The Illusion of Oil Return Predictability: The Choice of Data Matters!
Research by Thomas Conlon, Ph.D., Michael Smurfit Graduate Business School, University College Dublin, Ireland; John Cotter, Ph.D., Michael Smurfit Graduate School of Business, University College Dublin, Ireland; and Emmanuel Eyiah-Donkor, Ph.D., Rennes School of Business, France.
This article re-examines the previously documented evidence of crude oil return predictability from several popular economic predictors and technical indicators and their combinations. It shows that monthly average oil returns are forecastable, in line with evidence documented in previous studies. On the contrary, no evidence of predictability is found for end-of-month oil returns. The authors conclude that the evidence of oil return predictability documented in previous studies may be misleading, as it stems from the use of within-month averages of daily oil prices in calculating monthly returns whereas end-of-month returns are more relevant for risk management and investment decision making as reflecting actual change in asset value.
Read Article