<r>Yes, this is one very important article, but if read alone, you will get very pessimistic (or very cycnical) about credit modeling...There are other approaches:- structural models that lead to a given copula of default in a single horizon setup (see Vasicek, KMV model of the loss distribution- de...

<t>For the gaussian copula the tranchelet that is correlation neutral and the maximal wizou (drho dS), or with maximal correlation position and minimal wizou, are around expected loss when correl is near 0, but go opposite ways when rho goes to 1. For LHP portfolio,K=(1-R)Phi((1-rho)^{+- 1/2} Phi^{-...

<t>Thanks Android, for this contributionyou should be extra carefull with the wording of the "time and strike dimensions" section, the wording is very ambitious, for what is being done.the question of whether to use "BC curve by strike" or "BC surface by time/strike" was asked in Dec 2004, and answe...

Hi Stringbean, your point is taken, if you have such a model in mind, there is no point wasting our timediscussing about BC. Let's start another thread, and discuss the merits of this other model.

<t>Quote also have a question for Sebastien: is R_tilda dependent on time, or not? If not, I am probably misunderstanding something in the paper, since it's not clear to me how it's possible to enforce both the two properties at the top of page 7, for each point in time in the case R_tilda is time-i...

<t>cjlnyc, Indeed, the 30%-100% expected loss is determined by the 0%-100% and the price 0%-30%.The running spread has a slight dependence in terms of timing of the loss, but this is very small. Still, the base correlations obtained will be much lower allowing for calibration, and deltas are lower o...

<t>Quote I found that even if it can reduce the implied base correlation, it doesn't help at all with the issue of producing a super senior ATM spread that matches market quoteActually this is a model question, not an implementation question :-)I agree with the premise that it reduce the implied BC,...

<t>QuoteOriginally posted by: BillyTheKidHi,Concerning BNP Model, I have a question about how they price the CDO. In their paper we can understand two things:1>> At the begining I understand that they passe from Pi{default probability} to {Pi_tilde}. And Instead of pricing with the couples (Ri, Pi) ...

<t>QuoteOriginally posted by: BillyTheKidHello sebapi747,I dont agree with you when you say that the default and the recovery in this model are independent: they are strongly correlated through N_t poisson process.Concerning the low base correlation for the equity tranche like (september 2005), migh...

<t>QuoteOriginally posted by: lhjensenQuoteOriginally posted by: katastrofaKrekel's model is simpler to implement and numerically stable.I'm trying to implement Krekel's model for my master thesis and I have some difficulties which is more in relation to calculation of the loss distribution using (A...

<r>QuoteOriginally posted by: BillyTheKidHi,Here is a new random recovery approach for cds and cdo pricing. It gives good results! It reduces the base correlation more than Krekel and BNP approaches by 10%. <URL url="http://sharaf.ifrance.com/CDO/RecovSto.pdfBillyThanks">http://sharaf.ifrance.com/CD...