Yesterday, data from TransUnion came out showing that the types of loan modifications favored during the foreclosure crisis, all of which fell short of principal reductions, provided only temporary relief and performed badly over time, with a very high re-default rate. So what about the flip side? What do we know about principal reductions and the re-default rate?

There haven’t been that many, so the sample size is pretty low. But as Amherst Securities Group showed in a new set of research, what we know from that sample size is that principal reductions perform quite well.

Reducing the amount struggling homeowners owe on their mortgages is proving to be a more effective way to prevent foreclosures than other methods, such as reducing interest rates or postponing payments, a new report finds.

In a report presented this week, Amherst Securities Group said that when principal reductions brought mortgages near the home’s market value, borrowers were substantially less likely to fall behind on payments again and lose their homes.

Only 12% of borrowers who received principal reductions re-defaulted in 2011, Amherst found. That’s compared with 23% of borrowers who received mortgage modifications with interest rate reductions (but no principal reduction) and 30% who received forbearance, which postpones their debt repayment.

So that’s the data. When it comes to homeowners keeping their homes, principal reduction works, and crappy HAMP-style loan modifications do not.

However, when it comes to banks and servicers, principal reductions do not work, as they eat into profits. What’s more, crappy HAMP-style loan mods work decently, because they often forbear principal or capitalize late fees in ways that increase profits. What often works best is foreclosure, because the servicer gets first in line to recoup its costs in a post-foreclosure sale. So that’s why we get the policies we get.

Principal reduction as a percentage of overall modifications are up this year at the big banks, though overall modifications are down. However banks cannot reduce principal on loans owned or backed by Fannie and Freddie, and at this point that’s most loans. Gary Peters has a bill in the House, with Republican support from John Campbell, to implement a principal reduction program at Fannie and Freddie where the taxpayer shares in the upside appreciation of the home upon eventual sale. The overseer of Fannie and Freddie, the Federal Housing Finance Agency, has a report showing that, under current rules, principal reduction would save them $1.7 billion. But FHFA has been resistant.

There’s also the matter of principal reductions from the foreclosure fraud settlement, but as I have demonstrated, there’s less there than meets the eye.

The larger point, proven by the evidence, is that principal reductions are win-win. Investors and homeowners benefit. Servicers have screwed-up priorities and financial incentives, and FHFA is driven by ideology. So we’re not getting that win-win housing policy. The clear next step is not only to toss out the leadership at FHFA, but to break the servicing model which is holding back the housing market once and for all, so we can allocate these losses from the bubble collapse properly.