Cross market predictions for commodity prices

Abstract

This paper shows that commodity prices can be predicted from cross-market information by establishing long-run cross-market commodity price equilibrium models, which are characterized by a linear relation between prices across different markets. Using data from five representative commodity markets (oil, copper, gold, corn, and cattle) during the period 2005–2018, we demonstrate that oil and industrial metal markets have formed a long-run price equilibrium with other markets across different commodity families. However, agriculture and gold markets do not tend to have long-run price equilibrium relations with other commodity markets. Furthermore, we show that the absence of a price equilibrium is due to the cross-market liquidity interference effect. After we control for the liquidity effect, long-run cross-market commodity price equilibrium relations are reestablished for agriculture and gold markets. These results can aid in demonstrating that liquidity can capture most of the missing information that is not reflected in price dynamics in less liquid markets, such as agriculture and gold markets. Therefore, less liquid commodity price predictions require both prices and liquidity levels from cross-markets, while liquid commodity prices (oil and metal) can be predicted based solely on cross-market prices.