Study On Monte Carlo Method For Pricing Exotic Options In Binomial Model

Currently, financial derivatives market is growing, and complex exotic options are created by financial engineers to meet the needs of the customers, but their analytical solution is not easy to be obtained, and there are only several options with special characteristics can be valued by the existing numerical methods.The purpose of this paper is to find out a method which suits most exotic options and achieves the precise requirement for applications in the practice. By combining the two methods of binomial model and Monte Carlo, our new method is obtained.This paper consists of five chapters, and the main work is introduced as follows:In chapter one, we will introduce the research background and status and put forward the main idea of this paper.In chapter two, based on the studies of the characteristics of various existing exotic options, we get the essential characteristics used for the classification of exotic options, and describe them in this paper briefly. Finally according to these features, we complete the classification of some common exotic options.In chapter three, four kinds of existing common numerical pricing methods for options are given to explore the advantages and disadvantages. They are the pricing method in binomial model, the pricing method in trinomial model, fuzzy number method and Monte Carlo method. By analyzing the applications of them, we obtain the bases for the combination.In chapter four, based on the binomial model, this paper improves the Monte Carlo method and achieves a new one, the Monte Carlo method in the binomial model. We will prove that the new method is universal by several examples. The new method overcomes the problem that Monte Carlo method cannot be used to price an option which can be executed in advance with the help of Longstaff s theory.In chapter five, using Matlab software this paper completes several pricing problems of exotic options, and proves that new method achieves the precise requirement for application in the practice. So far, this paper has proved that the new method is universal, have a good precision and can apply to the actual requirements.