wilmott.com

Hi guys, I am wondering if I could pick your brains on the ISDA CDS Model....Suppose if I sell 5y CDS protection on a particular name, and the spread is say 700 bps, but the standard coupon is, say 250. That means that uner the SNAC protocol, I should receive 450 bps worth of upfront payment as well as 250 bps running for 5yrs. However, let say that suppose I have the option of either taking the upfront and receive 250 running or receive no upfront but take the full 700 running, under what circumstances would either be optimal? Presumably, I would expect that if there is a very high likely hood of default in the near term, it would be bext to accept the upfront even though they are "equivalent" according t the model. It would also depend on the term structure of the credit curve I assume? So that, if we have the full credi structure, it would be possible to determine whether taking "upfront plus standard coupon" or "no upfront and full coupon" is optimal?Thank you for your time on and attention on this....

I think there is a kernel of an interesting question lurking here, but it is a little hard to get at. First, you don't have the option to choose between receiving the quoted spread or the corresponding upfront. And 250 is not on the ISDA menu of spreads... But, you could imagine a situation where a high yield-ish (or should that be high-ish yield?) name might simultaneously trade with 100 and 500 bps running. In this case you would in fact have a choice that is roughly equivalent to the one you posit. But, you would not see a single spread quote that would apply to either. The contract with 100 bps running would be quoted as a spread, which will then have to be painstakingly translated into the actual upfront amount, whereas the 500 bps contract would be quoted as a price (100 minus the upfront amount). Each of these quotes will presumably (if dealers do their job well) fully incorporate all public information about the credit term structure. So, unless you have private information about that, the two alternatives should be priced consistently.

Could the CVA-adjusted prices of these CDSs (assuming they can be both traded with the same counterparty) be different?

The two contracts will definitely have different counterparty risk profiles, so if the quotes are consistent on a fully collateralized basis they will not be (in general) in the presence of counterparty default risk.

The ISDA CDS model is a very simple model, it is just used to compute the upfront corresponding to a market quote,but even on markitMarkit closing price, you can see that several credit entities have quotes for different CDS coupons : Volvo is quoted for 25, 100 and 500 margins and the corresponding ISDA model prices are 203, 201.5 and 192.The ISDA model follows a Bond logic : price a financial asset using only one parameter, and for a Bond the yield is ok to price, but not for relative value analysis, it is the same for CDS.To get all these prices right, you will need to have a curve of par CDS and an asumption for the recovery in line with the market (and not the very simple recovery assumption of the ISDA)