Information transmission and entropy-based network between Chinese stock market and commodity futures market

Abstract

This paper measures the information transmission and relationships between Chinese stock market and the commodity futures market on different time scales novelly by combining the transfer entropy from information theory and the multi-scale analysis technique, the complete empirical mode decomposition based on adaptive noise (CEEMDAN). Pairwise information transfer and average information inflow and outflow of assets and markets are investigated. Information spillover network is built based on the net transfer entropy and the direction of information spillover is identified. The results show that the direction and magnitude of information spillover have changed as types of assets and the time scale have changed. There is heterogeneity in the relationship between stock market and the metals, energy and agricultural product futures markets. Lastly by cluster analysis using minimum spanning tree based on transfer entropy, it shows that the methanol futures and information industry form the core of the network. The study finds that with the increase of time scale, the relative information transmission between the commodity futures market and the stock market gradually decreases, and the stock market is in a more influential position.