Studies of the price and world market for gold in past academic, fi nancial and semipopular literature have neglected production technology and have concentrated primarily on how the price of gold infl uences gold production. With respect to gold production, classical theorists assumed gold supply to be completely elastic at the market price. Classical economists ignored the question of exhaustion, probably because of ever-increasing gold stocks during the nineteenth century (Rockoff, 1984). This implied that changes in non-monetary factors, such as real incomes and tastes, could not affect the price of gold in the long run. Supply was only a function of the real price of gold, with the nominal price set by the monetary authorities. Bordo and Ellison (1986) relaxed the assumption of constant costs for gold producers under a gold standard. Attempts to build forecasting models for the price of gold after the gold standard have had mixed results and have generally ignored supply considerations1 (Salant and Henderson, 1978; Flood and Garber, 1980; Marsh, 1983; Baker and Van Tassel, 1985; Sherman, 1986).