by Tanta on 7/14/2008 03:57:00 PM

Sez Bloomberg (thank you, Brian!):

July 14 (Bloomberg) -- More than two of every five subprime borrowers whose mortgages were reworked in the first half of 2007 are defaulting anyway, Moody's Investors Service said.

Among subprime adjustable-rate mortgages modified in the first half of last year, 42 percent were at least 90 days late on March 31, the ratings firm said in a report today.

Modifying loans granted to consumers with poor credit records has gained favor as record numbers fail to keep up with payments and home prices tumble. Loans reworked more recently may perform better than ones modified in early 2007 because lenders are increasingly lowering interest rates and offering changes to consumers with fewer missed payments, Moody's said. That's different from 2007, when lenders focused on enforcing repayment plans.

I have not yet gotten my hands on the detailed Moody's report on this subject. However, I did look at Moody's press release, and it doesn't exactly attribute the issue for the early 2007 modifications to repayment plans. Per the press release (no free link), the majority of modifications done in the first half of 2007 involved simple deferral of principal/capitalization of past-due interest (that is, the borrower got "brought current" by having past-due interest added to the loan balance) without other changes in the loan terms. Modifications like that result in the same or even a slightly higher monthly payment than under the original loan terms.

If modifications processed in the second half of 2007 and later mostly involve 1) significantly lowered payments and 2) significantly less delinquent loans, then certainly theory predicts they will have a better re-default rate. On the other hand, it's early to be confident that the 40% redefault rate on the earlier mods will hold; generally speaking you need at least two years of "seasoning" on a group of modifications before you get useful numbers on re-default. That said, Moody's servicer survey from last year predicted a redefault rate of around 35% based on past experience of the servicers. It will be curious to see how high that redefault rate can go before the "least loss" models tip back toward foreclosure. So far Moody's is simply saying that the impact on cumulative losses of the early modifications has been "modest." That suggests to me that it may not take a much higher redefault rate for this "modest" lowering of cumulative losses to disappear.

I know that I for one am looking forward to Hope Now and the OCC to start reporting on re-defaults in their metrics (although I'm not going to hold my breath). Moody's is reporting strictly on subprime ARMs; while these have been the focus of modification efforts, we still need to know what's happening in the prime and Alt-A segments.

Otherwise, Moody's reports that as of March 2008, nearly ten percent of subprime ARMs with a reset date in the preceding 15 months had been modified.