Does knowledge of the cost of carry model improve commodity futures price forecasting ability?: A case study using the London Metal Exchange lead contract

Abstract

The use of futures prices to predict commodity cash prices is important both to practitioners and researchers yet the literature provides conflicting results on the ability of futures prices to predict cash prices. Brenner and Kroner [Journal of Financial and Quantitative Analysis 30 (1995) 23] argue that if the cost of carry model applies to commodity futures pricing then current futures prices may not accurately predict subsequent cash prices. Inventory, cash price return variance, cash price return first order auto-correlation and interest rates are used to proxy carrying costs in a test of the ability of commodity futures prices to predict cash prices. Various predictive models relating futures price to cash price are described, including univariate and multivariate error correction models. London Metal Exchange (LME) lead cash prices, lead futures prices, lead inventory and UK treasury bill rates are collected over the period 1964 to 1995. Analysis of this data confirms the importance of the cost of carry model elements as well as futures price in forecasting cash prices.