Hi, long time no see. How's doing, every WILMOTTer? I'm facing now a problem about constructing a Excel-Prototype for Valuation of credit derivatives, in particular iTraxx-index-swap and nth-to-default basket CDS. I have totally no idea about what i shall do, although after reading several papers and thesis, since there are lot of parameters, who can definitely not be constant (Recovery rate, Intensity, PD) and the BOSS had no time for even a little explanation and i couldn't even reach anybody from our department onto the mobilphone. The deadline my be on this friday, that means one day before the weekend and one day before the EM 2008!!! OMG. Some particulary Q regardly to (regular) CDS are listed below: 1. Can we set the recovery rate just be about 40% as the common US corporate recovery rate. I think yes because it is commonly used. 2. I saw the CDS pricing formula, which contained a deterministic Intensity. But in the subsequent paragraph in the same paper the author talk about the calibration of Intensity from CDS market quotations. Does that mean, that we can easily use a calibrated and constant intensity in the pricing formula of CDS? 3. A relationship btw. basket CDS and regular CDS. We shall use one factor gaussian copula model to model the default time and thus incorporate the correlations under the reference entities in the basket. J. C.Hull in his book "Option, Futures and other derivatives", 6 th. edition, chapter 21, Page 519-520 said: ... ... The Prob. that the nth default will happen btw. times T1 and T2 conditional on M is P(n,T2| M)- P(n,T1| M). This gives the prob. distribution for the time of the nth default conditional on M with M ~ N(0,1). By integrating over the distribution for M we obtain the unconditional prob. distribution for the time of the nth default. With this distribution in hand, we can balue a nth-to-default CDS in exactly the way as a regular CDS... 4. One after all, what u ppl will deal with that? set all of the parameters be deterministic even stochastic, or let some of them be constant? What does it exactly mean. How can we incorporate the unconditional prob distribution for the time of the nth default into the pricing formular for a regular CDS (i"m poor in statistic although i was a mathimatican...) I'll be thanksful to anybody, who can give me, even a abstract, direction or guide, what i shall do, and what did normally a bank/financial inst/ consulting firm when facing that kind of task. Actually i'm a new fish and i used to use MATLAB when i had to do simulation. BUT EXCEL, it's not my lover... I'm waiting for a heartwarm guy with a nice answer, even if it's not 100%-correct. Thaks in advance before the beginning of EM. What a shame that we can not enjoy the games of England National Team. C U 2010 in South Africa.

let'ssee if I can help you a little bit (i'm also a new fish in this world).1-Yes...Use 40% (or any other deterministic recovery. it will be ok for most of the applications (CDS/CDO). 2-If you are going to valuate products without optionality and/or no fwrd start a deterministic intensity will be ok. Usually the value of the deterministic intensity can be easily obtained from CDS quotes (there are some threads in willmot regarding this issue)3-One factor Gaussian copula with base correlation is the standar in the market (by now ) ..anyway take a look to "all your hedges in one basket"..the last part explain quite well how to procedd in the independence case. If you have to do something fast the easiest thing is to considers all names in the basket equal so you only have to care about the first dflt but not about which is the name defaulting first (the later obkligues you to condition to each one of the possible names...)4- for things just like CDS/CDO/nth to Dflt I consider it to be a goo approach as most of the results are only going to depend on how things are seen from today. If i had to do this i will proceed with this way (deterministic recoevery/dflt intenisity/correlation)Hope this can help you. anyway i hope that someone more experienced than me could reviewe my comentRegardswe will enjoy the play of the spanish national team (ironic)

Thanks a lot Arlequant. I'll try it using deterministic/constant recovery rate/ intensity/ correlation as u suggested. A kindly answer from anybody else will be also appreciated.