High-frequency Trading Model Of Stock Index Futures

April16,2010, the CSI300stock index futures enriches investors’ investment strategy.Affected by the legend story of James Simmons and the rise of quantitative investment abroad,the quantitative investment has grown up in domestic in recent years. More and moremathematical models such as: statistical arbitrage model and the conditions for a long periodof models, are widely used in the stock index futures market. However, quantitativeinvestment in the domestic market has just started, for now, what’s common in the domesticmarket is trading models based on the Moving average, breakthrough, regression and othertraditional technical indicators, and trading models based on pattern recognition.However, the application of the above two types of trading strategies are only to predict theprice, volume and direction. And they are often independent when used, do not play theirrespective advantages. Therefore, this paper proposes a new trading strategy base on themarket microstructure theory: High frequency trading strategies based on the concept offiltering. Then, through the traditional technical indicators and pattern recognition theory, astructured High-frequency trading model of Stock index futures is build. The model mainlyincludes intraday trading strategy, building prediction model and strategy evaluation. TheModel take two regression models to predict the rate of return of double Moving average highfrequency trading strategies, and use Sharpe ratio to test the model.Finally, using the historical data to the model, we find that the high-frequency tradingstrategies based on the concept of filtering can significantly increase the Sharpe ratio ofstrategy, effectively control the risk. At the same time the empirical result shows that thebasic strategy to join compound stop conditions can effectively reduce the risk, reduce thebiggest loss of the policy.