Hello again, There's seems to be some inconsistencies in the method used to derive u,d, and risk neutral p for the binomial model in past papers for CID (April 2004, Q8) and 109 (April 2002, Q6). In CID the variance of S_t is assumed to be: in 109 it uses the log normal variance: Hull uses the same as CID. So which is the "right" answer? I think 109 is probably more theorectically correct, but Hull is the bible...