Betting on yourself
Robert Waldmann of Angry Bear has a fascinating post exploring the possibility that sharp movements in the value of Lehman senior debt could be explained by the possibility that Lehman had sold Credit Default Swaps on itself. Since a CDS is insurance against the possibility of default on debt, this is a no-lose bet for Lehman. If the firm survives, they collect the premiums and pay nothing and, if it doesn’t the losses are borne by the creditors. And, as Waldmann points out, it’s not crazy to buy such a CDS, since it will retain some value in bankruptcy. If you’ve already sold a lot of Lehman CDS yourself, there’s a significant hedging benefit. So both parties benefit, and the losers are the existing bondholders. Waldmann has an interesting optimization exercise to show that optimal (for Lehman) use of the CDS option could explain the collapse in the value of Lehman bonds.
Thinking about this, I’m more and more convinced that Warren Buffett’s description of derivatives as financial weapons of mass destruction applies in spades to CDSs.
For a start, it seems obvious that allowing firms to sell CD swaps on themselves is a terrible idea, but Waldmann says it’s not illegal and has happened in the past. But even if you could stop the most obvious version, there are plenty of ways around it. Suppose for example that Bank A and Bank B sell lots of CD swaps on each other. Then if one gets into trouble so does the other, and both default, so the CDS are reduced to their bankruptcy value. As long as both banks survive, it’s money for jam.
What this means is that it’s not possible to value a CDS (and therefore any kind of debt for the issuer) simply by looking at the risk of default for the insured firm, and then considering the risk of default for the issuer. You need to know the correlation between the two, and this requires not detailed knowledge of the entire asset position of both parties, including their correlations with other parties.
Presumably, at some point, bond markets will start to anticipate this risk. But, in the absence of detailed knowledge of all the counterparty risks involved, the only real option is not to buy corporate bonds of any kind. That’s pretty much what’s happened in recent months.
A centralized CDS clearinghouse might address these risks to some extent. Alternatively, if the risks were ignored, it could create yet more systemic risk by allowing arbitrage between supposedly identical CDS instruments with radically different counterparty risk. I’ve seen a bunch of proposals for such a market to be established and some announcements that its coming Real Soon Now, so perhaps we’ll get more information on this point before long. Then again, perhaps not.