摘要
金融资产的未来性决定了人们在作出决策选择时必须面对不确定性,当投资者面临不确定性时,概率分布最好地描述了不确定性。通过2007年1月4日至2011年8月30日期间1029个交易日的实证研究,发现我国股市收益率序列具有有偏、尖峰厚尾特征,而不满足正态分布。随后选择t分布、偏t分布与非对称Laplace分布来拟合我国股市收益率序列分布,结果显示偏t分布能很好地描述这些分布特征,说明偏t分布拟合我国股市收益率序列分布更具有合理性和有效性。
The future features of financial assets make people have to face uncertainty when they make decision. When investors face uncertainty to make choice, probability distribution is best way to describe uncertainty. The empirical results using returns of Shanghai Stock Exchange, indicates that there exist obvious differences between return series and normal distribution,and the empirical distribution is characterized by asymmetry, leptokurtosis and heavy tails. In order to describe these features, applies t distributions, skewed student-t distribution and asymmetric Laplace distribution to analyze the empirical distribution of return.The result shows that return series conform well to skewed student-t distribution provides a better fit than t distribution and normal distribution.
出处
《现代计算机》
2011年第21期9-12,共4页
Modern Computer
关键词
T分布
偏T分布
非对称LAPLACE分布
Distribution
Skewed Student-t Distribution
Asymmetric Laplace Distribution