Predictability of crude oil prices: An investor perspective

We forecast the density of crude oil futures returns using both macroeconomic variables and technical indicators over the period January 1986 through December 2015. The macro variables reflect oil market fundamentals while the technical indicators are constructed based on the popular moving average rules. Several combination strategies over both constant and time-varying parameter models are employed to generate density forecasts. The out-of-sample result shows statistical and economic significance of the predictability. Forecast combination over technical indicators generates more accurate density forecasts than the combination over macro variables. Technical indicators also perform better in terms of Sharpe ratio and certainty equivalent return for risk-averse investors who seek a trade-off between risk and return in the oil market. Technical indicators can better predict oil return density during the expansion period, while macroeconomic variables generate more accurate out-of-sample forecasts during the economic recession period, providing complementary information over the business cycle.