Uncertainty and U.S. Stock Index Returns: A Comparative Analysis of Economic Policy Uncertainty, Geopolitical Risk, and Oil Price Uncertainty for the S&P 500

Abstract

This study examines how three prominent forms of macro uncertainty—Economic Policy Uncertainty (EPU), Geopolitical Risk (GPR), and Oil Price Uncertainty (OPU)—affect the S&P 500 monthly log return (SP500return). A four-variable vector autoregression (VAR) is estimated after verifying stationarity in levels and selecting the lag order with information criteria. The empirical workflow includes unit-root tests, lag-length selection, system estimation, stability diagnostics based on characteristic roots, Granger causality tests, and orthogonalized impulse responses identified under a Cholesky ordering in the sequence EPU, GPR, OPU, SP500return. All series are I(0); information criteria favor VAR(4); and the system is covariance-stationary. Results show that EPU contains forward-looking predictive content for aggregate equity returns—higher EPU today is followed by lower SP500return. OPU delivers a short-run negative effect that mean-reverts by the mid-horizon, whereas GPR exhibits weak or statistically insignificant mean effects at the index level, consistent with transmission mainly through volatility and sectoral reallocation channels. Portfolio guidance follows: monitor EPU in macro-risk dashboards; during OPU upswings, trim exposures to cost- and rate-sensitive sectors and consider commodity or inflation-linked hedges; for GPR, prioritize volatility targeting and sector rotation over directional index positions. Future work can employ time-varying-parameter or regime-switching VARs, sign- or narrative-restricted SVARs, and local projections to capture state dependence and nonlinearities.