Joshi: Rapid computation of drifts in a reduced factor LIBOR market model
Posted: October 23rd, 2006, 8:43 pm
HiI'm implementing Joshi and Rebonatos Displaced diffusion, stochastic volatility LIBOR market model.As I'm pricing ratchet caps, I'm using the "very long jump procedure" i.e. evolving all the forward rates to their reset date in one jump using the total terminal covariance matrix (TOTC)The TOTC is (almost allways) of rank n (number of forward rates) regardless of the number of factors, F.Is it still possible to use, in some way, the procedure of Joshi in "Rapid computation of drifts in a reduced factor LIBOR market model"?As I see it, for this to be posssible, I need to find an n*F matrix, A, so TOTC = A*Transpose(A). Is this possible?Or do I have to work with a full factor model, when using the very long jump procedure.Christian