Red Herrings

from the WSJ, Buyers Should Pay for Bond Ratings:


By ERIC DINALLO

There has been a great deal of justified criticism of the credit-rating agencies that gave triple-A and double-A ratings to billions of dollars of debt securities that clearly did not deserve these high ratings. Everyone agrees that something needs to be done to prevent inflated ratings. But what?

A recent report by the Group of 30 (international financial experts led by Paul Volcker) recommended that regulators encourage the development of payment models that "improve the alignment of incentives" and permit rating users to hold rating providers accountable. Similarly, Securities and Exchange Commission head Mary Schapiro recently called for an examination of " how the rating agencies are compensated, how they manage conflicts of interest, and what role they should play in our markets."

The insurance industry and its regulators can lead the way by implementing the only effective proposal: self-funded, independent buy-side ratings. Ratings, that is, that are paid for by the investors who use them.

Rating agencies' failures are not rooted in a lack of talent or insight, but rather in a fundamentally flawed business model. Those who issue the securities also pay for their ratings. This structure has created powerful incentives to bias ratings to keep debt securities' sellers satisfied and the rating fees flowing.

It is certainly true that agency mis-ratings and misaligned incentives are big parts of the story of the financial system meltdown. On the other hand, the claim that mandating that buyers pay for bond ratings will solve anything withstands little scrutiny and betrays a stunning ignorance about the manner in which regulation and markets work and interact.

The argument, simply put, is that so long as rating agencies are paid for by sellers, sellers will pressure agencies to artificially inflate ratings. If only agencies were paid by buyers, there would be no such pressure.

The problem with this argument is that buyers well know that rating agencies are paid for by sellers, and if they perceive or suspect that a seller is in the habit of pressuring or an agency is responsive to such pressure, they will question the accuracy of the rating, which -- if buyers are basing purchasing decisions on their faith in the rating -- would reduce demand and therefore profits. Agencies and Sellers knew, then, that they were playing a dangerous game if they pressured and responded to pressure.

While it is certainly imaginable, it is not clear to me that this pressuring and being pressured actually happened. We have seen emails from dealers about pushing products they knew to be crap, but if there has been similar from rating agencies, I missed it.

If it did happen, the question is why sellers and agencies -- rightfully it turns out -- where unafraid that the perception of tainted ratings would not affect demand. Phrased differently: Buyers too saw what was going on and were unconcerned. Which argues that buyers were more interested in getting the high rating then whether or not the high rating was meaningful. If that is the case, ratings will be no more reliable if buyer purchased.

It is, of course, hardly true that rating agencies' failures were not rooted in a lack of talent or insight. As compensation at the agencies substantively lags that of both the buy and sell side, people with more talent and insight are (to put it mildly) less likely to choose to work for a ratings agencies. Given the talent gap, it would be surprising if an investment manager -- constrained by regulatory and, perhaps, client mandates which gave ratings weight in portfolio construction -- was not simply interested in getting the high rating, and less concerned about whether or not that rating reflected some reality.

Even we pretend that agency raters are as insightful and talented as investment managers, given any reasonably liquid market, savvy investor managers are going to trust the rating implied in the price far more then the agency ratings.

Finally, and above all, no amount of resolving conflict of interest gets to the heart of the square, the regulatory regime is trying to circle: Risk, by definition, defies easy measurement and management. As noted previously on this blog, were that not the case, we would be better off with command economies rather then free markets. It does not strain the imagination, of course, to suspect that a regulator, in his heart of hearts, prefers command economies to free markets.

Consider that in 2003, on the equity side, regulators entered into settlements with Wall Street firms to resolve conflict-of-interest issues between their research and investment banking divisions. Like the rating agencies, equity research analysts held themselves out to be objective in their analysis. But they were paid by the issuers and their bankers. The regulators' investigations demonstrated that the firms and their client-issuers pressured equity analysts to provide bullish recommendations on their worst stocks.

This analogy is illustrative. Reasonably informed investors understood that Blodget's advice was tainted, and took it with an appropriate grain of salt. Given the decline in equity research since these conflict of interest issues were resolved, its reasonable to claim that the value of the "research" to the market was, all along, to the sell, not buy, side.

In the case of ratings, their value to the market is likely to regulators, not investors. It gives regulators some measure of the safety of the companies they regulate. Investors who trust rating agencies to measure risk better then their investment managers, ought hire a different manager.

The solution is for investors to buy and control publicly available bond ratings. Insurance regulators, who use ratings to determine capital reserves for insurance companies, can contract with rating agencies on a competitive basis to provide public ratings of issuers and their securities. This approach would solve the conflict-of-interest problem, because the primary users of the ratings are the ones who will be paying for them.

To fund a buy-side proposal, insurance commissions could collect a small fee from insurance companies that hold nearly $3 trillion in rated bonds, making them the largest industry sector that relies on credit ratings. The New York State Insurance Department estimates that for less than two basis points (0.02%) per year on that $3 trillion, insurers in partnership with insurance regulators can purchase transparent, conflict-free and cost-effective ratings. Buyers have a strong incentive to pay into a system that ensures the independence and accuracy of their ratings.

This paragraph is near perverse, Dinallo substitutes himself for the buy-side. His proposal, unsurprisingly, is to give himself -- a regulator -- control over selecting rating agencies. He would fund this control via a tax hike.

Rating agencies, of course, are already a (federal) regulator construct. His proposal then is less about buyers taking control from sellers and more about state regulators seizing control from federal regulators -- who currently govern rating agencies.

There is no reason to believe that state regulators will be more effective in policing rating agencies then federal regulators.

And, Dinallo's claim that his system would be conflict free is so obviously untrue it is unlikely he really believes it. Rating agencies, after all, rate public debt as well. States and municipalities have long complained about what they see as unfair treatment from rating agencies. The more control state regulators have over rating agencies the more positive rating agencies are likely to be toward state issuances.

Ratings will never be flawless -- no institution can have perfect foresight. But buy-side ratings will be conflict-free, and the process will be controlled by the investors that bear the long-term risk of the rated securities. Rating agencies with poor track records, errors or conflicts will not be trusted to serve and protect policy holders. And rating agencies will bid for contract renewals based on merit, so that they remain independent of the issuers they evaluate.

The largest obstacle to effective market regulation is, to my mind, regulators -- like Dinallo -- who, simply, do not believe in markets, and do not take seriously the choices of market participants. One has to have precious little faith in markets to believe that companies with poor track records will be trusted or long survive.