中国股票市场存在特质波动率之谜吗?——基于分位数回归模型的实证分析  

Is there idiosyncratic volatility puzzle in Chinese stock markets: A quantile regression analysis

作　　者：熊和平[1] 刘京军[2] 杨伊君 周靖明 XIONG He-ping;LIU Jing-lun;YANG Yi-jun;ZHOU Jing-ming(Economics and Management School of Wuhan University,Wuhan 430072,China;Lingnan College of Sun Yat-sen University,Guangzhou 510275,China)

机构地区：[1]武汉大学经济与管理学院,武汉430072 [2]中山大学岭南学院,广州510275

出　　处：《管理科学学报》2018年第12期37-53,共17页Journal of Management Sciences in China

基　　金：国家自然科学基金资助创新研究群体项目(71721001);国家自然科学基金资助项目(71771220;71571195)

摘　　要：选取我国沪深A股所有股票作为研究对象,采用OLS回归残差标准差提取和GARCH(1,1)加权平均等两种方法估计特质波动率,并利用Fama-MacBeth横截面回归法和分位数回归法对特质风险与股票预期回报之间的相关关系进行了实证研究.发现:OLS回归结果表明我国股票市场的特质波动率与股票预期回报之间呈现负相关关系,但在统计上不显著;分位数回归则表明我国股票市场的特质波动率风险与股票预期回报之间的关系是随着分位水平的变化而变化的,特质风险在低分位水平下与股票预期回报呈显著负相关关系,而在高分位水平下则与股票预期回报之间呈显著正相关关系.Using the samples of Chinese A-share listed companies,this paper empirically investigates the relationship between idiosyncratic volatility and stock cross-sectional return and throws light upon the question"Does idiosyncratic volatility puzzle in China". Traditional OLS regression residual standard deviation and GARCH model are used to estimate idiosyncratic volatility,and both the Fama-MacBeth cross-section regression and quantile regression method to investigate the relationship between idiosyncratic volatility and stock cross-sectional return. The OLS regression analysis shows that idiosyncratic risk is negatively correlated with stock expected return,but the relationship is not statistically significant,which means that idiosyncratic volatility puzzle does not exist. The quantile regression on the other hand gives a more comprehensive description of the relationship between idiosyncratic risk and stock expected return. At the low quantiles the relationship is significantly negative while at high quantiles the relationship is significantly positive.

关 键 词：特质波动率 股票横截面收益 分位数回归 

分 类 号：F830.9[经济管理—金融学]