A0153
Title: Scenarios for macroeconomic risk
Authors: Esther Ruiz - Universidad Carlos III de Madrid (Spain) [presenting]
Abstract: The interest in developing econometric tools to estimate densities of key macroeconomic variables is clear. Point forecasts are insufficient for informed decision-making. Densities measure macroeconomic vulnerability, crucial for resilience policies. Recently, econometricians and policy-makers have been interested in constructing realistic scenarios that can help understand economic resilience by providing early warnings of adverse, low-probability but potentially catastrophic events. The aim is to propose estimating conditional densities under stressed-factor scenarios using factor-augmented quantile regressions (FA-QR) with factors extracted from dynamic factor models (DFMs) via principal components (PCs). Severe yet plausible stress scenarios are based on the joint distribution of these factors. The results are illustrated by calculating growth-in-stress (GiS) for US growth and the 5\% quantile of stressed growth densities, and show GiS as a useful complement to growth-at-risk (GaR) for scenario analysis. The COVID-19 shock provides a natural environment to assess US growth vulnerability. Scenarios for US inflation are also obtained when domestic, international, and temperature factors are stressed, showing that, depending on the estimation method of factors' uncertainty, worst inflation-at-risk (IaR) scenarios can differ, with important implications for monetary policy design.
