摘要
本文通过概率坐标图发现上海股市收益呈明显的曲线,说明收益具有非正态性,随后的参数及非参数检验从统计上验证了这一点.我们又分别利用GED、ARCH及TPN三种模型实证拟合了上海股市收益的分布形式,由AIC可知它们较正态分布对实际都更具刻画力,尤以TPN最佳.
After it's found that return in S. S. E. is a curve rather than a line in probability plot, non - normality of return in S. S. E. is tested by both parametric and non -parametric methods. For searching more appropriate form of return's distribution, GED, ARCH and TPN are estimated empirically. According to A1C, the three models are superior to normal distribution and TPN is the best.
出处
《系统工程》
CSCD
1998年第1期21-25,8,共6页
Systems Engineering
基金
国家自然科学基金"九五"重大项目"金融数学
金融工程与金融管理的研究"子课题"金融风险分析
防范与控制"(79713007)资助