September 13th, 2007, 7:51 pm
How to do Monte Carlo Simulations with Local volatility surface??I have the diffusion process dS/S- =adt + b(S,t)dz (1)What is the good discretization process (1)? I realised that in the case of CEV (b(S,t)=S^(c/2)) for example, by change of variable we have a square root diffusion process and due to CIR we can discretise this easily. Do we have any discretization in the general case of (1)?Thanks.