Frannie's Future

The NYT talks today of "Mr. Paulson's long-term desire to wind down the companies' portfolios, drastically shrink the role of both Fannie and Freddie and perhaps eliminate their unique status altogether". It's not a bad idea, and I like the way he's going about it: let them increase their portfolios modestly from now until 2010, and thereafter force them to downsize dramatically.

They say that reforms only come in a crisis, never during the years of plenty -- but in this case the marginalization of Fannie and Freddie during the housing boom would probably, on net, have been a bad thing. Insofar as anybody was sticking up for old-fashioned underwriting standards, it was the GSEs, and I suspect that if they hadn't been around, prime underwriting would have suffered the same race to the bottom as subprime.

So what will the world look like once Fannie and Freddie shuffle slowly off the stage? Richard Green says that their great achievement was "the transfer of interest rate risk from households to investors" -- which is indeed a clever and useful thing to do. And the absence of US-style high loan-to-value fixed-rate mortgages elsewhere in the world, even in countries with well-developed mortgage securitization apparatuses, would seem to imply that if and when the GSEs aren't around any more, those cheap fixed-rate mortgages won't be around either. Sure, you'll be able to get a fixed rate -- but you'll pay for it with a significantly higher interest rate than that available on a floating-rate mortgage. And as a result, the fixed-rate mortgage won't be nearly as common as it has been until now.

In the meantime, however, the agencies will continue to operate -- and Nouriel Roubini, for one, is most unhappy with the structure of the bailout. I agree with this:

While not hitting the unsecured debt holders may have made some sense (as a lot of the agency debt is held by foreign central banks, sovereign wealth funds and other investors who would have fled the agency market if they had been subject to a haircut) not touching the subordinated debt of the GSE makes no sense; that is another additional bailout of a category of agency creditors that adds to the fiscal cost of the bailout.

I'm a bit unclear on who exactly owns Fannie and Freddie's subordinated debt, but I suspect a lot of it is in the US banking system, and wiping it out could have had a much harsher effect on that system than simply imposing a haircut on preferred shares. Still, if you're going to add to the fiscal cost of the bailout, then it's always preferable to do so in a truly transparent way: make the sub-debt holders take some pain, and then bail them out if necessary.