Optimization Of Monte Carlo Method In Pricing Autocallable Structured Products

Structured products are instruments combined with traditional financial assets and financial derivatives.Their main application is to create the flexibility of returns and risk and provide investors with a specific exposure.Unlike portfolios with the same composition,structured products are packaged portfolio,which are usually packaged into accord with the law and can be invested in the form of at any time.In recent years,the domestic financial market turmoil.The new type of autocallable structured products named snowball options with high coupon and big probability of guaranteeing principal become the focus of investors.The autocall feature causes the product to be redeemed if the reference asset’s value rises above a pre-specified price.which reduce the product’s duration and maturity.The payoff of an European option is determined by the final price of the underlying.However,the payoff of snowball options depends on not only the spot and final price of the underlying,but also the underlying path within the period.Therefore,snowball options belongs to the path-dependent options.Pricing path-dependent exotic options usually involves the issue of solving the multiple integral so it is difficult to obtain the analytical closed-form solutions.In practice,we need to estimate it by numerical method.The traditional numerical method in solving the problem of high-dimensional would produce "curse of dimensionality",which causes an exponential increase in the required amount of calculation in order to achieve same precision in high dimension.The Monte Carlo method can solve this problem.The application of Monte Carlo method is relatively simple,but its weakness is that computational efficiency is too low.So in this paper,we research the application and optimization of Monte Carlo method in pricing snowball options under the hypothesis of B-S model.We improve the effectiveness of the simulation by using three method to generate sample paths including random walk,brown bridge and principal component analysis,and using in combination with the Monte Carlo variance reduction techniques and the Quasi-Monte Carlo method.We are aiming to find a optimal Monte Carlo method in pricing snowball options.