September 8th, 2010, 7:33 pm
BGM is probably overkill for equities. It has some use in commodities where forwards are substantially decorrelated, but in equities, 1-factor models (or 2-factor models as with Heston) are typically sufficient. That said, Heston-type models get prices for complex structures like Napoleons wrong, so forward skew models where the variance is a function of the spot process have been developed which are similar in some sense to BGM models.There's a paper by Zilber called 'A Market Model of Stochastic Smiles' that might be interesting to you.