Forecasting Economic Policy Uncertainty: A Geopolitical Risk Perspective

Abstract

This study investigates the predictive power of Geopolitical Risk (GPR) on Economic Policy Uncertainty (EPU) using an augmented Heterogeneous Autoregressive (HAR) model across multiple forecast horizons. Employing a comprehensive dataset spanning from April 1987 to December 2023, the research aims to enhance the understanding of EPU’s determinants by incorporating the dynamic and often preceding nature of geopolitical tensions. By integrating GPR into the forecasting model, the study reveals that GPR contributes significant predictive content to the EPU, outperforming the baseline HAR model that excludes GPR. This is evidenced by the notable reduction in Mean Squared Forecast Error (MSFE) across all forecast horizons, indicating that GPR not only impacts EPU in the short term but may also have long-lasting effects. The Cumulative Difference of Squared Forecast Errors (CDSFE) analysis further highlights the model’s outperformance, with positive and statistically significant R² percentages ranging from 7.501% to 8.293% across the 1, 3, 6, and 12-month forecasting intervals. These findings highlight the importance of considering geopolitical factors in economic policy analyses and suggest that GPR can serve as an effective leading indicator for EPU. The implications of this study extend beyond the academic interest, providing practical insights for policymakers and investors seeking to mitigate risks associated with policy uncertainty. As geopolitical events continue to influence the global economic landscape, incorporating GPR into economic forecasting models could prove invaluable in strategic economic planning and decision-making processes.