Morning graph

From a new dataset of CDS prices I'm working with -- this suggests that even though the credit crisis began in the summer of 2007, the market didn't pick up on Lehman's problems until Spring 2008, and even then really didn't see a problem until a few weeks before the petition date. Aren't these things supposed to help with price discovery?

I've got Morgan Stanley on the graph too for context -- Lehman's a bit more volatile, but not notably different.

Comments

[Adam Levitin]

Quite interesting. But are we sure we really know which way the causation does with CDS pricing? Conventional thinking is that CDS spreads reflect risk on the reference asset. But CDS are also a way of shorting an asset, so as shorting demand grows, that could result in other investors getting worried and selling off of the asset itself, which would place downward pressure on the asset price.

I wonder how much of Lehman CDS pricing can be explained by having one big sucker around (AIG)? If AIG was willing to write CDS on Lehman like a drunken sailor (mariners having a well-known propensity to spontaneously write CDS when inebriated) wouldn't that mean that the AIG price would become the market?

Also it's worth looking at the ABX, an index of CDS on subprime mortgage backed securities. The ABX (price of CDS on subprime mortgages) spikes at the time of Lehman's bankruptcy. It's not as if Lehman's collapse resulted in worse mortgage performance, and Lehman wasn't guaranteeing any of the mortgages in the ABX reference pools. The only explanation that's left, I think, is that CDS prices are as much a function of counterparty risk as of reference asset risk.