基于高频数据的统计套利组合策略研究  

Research on the combination strategy of statistical arbitrage based on high-frequency data

作　　者：周明华[1] 周婷婷[2] 张敏凯 

机构地区：[1]浙江工业大学健行学院,浙江杭州310023 [2]浙江工业大学理学院,浙江杭州310023

出　　处：《浙江工业大学学报》2017年第3期336-341,共6页Journal of Zhejiang University of Technology

摘　　要：采用沪深300股指期货当月连续合约与次月连续合约的1min高频数据,借鉴组合思想,引入伯努利随机变量,将GARCH(广义自回归条件异方差模型)模型与Ornstein-Uhlenbeck模型进行组合,设计出新的统计套利组合策略,并在改进的统计套利策略的基础上进行套利.我们采用动态交易的方式对改进后的交易策略的实际交易效果和模型的有效性进行检验,实证检验结果表明:对单一套利模型进行组合是必要的,组合后的整体套利收益水平相较单一模型套利收益来说明显提高了.On the basis of the thought of combination strategy, a new combination strategy of statistical arbitrage which combine GARCH （ generalized autoregressive conditional heteroskedastic） model and Ornstein-Uhlenbeck model by introducing into Bernoulli random variable is developed, then we arbitrage on the basis of the improved statistical arbitrage strategy. The validity of the model and the actual trading effects has been tested by the dynamic methods. The results of empirical test show that it is necessary to combine the single arbitrage models, and the overall income level of the combination strategy compared to that of single strategy is significantly improved.

关 键 词：高频数据 GARCH ORNSTEIN-UHLENBECK 组合策略 伯努利变量 

分 类 号：F832.48[经济管理—金融学]