基于广义Hurst指数和蚁群优化算法的配对交易策略研究  

Research on Pairs Trading Strategy Based on Generalized Hurst Exponent and Ant Colony Optimization Algorithm

作　　者：孙景云 马小雯 SUN Jing-yun;MA Xiao-wen(School of Statistics and Data Science,Lanzhou University of Finance and Economics,Lanzhou 730020,China;Center for Quantitative Analysis of Gansu Economic Development,Lanzhou 730020,China;Gansu Key Laboratory of Digital Economy and Social Computing Science,Lanzhou 730020,China)

机构地区：[1]兰州财经大学统计与数据科学学院,甘肃兰州730020 [2]甘肃经济发展数量分析研究中心,甘肃兰州730020 [3]甘肃省数字经济与社会计算科学重点实验室,甘肃兰州730020

出　　处：《兰州财经大学学报》2024年第1期88-100,共13页Journal of Lanzhou University of Finance and Economics

基　　金：国家自然科学基金“协整和误价市场中几个最优配对交易模型的研究及应用”(72061020);甘肃省自然科学基金“市场波动率时变特性下最优动态配对交易模型的研究”(21JR1RA280);2022年陇原青年创新创业人才项目“多源大数据驱动的甘肃省旅游需求预测及游客感知研究”。

摘　　要：以上海期货交易所的8种金属类期货为研究对象,首先结合传统协整理论和一阶矩广义Hurst指数方法筛选出具有较强均值回复特性的期货对作为最优配对组合,然后对选出的最优配对组合设计交易策略,利用蚁群智能优化算法确定多空交易的最优开仓阈值。在实证回测阶段,对历史数据采用滑动窗口方法进行多次样本内及样本外回测,并与基于均值和标准差的传统固定阈值策略进行对比。结果发现,螺纹钢-热卷、热卷-铝期货配对为不同样本期下的最佳配对组合。从交易效果看,采用蚁群智能优化算法所确定的上、下开仓阈值策略相比于传统阈值策略,在多空交易中获得了更高的年化收益率和夏普比率。This study focuses on eight metal futures contracts listed on the Shanghai Futures Exchange.Initially,leveraging traditional cointegration theory and a first-order generalized Hurst exponent method,futures pairs demonstrating robust mean-reverting characteristics are identified as optimal pairs for trading.Subsequently,trading strategies are devised for the selected optimal pairs,employing an ant colony optimization algorithm to determine the optimal opening thresholds for long-short trading.During the empirical back-testing phase,the research utilizes a sliding window approach for multiple in-sample and out-of-sample backtests on historical data,comparing the results with traditional fixed threshold strategies based on mean and standard deviation.The findings reveal that the steel rebar-HRC and HRC-aluminum futures pairs are the best combinations for different sample periods.In terms of trading performance,the strategy employing the ant colony optimization algorithm to determine the opening thresholds outperforms traditional threshold strategies,achieving higher annualized returns and Sharpe ratios for long-short trading.

关 键 词：协整理论 广义HURST指数 蚁群优化算法 多空交易 

分 类 号：F832[经济管理—金融学]