摘要
本文在投资组合难以充分分散化的前提下,综合考虑个股的总偏度、系统性偏度和特质性偏度对个股定价的影响。为了克服样本极端值对偏度测度的影响,通过混频分位数回归先估计个股和股指收益率的条件分位数,再借助Cornish-Fisher展开式间接估计个股和股指的总偏度,这是一种稳健的估计方法。基于构造的个股和股指的稳健总偏度将个股的稳健总偏度分解为稳健系统性偏度和稳健特质性偏度。借鉴Zhu等(2021)对赌博偏好的检验方法,实证发现,在横截面上,基于偏度的排序分组检验显示对个股收益率的预测能力强弱顺序依次为稳健总偏度、稳健系统性偏度、稳健特质性偏度、样本矩总偏度;在时间序列上,基于含偏度的因子模型检验显示对个股收益率及股市异象的预测能力强弱顺序依次为稳健总偏度、稳健系统性偏度、样本矩总偏度、稳健特质性偏度。实证结果表明,偏度是A股定价的重要因子,但稳健总偏度优于样本矩总偏度,稳健系统性偏度优于稳健特质性偏度。
Under the premise that it is difficult to fully diversify the portfolio,this paper comprehensively considers the influence of total skewness,systematic skewness and idiosyncratic skewness of individual stock on its pricing.In order to overcome the influence of sample extreme value on skewness measure,this paper estimates the conditional quantile of individual stock and index return by mixing quantile regression,and then estimates the total skewness of individual stock and index indirectly by Cornish Fisher expansion,which is a robust estimation method.Based on the constructed robust total skewness of individual stock and index,the robust total skewness of individual stock is decomposed into robust systematic skewness and robust idiosyncratic skewness.Based on Zhu et al.'s(2021)test of gambling preference,this paper empirically finds that:On the cross section,the rank grouping test based on skewness shows that the order of forecasting ability of individual stock return is robust total skewness>robust systematic skewness>robust idiosyncratic skewness>sample moment total skewness;in the time series,the factor model test including skewness shows that the order of forecasting ability of individual stock return and anomalies is:robust total skewness>robust systematic skewness>sample moment total skewness>robust idiosyncratic skewness.The empirical results show that skewness is an important factor in A-shares pricing,but the robust total skewness is better than the sample moment total skewness,and the robust systematic skewness is better than the robust idiosyncratic skewness.
作者
夏仕龙
Shilong Xia(School of Economics,Sichuan University,Chengdu,China)
出处
《南大商学评论》
2021年第2期108-122,共15页
Nanjing Business Review
基金
2019年度教育部人文社会科学研究青年基金西部和边疆地区项目“不确定环境下我国宏观经济政策协调搭配研究”(19XJC790015)
关键词
偏度
资产定价
因子模型
混频分位数回归
Skewness
Asset Pricing
Factor Model
Mixing Quantile Regression
