A Risk-Sharing Postscript

This Wednesday, five respected housing policy experts—Jim Parrott, Lew Ranieri, Gene Sperling, Mark Zandi and Barry Zigas—together released a paper titled “A More Promising Road to GSE Reform.” In it, they propose to “merge Fannie and Freddie to form a single government corporation, which would handle all of the operations that those two institutions perform today, providing an explicit federal guarantee on mortgage-backed securities while syndicating all noncatastrophic risk into the private market.”

I will have more to say about the overall “Promising Road” proposal later, but since my previous post was highly critical of Fannie Mae’s Connecticut Avenue Securities transactions, and programmatic securitized risk-sharing forms the core of the secondary market system put forth in this paper, I wanted to briefly comment on that aspect of it.

We just now are crawling out from under the ruins of our last experiment with securitized risk sharing—the collateralized debt obligations (CDOs) that were supposed to be the answer for how to finance the riskier tranches of private-label securities (PLS)—and already we see a proposal to give a variation on that theme another try.

I’m generally not one to take shots, but I also can’t resist irony. The “Promising Road” paper was put up on a website hosted by Moody’s—the same agency that, along with Standard & Poor’s, told us that the risk on individual low-rated tranches of PLS was not correlated but independent, so that you could put a pool consisting of nothing but these low-rated tranches together in a CDO and safely rate 80 percent of the new security Aaa. That worked terrifically well for as long as the market was spiraling upwards—then suddenly it didn’t work at all.

If past experience weren’t enough, there are warning signs everywhere about the new risk sharing deals Fannie and Freddie have been doing. The most prominent one is the makeup of the investor base. The vast majority of investors in risk-sharing securities are leveraged—mainly hedge funds, but also commercial banks. The “private capital” they bring to the table isn’t equity; it’s debt. Take the hedge funds. When they think credit risks are low, they view risk-sharing transactions as an opportunity to borrow at “LIBOR plus a little” and invest at “LIBOR plus a lot,” with little chance that credit losses will erode enough of their spread to prevent them from earning double-digit returns for their investors and “two and twenty” fees for themselves. The minute they think credit losses might spoil this trade they’ll stop doing it, and if they think credit losses are about to spike they’ll sell the deals they own, roiling the market for credit risk at exactly the wrong time in the housing cycle. And if they can come up with a way to short these deals—as with synthetic CDOs the last time around—they’ll do that, too.

The daunting problem to overcome with securitized risk sharing is that there is a limited universe of investors with an appetite for these transactions, and their appetite varies greatly depending on where we are in the cycle. Few of those investors know much about mortgage credit risk themselves, and many—as was the case with the buyers of CDOs—know next to nothing. Investors correctly assume that the institutions packaging and selling mortgage credit risk know more about it than they, the investors, do, so they require a significant premium to accept it, even in good times. And in bad times they want nothing to do with it.

There is no getting around the fact that for capital markets investors, credit risk-taking isn’t a business; it’s a trade. It always has been, and it always will be. That’s the compelling argument for having specialized institutions like Fannie and Freddie, and the mortgage insurers, grade, price, hold and manage mortgage credit risk. These institutions know the risks, and are able to diversify them across product type, loan characteristic, geography, and time. They’re closely regulated—unlike capital markets investors, whose activities in the credit risk market aren’t regulated at all—and they can and should be required to hold capital adequate to their risks. They don’t get surprised when the cycle turns on them, and, more importantly, don’t cut and run when that happens. Credit risk management is their primary business, and they’re in it for the long haul. The financial system benefits from that.

The authors of “Promising Road” greatly underplay the dangers of making a $10 trillion market crucial to the functioning of the U.S. economy dependent upon the mercurial attitudes and preferences of leveraged investors like hedge funds. They confidently state, “Mortgage rates in the proposed system would be no higher on average through the business cycle than those in the current system (see Box 1).” Although Box 1 might tell us that, the markets and our experience tell us something very different. If we truly do go from our last securitized risk-sharing disaster to another one, we’ll have no one to blame but ourselves. We’ve seen this movie before.

I’ll have my own proposal for mortgage reform on this site by the end of next week.