Research On A-share Quantitative Trading Strategy Based On Statistical Arbitrage

This paper studied the quantitative trading strategy on basis of statistical arbitrage for A Share Market of China with 12 band-stocks' daily trading data which ranged from Jan 1st,2014 to June 30 th,2015.To begin with,we used co-integration theory to select BOC(601988.SH)and ICBC(601398.SH)as paired stocks which had the best co-integration relationship.Then we made an original pairs trading model with the 2 stocks mentioned above.When studying the method to improving the original pairs trading strategy,we noticed that the Bollinger Bands were designed on basis of statistics and they could describe asset price's normal fluctuating range dynamically by using the relative position and volatility rate of asset price.Therefor we decided to apply Bollinger Bands' breakthrough strategy to describe the normal fluctuating range of paired stocks' spread dynamically and improve the mechanism of touching off the arbitrage trading.Finally,we designed a new type of pairs trading model on basis of co-integration theory and Bollinger Bands' breakthrough strategy.This new pairs trading model can change the thresholds of opening and closing position adaptively according to market quotation.That means our new model can catch more arbitrage opportunities and makes considerable profit because of its better mechanism of touching off the trading.Besides,this new model can avoid the problem of over-fitting and therefore has excellent stability and maneuverability.Compared to the original pairs trading model,the annualized return of our new model has reached to 10.76% in the back-test,which confirmed that improving the original pairs trading model by applying Bollinger Bands' breakthrough strategy has acquired ideal result and showed more practical significance.