable verification of such factual elements (which verification may be based on a sampling technique that does not amount to an audit) from other sources that it considered to be competent and that were independent of the issuer and underwriter.
That seems pretty reasonable to me, but unless court precedent changes, it could be deemed unconstitutional. When it comes to the right to free speech, facts don't much matter. That, of course, demonstrates the lunacy of giving the rating agencies this protection in the first place.
At last, there is one tiny provision I quite like. The Senate draft has some ideas for due diligence:
The issuer or underwriter of any asset-backed security shall make publicly available any third-party due diligence report obtained by the issuer or underwriter.
I'm not sure if the document intends to do so, but this appears to pave the way for investors to begin to wean themselves off of the rating agencies. As an investor, if I know the results and context of the due diligence, then I can more readily make an independently determination of the credit quality of the bonds being sold. Investors should be privy to as much information as the rating agencies so that they don't need to rely on them. In such a world, the rating agency problem could take care of itself, because they may become largely irrelevant.