统计研究 ›› 2017, Vol. 34 ›› Issue (10): 66-76.doi: 10.19343/j.cnki.11-1302/c.2017.10.006

• 论文 • 上一篇    下一篇

中国股市高阶矩风险及其对投资收益的冲击

史代敏等   

  • 出版日期:2017-10-15 发布日期:2017-10-25

The Impact of Market Higher-moment Risks on the Return of Stock Investment

Shi Daimin Tian Lemeng Liu Zhen   

  • Online:2017-10-15 Published:2017-10-25

摘要: 本研究首先建立NAGARCHSK模型,推算市场收益率的条件高阶矩序列,在此基础上建立引入高阶矩风险的收益-风险时变四因子状态空间模型,并基于2000-2016年中国股票市场的收益率数据,实证探究不同时期市场高阶矩风险对投资收益的冲击。结果显示:我国股票市场收益受到高阶矩风险的影响,并且条件高阶矩序列表现出时变和波动聚集的特征,大规模的全球性金融危机和国内市场的重大风险事件均会使股市收益的条件高阶矩序列出现持续的异常波动。在未出现极端金融危机的稳定时期,市场收益率的条件方差会趋于对投资收益产生正向影响,条件偏度和条件峰度对投资收益的影响在正向和负向之间不断交替,增加了投资收益的不确定性。然而在全球性的极端金融危机时期,市场收益率的条件方差会转而对投资收益产生负向影响,条件峰度则会对投资收益带来持续的正向影响。

关键词: 高阶矩风险, 投资收益, NAGARCHSK模型, 状态空间模型

Abstract: Firstly, we develop the NAGARCHSK model to compute the conditional higher-moment series of the stock market return, with which we develop the time-varying four factor state space model based on higher-moment risks information, so that we can analyze the impacts of higher-moment risks on the overall return of stock investment in different periods. The results of the empirical analysis during the year of 2000 to 2016 reflect that China's stock market is affected by the higher-moment risks, and the risks of higher-moment is time-varying and volatility clustering. Meanwhile, global financial crisis and major risk events in domestic market could cause abnormal fluctuations of the conditional higher-moment series. In most cases, the conditional variance of the market return has a positive impact on the stock investment return, the conditional skewness along with the conditional kurtosis of the market return tend to increase the uncertainty of the return of stock investment. However, during the period of extreme financial crisis, the conditional variance will turn to have the opposite influence on the return of stock investment, while the conditional kurtosis will turn to have the continuous positive influence on the return of stock investment.

Key words: Higher-moment risks, Investment Return, NAGARCHSK Model, State Space Model