HiI was trying to price an inflation option in a bs framework. The option I am pricing is a single 10y caplet ZC inflation, i.e. the pay-off is Max(Index(10y)/Index(now) - K, 0). My basic model is based on Jarrow-Yildirim, i.e. real + nominal + infaltion as FX, the idea is since dI/I = break-even related drift term + fxvol related stochastic term, so we could model dI/I as a lognormal process and price the option is a bs framework, with an "equivalent" bs lognormal vol. using simulations, I get an approximation as "equivalent" bs vol in this case as given byvol(zc)^2 = vol(fx)^2 + vol(be)^2 X T^(3/2), where vol(zc) is the "equivalent" bs vol, vol(fx) is the lognormal inflation index vol, and vol(be) is the normal vol of the 10y zc breakeven rate. But I am not quite sure I got it right. Any insight highly appreciated. thanks vm