Reliable Approach to Forecast Prices of Precious and Base Metals

Getting reliable and trustworthy estimates of future metal prices is important. This manuscript applies a pricing model based on geometric Brownian motion simulation to test the reliability of expected price forecasts of silver, aluminum, copper, iridium, nickel, lead, palladium, platinum, rhodium, ruthenium, tin and zinc. Expected prices were estimated by totaling up the product of simulated prices and associated probabilities at the monthly, quarterly and annual frequencies, with historic mean and standard deviation based on a rolling twenty-year window as proxies for drift and diffusion. Results indicate that one-period ahead forecasts based on higher number of simulations are more reliable than those based on only one simulation. Besides monthly forecasts and quarterly forecasts may be more trustworthy than those at the annual frequency.