- Leverage effect
Asymmetric GARCH models were developped for equity stocks to take into account the larger response of the conditional variance to negative price shocks. We show that these asymmetric GARCH models are also relevant for modelling commodity prices. Contrary to the equity case, positive shocks are the main contributors to the conditional variance of commodity prices. The theory of storage, by relating the state of the inventories of a commodity to its conditional variance, is a serious candidate to explain the phenomenon, as positive price shocks for commodities usually serve as proxies for the deterioration of the inventories. We find that this inverse leverage effect, or “inventory effect”, is relatively robust, for different subsamples, for diverse types of commodities and for different ways of specifying the asymmetry, though weaker than the leverage effect for equity stocks. Appropriately specifying the asymmetric conditional variance of commodities could improve risk management, hedging strategies or Value-at-Risk estimates. Incidentally, the inventory effect sheds some new light on the debate about the origin of the leverage effect.