Estimating the output gap in times of COVID-19

We extend a standard multivariate filter used to estimate the Output Gap (OG) in Chile to account for large economic shocks, such as those observed during the COVID-19 crisis. We propose two methodological extensions. First, we introduce exogenous supply shocks in the dynamics of potential output growth. Second, we add "system" prior distributions on a set of characteristics of the model. We show that these extensions improve several desirable properties of Chile's OG, such as stability, accuracy, consistency and also reduce the uncertainty surrounding the estimates. Next, we generalize our results by applying our methodology to a panel of countries, showing that the extended filter yields a tighter relationship between inflation developments and the OG, when compared to its standard counterpart.