CDS Didn't Bring Down Bear and Lehman

I'm thinking I should institute some kind of CDS Demonization Watch: this meme is only going to grow and grow. And it's spreading, too, into the wonkier areas of the financial press -- people and publications on the ought-to-know-better list, such as Nathaniel Baker, of The Deal:

The bankruptcies of Lehman Brothers Holdings Inc. and Bear Stearns Cos. among others were not caused by personal bankruptcies but by haywire derivatives contracts. Specifically, the credit default swaps financial institutions were relying on to protect them from subprime exposure turned out to be worthless. Many financial institutions on Wall Street and elsewhere might have even profited from the subprime crisis, had there been a settlement mechanism in place for CDS, as there is for options and other derivatives.

For "personal bankruptcies", here, read "foreclosures", which are much the same thing, and you've got yourself an almost perfectly wrong-headed argument. Did a wave of foreclosures help to bring down highly-leveraged institutions with significant real-estate exposure, among them Bear Stearns and Lehman Brothers? Yes. Did "haywire derivatives contracts" in general, and CDS in particular, play a much bigger role? No.

How do we know this? Well, just look at the magnitude of the exposures that Baker is talking about. Back in January, Bernstein Research analysts totted them up, and came to the conclusion that Lehman's unsecured exposure to triple-A counterparties in general -- not just the monolines -- was $4 billion: large, but certainly not large enough to bring down a bank with a balance sheet of over $600 billion. Bear Stearns's exposure was smaller still, just $330 million. What fraction of that exposure eventually turned up on the banks' income statements as a mark-to-market loss? That I don't know, but it's not necessarily very large: remember that AIG's troubles only really snowballed after Lehman and Bear had gone under -- and AIG was by far the largest triple-A writer of CDS.

And it's simply silly to assert that a CDS settlement mechanism could singlehandedly have seen Lehman and Bear make money, rather than lose it, from the subprime crisis. Does Baker think those banks fully hedged all their real-estate exposure in the CDS market? Even Lehman, whose exposure was primarily commercial, rather than residential? Of course they didn't: they drank the kool-aid as much as any of the clients to whom they were trying to sell mortgage-backed securities.

It's worth emphasizing that the CDS demonization meme, at least in this form, is a dangerous one -- because it implies that it wasn't really the banks' own fault that they went bust, and that the implementation of a CDS exchange could in and of itself bring the amount of systemic risk down substantially. Neither is true. By all means fiddle around with the CDS market; it might well do some good. But don't try and pretend that if we'd only done so sooner, Bear and Lehman might now be thriving.