Testing rational expectations in primary commodity markets

The standard method used to test the rational expectations hypothesis (REH) in primary commodity markets is by means of a structural approach. In this paper, a parsimonious vector error correction model in the price and stock equation is derived that maintains almost complete information of the underlying structural model. The empirical section utilizes 1955–2000 US copper data to investigate the properties of the model extended to the macroeconomic variables. The estimation results are statistically robust and are in keeping with economic theory. Three different results are found: (i) price adjustments depend on the shortrun dynamic of the stock equation, whereas the long-run dynamic is statistically rejected; (ii) the over-identification restrictions, including the test for the REH, are not rejected; (iii) the forecast simulations on price are well performed.