Hello,How can I (very approximately) create a log-spread option? i.e. max(X - k, 0) using the below options. where X = ln(m1/m2) and m1 is month1, m2 = month2Open to dynamic hedge but static hedge preferred ln(m1/m2) or m1/m2 -1 is still OK.Givens:- My market has abundance of options on m1, m2 as well as max(m1-m2, 0) for delta = 10, 25, 40, 50- if you need more info on the distribution of my market: let us assume it is a) interest rates and b) commoditiesPrefer rough approximation as I want to trade this. Elegant solution is probably impossible without strong assumption on underlyings.

are you willing to assume some things? like geometric brownian motion for the two stocks with some correlation?because then you just have max(ln(m1)-ln(m2) - k) and you'd know the distributions of ln(m1) and ln(m2) . you can back out the correlation from the max(m1-m2) option. Then you can hedge correlation with max(m1-m2) option and deltas with the stocks ?