For those who haven't read it yet, the FT reports today that Moody's 
Investors Service gave incorrect AAA ratings to billions of dollars 
worth of a complex debt product because of a bug in its computer 
models. 

Some senior staff within Moody's knew that products rated the previous 
year (i.e. 2007) had been given the top rating, and that after a 
computer coding error was fixed, their ratings should have been as much 
as four levels lower.

As an extension, and my personal opinion, this is also why a broadly 
used backtest-engine should be preferred to the usual in-house one. 
Then again, I don't think all public backtesters fit the bill (you know 
what I mean).

PS


Reply via email to