wilmott.com

Hi,Would anyone please shed some lights on how to convert the risk neutral default probability to physical default probability?note : The underlying model does not necessarily have to be Merton model.Thank you,Henry

I don't know but there have been many forum discussions, for examplehttp://www.wilmott.com/messageview.cfm?catid=8 ... BTABLE=You might want to track down the blog link to "R-probabilities", as this is relevant to your quest.

Hi, if you do not consider any particular model, such as KMV which is based on historical experience of mapping Q->P, I don't think this is possible. Every market model will give you a different set of risk-neutral probabilities.If, for example, you bootstrap Q-default probabilities from CDS quotes there is no way how to map them to empirical ones. A long time ago I did some exercise where I tried some kind of Q->P convesion but without any reasonable success. You should also bear in mind that Q-probabilities are usually recovery-rate (RR) dependent (higher RR -> higher Q prob of default) while P-probabilities are, say, RR independent.

Don't think it's possible neither. Take BS model as an example.The risk neutral drift is "r" (interest rate). The physical drift "u" is not even used in the PDE.I believe it's similar case when it comes to default probability model like credit spread.Just my 2 cents.