定价Kou跳扩散美式期权模型的一种有效算法

An Efficient Algorithm for Pricing American Option Under Kou Jump Diffusion Model

针对Kou跳扩散模型美式期权定价问题,空间方向采用中心差分格式离散,时间方向采用Rannacher格式离散,并利用简单有效的递推公式近似积分项.采用模超松弛迭代法求解美式期权离散得到的线性互补问题,分析了离散矩阵的性质和算法的收敛条件.数值实验验证了理论分析并表明所构造的方法是有效稳健的.

For American option under Kou jump diffusion model,the central difference scheme is used to discretize partial differential operators in spatial direction,Rannacher discrete scheme is used for time derivative,an easy-to-implement recursion formula is employed for the approximation of integral term.For the resulting linear complementarity problems obtained from the discretization of American option are solved by the modulus-based successive overrelaxation(MSOR)iteration method.The property of the system matrix and convergence of the MSOR method are analyzed.Numerical experiments conform theoretical analysis and further show that the proposed method is efficient and robust.