Shanghai Copper Futures Pricing

In this article, we research copper futures including the London LME, New York COMEX and Shanghai copper futures with qualitative, quantitative and mathematical modeling method. We conduct an empirical analysis on copper futures pricing in Shanghai Futures Exchange, and forecast the futures prices in and out the sample with the stochastic process, differential equations, state space model and Kalman filter method.With qualitative analysis, we conclude that the copper price will be subject to many variables of the real economy and virtual economy in the framework of the global macroeconomic. At different stages of economic development, the dominant factors to affect the price of copper futures are different. With the development of the global economy, as well as the limited copper resources, the copper price in the long term will maintain the overall upward trend. With the evolution of the economic cycle, the price of copper at different stages will experience periodic rise and decline. But, in the short term, what factors and which variables on the price of copper are more direct and more effective? We can only explore the key variables in a unique economic environment.However, the methodology of qualitative analysis cannot tell us some other problems, such as, how and how much the copper price changes. To predict the copper futures price effectively, we must apply the method of mathematical modeling to empirical research.In the beginning of the quantitative analysis, we conducted an empirical analysis on copper futures pricing in Shanghai Futures Exchange (SHFE) with Schwartz (1997) two-factor model. The empirical result shows that, there is weak correlation between the copper spot price and convenience yield in China’s copper market. This suggests that further research should include other underlying factors, which may play more important role than convenience yield in emerging markets. Furthermore, in-sample forecasts errors become larger. This may indicates that the performance of the pricing model may be not bad when the fluctuation of copper price is low or parameters are stable.Then, considering the influence of the foreign copper futures on the SHFE copper futures, we add the volatility rate of foreign copper futures price as a third factor into Schwartz (1997) two-factor model, and establish three-factor model including the spot price, convenience yield and the international copper futures price rate. The conclusion is that, in-samplc forecasts errors of the SHFE copper futures with the three-factor model, is farless than the two-factor model. Out-sample forecasts result is excellent; the range of the average error rate is from-0.54%to0.43%.Overall, the three-factorf model including the spot price, convenience yield and the intemational copper futures price rate can be better portrayed the volatility characteristics of the SHFE copper futures pricc, and the effect of pricing and forecasting capabilities have performed better.