David McNew/Getty Images
A recent ProPublica/NPR report on Freddie May refusing to refinance mortgages for struggling homeowners shows that the market is still coming to terms with new ways of measuring risk.
Jesse Eisinger and a ProPublic co-author, Cora Currier, along with NPR's Chris Arnold, have followed up on their original story about Freddie Mac allegedly betting against homeowners being able to refinance their mortgages. To summarize without getting too deep into risk-mitigation instruments and complex financial jargon, Freddie was using these things called "inverse floaters," and more or them than Eisinger originally reported ($5 billion), to...well, what exactly? Eisinger argues that they were being used to bet against homeowners refinancing out of high interest rate mortgages — a neat trick, given that Freddie could set the refinancing rules.
Some bloggers, myself included, have asked whether this really what was going on. Eisinger posted a lengthy comment on my blog and also Felix Salmon's blog at Reuters, helpfully addressing many of the issues that the debate over the story has raised. Felix fires up his analogy-o-matic and provides a good, simple explanation of what Freddie was up to (it involves, cleverly, a real-estate hook). Ultimately, he agrees with Eisinger:
It seems to me that Eisinger is right and that Freddie is violating the spirit of the Treasury’s instructions. Treasury wants Freddie to sell down and derisk its balance sheet. Freddie, in response, started selling down its balance sheet, but kept as much risk as it possibly could, in the form of inverse floaters.
And does anybody really believe that Fannie and Freddie should be taking on more risk, in relation to the size of their balance sheets? I’m sure that doing so is good for the annual bonuses of someone getting paid $2.5 million a year to run Freddie’s mortgage portfolio. But it’s unlikely to be a good idea for anybody else.
I now think that what this all boils down to is whether Freddie was taking on greater risk to "bet against" a refi process that it could control (very bad) or whether Freddie was using inverse floaters for what I understand they're typically used for, which is to work in a pairing with a regular floater to make money on interest rate movements (not nearly as bad). I don't think the refi/no-refi of the Silversteins, the couple profiled in the original ProPublica story, is even relevant anymore. It would be great if they could get a refi — although who knows if a bank, rather that Freddie, would consider them creditworthy.
The bigger issue — and this is one for the FHFA, Freddie's regulatory master, to deal with — entails Freddie's motives. I'm not sure that a taxpayer-owned entity should be slicing off the riskiest portion of its portfolio to play the interest rate environment. This is quite apart from whether anyone at Freddie or the FHFA could have foreseen that doing so would make it look as if Freddie, by tightening refi rules, was stacking the deck in its favor.
Anyway, it sounds like it's game over for Freddie and the inverse floaters. Which if I'm right means that the agency can get back to not doing a refi for the Silversteins, for perfectly normal reasons.