The Treasury's Making Home Affordable Program (HAMP), designed to prevent foreclosures, continued to struggle in July. Fewer trial modifications were started and made permanent last month than in June, according to the latest report (.pdf). There were also more cancellations. Although the program continues to slow, foreclosures remain high.
HAMP's performance last month can be summed up through this chart:
You can see that the gold line for trials started and the green line for modifications made permanent both ticked downward last month. New trials hit a new low, while monthly loans made permanent nearly matched the December 2009 level. The red line shows total cancellations in July eclipsed June's number. This indicates more borrowers with either a trial or permanent modification left the program last month. Obviously, these trends are all the opposite of what the Treasury wants to see.
As for cancellations, they're becoming increasingly significant. At this point, a whopping 47% of trial modifications started have been cancelled. In July, there were nearly twice as many borrowers with modifications cancelled as there were borrowers who obtained trials or permanent modifications combined. The number of permanent cancellations, in particular, was the highest we've seen yet at 4,089. The previous high occurred in May with 2,613. So you can see how big of a number July's was compared to prior months. Remember, borrowers have to pass through a number of filters before they achieve permanent status, so very few cancellations should occur at that point.
The Treasury's Take
Defending a program performing so poorly is sort of like trying to argue that the Pittsburgh Pirates are actually a pretty good baseball team despite their 40 and 81 record. Nonetheless, on a conference call Friday, the Treasury did its best to explain reasons why the foreclosure prevention effort really wasn't as bad as it looks.
First, the officials said that the high cancellations aren't likely to continue forever. They explained that many borrowers initially qualified for trial modifications based on stated information. When they were required to actually prove things like income and primary occupancy, many failed to do so and were kicked out of the program. The renewed effort on documentation this spring caused a steep rise in cancellations that the Treasury believes will subside in coming months.
Second, Treasury officials argued that HAMP served as a novel approach to mortgage modification that all other private servicers used as a model for their own proprietary programs. They stressed that many borrowers who fail to ultimately qualify for a government modification manage to obtain one in precisely one of those private programs that HAMP made possible. It seems this would be a case of the student becoming the teacher, however. Why are servicers' proprietary modifications able to catch struggling borrowers that slip the holes in HAMP's net? Why are private sector programs succeeding where HAMP is failing?
The Delinquency Revision
If you follow HAMP, then you may recall the controversy last month, when the Treasury realized that a chart detailing HAMP modification delinquency was inaccurate. It turns out that officials quietly revised (.pdf) the figures earlier this month. It's unsurprising that they didn't announce the revision with much fanfare: the new numbers were worse. Here is the revised chart:
And here's the original:
Certainly, the revised chart looks a little bit more reasonable. For the oldest vintage shown, Nine months after becoming permanent, the delinquency rates are now stated as 19.6% for the 60+ days delinquent bin and 14.9% for the 90+ bin. That's much worse than the figures of 7.7% and 2.4% first reported, respectively.
But in the context of re-default, the revised numbers still aren't too bad. Analysts expected the default rates to be in excess of 50%, which makes 15% look pretty great. Of course, there's still plenty of time for these loans to fail, as the modifications tracked here aren't even a year old. Still, this performance is relatively good so far.