Only 0.1% of Government Mortgage Modifications Cut Principal?

One of the chief criticisms of the government's Home Affordable Mortgage Program (HAMP) is that it hasn't been doing enough to convince underwater homeowners to rework their loans. The key to doing that is principal modifications. The Treasury essentially conceded this point in March when it announced a new initiative aimed at reducing more principal. Yet, a new report indicates that principal is rarely touched: it says as few as 0.1% of modifications rely on this method. Is HAMP really ineffective at lowering principal?

Is The Portion Really 0.1%?

Shahien Nasiripour at the Huffington Post brings the heat.* He cites the following chart from a report (.pdf) released this week by the Office of the Comptroller of Currency:

OCC Chart HAMP 2010-06.PNG

(Note: Presumably, the date for Dec. 31 above is a typo, as it must be the year 2009, not 2010.)

The line that he is interested in is "Principal Reduction." As you can see, during these two quarters, the OCC found that only 0.1% of HAMP modifications utilized principal reduction. This is incredibly low. But the statistic doesn't provide the whole picture.

According to a Treasury official I spoke with, the OCC wasn't using the full dataset, but a subset of the modifications made permanent over the time periods shown. And this is confirmed by the HAMP monthly report data. According to those reports, 62,196 and 163,863 modifications were made permanent in Q4-2009 and Q1-2010, respectively. But the chart above shows samples of only 20,679 and 99,980 for the associated quarters. The Treasury source says that some servicers were left out. If these servicers were the ones more more commonly providing principal reductions, then considering the full universe of modifications might produce a much different result.

So simply combining the OCC's limited performance data to conclude that only 120 loans out of the sample of 120,659 results in a 0.1% portion of principal reductions isn't necessarily right. In fact a whopping 105,400 loans aren't even considered. For all we know the principal reduction rate could be as high as 47%, given those unaccounted for loans.

Of course, it isn't. The Treasury admits that it's below 10%, but won't reveal the specific portion of modifications that cut principal. It could help its image a little by providing this stat to contradict the OCC's claim. But even without the Treasury revealing the number directly, it's certainly plausible that the portion could be higher than 0.1%.

Consider the Timing

It's also a little unfair to judge the Treasury on its principal reduction performance through March. It was only at the end of that month that it announced its intention to begin a new principal reduction program. The new incentives for servicers didn't take effect until June 3, and they won't be required to consider principal reduction until October 1. The assertion that the Treasury should have pushed for principal reduction earlier is a legitimate complaint, however. But since we know it didn't, a low portion of its modifications utilizing this method isn't shocking.

Principal Reduction Stats Not in HAMP Report

Principal reductions appear to be reported by the Treasury in the following chart each month (this one from March):

modification step chart 2010-03.PNG

They were 27.6%, right? Wrong. The "Principal Forbearance" actions reported here, in fact, aren't principal reductions. A Treasury official explained that this modification strategy takes part of what homeowner used to owe and places it in an account that doesn't carry interest. So it's really just another form of interest reduction.

If HAMP modifications do begin utilizing principal reductions after June, the Treasury should strongly consider adding in this data to its monthly report going forward. If its policy change is successful, then this additional data could help to silence some of its critics.

*Note: Just to be clear, Nasiripour does mention in his article that the Treasury disputes the data set as being non-representative as well. I didn't mean to imply that he necessarily accepted the OCC rate as fully representative of the entire population of modifications, just that he was the only one I had seen to break the story.