V/S长记忆检验的有效性——以上证50指数及其成分股为例  

Study on the V / S Test Validity of Long Memory——As the SSE50 index and its 50 constituent stocks for example

作　　者：管河山[1] 王谦[1] 刘春[2] 

机构地区：[1]南华大学经济管理学院,湖南衡阳421001 [2]南华大学研究生处

出　　处：《南华大学学报（社会科学版）》2015年第3期54-58,共5页Journal of University of South China(Social Science Edition)

基　　金：教育部人文社科青年项目"海量金融时间序列数据平稳性检验方法研究"资助(编号:13YJCZH044);湖南省研究生创新项目"大数据时代金融时间序列长记忆性检验方法研究"资助(编号:CX2014B397)

摘　　要：文章采集上证50综合指数及其成分股数据进行V/S长记忆性检验,实证发现:V/S方法具有较好的稳健性,但其计算复杂度较高,不适合处理海量的数据;采用日、周或月三种采样间隔所得到数据的检验结果有显著差异。增大采样时间间隔虽然可以减少数据量、简化计算,但将导致检验结论不一致;上证50综合指数检验结果与其成分股数据检验结果之间存在明显区别,由此可见,综指与其成分股的长记忆性判断不可混为一谈。实践中应谨慎对待V/S检验结果。The paper did V/S Test with the data of SSE50 index and its 50 constituent stocks in Shanghai Security Exchange and also found that ： The V/S method has high complexity and time-consuming in calculation, so it may not be suitable for handling tnassive amounts of data;under different collected ways of data（ day, week and month）, the results are significantly different in the V/S test of long memory ;increasing the sampling interval can reduce the amount of data and simplify the calculation, but it will lead to an inconsis- tent conclusions of the test; the long memory test results of 50 constituent stocks of SSE50 index are not very consistent, thus, we can- not confuse the results of them. In a word, the long memory test results still should be treated seriously.

关 键 词：时间序列 长记忆性 V/S 有效性 

分 类 号：F276.6[经济管理—企业管理]