摘要
文章提出了一种新的非完全复制指数跟踪方法。该方法基于选择具有一定优良性质的股票组合的基础,构建跟踪组合来跟踪指数。新方法突破了文献已有方法事先指定股票来构造指数跟踪的传统框架,更有实际价值。实证分析表明,采用该方法构建的跟踪策略通过选择较少的股票同时得到的跟踪误差也较小,且优于目前比较流行的分层抽样策略。最后应用所给的方法对沪深300指数进行了股指期货的期现套利研究,结果表明该策略比文献已有的方法存在更大的套利空间。
This paper provided a new strategy to construct a portfolio to replicate the reference index.The strategy tracks the index based on a more excellent stock selection technique.The method we proposed broke the traditional framework which should pre-designate a stock pool to construct portfolio in literature.The empirical results show that this method brought relatively small tracking error even with fewer stocks,and simultaneously the strategy performed better than the current stratified sampling strategy.We applied this method in the analysis of the future-spot arbitrage of CSI 300 stock index futures,and found out this strategy get more arbitrage return than the existing method.
出处
《数理统计与管理》
CSSCI
北大核心
2014年第3期508-518,共11页
Journal of Applied Statistics and Management
基金
国家自然科学基金资助项目(70221001
70331001
71003100)
教育部人文社会科学研究一般项目(11YJC630270)
中央高校基本科研业务费专项资金资助项目(11XNK027
10XNF020)
关键词
指数跟踪
高维选元
期现套利
index tracking
stepwise search feature variables
future-spot arbitrage
