Enhancing equity return forecasting through oil price fluctuations: An analysis of economic constraints and predictive strategies

This study investigates the effectiveness of constraint-based forecasting methodologies in enhancing equity return prediction accuracy by using equity market information across multiple economic regimes. We develop a comprehensive framework comparing unconstrained forecasting with three innovative constraint strategies: Campbell-Thompson (CT), Parameterized Time-Varying (PTV), and zero-yield (ZY) constraints, analyzing both original oil returns (OR) and adjusted positive oil returns (APOR) using extensive data spanning 1927-2024. Our empirical analysis reveals that the ZY constraint, based on the three-sigma rule, consistently outperforms traditional forecasting methods, achieving remarkable out-of-sample R-squared improvements of 1.459% for univariate models and 1.777% for forecast combinations, respectively. Contrary to theoretical expectations, the OR demonstrates superior and more reliable predictive performance compared to the APOR across all constraint methodologies, with the OR achieving positive forecasting results, while the APOR consistently produces negative out-of-sample performance despite stronger in-sample relationships. The economic utility analysis demonstrates substantial practical benefits, with Sharpe ratio improvements ranging from 10.7 to 21.7 basis points annually and certainty equivalent return gains exceeding 145-365 basis points, translating into economically meaningful benefits for portfolio management applications. We identify discount rates and sentiment channels as the primary transmission mechanisms underlying oil-equity predictability, while cash flow effects show minimal empirical support. The regime-dependent analysis reveals superior performance during periods of economic expansion and geopolitical uncertainty. These findings establish constraint-based forecasting as a valuable methodological innovation for commodity-equity predictability analysis, with important implications for energy portfolio management, macroeconomic forecasting and monetary policy formulation. • We forecast equity market returns using oil price fluctuations under various economic constraints. • Constraint strategies improve the predictive power of symmetric oil returns over asymmetric ones. • A novel restriction based on the three-sigma rule improves prediction accuracy over other methods.