Credit ratings agencies: the problem that hasn't gone away, as much as we've tried to ignore it
Okay, so maybe that's not entirely fair. The SEC has signed off on some new rules and proposed some additional ones to try to temper conflict of interest at the companies that told us all those new-fangled mortgage-related securities were unlikely to ever go bad. Oops! And now SEC head Mary Schapiro is talking about yet another batch of rules that would specificially target ratings shopping—the tendency of debt issuers to pay for ratings from companies that tell them (and potential investors) what they want to hear. In front of the House Financial Services Committee this morning, she said she's creating a special branch of examiners dedicated to overseeing ratings agencies.
Great. This is much more than the Obama Administration's meager response. Although, unfortunately, none of it gets at the larger issue of how dependent our financial system is on ratings agencies. And nefariously so. Why do we call Moody's, S&P and Fitch "agencies"? They're profit-seeking companies acting in the best interest of their shareholders... and yet when one of them says a piece of debt issued by a corporation or a city or a bankruptcy-remote special purpose entity is likely to default at a particular rate that opinion carries the weight of a government imprimatur. More than a year ago Justin wrote that maybe part of the solution is undoing this ratings dependence. Everyone from pension managers to mutual funds to bank regulators to the Federal Reserve depend on ratings agencies to tell them how decent certain bonds are. They shouldn't.
What would be the alternative? Well, see, that's where everyone trips up. Because to not depend on the small group of companies we've all agreed to depend on, market participants are going to have to start coming to their own, independent judgements. If that sounds like it costs money and requires some level of expertise, that's because it does. Professional investors would have to professionally invest. Regulators would have to regulate. And that's all so difficult.
There has been a lot of talk about problems with the issuer-pay model—the fact that the people issuing the debt pay to get it rated. Paul Kanjorski, the Congressman who chairs the House of Representatives capital markets subcommittee, has recently stirred up that argument, partly, it seems, because David Einhorn is shorting shares of Moody's. It's a fine thing to talk about, reforming issuer-pay, but let's not lose sight of two things.
First, if you make investors pay for ratings, you're swapping one set of conflicts for another. Investors have a desire to own highly-rated bonds just like issuers have a desire to sell them. Maybe investors wouldn't color the judgment of ratings agencies as much as issuers—but switching from one to the other doesn't magically get rid of the entire problem.
Second, we shouldn't be beholden to ratings agencies in the way that we are. Bloomberg recently had a story about how the Federal Reserve's TALF encourages ratings shopping—the very thing the SEC is now targeting. It's a complicated, mixed-up world out there in ratings land, and the only thing that will really start to extricate us from it is to simply care less about the opinions of Moody's, S&P and Fitch in the first place.
Barbara!
-
1
"What would be the alternative? Well, see, that's where everyone trips up. Because to not depend on the small group of companies we've all agreed to depend on, market participants are going to have to start coming to their own, independent judgments."
This might be true, but I'm not sure that it is. What, for example, prevents the SEC from setting up an expert system (with industry participation) to rate bonds and publishing the results on its website? I don't necessarily believe that the effort would be technically impossible; but I know for certain that the pushback from Moody's et al. would be monumental. It may not ultimately work, but I don't see anybody even trying.
-
2
Letting the buy-side do the rating bit may work out better. Yes, they want to hold highly rated instrument but finally their performances are judged on the alpha created on how they work their money. Unlike the sell-side that creams the fees from one tranche and moves on to the next con-job.
Thoughts?Sincerely
Subrata -
3
There's definitely a solid argument to having the buy-side pay for ratings, and you lay if out well, Subrata. There already are ratings firms with this model, and their marketing line is that they provide better information since they're not in any way dependent on the sell-side. The hiccup with having this be the predominant system overall is that then ratings would be privately owned and the market would have a lot less information about companies—good for alpha generation, bad for the public domain. Plus, there's an argument that, with the way technology is today, the ratings would leak and become public anyway, and then there wouldn't be an incentive for anyone to pay.
One possible way around this, as pneogy points out, is to pool resources and use that communal pot of money to finance ratings. This could be government-mediated—the imagined division of the SEC, which might be paid for with a small tax on securities issuance or trades—or industry-centric. For instance, the National Association of Insurance Commissioners, a trade group of state insurance regulators, has a Securities Valuation Office that determines the quality of various assets held by state-regulated insurance companies. With a staff of about 50, that office rates about a quarter of the $2 trillion in assets held by state-regulated insurers. It costs $850 a year for a state regulator to belong, and then the cost per rated issue is $2,800—a fraction of how much a rating costs at one of the Big Three. Oh, and the SVO is kind of good at what it does. A few years ago the SVO started warning about hybrid securities, and the investment community got all up in arms—until hybrids collapsed and the SVO was proven right. So, yes, it's technically possible.
-
4
Barbara,
Thank you for the detailed info on the SVO. I am unsure about the technical feasibility of "an expert system" for bond evaluation - a sort of super Fair Isaac system for evaluating bonds - that would be less subjective and remove some of the perverse incentives. Using the SVO for bond evaluation is clearly technically possible.
Most Popular »
- Tennessee Mayor Accuses Barack Obama Of Hating On Charlie Brown, Peanuts
- Wii Fit Plus Review
- Obama Shifts Date of Copenhagen Visit
- NV Sen Poll: Reid In Trouble
- The PlayStation Turns 15, We Reminisce
- 'Forgotten Man' II: Two-Thirds of Jobless Blue-Collar
- 135 Money-Saving Resources and Tips, Special Holiday Season Edition
- Twitter App Showdown: Echofon Pro vs Tweetie 2
- False Economy: Think You're Saving Money? Think Again
- Loving The Joke
- How Strong Is the Evidence Against Amanda Knox?
- Will Federal Spending Mistrust Mean the End of Obama's Audacity
- Amanda Knox, Convicted of Murder in Italy
- Nicolas Sarkozy: A French Paradox
- India, Pakistan and the Battle for Afghanistan
- Hate Your Job? Here's How to Reshape It
- Amanda Knox Testifies: The Murder Trial That Has Gripped Italy
- Helicopter Parents: The Backlash Against Overparenting
- Astronomers Spot Planet-Like Object GJ 758 B in Orbit
- Foxy Knoxy Case Still Roils Italy














RSS