An information diffusion-based model of oil futures price

Inspired by the increasing evidence of financialization/speculation in commodity pricing, this paper constitutes a first attempt to build an information diffusion-based asset pricing framework for the oil futures market. With gradual information dissemination, slowly decaying uncertainty about the asset's future fundamentals generates persistent conditional volatility and a drift in asset return. Volatility-based proxies for information flows are proposed to examine empirically the asset pricing implications. The results confirm a significant intertemporal relationship between return on the price of oil futures, information diffusion and volatility components. An important implication of our study is that the slow diffusion of information generates predictability in price dynamics. A forecasting model is then constructed and tested in relation to our theory. It is found that the lagged series of the pricing factors possess significant predicting power for returns.

► Oil future contracts are modeled as financial assets in an information diffusion framework. ► The theoretical model yields a three-factor asset pricing formula. ► Empirical investigation confirms the validity of our theoretical model. ► Our forecasting model has better predictive accuracy than the GARCH and ARMA models.