Awaiting the Lehman CDS Auction

Here's an interesting theory: the reason that bank lending has ground to a halt is that everybody's waiting until the results of the Lehman CDS auction, which is currently scheduled for October 10. "Banks are hoarding cash," says Elizabeth MacDonald, "in expectation of expected payouts on anywhere from $200bn to $1 tn-no one knows the amount, adding to volatility-for these defaulted credit derivatives".

I think there might be a germ of truth to this, but that it's largely a second-order, rather than first-order, effect. And I have a slightly peculiar reason for being a little hopeful: the banks are not triple-A rated, and never were.

If I recall correctly, back in the early days of the derivatives market, banks would set up bankruptcy-remote triple-A subsidiaries to buy and sell derivatives; the reason was to reassure their clients about counterparty risk. But clearly the derivatives market couldn't really take off if everybody had to be triple-A. So a lot of derivatives trading moved to exchanges, and other protocols sprung up in the OTC markets -- mainly involving the posting of collateral.

But some triple-A writers of derivatives remained. And as Bill Ackman said at the Value Investing Congress yesterday, all of the big blowups in the CDS market -- AIG, MBIA, Ambac, etc etc -- were caused by triple-A entities. (He ignored Bear Stearns, but as far as I know, Bear didn't lose money on its CDS portfolio.)

Ackman's point is that the discipline of posting extra collateral when the CDS price moves against you actually worked. And the only place where it failed was the one place where it didn't exist: customarily triple-A entities didn't have to post collateral, precisely because they were triple-A.

Which brings us to Lehman. Is it really the case that a lot of banks wrote credit protection on Lehman brothers, without hedging their exposure? I doubt it, somehow. Banks have more than enough counterparty risk with other banks as it is, they have no need to gratuitously add to it by writing CDS.

Now there are people who made money betting against Lehman Brothers by buying default protection. And since the CDS market is a zero-sum game, there must therefore be people who lost money by selling that protection. The $400 billion question is whether they have the wherewithal to make good on their obligation. (And remember that $400 billion is a gross number: the net exposure -- the total amount that some people made and others lost -- is much smaller.)

I'm optimistic on that front: I think the answer is yes, although it might well involve selling collateral and other securities in order to come up with the cash. So there could be some nasty liquidation events on or around October 10. But I suspect that a lot of the exposure to Lehman came from synthetic bonds, CDOs of CDSs, and that kind of thing -- in other words, it resides on the buy-side, not on the sell-side.

It's always possible that some hedge fund somewhere will find itself going bust as a result of writing protection on Lehman -- but so far the big hedge-fund returns on CDS have been positive (Paulson, Lahde) and not negative. I'm holding out hope that the same will hold true on October 10.