In recent years as an important numerical analysis method for financial derivative securities pricing,Monte Carlo Simulation gets wide application and development.

Embody the characteristics of foreign investment in the market-currency derivative securities increasingly preferred by investors.

With the growth in globalization of investments in recent years, the quanto derivative securities have gained wider popularity.

To precisely price the financial derivative security is imperative necessity of avoiding the risk in the financial market.

In this paper, we mainly discuss the derivative security pricing problem for the untradable underlying assert.

To the bonds issued by the counter party and derivatives which ranks equally with the bonds, the rec overy proportion of the expected no_default value is same.

Its basic idea is to construct an instantaneously risk-free investment portfolio by appropriately selecting the investment weight of derivatives in the uncertain securities market, and, according to the important relation that value increment of the risk-free portfolio should equal the product of market interest rate and initial value, study the pricing m.

By using no-arbitrage principle, neutral-risk principle, this paper chiefly analyses and discusses the problem on representing the valuation of financial derivative securities as mean and high-dimension integral, and gives a good theoretical foundation for further research on valuation of financial derivative securities.

Methods for constructing standard Winner Process can have a very important influence on estimation result of Monte Carlo simulation in the course of pricing financial derivative securities.

In recent years,as an important method for pricing financial derivative securities,artificial neural network methods have drawn more attention in financial academic field and wide application in practice.