An assessment of time series methods in metal price forecasting

The international price for metals is pivotal in the profitability equation for mining companies. If producer prices rise, assuming production levels and costs remain the same, profits are expected to increase. Accordingly, producers welcome any means by which price instability and unpredictability can be reduced. The paper analyses the ability of two user-friendly time series forecasting techniques to predict future lead and zinc prices. The conclusion is that price forecasting is difficult. It should, however, be acknowledged that whilst neither of the two models are definitive, they are useful for the mining company vis-a-vis its planning process. In particular, the results from the analysis in this paper suggest that ARIMA modelling provides marginally better forecast results than lagged forward price modelling. The methodologies employed in this paper have a broad based application to base metal forecasting by mining companies in general, that is, the applications are transferable. (c) 2005 Elsevier Ltd. All rights reserved.