我国股票市场收益率偏度对风险补偿的影响研究

A Study about the Impact of Return’s Skewness on Risk Premium in Chinese Stock Market

收益率与波动率之间的关系(风险补偿)在金融资产定价和风险管理中至关重要,但二者之间的研究却一直没有定论。金融理论表明为正向关系,但实证却往往得到相反的结果。众所周知,尖峰厚尾以及负偏是收益率的典型特性。这些内在属性对风险补偿无疑有着重要的影响。本文研究表明,收益率的负偏和尖峰厚尾会导致风险补偿系数减小,进而导致风险补偿系数在实证当中的不确定性。虽然各个行业的夏普比率相差无几,我国股票市场的第"Ⅰ"类、第"Ⅱ"类风险补偿以及风险价格在不同行业之间存在较大的差异,且三者随着投资期限的增长,分别有下降、上升和轻微上升的趋势。

As the first fundamental law of finance,the relationship between return and volatility(risk premium)is far from being conclusive,the so-called“Return-Volatility Puzzle”.Financial theory shows a positive relationship,but empirical results tend to be opposite.As we all know,kurtosis,thick tail and negative skewness aretypical featuresof returns.These inherent and endogenous attributes undoubtedly have an important impact on risk compensation.Besides,the non-normal or asymmetrical distributional features can produce ineffectiveness or inaccuracy for those estimators of coefficients with wrong distributional assumptions regardless of linear or nonlinear tests of the relationship.Based on the series of studies of Theodossiou(1998^([20]),2016^([13]))about SGT(Skewed Generalized t)distribution,the relationship between volatility and return of China’s stock market is studied from the perspective of the inherent asymmetry(skewness and kurtosis)of returns.Under the framework of Theodossiou(2016)^([13]),the risk compensation is decomposed into two parts:“pure”risk compensation(type-I risk compensation,caused by volatility)and type-II risk compensation that is a nonlinear function of the parameters of skewness and kurtosis in SGT.type-II risk compensationcan be positive or negative,which surely can result inconclusion of empirical works.Besides,considering the impact of investment terms,the“risk price”is further considered under different investment terms.In this paper,six indexes of Shanghai Stock Market and 20-years dataare used to represent Chinese stock market.The empirical results show that i)the negative skewness andkurtosis will lead to the reduction of risk compensation coefficient of volatility,leading to the uncertainty of risk premium coefficient in empirical research,whichpartly explains why the relationship between volatility and return is not stable and inconclusion;ii)because of the generality and flexibility of SGT,the estimators obtained from SGT are more stable and reliable than that of GARCH models with normal distribution;iii)although the Sharp ratios of different industries are almost the same,there are great differences among the type-I and type-IIrisk premium in different industries in Chinese stock market.iv)there are obvious leverage effect and clustering effect in Chinese stock market.v)with the increase of investment term,the type-I,type-II compensation and risk price have respectively decreasing,increasing and slightly increasing trends.From the perspective of the endogeneity of skewness and kurtosis,separating risk compensation from traditional risk compensation not only explains the contradiction between theoretical models and empirical models,but also explains the“volatility puzzle”.