基于Markov状态转换的高频统计套利策略研究  

Research on High-Frequency Statistical Arbitrage Strategy Based on Markov State Transition

作　　者：闫舒悦 王辉 陈晨[1] 

机构地区：[1]南京信息工程大学,江苏 南京

出　　处：《应用数学进展》2022年第11期8200-8211,共12页Advances in Applied Mathematics

摘　　要：做空与对冲机制的放开使得配对交易策略可行,协整配对的套利策略被有效促进和发展。本文引进Markov状态转换模型构建交易策略,它能够较好地解释市场状态变动的过程,并结合OU过程求得满足期望收益最大条件下的最优入场交易信号。研究表明,Markov模型将市场分为“高波动”和“低波动”两种状态,高波动的存在客观上影响了配对交易的收益。同等风险下,基于马尔可夫状态转换模型策略的配对交易收益要比基于协整模型的收益情况更优,收益率的波动更稳定风险更小。The pairing trading technique is practicable thanks to the liberalization of the short and hedge mechanisms, and co-integration pairing trading is effectively promoted and developed. In order to build a trading strategy, a Markov switch model is introduced in this study. This model can provide a better explanation of the process of market state change, and combined with the OU process obtain optimal entry trading signals that satisfy the greatest expected return. In conclusion: the market is divided into two states by the Markov model: "high volatility"and "low volatility", and the presence of high volatility objectively affects pair trading returns. The returns of pairing trading using Markov switch model are better than those using a co-integration model at the same risk;volatility of yield is more stable and less risky.

关 键 词：配对交易 协整 高频 Markov转换模型 最优阈值 

分 类 号：F832[经济管理—金融学]