Hi all,Has someone already study HJM with Lévy processes?I found some articles about it, Eberlein & al., Nele Vandaele etc...I am trying to show that implied vol fits better in Lévy than in HJM. I want to see if I can reach SABR smile.The code for Gaussian HJM and calibration is done already (C++ & Matlab version).Therefore I need to price swaptions. I would rather use Variance Gamma or NIG processes.Would you use FFT methods? (Peter Carr methods applied to Lévy processes)Would you use some approximations as suggested by Vandaele?Which numerical methods would you use?Any help (as little as it could be) is the very welcome!Thanks a lot.

Anyone?I know that Lévy processes are not quite used in IRD (it's more for CDO) but did someone study this kind of models applied to HJM already ?Thanks a lot

I am working on L\'evy HJM.I suggest you use the Fourier based method discussed in the paper of Eberlein, Glau and Papapantoleon (2010).I have coded it in matlab also.Mesias