Testing Hybrid Forecasts for Imports and Exports

Abstract

This article discusses the use of hybrid forecasting, which combines model-based forecasts with judgmental adjustment, to improve import and export forecasts during exceptional times like the COVID-19 pandemic. The study finds that hybrid models performed better than pure model-based forecasts during the pandemic's lowest point but did not significantly improve forecasts during the recovery or over longer forecasting periods. The article emphasizes the importance of forecasting imports and exports, as they contribute significantly to GDP and provide valuable information to policymakers. It also examines the pros and cons of model-based, judgmental, and hybrid forecasting methods. Additionally, the study compares the performance of private-sector forecasts and pure model-based forecasts during the pandemic, finding that private-sector forecasts outperformed pure model-based forecasts for short-term forecasts during the peak of the pandemic but performed similarly outside of that period. The results suggest that judgmental adjustment may be more useful for short-term forecasts during uncertain times but may not significantly improve forecasts during normal times or longer horizons. The study also highlights the importance of continuously monitoring multiple models and using model averaging for forecasting during extreme events. [Extracted from the article]