Analysis Of Pairs Trading Strategy Based On A-share

Since the Shanghai Stock Exchange and Shenzhen Stock Exchange started the margin trading business at the end of March 2010 officially,the domestic stock market has officially owned the sell-short mechanism,which has provided a good platform for the application of pairs trading in China.Pairs trading strategy is an arbitrage statistical strategy,which requires to find two stock with long-term equilibrium relationship.When the spread between the two stocks deviates from the long-term equilibrium level,long the relatively undervalued stocks,short the relatively overvalued stocks,and then wait for the spread to return to equilibrium to earn the profit of spread convergence.This paper takes the component stocks of Shanghai and Shenzhen 300 index as the research target,takes the stock data from the beginning of 2017 to the end of 2020 as the training sample,and takes the stock data from the beginning of 2021 to the end of 2021 as the back test sample,to verify the feasibility of pairs trading in the domestic market through empirical analysis.Pairing trading can be divided into two steps:one is to screen stocks,and the other is to build trading strategies.In screening stocks,this paper selects the qualified stock portfolio through correlation analysis and cointegration test,and uses the error correction model to calculate the spread series and pairing ratio.In the construction of trading strategy,it is divided into fixed threshold strategy and time-varying threshold strategy.The fixed threshold strategy believes that the standard deviation of spread series is always stable,so a certain multiple of the standard deviation of spread series is used as the trading signal of opening and closing positions.The time-varying threshold strategy considers that the standard deviation of the spread series is always changing,and uses GARCH model to fit the time-varying standard deviation series of the spread series,taking a certain multiple of the time-varying standard deviation series as the trading signal of opening and closing positions.In this paper,we use one standard deviation as the opening threshold,two standard deviations as the stop loss threshold and zero standard deviation as the closing threshold to simulate pairs trading with two trading strategies.Through the back test of the two trading strategies,the two trading strategies have good benefits in the training interval and back test interval,and the following conclusions are obtained:first,after considering the cost,the cumulative rate of return of the fixed threshold strategy in the training interval is 36.81%and the time-varying threshold strategy is 47.02%.The cumulative yield of the fixed threshold strategy in the back test interval is 7.34%,and the time-varying threshold strategy is 15.21%.Compared with the fixed threshold strategy,the time-varying threshold strategy increases the cost in all aspects due to the increase of transaction times,but the overall net profit is still much better than the fixed threshold strategy.Second,in the margin account,due to the increase in the number of transactions,the time-varying threshold strategy,on the whole,the risk on the margin is greater than the fixed threshold strategy.During the training period,a total of 4 margins were added,while the fixed threshold strategy only added one margin.Third,compared with the fixed threshold strategy,the time-varying threshold strategy introduced into GARCH model also has a significant improvement in sharp proportion and transaction times,but it is not as good as the fixed threshold strategy in maximum pullback.