Title: Macroeconomic forecasting during disaster recovery
Authors: Jeffrey Campbell - University of Notre Dame (United States) [presenting]
Abstract: An empirical model of ongoing disaster recovery is provided to augment a preexisting Box-Jenkins forecasting model. The model sums the original series with possibly random disaster shocks, each of which follows a first-order autoregression. Application of the Kalman filter yields forecasts of the disaster recovery's expected duration and estimates of the counterfactual time series without the disaster. Applications to Japanese IP following the Tohoku Earthquake, Israeli IP following the Yom Kippur War and the U.S. and European macroeconomic time series since March 2020 illustrate the model's usefulness.