Trend Decomposition, informativeness and co-integration of futures market price discovery

In this paper Beveridge-Nelson Decomposition is used to examine trend content of soybean purchase price of oil plant in northeast China and soybean future prices of DCE and CBOT, in order to uncover full information in soybean´s price discovery process and avoid contemporaneous correlation. It was found that DCE price was consistently more imformative about the fundamental price of soybean because of its lower noise level. Co-integration analysis and error correction model proved that there is a long-term equilibrium among three sample series; however, no evidence shows co-integration between two futures prices. Both DCE price and CBOT price are Granger Causes of domestic spot price. Price discovery process in DCE is independent of CBOT while price in CBOT sharing information from DCE, due to current situation in Chinese soybean spot market.