Distributional predictability between commodity spot and futures: Evidence from nonparametric causality-in-quantiles tests

Abstract

Unbiasedness and informational efficiency of futures markets under different market conditions is a claim that still remains unsettled in the theory of non-arbitrage and asset pricing and in empirics as well. This study investigates this claim using a novel causality-in-quantile model of Balcilar et al. (2016) for two energy commodities, crude oil and natural gas, and two precious metals, gold and silver. The model estimates causalities-in-mean and -variance between spot and futures market in a time varying context conditioning on the states of the markets represented by the quantiles of the conditional distribution of the dependent variable. The causality in return is asymmetric and unidirectional from futures to spot market for all commodities. That means the predictability of the futures market, due to its informational efficiency, is strong in the normal market and declines when the spot market enters into extreme bearish and bullish conditions. The causality-in-variance is bi-directional in the normal to bull markets except for natural gas where it is unidirectional from futures to spot only. It is a confirmation of the risk management and price discovery role of futures market. Lack of causality in the bear market entails some kind of disconnectedness between the spot and futures markets in a bad market where intervention is called for from the exchange and regulators to restore stability in the spot and futures market dynamics. Although economic uncertainty is found to have no impact on the causality-in-mean except gold; however, the causality-in -variance is influenced in the case of gold and crude. This is a kind of reaffirmation of the fact that under economic uncertainty, futures contracts are used for hedging under different market conditions. However, the causality between commodity spot and futures are resilient to exchange rate.