The Relationship Between CSI300 Index Futures And CSI300 Index And The Study Of Trading Strategy

With the rapid development of financial industry and the increase of risk and uncertainty, various financial derivatives innovation emerges in endlessly. It is believed that stock index futures is the most successful financial innovation and one of the most actively traded derivatives. Index futures has functions such as price discovery, arbitrage, hedging, market stabilizing so that it is favored by investors and governments. However there exist potential risks due to high leverage of stock index futures. To improve system of Chinese capital market,China Financial Futures Exchange released CSI300 index futures officially traded on April 16 th, 2010 when CSI300 index had a sharp decline. It makes investors doubt whether index futures has the ability of stabilizing the stock market and price discovery. Therefore this paper studies the relationship between CSI300 index futures and CSI300 index and puts forward an appropriate trading strategy based on the relationship.One-minute high frequency data of CSI300 index and CSI300 index futures are selected from April 16 th, 2010 to April 15 th, 2011 as sample data. The results are as follows:(1) For efficiency and accuracy, this research takes natural logarithm of CSI300 index futures and CSI300 index. After unit root test, it shows that CSI300 index and CSI300 index futures are integrated of order 1. The first difference of logarithm means that yield is stationary and can be predicted with methods of time series.(2) We show that there exists a long-term relationship between CSI300 index futures and CSI300 index with threshold autoregressive model.Furthermore, including efficient energy parameter can reflect a stationary trend and thus improve the threshold autoregressive model. We also show that CSI300 index futures and CSI300 index have a long-run relationship. Given this long-run relationship, that causal relationship is tested with threshold error correction model and it shows that CSI300 index futures leads the price about 15 to 28 minutes earlier, while CSI300 index has a weaker support to lead CSI300 index futures about 8 to 47 minutes.(3) Based on the results above, this research predicts the trend of CSI300 index futures through AR model, MA model and ARIMA model by selecting the data from April 16 th, 2010 to April 15 th, 2011. As it is shown in the results,accuracy of MA model is better than ARIMA model and AR model, but the difference is small. Compared with the trend of stock index of lag order about 18,20, 22, 24, the predictive data and real data of CSI300 index futures, from April16 th, 2011 to April 9th, 2013, have the same trend so that the prediction is effective. Therefore investors can make a trading strategy in advance to get profit from trend of stock index by predicting trend of stock index futures.