In our efforts to understand what went wrong at the credit rating agencies, we visited Frank Raiter. He worked at Standard and Poors but is now far from Wall Street, happily clearing land for a farm in rural Virginia and caring for a turtle in his office.
Raiter has been critical of his former employer, who he says rejected his requests to get better data and build better models. S&P disputes that.
In the interest of letting you all make up your minds, here's a link to video of an interview and lecture Frank Raiter gave at Cornell University.
And here's part of a response S&P gave to Congress after Raiter testified on Capitol Hill.
As we understand it, Mr. Raiter's claim is that S&P refused for commercial reasons to adopt an allegedly superior "new" model to analyze the credit risk of mortgage loans underlying certain RMBS transactions. The allegation is baseless.
While S&P did undertake work on developing equations based on large volumes of loan data earlier this decade, the effort ultimately did not bear analytic fruit. Contrary to Mr. Raiter's allegations, the reason was not budgetary or commercial. Instead, no analytical consensus was ever reached that the work produced results that could be relied upon in S&P's ratings analysis. In fact, despite continual testing and review, this "model" repeatedly produced fundamentally counterintuitive -- and, in the view of our analysts, insupportable -- results. For example, the results predicted that adjustable rate mortgages were less likely to default than fixed rate mortgages. Subsequent history has obviously dispelled that notion.
Moreover, contrary to other charges Mr. Raiter has made, S&P has repeatedly updated the models it does use to analyze these loans to reflect new information, including new risk factors we see. Indeed, since 2001, S&P's LEVELs model has been updated 16 updates to LEVELS we have implemented since 2001, 10 of which have occurred after Mr. Raiter left our company in 2005. A number of these updates specifically addressed the increased risk we saw with subprime mortgages. In retrospect, as we have repeatedly acknowledged, it is clear that some of these mortgages have performed worse than we forecasted they would. However, to suggest, as Mr. Raiter has done, that this in any way resulted from an unwillingness on S&P's part to try to take appropriate action is entirely unfounded.
You can read the Securities and Exchange Commission's investigation into the rating agencies.
And here is Frank Raiter's testimony before the House Committee on Oversight and Government Reform.