摘要
本文研究基于协整的配对交易的阈值选择问题。我们针对价差序列进行AR(1)建模拟合,进而采用数值算法估计交易持续期、交易间隔期和交易次数,最终将最优阈值的选择转换为利润最大化的问题。在对我国A+H股股价数据进行的实证分析中,本文显示这种最优阈值的确定方法是有效的,并且该方法在固定参数下的交易表现优于时变参数模型。
This paper studies the threshold choosing problem of pairs trading strategy based on co-integration analysis. We fit the spread series using AR（1） process, and estimate the duration, interval and number of transactions by applying numerical algorithms, and finally transferring the choosing of optimal threshold to the profit maximizing problem. Empirical analysis of pairs trading strategy based on the optimal threshold by using A＋H shares data shows that this method is effective, and can provide better performance in fixed parameter model than in time-varying parameter model.
出处
《投资研究》
2015年第11期79-90,共12页
Review of Investment Studies
基金
国家自然科学基金项目(71473092
70971051)资助
关键词
配对交易
最优阈值
协整
AR(1)过程
Pairs trading
Optimal threshold
C o-integration
AR（1） process
