两维数值微分与期权标的资产波动率的重构-计算数学专业论文.docxVIP

摘 摘 要 本文首先对正问题以及期权市场波动率的反演做了一些总结回顾，并对性形 化方法做了一些尝试，得到了一些结果．另外，提出了一种对两维散乱数据求解 两阶数值微分的方法．对于散乱数据点，通过Tikhonov正则化，可以重构原函 数，以及它的一阶和两阶导数．因为Dupire形式是期权价格对时间T和敲定价 格K的一阶和两阶导数的组合，因此，这种方法可以很容易的应用到通过Dupire 形式重构市场波动率．另外，证明了这种数值微分的方法的收敛性，并且给出了 一些数值例子，结果表明，这种反演方法是有效而且稳定的． 在本文的第一章中，简单推导了Black-Scholes方程，给出了正问题的数值解 法，并对期权市场波动率的反演做了一些总结和回顾．特别对线性化重构方法做 了一些推广和尝试． 在本文的第二章中，通过Tikhonov正则化给出了求一种两维两阶数值微分 的方法，并给出了唯一性证明、误差估计以及一些数值例子，将这种微分方法应 用到Dupire形式，重构了标的资产波动率． 关键词： Bla,ck-Scholes方程，波动率，线性化，Dupire，两维数值微分 Tikhonov正则化、误差估计 中图分类号：0241．5 ABSTRACTIn ABSTRACT In this paper，direct problems of option pricing and some methods in solving inverse problems about the volatility of underlying as8ets summarized Some results of linearized method provided．In addition，a new method of recovering the local volatility by numerical derivatives of 2-dimensions is proposed．Within the frame of method，we achieve the real function，first and second deriv- atives．The convergent rate and numerical examples also given The thesis falls into two chapters．In chapter 1，Black—Scholes equation is provided，and numerical method of the equation is introduced．We also snmul- rize sonde other methods in this chapter，which widely used to solve inverse problem about the volatility．Especially，the linearized method attempted． In chapter 2，a method of nmnerical derivatives of 2-dimensions is proposed． The first and second derivatives of 2-dimensions smooth function are approxi— mated by using the Tikhonov regularization method．It is proved that the approx— imate function can be chosen minimizer to COSL functional．The existence and uniqueness theory of the minimizer is established．With the known smooth． of 2-Dimensions function，convergence result is also proved．The numerical results provided to support the theoretical anal?,si8 of this work． Keywords：Black—Seholes equation，volatility，linearization，Dupire numerical derivative of 2-dimension，Tikhonov regulation，error estimation M R Subject Classification：0241．5 第一章 第一章 正问题的介绍与期权标的资产市场波动率的反