Hi everyone,I'm currently writing my master's thesis on the performance of different CDO pricing models before and throughout the credit crisis. My goal is to compare the actual market mid-spreads for the DJ CDX IG and DJ iTraxx Europe with the fair spreads determined by the pricing models. The first model that I'm trying to implement is (obviously) the one-factor Gaussian Copula. Right now I'm trying to calibrate the copula first using compound correlation and after that base correlation. Under the finite homogeneous pool assumption the conditional default probability for each entity in an index' pool boils down to this:My understanding is that I first need to estimate the hazard rate or the cumulative hazard rate (the entire integral of hazard rates) and then in a compound or base correlation framework the Copula's correlation parameter. As far as I know, the hazard rates can be estimated directly from tranche spreads of an index. After going through my literature, the forums and the web, I still can't find a well documented way of how to do just that. I've come across the "clean pool hazard rate",which is analogous to the single-name CDS hazard rate. Unfortunately, I believe, this is only valid when used under a large pool not finite pool assumption (as the pool then could be treated as a single name under the law of large numbers?).Could anyone tell me whether I'm generally on the right track and where (papers/books) I could possibly find an actual description of how to estimate (cumulative) hazard rates for my Copula in this synthetic CDO setting?Any hints would be greatly appreciated!THX

Hi feju,I may have misunderstood your post, but generally when pricing a CDO you need to work out the hazard rate for each particular reference in the portfolio. For corporate references you would typically get this using CDS spreads. Essentially you work out the probability of default that gives you a PV of 0. Then plug these into the copula, together with a correlation number.Estimating hazard rates for a corporate entity is discussed in Hull - Options, Futures and Other Derivatives. There are plenty of CDO related papers on the web, as well as some very useful threads here on Wilmott. I have learn't a lot from those threads - again, use the search function above and you should get plenty of results.Cheers,Donal