I will miss Neil Barofsky, who announced last month that he will leave his post as Special Inspector General of the Troubled Asset Relief Program, for many reasons. But one stands out: He recognized that dragging borrowers through a government-run battery of trial loan modifications -- as has been done under the Home Affordable Modification Program (HAMP) -- is a form of cruelty:

Failed trial modifications often leave borrowers with more principal outstanding on their loans, less home equity, depleted savings, and worse credit scores. And even in situations where they never missed a payment, such borrowers may face back payments, penalties, and even late fees that become due once their trial modification is cancelled. The impact of these added burdens becomes even greater when trial modifications are allowed to continue long past the three-month period called for by the program. While it may be true that some homeowners benefit from "temporary relief" of a trial modification even though the modification ultimately fails, Treasury's repeated references to the benefits of failed modifications ignores the real and often debilitating harm that such modifications have inflicted on many families, and appears to be little more than an attempt to define specific failures as successes.

In today's testimony [pdf] to the House Committee On Financial Services, Barofsky described the failure of HAMP, which is on track to produce permanent modifications of 700,000 to 800,000 mortgages – far fewer than the 3 million to 4 million originally promised. (Barofsky, giving a high estimate of the shadow inventory, refers to "13 million homes" actually headed toward foreclosure.)

Redefaults were the undoing of HAMP. To the extent the program can claim any success, it's that fewer loans modified under HAMP go bad again than go bad in the industry as a whole. But this slight good news is outweighed by the enormous amount of taxpayer money you have to pay in order to cover a reduction in outstanding principal -- the "cramdown" that is the only effective means of improving loan performance. The stuff Barofsky mentions about the three-month period is interesting if you follow redefault rates (as I do from time to time). There are important probability matters around how long you continue with a trial, and borrowers who go beyond 90 days delinquent are at almost-certain risk of permanent default. Barofsky also mentions 792,000 cancelled modifications and 152,000 modifications “in limbo.” It would be interesting to know whether any of those churn modifications are falsely depressing redefault rates.

Not that any of this stuff is meant to be looked at too closely. Barofsky also describes the steadfast refusal to adopt benchmarks for HAMP and tries to follow Treasury’s regularly changing criteria for success:

In recent months, Treasury's evolving defense of HAMP has featured the claim that HAMP has had a beneficial impact on private modifications that occur outside of the HAMP program. This too is a questionable measure of success. While Treasury may deserve credit for having had a positive, if inadvertent, impact on industry practice, according to a December 2010 COP report, "when pressed, Treasury acknowledges that there is no clear causal link between HAMP and proprietary modifications." Furthermore, while data suggests that proprietary modifications have generally improved from the homeowner's perspective since the launch of HAMP, the terms of such modifications are typically far less advantageous, often including some unfavorable terms for the borrower, higher rates of redefault, and broader imposition of servicer fees that are specifically prohibited in HAMP. In other words, it is odd for Treasury to celebrate modifications whose terms would largely be unacceptable from both the borrower's and Treasury's perspective in HAMP.

More on Barofsky and on Mortgage Metrics.