LAUSR

Abstract This study aims to analyze the suitability of factor models in describing the dynamics of real prices for four main non-ferrous industrial metals: aluminium, copper, nickel and zinc. For that purpose, using an extensive dataset of monthly time series covering the years 1980–2019, we extract four different common factors explaining commodity prices, exchange rates, financial and macroeconomic indicators. Next, we examine these factors as potential predictors of the movements of four metal prices with the use of two model classes: direct forecasts (DF) and factor augmented vector autoregressions (VAR). We show that for three out of four metals (aluminium, nickel and zinc) VAR models provide relatively good point and density forecasts, outperforming the random walk benchmark as well as DF models.