摘要
本文针对正态分布在拟合厚尾分布上的不足提出了两种修正方法——参数修正方法和非参数修正方法,经过实证分析我们发现经过修正的分布能更好地拟合厚尾分布,相应的VaR值的预测效果也更好,其中非参数修正得到的VaR预测效果最佳。
In order to defect the flaw of the normal distribution when fitting the heavy tail distribution, in this paper, two modified methods were presented, which were Parametric and Nonparametric Modified Methods. By empirical analysis, the modified distribution can fit the heavy tail distribution better, and the forecasting effect of the VaR is better too, and the best forecasting result is that resulting from the nonparametric modified method.
出处
《数理统计与管理》
CSSCI
北大核心
2007年第5期867-874,共8页
Journal of Applied Statistics and Management
基金
国家自然科学基金10471135
教育部博士点基金
中国科学院和中国科技大学创新基金
关键词
非参数方法
均值回归
厚尾分布
在险价值VAR
事后检验
nonparametric methods
mean regression
heavy - tail distribution
Value at risk （ VaR ）
back - test methods
