Pairs Trading Based On Partial Cointegration Theory

As a new theory of cointegration test,partial cointegration can effectively increase number of pairs and trading frequency in pairs trading.However,it still has disadvan-tages of transaction threshold selection,less risk control and fewer trading opportuni-ties.To overcome these problems,first,this paper proposes intelligent optimization algorithms represented by genetic algorithm to calculate the optimal transaction thresh-old,and compared with the original fixed transaction threshold.Then,this paper change the setting of fixed stop loss.Propose the optimal transaction threshold with stop loss conditions,and compare it with the optimal transaction threshold model with fixed stop loss and no stop loss.Next,this paper change the original one-way opening and closing position to two-way opening and closing position to improve trading opportunities and overall profitability.In addition,this paper conducts a comprehensive and sufficient empirical test.Empirical test is not only carried out in the matching stocks of Shanghai and Shenzhen 300,China Securities 500 and S&P 500 industries,but also in the extreme stock market and high-frequency futures market by testing the commonness and differ-ence of multiple intelligent optimization algorithms-genetic algorithm,particle swarm optimization algorithm and differential evolution algorithm in pairs trading,which re-flect the stability and superiority of the improved method proposed in this paper.