Commodity Storage and Interseasonal Price Dynamics

This paper continues Deaton and Laroque's search for a variant of the nonlinear rational expectations commodity storage model that can explain the observed behavior of primary commodity prices. Using numerical functional equation methods and Monte Carlo simulation techniques, we demonstrate that the failure of Deaton and Laroque's model to explain high autocorrelation in primary commodity prices is attributable to the assumption of a constant returns to storage technology. Our findings suggest that a classical cost of storage function with "convenience yield" can give a satisfactory explanation for the high autocorrelation.