I have some troubles with understanding of composite forward and options.So, we have underlying security S settled in USD and forex rate FX from USD to EUR (USDEUR). Then we move to a risk neutral world with respect to zero-coupon bond, that pays 1EUR at time T. In this word expected value of S (denominated in USD) is just vanilla forward of S multiplied by exp(alpha * t), where alpha is quanto adjustment. So, the question is how could i represent the value of compo forward? I.e. the expected value of S * FX in chosen word. I guess it should beE (S * FX) = E(S) * E(FX) * exp(alpha * t) , but i dont know why If it's true, compo forward just equals to F * forwardFX, with F - vanilla forward.Second, i try to evaluate compo option by FD or trees. What values of risk-free rate, div rates and volatility should i use? I guess volatility should be compoSigma = sqrt(sSigma^2 + fxSigma^2 + 2 * cov) and EUR risk-free rate should be taken. But it's not enough!Many thanks in advance.

think about how you would hedge a composite forward versus how you would hedge a forward contract in the domestic currency

daveangel, thanks for replyRegarding your question - i know how to hedge it (it's just static hedge with borrowing USD, buying S and forward forex) and i know that compo forward value is product of vanilla forward and forwardFX, but i try to deduce this formula in martingale world. So, we have to dependent processes S(t) and FX(t). Why isE (S * FX) = E(S) * E(FX) * exp(alpha * t) in world which is risk-neutral with respect to zero-coupon bond, that pays 1EUR at time T??Usually, when we have two dependent variables, the math. expectations of their product is the sum of covariance (in compo case equals to quanto adjustment or alpha) and product of expectations. But in this case, it looks like the drift of compo process is increased by alpha(cov). Please, clarify it for me!P.S and also about compo process diffusion

I am sorry I am not an expert on martingles so can't help you with that. But it should be clear that the covariance does not affect the composite forward as the hede is purely static. You can apply Ito and get the result.