随机波动率模型下的单位根检验及对我国股市的实证研究  

Unit Root Test for Stochastic Volatility Models and An Empirical Study of Chinese Stock Market

作　　者：孙便霞[1] 韩鹏[1] 王明进[1] 

机构地区：[1]北京大学光华管理学院,北京100871

出　　处：《数理统计与管理》2008年第1期148-155,共8页Journal of Applied Statistics and Management

摘　　要：判断股市收益率波动的持续性问题通常归结为检验波动过程中是否存在着单位根。针对传统的单位根检验方法功效偏低,且检验的真实显著性水平容易失真的问题,本文采用随机波动率模型并利用改进的PP检验、KPSS检验,以及SIMEX检验方法对中国上证综合指数进行了实证研究.三种检验方法的结果都表明,股权分置改革后的上证综指收益率波动过程有单位根,波动呈现出很强的持续性.The existence of unit root in the volatility process of stock return is regarded as the evidence of persistence of the volatility process, which is essential for people to learn the characteristics of stock return volatility and to price its derivatives. Since the traditional unit root testing procedures have low power and may produce size distortion, the paper adopts the adjusted PP test and KPSS test, and SIMEX test to study the volatility of Shanghai Composite Index（SCI） based on stochastic volatility model. The results of these three tests indicate that the volatility process of SCI after shares separation reform has a unit root and displays strong persistence.

关 键 词：单位根 随机波动率模型 Perron-Ng检验 KPSS检验 SIMEX检验 

分 类 号：F830[经济管理—金融学]