Forecasting Output Growth and Inflation: The Role of the Great Recession

Abstract

This paper examines the local forecasting performance of financial and macroeconomic variables in predicting output growth and inflation relative to autoregressive models from 1959 to 2020, with a focus on changes in the relative forecasting performance triggered by the onset of the Great Recession. Key findings highlight multiple shifts in predictability among commonly used economic models, like the composite leading indicator, and in variables recently attracting attention, like consumer expectations, in forecasting output growth; those models exhibit superior predictive performance during the Great Inflation and Great Recession, but not during the Great Moderation. Financial variables emerge as important in forecasting during the Great Recession. Additionally, labor market variables became relevant in predicting both real economic activity and inflation during the Great Recession period, aligning with the increased attention that the labor market has recently attracted from macroeconomists and policymakers. Lastly, oil prices also exhibit pockets of predictability for forecasting inflation during the Great Recession. Our study contributes to a comprehensive understanding of forecasting dynamics of economic models, and the shifts introduced through the Great Recession experience.

Keywords: Great Recession, Pockets of predictability, Output growth forecasts, Inflation forecasts, Local forecast evaluation

JEL Classification: E37, C22, C52, C53