Modelling the Price of Industrial Commodities

Economic Modelling, Vol. 16, 1999

Posted: 24 Nov 1999


This paper connects two branches of economics in a study of the price formation of industrial commodities. The first branch is the theory of irreversible investment under uncertainty, the second the theory of competitive speculation in stocks. A model which combines these two elements is developed and applied to the aluminium market.

The theory of irreversible investment under uncertainty deals with problems of investment or planning where the investment outlay is irretrievable, once spent. The combination of irreversibility and uncertainty has been shown to change the usual Marshallian action triggers, measured as deviations from variable cost, by a factor of as much as 2 or 3, leading to inertia, or "hysteresis", in responses of supply to changes in price. The approach has ranged from the point of view of the isolated investor, through industry-wide and general equilibrium.

Inventory speculation is a key feature of trade in commodities. Indeed, the price of a product such as aluminium on the LME (London Metals Exchange) is commonly the reference price in contracts in the industry. In the usual model of fluctuations in the price of industrial raw materials, demand and supply have low price elasticities in the short-term and demand is subject to random shocks. Therefore, the market price of these goods can fluctuate dramatically. Stock speculation introduces an asymmetry in these fluctuations. When supply exceeds consumption demand and prices are low, investment in inventories occurs because of expectations of future price rises. The effect of inventory speculation is therefore to keep prices higher than otherwise. If, on the other hand, consumption demand at the current price is in excess of supply and stocks are low, the price may shoot up. Thus, inventory speculation does not even out price peaks to the same degree as the troughs.

The contribution of this paper is to tie the two models outlined above together with the aim of modelling the evolution of industry supply, speculative demand, prices and consumption. The result is a combination of the Dixit entry/exit model with a version of the Deaton-Laroque/Williams-Wright model. This is motivated by the failure of the latter type of models to generate cycles of sufficiently long periodicity to account for observed movements in the prices and inventories of primary commodities. The predictions of the model are then compared to real data on the aluminium market.

In a continuous time version of the model, an endogenous price zone arises from the opening and closure of capacity at boundaries of the price zone. Dynamics of the price process within the price zone are determined by competitive speculation. A discrete time version of the model is scaled to the dimensions of the aluminium market and solved numerically. Simulated prices and inventories exhibit characteristics similar to the real series: volatile prices, asymmetric adjustment of production to price changes and long inventory cycles.

Note: This is a description of the paper and not the actual abstract.

JEL Classification: E22, E23, D84, D92

Suggested Citation

Baldursson, Fridrik Mar, Modelling the Price of Industrial Commodities. Economic Modelling, Vol. 16, 1999, Available at SSRN:

Fridrik Mar Baldursson (Contact Author)

Reykjavik University ( email )

Menntavegur 1
Reykjavik, 101
354-8256396 (Phone)
354-5996201 (Fax)


Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Abstract Views
PlumX Metrics