The Research On Forecast For SHFE Copper Price And Hedging Strategies

Due to the excellent physicochemical properties, copper is widely used in many industries and becomes the most important mental in the whole national economy. As the largest consumer for copper in the world, China relies on the import to satisfy almost 30 percent of its demand for the refined copper. However, the price of the copper is affected by kinds of factors, such as the demand and supply, the global economic situation, the money policies and so on, which accounts for the remarkable fluctuation in the copper price. Consequently, the accurate forecast for the copper price is of great significance for the foreign trade for the copper.Based on this topic, the paper has used three kinds of single models, such as the ARIMA model, GM(1,1) model and BP neural network model, to forecast the SHFE copper price. Based on the result from the above-mentioned single models, we have built two kinds of combination forecast models,the optimal weighted linear combination model and the nonlinear combination forecast model based on the BP neural network, to improve the precision of the forecast model. In addition, by applying nine kinds of data in the model, we has tested the feasibility of the model in different market circumstances with different amounts of the data.The research has drawn the following conclusion:(1) either of the two kinds of combination forecast models has a better forecast precision over the three single forecast models in the different market circumstances with different amounts of the data. (2) Between the two kinds of combination forecast model, the nonlinear combination forecast model based on the BP neural network has a better feasibility in the most conditions than the optimal weighted linear combination model. (3) The combination forecast model is feasible in the most conditions except the reversal market in the long run.To draw a conclusion, this paper gives a combination forecast model by combinating three kinds of the single model in a linear or nonlinear way, which can help to make a price and hedge against the price volatility in the copper trade.