Friday I submitted a Proposal on Transforming CDS. I encourage you to read it if you hold just a passing interest in the current financial crisis. But the talk of moving CDS to exchanges has been gaining momentum, touted as the cure by SEC Chairman Chrisopher Cox.
If the man who advocated for deregulation of all things financial, and who was appointed to head SEC for the purpose of making it irrelevant, touts The Cure for the crisis that his agency helped create, we need to be automatically suspicious.
But let's go beyond philosophy, go into details and make a concrete case.
Exchanges reduce counterparty risk by being the sole counterparty; in turn, they protect themselves by imposing margin requirements on their client accounts. So far, in almost everything exchange-traded, this mechanism has succeeded. But this does not make it a fit for CDS due to its inherent high leverage on the seller side.
Say I sell on the exchange $1B CDS on XYZ, which is a decent credit and premium is $10M per year (100 bps). The exchange finds you as the buyer. Since you're liable for the $10M annual premium payment, it requires you to put up collateral of $10M. No problem.
But how much collateral does the exchange require me to post? Obviously the amount should be tied to the notional, NOT PREMIUM, since my liability is tied to the notional -- a fraction of it, but the fraction could range from 0.1% to 91%, as illustrated by recent Fannie and Lehman CDS auctions. 100% of notional? There's no way any seller could afford to sell under such collateral requirements, not at a price buyers would buy. Remember, my collateral would be 100 times of my annual proceeds from this transaction. Even 10 times premium would be too high for the seller. Even at 2 times sellers would cry for pain.
But 2 times premium is 2% of notional, and assuming recovery of 40% (another laughable character of CDS standard models), 1/30 of the seller's liability in case of default.
There may be some remedial measures, such as increasing collateral ratio as the credit deteriorates. But, unlike naked equity options, the payout for CDS is by nature sudden-death, a jump event. The day before Lehman bankruptcy, Lehman CDS was trading at 20% upfront. The next day, you're looking at 80% payout.
Convergence between collateral and potential liability on CDS sellers is not possible. Recovery rate upon default is an unsolved problem. Putting CDS on exchange forces the exchange to take the recovery risk.
Who fills the hole? The Exchange. What happens if The Exchange fails? Complete, certain failure of CDS market. In contrast, the fallout from Lehman bankruptcy we've seen so far is less than 100% in both probability and scale.
If we let the old players and the old guards design the new system, the exchange would be guaranteed by the government, with the exchange taking the profit in good times and the government taking unknown, undetermined risk of recovery rate for no return when disaster strikes. And the collateral requirement of sellers would be pathetically inadequate. Future systemic failure is assured.
Will the society allow this travesty to repeat itself?
Posts
4 comments:
What happened to the LEH settlement? No bank failures? No hedge fund collapse?
Three possibilities, Bob.
1. Nobody had sold huge sums of Lehamn CDS.
2. Buyers actually did their daily collateral adjustment and sellers actually hedged. So in the end nobody fell.
2. Somebody did get severely or fatally injured but it's not public yet.
I remain doubtful of the first two scenarios.
Thank you for your reply.
It is as it seems. Opaque. No one willing to admit defeat or not allowed to admit defeat for fear of others following suit.
Actually, now I think I may have missed a delay factor. Sellers who had to pay could've made the payment on 1021 with whatever cash (capital or loan, of course more likely the latter), and then proceeded to unwind and raise cash. I suspect the latest leg of the Great Unwind of equities, commodities, commercial papers, and carry trades since 10/21 has a lot to do with this undercurrent.
Post a Comment