Beyond Oil Prices: Detrended Oil Prices and Stock Return Predictability

ABSTRACT

Fluctuations in crude oil prices exert a substantial influence on global financial markets. While the level of oil prices has been widely studied, we argue that deviations from their long‐run trend carry distinct information for asset pricing. We construct a measure of detrended oil prices (DOP) and examine its ability to predict stock returns from January 1986 to December 2023. DOP emerges as a robust negative predictor of aggregate stock market returns and outperforms conventional oil‐price measures in forecasting accuracy. Its predictive power is economically meaningful and statistically significant both in‐sample and out‐of‐sample, persists across short‐ and long‐run forecasting horizons, and holds across a broad sample of oil‐importing and oil‐exporting countries. Economic analysis suggests that DOP is mainly associated with demand‐side oil shocks, contains information about both cash‐flow and discount‐rate news, and is consistent with investor underreaction.