Research On The Cyclical Fluctuations And Forecasting Model Of Copper Futures' Price Based On Wavelet Theory

With the increasing needs of copper in China, the domestic consumption of copper has ranked first in the world. Copper futures was becoming an important means of investment and a hedge. However, due to the interference of various factors, it is difficult to predict the copper futures price and the risk of futures trading. Therefore, in order to avoid the risk of international trade and provide supporting information for the investors and producers, this thesis focused on the application of wavelet theory in the sequences processing in copper futures price and aimed at finding a more accurate prediction model for the domestic producers and investors.As the LME copper price is the main basis for decision-making of domestic enterprises, we use the LME three-month copper futures price as research subject. First of all, we studied the factors which influence the copper futures market and then researched the period of copper futures market by wavelet transform and decomposition. After that we compared the periodic components of copper futures price and the corresponding periodic component of China's economic cycle so as to reveals the linkage between them. Finally time series reconstruction model was established by the different decomposition parts of the time series, which is based on wavelet analysis and ARMA model. The forecast results show that the combined model which established by this paper is more accurate than the traditional AR-GARCH model.With the wavelet analysis tool's wonderful time-frequency features and local analysis function, the result of this thesis will supply a better condition to predict the price of futures and to invest in futures market.