The relative efficiency of commodity futures markets

The ability of futures markets to predict subsequent spot prices has been a controversial topic for a number of years. Empirical evidence to date is mixed; for any given market, some studies find evidence of efficiency, others of inefficiency. In part, these apparently conflicting findings reflect differences in the time periods analyzed and the methods chosen for testing. A limitation of existing tests is the classification of markets as either efficient or inefficient with no assessment of the degree to which efficiency is present. This article presents tests for unbiasedness and efficiency across a range of commodity and financial futures markets, using a cointegration methodology, and develops a measure of relative efficiency. In general, the findings suggest that spot and futures prices are cointegrated with a slope coefficient that is close to unity, so that the postulated long-run relationship is accepted. However, there is evidence that the long-run relationship does not hold in the short run; specifically, changes in the spot price are explained by lagged differences in spot and futures prices as well as by the basis. This suggests that market inefficiencies exist in the sense that past information can be used by agents to predict spot price movements. A measure of the relative degree of inefficiency (based on forecast error variances) is then used to compare the performance of different markets. © 1999 John Wiley & Sons, Inc. Jrl Fut Mark 19: 413–432, 1999     

Valuation of futures and commodity options with information costs

In this article, futures and commodity options are analyzed in the context of Merton's (1987) model of capital market equilibrium with incomplete information. First, following Dusak (1973) and Black (1976), the conditions under which Merton's model can be applied to the valuation of forward and futures contracts are proposed. Then an application to futures markets is given. We provide a partial differential equation and the formulas for European commodity options, futures contracts, and American options in the same context. The models are simulated and compared to standard models with no information costs. We find that model prices are not significantly different from standard model prices. However, our models correct for some pricing biases in standard models. In particular, they reduce the overvaluation bias for European and American commodity options. It seems that the costs of gathering and processing information regarding the option and its underlying asset play a role in explaining the biases observed in standard models. This work can be applied to other futures markets. © 1999 John Wiley & Sons, Inc. Jrl Fut Mark 19: 645–664, 1999