摘要
运用函数型主成分分析方法,对中国沪市89支股票的月度收益率进行分析,其结果表明函数型主成分方法能够较为准确地捕捉到月度收益率的时间波动特征,特别是它在时间上的变化方向和形式,为股票收益率的建模和预测提供科学依据。
This paper studied 89 monthly yield curves in Shanghai Stock Market based on functional component analysis, and found that the fluctuation of yield curves, especially its direction and shape, can be exactly illustrated through functional component analysis. This forms an important basis for the modeling and foreeasting of stock yield curve.
出处
《统计与信息论坛》
CSSCI
2009年第3期52-56,共5页
Journal of Statistics and Information
基金
教育部人文社会科学规划项目<数据挖掘中关联规则的统计研究及应用>(06JA91003)
关键词
股票收益率
函数数据
主成分分析
stock yield
functional data
principal component analysis
