Pricing of American Put options

Abstract

Abstract In this thesis we have compared two numerical methods for pricing an American put option, the binomial model and a finite difference scheme. The two methods have been compared with regard to accuracy and computational speed. Thus, the best model is the one with the highest computational speed, which at the same time generates accurate results. We compare their exercise boundaries, see how they are affected by the amount of iterations and grid size, and analyze the theory behind the two methods. Since option pricing depends on market parameters, the methods have been compared multiple times, for different market parameters to obtain a complete comparison for multiple aspects. The thesis also includes the derivation of the binomial model algorithm as well as the theorem and a sketch of the proof for the boundary value problem of the partial differential equation used to price American put options.