美式期权定价的拟蒙特卡罗模拟及其方差减小技术

发布时间：2018-01-31 20:55

本文关键词： 美式期权 Faure序列控制变量技术对偶变量技术　出处：《北京化工大学学报(自然科学版)》2014年03期 　论文类型：期刊论文

【摘要】：在研究拟蒙特卡罗方法的基础上,选择对美式期权定价有较好稳定结果的Faure序列代替原来的伪随机序列,得到了优于LSM模拟的结果,使模拟方差得到缩减。在此基础上,综合考虑蒙特卡罗模拟的方差减少技术,把控制变量技术与对偶变量技术相结合的复合方差减少技术运用到美式期权定价的过程中。通过模拟结果发现,控制变量技术对LSM模拟和LSQM模拟都能产生较理想的方差缩减效果,结合使用控制变量技术的LSQM模拟无论是在计算效率还是结果的稳定性方面,都比使用控制变量技术的LSM模拟好。如果再将控制变量技术与对偶变量技术相结合的复合方差减少技术与LSQM模拟方法相结合,得到的模拟结果则更精确。
[Abstract]:On the basis of studying the quasi-Monte Carlo method, the Faure sequence, which has a good stable result for American option pricing, is chosen to replace the original pseudorandom sequence, and the result is better than that of LSM simulation. The simulation variance is reduced. On this basis, the Monte Carlo simulation variance reduction technique is considered synthetically. The compound variance reduction technique which combines the control variable technique and the dual variable technique is applied to the process of American option pricing. The control variable technology can produce ideal variance reduction effect on both LSM simulation and LSQM simulation. The LSQM simulation using control variable technology is not only in computational efficiency but also in the stability of the result. It is better than the LSM simulation using the control variable technology. If the combination of the control variable technology and dual variable technology is combined with the LSQM simulation method. The simulation results are more accurate.
【作者单位】：北京化工大学理学院;
【分类号】：F224;F830.9
【正文快照】： 引言美式期权是一种非常重要的期权,它具有可以提前执行的特点。美式期权定价经过了漫长的发展改进过程,直到1993年Tilly[1]成功的利用MonteCarlo模拟对基于可支付红利的美式看跌期权进行定价,终于打破了美式期权不能有效的运用仿真模型定价的传统观念。Broadle等[2]考虑到上

【参考文献】