“特质波动率之谜”与估计模型有关吗?  

Idiosyncratic Volatility Puzzle and Its Estimation Model

作　　者：陆静[1,2] 张银盈 LU Jing;ZHANG Yin-ying(School of Economics and Business Administration,Chongqing University,Chongqing 400030,China;Innovation Institute of Corporate Finance and Accounting Governance,Chongqing University,400030,China)

机构地区：[1]重庆大学经济与工商管理学院,重庆400030 [2]重庆大学公司财务与会计治理创新研究院,重庆400030

出　　处：《中国管理科学》2022年第9期36-48,共13页Chinese Journal of Management Science

基　　金：国家自然科学基金资助项目(71973018)。

摘　　要：以CAPM、Fama-French三因子和五因子模型为均值方程,分别采用无条件标准差,以及GARCH、EGARCH条件方差为基准计算出中国股票市场的特质波动率,研究了特质波动率与股票收益率之间的关系。研究发现,关于高特质波动风险对应低预期收益的定价异象来源于特质波动率估计模型的差异。具体地,当使用无条件标准差方式估计特质波动率时,存在定价异象,而使用GARCH、EGARCH等条件方差模型估计特质波动率时,则不存在定价异象。该结果在改变残差估计均值方程以及控制规模、流动性等其它变量后依然稳健。本文的研究有助于解释长期困扰在资产定价领域的“特质波动率之谜”。Traditional asset pricing theory holds that stock returns are mainly related to system risks and non-system risks can be offset by diversification investment.In recent years,some scholars have found that there is a negative correlation between non-systematic risks and expected returns in the stock market,which is contrary to the classical risk pricing theory,thus triggering a discussion on Idiosyncratic Volatility Puzzle.Different scholars have conducted more empirical tests based on different capital markets,data intervals,control variables,and idiosyncratic volatility estimation methods to explore whether the“Idiosyncratic Volatility Puzzle”is widespread in capital markets.However,conclusions are not consistent,and there are three controversies:positive correlation,negative correlation,and irrelevant.Through literature analysis,it is believed that different idiosyncratic risk pricing anomalies come from different idiosyncratic risk measurement models,and based on this,research is conducted.Specifically,CAPM and FF-3 models are used as the mean equation for estimating residuals,and conditional standard deviations of monthly internal unconditional standard deviations,GARCH and EGARCH models are used as idiosyncratic volatility.When the relationship between idiosyncratic volatility and stock expected return rate is discussed,the proxy variable of expected idiosyncratic volatility is the realized idiosyncratic volatility in t-1.Firstly,the portfolio is divided into quartiles according to the size of the trait volatility,and it is rebalanced once a month to calculate the tradable market capitalization weighted return,total market capitalization weighted return and equal weighted return within the portfolio.Meanwhile,the Jason alpha value of the portfolio is calculated respectively according to CAPM and FF-3 model,and the relationship between it and the trait volatility is discussed.Taking A-share listed companies in Shanghai and Shenzhen stock markets from 1997 to 2018 as samples,based on the existing research,C

关 键 词：特质波动率 预期收益 资产定价 资本市场 

分 类 号：F830.0[经济管理—金融学]