This post first appeared on November 16, 2007
A colleague was so kind as to send me the text of a speech given at the Graham & Dodd breakfast a few weeks ago by David Einhorn, CEO of hedge fund Greenlight Capital.
The speech has gotten play only in some personal-investment-oriented blogs like Seeking Alpha and Naked Shorts. Even though they are fine sites, it’s nevertheless a shame the speech hasn’t gotten broader interest. First is that most of the commentary focused on the general thrust of his argument but failed to pass on key bits of information that are likely to be news to most readers. The second it that there is still a considerable gap between what market participants know compared to legislators, regulators, economists and commentators. Yet somehow the industry types get a serious hearing when they reach an audience (i.e., when they are lobbying), but when people with insight but no axe to grind have something to add to the debate, too often it gets short shrift.
It was sobering for me to learn a few things from the speech. On the one hand, as someone who is not in the marketplace (and lacks access to a Bloomberg terminal), I am in the same position as a journalist: I am only as good as my sources. But on the other, some of the things that I picked up seemed fundamental, yet I haven’t seen them either in the press or the academic articles I’ve read.
Einhorn tells us that the securitization process is flawed (“securitization is a mediocre idea” is the nicest thing he had to say about it) and that the rating agency role is in need of root and branch reform. The entire speech is very much worth reading, and his comments are pointed and insightful. However, he does not fully draw out the implications of what he found.
The rating agencies’ role was even more deeply compromised that most commentators recognize. Their practices made it cheaper to issue the very sort of paper that is was most profitable for them to rate. They did it by letting the creators and sellers of structured credit products play on the popular perception and regulatory fiction that all AAAs are created equal, when in fact, the more complicated the paper was (and therefore more costly to rate) the more risk it was allowed to carry in each rating class.
Some key observations:
The credit problems we have are experiencing are not the result of subprime contagion, but of charging too little for risk-bearing. There was concern in some quarters about systemically low credit spreads, yet participants were confident liquidity would always the abundant. Einhorn believes that the complacency about risk was at least partly due to the securitization process, in which most investors looked heavily to the ratings in their evaluation process.
But just as some animals are more equal than others, so to are ratings:
Consider municipal bonds. According to S&P’s long-term data the 10 year default rate on an A rated municipal bond is 1%, while a corporate bond’s default rate is 1.8%, and a CDO’s is 2.7%. An A rated muni has the same change of default as and AA/AA- rated corporate and a AA+ rated CDO. When municipal bonds default the expected recovery rate is 90% compared to 50% on corporates and CDOs.
This isn’t an accident. About a decade ago, Moody;s said, “No matter what types of instruments the ratings apply to, no mater where the issuer resides, and no matter what the currency or market in which the security is issued, Moody’s ratings are intended to have the same relative meaning in terms of expected credit loss.”
Without much fanfare, the rating agencies abandoned this process…..Instead, for each type of bond, they use a different rating scale with a different so-called “idealized default rates” for each rating. The idealized default rate for a municipal bond at a given rating is less than the idealized default rate for a corporate bond, which is less than the idealized default rate for an asset backed security which is less than the idealized default rate for a CDO….
Nomura Securities pointed out that hypothetically, if you took an AA+ rated asset backed security and repackaged it all by itself and called the repackaged instrument a CDO, it becomes AAA because the CDO has a higher idealized default rate than the asset backed security.
Einhorn points out that the rating agencies’ fees are correlated with their willingness to look the other way with credit losses. CDOs carry the highest fees, ABS next, and regular corporate ratings are cheaper still. But of course, the agencies will argue that the more complex structures are more labor intensive to rate and therefore warrant higher charges.
Regardless, this system, of higher idealized default rates for securitizations has the effect of promoting securitizations. And the higher idealized default rates promoted CDOs.
For a financial institution, the very fact that the loans with the same loss profile will get a higher rating in an ABS than if they sit on their balance sheet means that it is uneconomic for them to keep them on their balance sheet (note that this is independent of the traditional impetus for securitization, namely, the cost of deposit insurance). The rating agencies have stacked the deck via their rating process to make it even cheaper to securitize assets than it would be if investors knew the true exposure to loss.
Now mind you, we are NOT saying that securitization would not have happened without the rating agencies gaming the scorecard. Back in the early 1980s, when AAA still meant AAA, McKinsey was showing its banking clients charts illustrating how securtization had a significant cost advantage relative to traditional lending (message: investment banks are and will continue to eat your lunch). But the rating agencies’ actions played a significant role in the underpricing of risk, and almost certainly made the ABS and CDO markets larger than they would have been with proper risk measurement and disclosure.
A number of observers have criticized investors for being so dependent of rating agencies. Aside from regulatory requirements that mean many investors need to consider rating in their evaluation process, Einhorn gives us another reason: at least as far as structured credits were concerned, the rating agencies had far better information about the deal than end investors:
Have the rating agencies developed an expertise in analyzing these structures? Perhaps, but more pertinent, they are the only ones who can evaluate them, because they are the only ones with the detailed information about the structures. The underwriters give the rating agencies much more information that is contained in the prospectus. In their evaluation of corporate credits, rating agencies are exempt from regulation FD. This means that they can receive confidential information not available to other market participants. This is kind of like a confessional where the priest delivers a public opinion on the extent of your virtues or sins – and your spouse has to guess what a AAA or BBB means about your fidelity.
In the recent crisis, the rating agencies say they shouldn’t be held accountable for their opinions because they are…..nothing more than journalists engaged in free speech. What would it be called if you paid a leading publication to do a story on you and you could putt it before press if it were unflattering? Funny, that is not free speech but “for-profit speech.” According to the rating agency party line, if people disagree, they are free to ignore the ratings. However, a credit rating is not an ordinary opinion. It is a “special” opinion – an insider opinion, because it is based on a different information set. I can’t replicate a rating analysis because I am not privy to the information. Therefore, I am not on an equal footing to be able to decide whether I agree or disagree with a rating agency. Since they know more than I do, the presumption has to be to agree. The rating agencies are structurally set up to have “insider opinions.” They just don’t want legal liability for having issued conflicted and flawed insider opinions.
Incidentally, this lack of information has made it harder for the market to find a clearing price of distressed pieces of structured deals. Without enough information in the market – other than a credit rating – it is hard for informed buyers and sellers to know what to do, once the credit rating comes into doubt.
There is not justification in the credit markets for having the rating agencies have access to deal structure information that is kept secret from investors. The only party that appears to benefit is the investment banks, who might have some structuring tricks they’d like to keep secret from their competitors. By contrast, there is an argument for rating agencies and analysts having access to corporate information of a strategic nature that they would be loath to reveal publicly. But equity analysts lost their Reg FD exemption and the world did not fall apart. There is far less justification for a Reg FD exemption of any kind in the debt business. Einhorn argues that the information that rating agencies possess should be made public.
A final tidbit from Einhorn:
In early September, a senior Moody’s executive….at a small private dinner….said, “Moody’s would never lower the credit ratings of a financial guarantor, because that would put the guarantors out of business.”