Friday, May 30, 2008

Broken Models

"The latest blow to the reputation of Moody's Investor Service in structured finance is the charge that it knew of a computer error that inflated the ratings of constant-proportion debt obligations, or CPDOs, and didn't disclose the problem. But Standard & Poor's gave the deal in question the same top rating--and a bit of a dodgy bit of software code wasn't to blame. The real problem is the rating companies' reliance on computer black-box models in the first place. ... Moody's, a unit of Moody's Corp., says it takes the questions about its models seriously, and has hired an outside law firm to review its CPDO ratings process. Still, the crisis of confidence in structured finance didn't stem from calculation errors. It was a reaction to the discrediting of the rating firms' overly theoretical and insufficiently critical approach--which investors blithely accepted at face value", my emphasis, WSJ, 22 May 2008.

"Moody's Investors Service's problems in rating complex debt instruments known as CPDOs show the difficulty in assessing new, untested products with the same scale that conveys the financial strength of blue-chip companies or the U.S. government. ... On Wednesday, Moody's said it retained law firm Sullivan & Cromwell LLP (S&C) to review its European CPDO-ratings process. The Financial Times reported that data errors were discovered after some of the early CPDOs were rated by Moody's, and those errors went unreported. ... Though [the SEC] can't regulate the specific methodology that a firm uses, the SEC can sanction firms for straying from their stated methodologies. ... Some former Moody's officials said data integrity has been an issue before. Sylvain Raynes, a former Moody's quantitative analyst, said he discovered during the late 1990s, a 'bug in the code' used to rate a deal backed by rental cars. In an interview, Mr. Raynes said Moody's should have downgraded the deal after discovering the error, but decided not to", my emphasis, WSJ, 22 May 2008.

I disbelieve Moody's. It "hired a law firm". Why? To review "process", not susbtance. If Moody's wants to fix its ratings, it could hire Joel Stern MBA (Chicago, 1964), a guy I believe knows something. Why did Moody's hire a law firm? Most likely: to fend off anticipated litigation. Hey Mike Garcia (MG), if Moody's knowingly and willfully let its ratings be comunicated to any third party while knowing they were wrong, can you indict it under 18 USC 1341 or 1343, mail or wire fraud? Get out "Ed McMahon's hermetically sealed mayonnaise jar", see my 14 and 28 April 2008 posts.

Does anyone remember Arthur Andersen, once auditor for Enron? I think it's time the SEC yanked Moody's and S&P's NRSRO status. This is an outrage. MG, start drafting those indictments. Now! "[D]ecided not to", how interesting. Hey MG, if we have two or more predicate acts within a ten-year period, relatedness and continuity and threats of continued criminal activity, can we state a RICO count? Remember RICO, 18 USC 1961-69? Will this Moody's and S&P outrage be left to the plaintiffs' bar to respond to with a civil RICO action? Hey Chris Cox, hey Walter Riccardi (WR), what if anything will you do about this? Oh, I'm sorry WR, you're leaving the SEC. Did you send S&C a resume and is the fix already in? Imagine, the SEC has time to harass short sellers, but will likely do nothing about this. I wonder if the SEC will harass Moody's short sellers?

No comments: