极端波动、跳跃和尾部风险——基于已实现波动率的股票市场风险动态预测  

Extreme Volatility,Jump and Tail Risk:Dynamic Forecasting Stock Market Risk based on Realized Volatility

作　　者：柳会珍[1,2] 顾岚[3] 胡啸兵[2] 

机构地区：[1]北京国家会计学院,北京101312 [2]西安交通大学经济与金融学院,陕西西安710061 [3]中国人民大学统计学院,北京100872

出　　处：《数理统计与管理》2014年第1期158-169,共12页Journal of Applied Statistics and Management

基　　金：中国博士后科学基金(20100471621)

摘　　要：本文基于跳跃扩散波动理论,利用非参数方法估计波动中跳跃成份,研究我国股市波动中跳跃的动态演变特征,将跳跃作为股市波动的重要因素纳入模型,建立我国股指收益率的非齐次自回归已实现波动率模型,利用条件极值方法对我国股市的波动风险进行动态预测.统计结果表明:在股市大波动时期跳跃发生更为频繁,而且和连续成份相比较,跳跃对股指波动的贡献占据主要地位。和传统的EGARCH模型相比,包含跳跃的已实现波动率模型对股指波动风险预测性能明显优于EGARCH模型风险预测,跳跃对股指波动风险预测具有显著的解释力.对股市历史数据的分析和预测表明,跳跃对于股指日收益率尾部行为具有重要的影响,大跳跃的发生导致尾部风险显著增大。本文研究结果对于政府监管部门监控股市和制定有效的调控措施加强股市风险管理具有重要的参考价值。Based on jump diffusion volatility theory, this paper estimates jumps components in volatil- ity by using nonparametric method and studies the dynamic evolution features of jumps in volatility of domestic stock market. Incorporating Jumps as important factors into the model, We build the heteroge- neous autoregressive model of realized volatility for stock index returns, and use the conditional extreme value approach to predict the volatility risk of stock market dynamically. Statistical results show that during large movements in stock market jumps occur frequently and compared with continuous compo- nents, the contribution of jumps to stock index volatility predominates. When compared with traditional EGARCH model, the forecasting performance of realized volatility model incorporating jumps is prior to EGARCH model, and jumps have significant explanatory power for the prediction of volatility risk of stock index. Analyzing historical data and predicting results indicates that jumps have important impact on tails behavior of daily returns of stock index, and occurrence of large jumps results in signif- icant increasing of tail risk. The research results in this paper will have important reference value for the government regulatory department to monitor the stock market and to formulate efficient regulation measures to strengthen risk management of stock market.

关 键 词：已实现波动率 跳跃 尾部风险 动态预测 

分 类 号：F830.9[经济管理—金融学] O212[理学—概率论与数理统计]