上证指数周收益率分布的分段高斯性  

The Piecewise Gaussianity of Weekly Return Rates of Shanghai Composite Index

作　　者：袁天姿 王家赠[1] 

机构地区：[1]北京工商大学理学院,北京100048

出　　处：《数学的实践与认识》2015年第17期93-100,共8页Mathematics in Practice and Theory

摘　　要：以往的研究显示股票收益率往往有厚尾的特性,在本文中,我们假设收益率服从不同的分布,所以我们将数据分段,用两种方法对数据进行分段分布拟合.第一种方法是假设这三段数据服从相同均值,但不同方差的正态分布;第二种方法是假设这三段数据服从均值和方差均不相同的正态分布.在这两种假设下,对于每一小段,用最小二乘原理得到了相应分布的参数值.最后利用K-S检验进行验证.结果显示,上证指数周收益率服从分段正态分布.Previous studies have shown that stock yields tend to have a fat tail. In this paper, it is assumed that yields obey segmented Gaussian distributions. Therefore, two methods are used to fit for segmented data. The first method is to assume that the three pieces of data subject to the normal distributions, which has the same mean value but different variances; further method is that both of the parameters are different. The principle of least squares is used to obtain the parameters of corresponding distribution. Finally, K- S test is used for verification. Results show that the yields, which are on the Shanghai composite index, obey segmented Gaussian distribution.

关 键 词：周收益率 Q-Q图 最小二乘原理 K-S拟合优度检验 

分 类 号：F832.51[经济管理—金融学] F224