Moody's archaeology is quite unconvincing
Last week, Moody’s published a report on the global credit crisis (Archaeology of the Crisis) that I found totally unconvincing. To begin with, the word “archaeology” in relation to such a recent event is perplexing, but that is a minor issue.
My most serious disagreement is with the section in the report on “Reduced risk traceability in the financial innovation process”. Moody’s argue:
Rating agencies were supposed to bridge some of the information asymmetries, but this proved to be some-what unrealistic when the incentive structure of (sub-prime) loan originators, subprime loan borrowers, and market intermediaries also shifted in favour of less information.
... The problem in the case of extreme complexity of inter-connecting financial systems is that it is hard to see how the level of information could reach levels adequate to enable reasonable risk management standards.
Risk traceability has declined, probably forever. It is extremely unlikely that in today’s markets we will ever know on a timely basis where every risk lies.
This leads to two conclusions. First, more capital buffers will be needed or required by counterparties and regulators. Second, not just more but more intelligible information is needed ...
Moody’ is trying to argue that rating agencies failed because the task that they were asked to do was impossible. The first problem with this argument is that they should have thought about this before they accepted the rating assignment and collected their fees. The second problem is that the solution of greater capital buffers (higher attachment points for various rated tranches) was available to the rating agencies all along, but they chose not to go this route.
The third problem with Moody’ argument is that it is becoming increasingly clear that the credit crisis that we are seeing today has nothing to do with financial innovation but is more like the familiar credit and banking crises of the past. For example, it is quite similar to the Japanese banking crisis of the 1990s which also had its origins in a property market collapse.
In August of 2007, it was possible to argue that the credit crisis was somehow related to the difficulty of valuing complex financial instruments like CDOs. One might have thought that credit correlations (which are difficult to estimate) might have been underestimated but there was no problem with the core credit assessment. However, the fall in prices of sub prime linked instruments has been so steep and persistent that it is now clear that the issue is not about correlations but about the value of the underlying credit. A good deal of this underlying value depends on macroeconomic variables like housing prices and GDP growth where there is little if any information asymmetry.
Today, the rating agencies can really take one of only two positions. Either they can admit that they made a grave error in estimating credit risk in a whole large class of credits without any significant attenuating circumstances. Or they can argue that the markets have got it all wrong and the credit quality of sub prime and alt-A housing loans is not as bad as the market thinks. With every passing week of new data coming in from the US, the second position looks increasingly untenable.
In 2001, when the rating agencies were severely criticized for their failures in relation to Enron, I argued that the most that one could accuse them of was some degree of complacence. I wrote: “Any criticism about the rating agencies must keep in mind that the agencies gave Enron a rating that reflected a high level of risk.” Moreover, there were significant information asymmetries between the rating agencies and Enron. The asymmetries in sub prime lending were much lower and the rating agencies had the law of large numbers on their side. I think therefore that a mea culpa would be more appropriate than the kind of archaeological sophistry that Moody’s has dished out to us.
Posted at 4:23 pm IST on Mon, 14 Jan 2008 permanent link
Categories: credit rating
Comments