Modeling and empirical studies of calendar spread arbitrage of real-time CSI 300 stock index futures

On April 16, 2010, the first stock index futures, CSI 300, was launched in China. Various studies on CSI 300 using simulation data have been performed. In this paper, real data are used to study the spread arbitrage of CSI 300. This paper introduces an arbitrage interval by considering transaction costs on top of previous models and actual situations in China. Co-integration analysis has been tested for validity and then was applied to index futures spread arbitrage through empirical analysis using CSI 300´s high frequency data sampled every 5 minutes. Various time points of arbitrage opportunities are verified. Finally, an ARMA model based on co-integration analysis is constructed to help predict more stable arbitrage opportunities.