The Study Of The Statistical Arbitrage Strategy And Application

Statistical arbitrage aims to get a return on investment by the use of numerical method of constructing portfolio at the same time of reducing market volatility.Its origin dates from the eightys of the 20 th century in Morgan Stanley's match trading.In the following more than ten years.It is very popular among the major brokerages and investment institutions and brought in abundant profit for them.The security market of our country started relatively late,until march 31,2010,the Shanghai and shenzhen exchange margin trading pilot officially launched,stock market of our country is the shorting mechanism.Statistical arbitrage catched the survival of the soil.Recently years,due to the change of market structure,the development of statistical arbitrage in the late incorporates a variety of factors,including trading experience,empirical observations,experimental analysis and so on.With the gradual development of time series analysis theory,analysis of price time series towards capital market problem by using the statistical arbitrage will be more and more perfect.Scholars in our country has also done research on statistical arbitrage,mainly focusing on the structure of the arbitrage model.This article expounds emphatically from two parts,theory and practice.First half focuses on theory,mainly discussing the related theory of statistical arbitrage and trading strategies.including cointegration theory,the stationarity test,common method,GARCH model and so on.The second half take ChangYuan electric power and HuaDian energy data as an example to analysis the performance of statistical arbitrage.In the theoretical part,we emphatically discusses the two aspects : one hand is how to choose stocks,we take the couple whose correlation coefficient is so high as the trading object,the other hand is how to make trading rules,we take 1.25 times of the standard deviation of the residual sequence as the threshold of the trade.Statistical arbitrage is effective in our country through the empirical test.Although there have been a loss,the cumulative annual return is still as high as 18%,which is a good choice to large-scale investment institutions.