摘要
针对股市收益分布的"尖峰肥尾"特征,引入了偏t分布作为新息分布。基于VaR方法,从风险估计的角度,利用ARFIMA(2,d_1,0)-HYGARCH(1,d_2,1)-skt模型对1996年12月17日至2007年7月5日期间的沪深股市收益进行了实证分析.实证结果显示:沪深股市具有显著的双长记忆特征;上海股市的日收益率和波动率的长记忆性均比深圳股市强;ARFIMA(2,d_1,0)- HYGARCH(1,d_2,1)-skt模型对我国股市收益具有较强的风险估计和预测能力。
Taking account of ＂high Kurtosis and fat tail＂ of daily returns of Chinese stock markets, skewed t distribution was introduced as errors distribution. Based on VaR method, ARFIMA（2, d1,0）- HYGARCH（1, d2, 1）-skt model was used to empirical analyze daily returns of Chinese stock markets from December 17, 1996 to July 5, 2007. As empirical results shown, Shanghai and Shenzhen stock market all present strong double long memory, and the long memory of Shanghai stock market is stronger than Shenzhen stock market. The results also indicated that the ARFIMA（2, d1,0）-HYGARCH（1,d2, 1）-skt model presented strong ability of VaR evaluation and forecast for Chinese stock markets.
出处
《数理统计与管理》
CSSCI
北大核心
2009年第1期167-174,共8页
Journal of Applied Statistics and Management
基金
江苏省教育厅高校哲学社会科学基金项目(08SJB7900020).
