How does one go about pricing these. I'm assuming that the default swaption type approach isn't suitable, because it's unlikely that all the credits will default. Is it different from an option on a stock index?

QuoteOriginally posted by: WibbleHow does one go about pricing these. I'm assuming that the default swaption type approach isn't suitable, because it's unlikely that all the credits will default. Is it different from an option on a stock index?what do you mean by default swaption approach?

Using something like Hull-Whites 2003 paper, valuation of credit default swap options

If you're willing to do this kind of assumptions, why not do them for index options as well? The model is wrong anyway. Before you ask, no, I don't have anything better in mind.There are people who know a bit about the issue... Nonius, kr, Jab, where are you? It seems things are evolving faster than expected.

What would be the flaws in assuming the credit index behaves in a log-normal manner and using straight B-S?

the flaw would be that experience suggests something completely differentand anyhow, the spread parametrization is not necessarily the best... spreads may go to a zillion but the different in MTM of your position between a million bps and a zillion bps is: 0.000000000000000000000000000000001.I would be more comfortable starting with the basket's expected loss parameter, model that as a beta, and then back-out the spread. But that's because I'm more CDO/basket based than anything. You have to understand that credit indices are not the shiny and sleek things that equity indices are... they have an inherent clunkiness to them because of the difference between equity market liquidity and credit. Even for equity options, you can't just go without a smile and hope for realistic results in some areas. For credit, I'm not sure I know what the consumer's motivation is just yet. Ask yourself a simple question: Who is doing more volume - hedge funds or insurance/pension? Ok, that's an easy question... and the point is that even with all the rinkydink complications of the CDO world, we will get better products based on that stuff than we will with CDS indices. If we are still struggling to make this stuff work well after all this time, there's probably an inherent issue. Ok, just my humble opinion.btw this Risk Guru shit is getting way outta hand... think I'm gonna set up some parallel logins.

wibble, kr's comments are insightful. to keep it simple would use some I-rate model and make tweaks as required by your needs