摘要
高阶矩是刻画资产收益涨跌非对称和"尖峰厚尾"现象中不可忽略的系统性风险。本文基于我国上证50ETF期权数据采用无模型方法估计隐含波动率、隐含偏度和隐含峰度,通过自回归滑动平均模型提取期权隐含高阶矩新息(Innovations),将它们作为高阶矩风险的度量,探讨其对股票收益的预测作用。研究表明:①在控制换手率和股息率等变量后,隐含波动率对于上证50指数和市场未来4周的超额收益有显著负向的预测作用;②隐含偏度新息越低,上证50指数和市场的超额收益越高,这种预测能力在未来1周和未来4周均显著,但随着时间的推移,隐含偏度新息的预测能力逐渐下降;③隐含偏度风险对于我国股市横截面收益也有显著的解释能力,投资组合在隐含偏度风险因子上的风险暴露越大即因子载荷值越大,则未来的收益会越低;④隐含峰度新息总体上与股票收益负相关。
Higher order moments are the non-ignorable systemic risk to depict the asymmetry between the rise and fall of assets’returns and"leptokurtosis and fat-tail"characteristics.Based on China’s Shanghai50 ETF option data,this paper estimates implied volatility,implied skewness and implied kurtosis with modelfree method and then extracts innovations of higher order moments with ARMA.Using the higher-moment innovations to measure the higher-moment risks and further explore the predictive ability of these risk variables.The research results indicate that:1)Implied volatility has significant predictive ability for four-week ahead excess returns of SSE 50 index and market after controlling turnover and dividend yield;2)The smaller the value of innovation in implied skewness,the higher the excess returns of SSE 50 index and market.Its predictive ability is significant in the following 1 week and 4 weeks,but the predictive ability of implied skewness innovation will become less significant as time passes.3)The implied skewness risk also has a significant ability to explain the cross-section returns,and the stocks with higher exposure to innovation in skewness exhibit lower risk compensation.4)Generally,implied kurtosis risk is negatively correlated with stock returns.
作者
王琳玉
倪中新
郭婧
Wang Linyu;Ni Zhongxin;Guo Jing
出处
《统计研究》
CSSCI
北大核心
2020年第12期75-90,共16页
Statistical Research
关键词
期权隐含高阶矩
股票收益
预测
定价因子
Option-implied Higher Moment
Stock Returns
Predictability
Pricing Factor
