1. (10 points) In the notes, our derivation of the unbiasedness of the importance sampling estimator relied on the property that 9(35) > 0 whenever f(:l':) > 0. If we instead relax this assumption to be that 9(93) > 0 whenever f (:13) > 0 and |h(:c)| > 0, would the resulting importance sampling estimator still be unbiased? Give a proof or counterexample. As a concrete example, consider the problem of pricing a European call option (discounted payoff h(S(T)) = e'TT[S(T) K ]+) using an importance sampling density 9 whose support is [K , oo)