作　　者： ; ; ;

机构地区： 华南理工大学工商管理学院

出　　处： 《管理科学》 2011年第1期82-89,共8页

摘　　要： 基于传统的可转债最小二乘蒙特卡罗模拟定价方法,通过使用随机Faure序列和方差减小技术,有效地降低模型估计结果的误差,使用考虑解释变量和被解释变量误差的全最小二乘回归方法代替普通的最小二乘回归方法,提出可转债的全最小二乘拟蒙特卡罗定价方法,并给出该定价方法的具体算法步骤。以2002年10月16日发行的燕京可转换债券为例进行实证分析,从可转债的理论价值、计算标准差以及模型的运行时间等几个方面与传统的蒙特卡罗方法进行比较。研究结果表明,使用全最小二乘拟蒙特卡罗方法进行计算得到的结果更为合理,且估计误差和计算时间都更少,从而验证了该方法在可转债定价应用上的有效性。 Based on the Least-Squares Monte Carlo method,the Total Least-Squares Quasi-Monte Carlo method for convertible bond pricing was proposed and the details of its algorithm were given.First,Faure sequences as well as variance reduction technique were used to effectively reduce the variance of the estimation results.Second,the Total Least-Squares Quasi-Monte Carlo method which took into account the errors of both the explanatory variables and explained variable was used to substitute the ordinary least squares regression method.At last,an empirical analysis of Yanjing Convertible bond issued on October 16,2002 was given to make a comparison between TLSQM method and LSM method,in aspects of the theoretical value,calculation errors and the running time.The results show that the theoretical value by TLSQM method is more reasonable,and TLSQM method has smaller estimation error and shorter computing time than LSM method,which proves TLSQM method is more effective for the pricing of convertible bonds in China.

关 键 词： 可转债定价 全最小二乘 拟蒙特卡罗 序列 对偶变量法

领　　域： [经济管理]