Research On The Application Of Statistical Arbitrage Based On Co-integration In SLB Market

Since China introduced the pilot margin trading in2010, the development scale of margin trading has growed significantly. The continued improvements in short mechanism create advantages for the application of statistical arbi-trage in our country. The first half of this pager describes the basics of margin trading, theories of statistical arbitrage and methods of co-integration test. Then systematically research the methods of building a statistical arbitrage model based on co-integration,and determining trading signals. For the short-age of the traditional statistical arbitrage model,build optimization model by improving the model and the trading signals.The optimization model adopts following methods:o Delayed opening strategy,o Advanced position strategy.o By using the GARCH model to deal with the spread sequence of trading portfolio, then use the time varying standard deviation instead of simple standard deviation to obtain more accurate trading signals,o Predict the long-term equilibrium relationship by moving average model.In empirical research,the data is the latest closing price from February21,2012, to February21,2014.we analyze the correlation, stationarity, co-integration and the speed of spread reversion of the underlying stocks in same industry. Ultimately, the pager select the Datong Coal and Pingdingshan shares from six industries that include banking,insurance,securities, coal, oil, as well as ETF. Based on the transaction data of Datong Coal and Pingding- shan shares in. recent years, we built a traditional statistical arbitrage model based on co-integration and optimization statistical arbitrage models. As we can be seen from theories and empirical analysis, the statistical arbitrage mod-el based on co-integration is feasible in China’s stock market.Applications of the delayed opening strategy, advanced position strategy,time varying stan-dard deviation based on GARCH and the moving average model in statistical arbitrage model are effective. The optimal model can capture more arbitrage opportunities, improving the efficiency of success arbitrage, reducing risk and improving return, etc.