Price Discovery in Commodity Markets: The Case of Metals

Abstract

This article provides a method of quantifying price discovery in commodity markets. By price discovery we mean the process by which commodity markets attempt to identify permanent changes in equilibrium transaction prices. We are interested in commodity markets in particular the dynamics of commodity prices and the existence of a liquid futures market. Our work departs from the previous literature in that it applies Gonzalo-Granger (GG) decomposition to show that future prices are information dominant in relation to spot prices, in the sense that they are the most important contributors to the revelation of the common factor. Unlike in the Hasbrouck methodology the GG decomposition offers a basis for robust estimation and testing despite the presence of high correlation across disturbances. The GG method, like the Hasbrouck method is appropriate to the study of price discovery in that only a minimal structure is imposed on the dynamics of the price series. We estimate and test common factor components attributable to spot and future (3 month) prices within 5 metals traded in the London Metal Exchange (LME). This involves our demonstrating that spot and future metal prices are cointegrated, suggesting that within the cost-of-carry framework the price differential is stationary. Our results show that the future price is infomationally more efficient for the most liquid LME traded metal contracts.