A Study On The Co-movement Effect Of Thermal Coal Futures And Coking Coal Futures Markets

In recent years, as the financial innovation is accelerated in China, more and more financial derivatives continue to emerge. As the most important derivative of black metal, Coal futures got much attention in the past years. In 2013, the thermal coal future, the coking coal future and iron future launched Chinese future markets which made the commodity future kinds more perfect. The Thermal coal future and coking coal future provided perfect tools to hedge and avoid the risk from the volatility of spot prices. There was a long-term phenomenon, "with the up with the fall", between the spot price of the thermal coal and the coking coal.This phenomenon provided a probability for investors to carry out the cross-bread training between their future markets. With this, we attempted to research the co-movement effect between thermal coal futures and coking coal futures markets existed or not. Besides, this paper tried to analyze the information transmission mechanism between them. This research can provide a reference for the cross-bread training and we can get some information about the running state of the coal future markets.The co-movement effect is examined from two aspects, the mean spillover effect and the volatility spillover effect. As the mean spillover effect, we apply Vector auto regression model, Cointegration test, Granger causality test and Impulse response analysis in the empirical research based on the 1 minute high-frequency data of thermal coal futures and coking coal futures. The result suggests that there is a long-term equilibrium between the price series of thermal coal futures and coking coal futures and the return series exist significant bidirectional mean spillover effects. As the volatility spillover effect, we apply the BEKK—GARCH model as the main econometric model. Otherwise, we use the methods of wavelet multi-resolution decomposition. With this method we can decompose the return series in different frequency domains. Then the researchers can analyze the different expression of the volatility spillover effect based on the econometric regression. The econometric regression result suggests that there were significant bidirectional volatility spillover effects between the return series of thermal coal futures and coking coal futures. Through the methods of wavelet multi-resolution decomposition we find the spillover effects express different in different domains. Specifically, in the d2、d4、d5 decomposition scale, the significant bidirectional volatility spillover effect exist but their expressions are different. In the d1、d3、al decomposition scale, the significant bidirectional volatility spillover effect do not exist...