海角大神

I don鈥檛 know if the President will say much about housing, but there are some important and potential helpful policy choices percolating in the background.

I鈥檝e long held that of all the stuff on the White House鈥檚 鈥渨e-can鈥檛-wait鈥� list鈥攖hings they can do to help the economy and jobs without going through that legislative death trap formerly known as Congress鈥攈ousing policy is the one with the greatest potential to actually move the needle.

And the most helpful policy in housing is the reduction of mortgage principal for underwater homeowners.聽聽 Research has clearly revealed that owing more than the value on your home is the strongest predictor of foreclosure, and housing finance analysts widely agree that principal reduction is the best medicine to avoid this outcome.

But what does any of this have to do with stuff we could actually do right now?聽 Good question.聽 The answer is that the Federal Housing Finance Agency, which regulates Fannie Mae and Freddie Mac, could quickly reduce the principal on millions of home loans they own or insure, without going through Congress.

So, why haven鈥檛 they done so?聽 Another fine question.聽聽 First, you need to recall that Fan and Fred are 80% owned by the US gov鈥檛 right now, and FHFA, as conservator, wants to protect the taxpayer.聽 That鈥檚 fine鈥攚e thank you, FHFA.聽

But鈥攁nd news accounts have been getting this quite wrong鈥擣HFA believes that loan forgiveness (principal reduction) would only save the taxpayers $20 billion while loan forbearance would save $24 billion (the latter modifies the loan, it does not reduce it).聽聽

In other words, the FHFA agrees that both types of loan adjustments would reduce defaults and thus reduce losses to taxpayers, with a slight advantage to forbearance, which, as I鈥檒l argue in a moment, is very likely incorrect.聽 I think if you did the analysis right, forgiveness would trump forbearance by a long shot.聽 But given the fact that reduction would clean this mess up a whole lot faster and more reliably than just changing the terms of the loans, and that taxpayers save either way, the path ahead鈥攖oward forgiveness, not forbearance鈥攕hould be clear.

Unfortunately, the FHFA is placing landmines in that path.聽 Based on a letter reviewing all this by FHFA acting director Ed DeMarco, news accounts like or are reporting that if Fan and Fred were to reduce the principal on a subset of the mortgages they own or insure, it would cost聽taxpayers $100 billion.

This $100 billion (it鈥檚 actually $102bn), however, is a gross number鈥攊t is the losses to the agencies, and the taxpayers, from all the mortgage defaults that FHFA expects to occur if they neither forbear nor reduce principal.聽 The relevant numbers, however, are the difference between the losses under a forbearance program ($78 billion), or a reduction program ($82 billion) and the cost of doing nothing.

The punch line, then, is that by their estimates, forgiveness saves the taxpayer $20 billion; forbearance, $24 billion.

But for a number of reasons, FHFA鈥檚 methods make forbearance look better than it really is.聽 This is some weedy stuff, but it matters:

鈥搕hey use a state level price index rather than a localized price level.聽 This approach averages across cities with huge price drops and those with normal price declines, and thus reduces the number of the deeply underwater borrowers.*聽聽 That in turn understates the impact of the policy most helpful to those borrowers: principal reduction.

鈥搕hey use FICO credit scores and debt-to-income ratios at the time of loan origination rather than where those measures are today.聽 Obviously, they鈥檙e worse today, so this makes the agencies鈥� book look better than it really is, and again, understates the benefits to principal reduction.聽 In other words, the way they do it artificially lowers their expected default rate, and so the policy that鈥檚 most effective against defaults for those with lower FICOs and higher DTIs gets less credit than it should.

鈥搕hey assume that all of their debt forgiven in their forbearance programs is repaid鈥�100% of it.聽聽 That鈥檚 not realistic and it significantly reduces the cost of this option. 聽 Simply building in a realistic default rate for debt that鈥檚 been pushed back to the end of the loan would raise the cost of forbearance relative to principal reduction.

Any one of these changes will sop up the $4 billion difference in an NY minute, showing forgiveness to dominate forbearance.聽 But even if the FHFA wants to stick with their numbers, reductions will go to work much more quickly and effectively to prevent defaults.聽

If they keep coming up with reasons not to do the right thing, the White House should do the right thing and replace DeMarco鈥攁 perfect good guy who believes 丑别鈥檚 doing the right thing here but isn鈥檛鈥攚ith someone who gets the urgency of the situation.

*Imagine a) that anyone with a home price decline of 30% is underwater and needs a loan mod, and b) a state has two homeowners in two different cities.聽 Homeowner A鈥檚 price went up 30%, homeowner B鈥檚 price went down 30%. Average them together across the state and no one needs a mod; use the local price index, and B should get one.