关于美式衍生证券定价的数值分析方法的分析与评述

Analysis and Review on Recent Advances in Numerical Methods for Pricing American Style Derivative Securities

利用现有定价理论 ,大多数美式衍生证券价格不能通过解析方法确定 ,因而必须通过数值分析方法来估计。目前 ,用于美式证券价格估计的数值分析方法主要有三类 ,即格点分析法、有限差分法和蒙特卡罗模拟 ,另外还有一些其它的有效方法。本文主要就这些方法在近些年来的应用与发展进行分析评述。全文共分五部分 ,第一部分是格点分析法 ;第二部分为有限差分方法 ;第三部分是蒙特卡罗模拟 ;第四部分为其他方法 ;最后是结论与展望。

According to current theories for pricing financial property,almost of American style derivative securities cannot acquire a close form formulation of their prices. Therefore,numerical techniques have been becoming very necessary. At present,there are three are three classes of main numerical methods: lattice ,finite difference,and Monte Carlo simulation. In addition to them,some other methods also have been proposed. This paper focus on an analysis and review of recent application and development in these methods. The whole paper includes five sections. Section 1 gives a survey of lattice method. Section 2 focus on finite difference method. Section 3 copies with Monte Carlo simulation. Section 4 discusses LBA and LUBA approximation. Section 5 concludes the paper.