Causal Relationships between CSI300 Spot and Its Index Futures Revealed by Granger Causality and New Causality

Granger causality has been widely applied to stock markets to study causal relationships between two time series variables. In this paper, we use Granger Causality (GC) and our recently proposed New Causality (NC) to find the causality relationship between the CSI 300 Spot and its index futures which covers the time period from April 7, 2015 to September 8, 2015. The stock market in this period in China experienced Crazy rise and rapid crash. We first use GC and NC methods to analyze the causality between the CSI 300 Spot and its index future for the all sample period. The results by both methods reveal that the causal influence from CSI 300 index futures to CSI 300 Spot is much greater than that from CSI 300 Spot to CSI 300 index futures. We then calculate the rolling causality using two methods. The GC results show that the causal influence from CSI 300 index futures to CSI 300 Spot is always much greater than that from CSI 300 Spot to CSI 300 index, this may not be true in practice. The NC results show that the causal influence from CSI 300 index futures to CSI 300 Spot is greater than that from CSI 300 Spot to CSI 300 index by 80%, this may be true in practice. Anyway, both methods demonstrate that CSI 300 index futures has a major causal influence in CSI 300 Spot.