Commentor Sandrew was feeling overlooked because I didn’t pick up on his comment on the Felix Salmon post, a comment that hit the key point in my later response to Salmon. Here’s his comment:
I think you may be focusing on the wrong transaction costs. To pierce the Coasean retort, one must demonstrate that frictions exist to impede the unwinding of CDS and the transfer of bonds. Pirrong argues that if such transactions are efficient, then the perverse incentives that bond hedging via CDS create can be efficiently diffused.
That said, I’m not convinced that the bond and CDS markets are as well lubricated as Pirrong seems to suggest. To me, the enormity of the CDS market is as much evidence for the presence of significant CDS unwind costs as it is for the popularity of the instruments. If unwinding or novating CDS was so cheap, why did the aggregate notional amounts balloon to umpty-trillion bucks over the course of the past decade?
Sandrew’s comment re the Coasean retort is correct. Re his second paragraph, well, I don’t really take a stand on how well “lubricated” the CDS market is, except to say that if the empty creditor problem is a really big deal, and the bad incentives resulting from hedged bond positions lead to huge costs, in my view it is likely that the CDS market is sufficiently lubricated to permit market participants to avoid such costs. I note that re the ballooning of aggregate amounts, etc., when it became apparent that there were large costs associated with failing to close offsetting positions, market participants acted with alacrity and did large amounts of tear-ups through TriOptima, etc. As a result, notional amounts have fallen dramatically (by about 40 percent if memory serves).
Thanks for the comment on Salmon’s blog, Sandrew. You are on point, and get props from me, if belated ones.
So, Sandrew–consider your head patted;-)