节点文献
扫描统计量——检测基金业绩持续性的新方法
Detect Investment Fund Performance Persistence With Scan Statistics
【摘要】 基金业绩持续性检测的一般方法是横截面回归与列联表分析,都是从基金行业层面整体检测持续性,不能有效的检测单只基金的业绩持续性。为此本文引入一种新的检测方法———扫描统计量,可以创新性地对单只基金进行有效分析。利用扫描统计量方法对我国基金业绩持续性进行了实证分析,发现部分基金存在持续性,给出了持续性强度的上限,为投资者买卖基金、基金业绩考核、风险监控提供了决策参考依据。
【Abstract】 Cross sectional regression and contingency table are widely used in the research of mutual fund performance persistence.These methods are ineffective when each individual fund is concerned.In this article we find that scan statistics can be used to deal with this task efficiently.This method is completely new in the research of persistence.We make an empirical research based on Chinese mutual funds,and find persistence does exist in some funds.We also report the upper limit degree of persistence.
【关键词】 金融学;
业绩持续性;
扫描统计量;
证券投资基金;
事件聚集;
【Key words】 finance; performance persistence; scan statistics; mutual funds; clusters of events;
【Key words】 finance; performance persistence; scan statistics; mutual funds; clusters of events;
【基金】 国家自然科学基金资助项目(70141015)
- 【文献出处】 运筹与管理 ,Operations Research and Management Science , 编辑部邮箱 ,2006年01期
- 【分类号】F830.91
- 【被引频次】51
- 【下载频次】442