January 5th, 2007, 4:02 pm
If you are talking about a stochastic vol. model, there is a paper by Dietmar Leisen,Stock Evolution under Stochastic Volatility: a discrete approach, J. Derivatives, 8, 9-27, 2000If you are talking about simply a time-dependent, but determinstic volatility, probably a variable time step keeping sigma^2(t) dt a constant will work, but the details are left to youas I am just thinking out loud.regards,