Sensitivities Of American Option Pricing By Malliavin Calculus And Monte Carlo Method

In this thesis,by means of Malliavin calculus,we obtain the Malliavin weights of sensitivities of American options,and then using Monte Carlo method and cubic spline approximation for continuity functions,we present a numerical method for computing the American option sensitivities.Through the data experiments and empirical tests,the following conclusions are obtained: compared with the generalized least squares method this method is more efficient;compared with European option American options are more sensitive to the initial price of underlying assets.The results may be helpful for the investors to adjust asset allocation,to reduce or hedge the risk.