Friday, June 13, 2008

OCC versus Hope(less) Now

(Tanta @ Calculated Risk) The Washington Post picks up the brewing controversy over the rather significant mis-match between the foreclosure and loss mitigation statistics reported in the new OCC Mortgage Metrics Report and the Hope Now reports we've been seeing since January. Of course they got seriously scooped on this by Housing Wire, who had this story on Wednesday, but I guess now that it's in the newspapers it's got legs:

John C. Dugan, comptroller of the currency, which oversees national banks, said his agency found "significant limitations with the mortgage performance data reported by other organizations and trade associations."

"Virtually none of the data had been subjected to a rigorous process to check for consistency and completeness -- they were typically responses to surveys that produced aggregate, unverified results from individual firms," Dugan said in a speech in New York on Wednesday. "That lack of loan-level validation raised real questions about the precision of the data, at least for our supervisory purposes."

Dugan said in an interview that he was referring to information provided by groups such as the Mortgage Bankers Association, which reports a foreclosure rate widely cited by regulators and the media. A report by the Office of the Comptroller of the Currency calculated that the rate was higher based on raw data it collected from nine of the country's largest banks.

Dugan's comments also raised questions about the accuracy of the reporting from Hope Now, an alliance of mortgage firms and banks that was formed to help financially troubled holders of subprime mortgages. Leaders of the coalition, which was put together by the Bush administration, contend they have aided more than 1 million homeowners. Those figures were self-reported by lenders in response to the kind of surveys Dugan has faulted. . . .

In an interview yesterday, Dugan tempered the strong language he used in his speech. "It was not intended to be a criticism of what they are doing," he said of MBA and other industry associations. Their figures, he added, "get you in the ballpark . . . but we wanted to have a much more specific level of detail."

Banks and mortgage firms have widely varying definitions for what constitutes a loan modification for a struggling borrower and even define subprime mortgages differently. The lack of standards leave the data open for interpretation or manipulation.
I have already gone on record accusing Hope Now of using "weird numbers," and I hope the OCC eventually comes up with a clean enough database (the OCC database currently has 20% of loans classified for credit quality as "other" because they're missing FICO scores) and better definition of credit quality (currently they're using FICO only, not such things as documentation type or LTV/CLTV) that its numbers can provide a better baseline for measuring loss mit activities. I am also willing to observe that if the OCC is only noticing here in Q2 2008 that big bank databases are sloppy and inconsistent, the OCC certainly should "temper the strong language" a touch.

Ultimately, this is going to come down to new regulatory rules on data reporting and management for supervised institutions, as it should. The industry will whine about regulatory burdens, as it always does, but it will be hard to avoid the conclusion that "voluntary" reporting via the MBA or Hope Now is not producing reliable numbers. It will of course also occur to some of us that the OCC's supervision of these banks over the last several years has apparently been based in part on data that it never until now seemed to realize needed some cleaning up.

No comments: