Reliable early warning signals are essential for timely implementation of macroeconomic and macro-prudential policies. This paper presents an early warning system as a set of multi-period forecasts of indicators of tail real and financial (systemic) risks. Forecasts are obtained from: (a) autoregressive and factor-augmented VARs with linear GARCH volatility (FAVARs), and (b) auto-regressive and factor-augmented Quantile Projections (QPs). We use a large database of monthly U.S. data for the period 1972:1-2014:12 to forecasts our tail risk indicators with each model in pseudo-real time. Our key finding is that forecasts obtained with autoregressive and FAVAR models significantly underestimate tail risks, while forecasts obtained with autoregressive and factor-augmented QPs deliver superior and fairly reliable early warning signals for tail real and financial risks up to a one-year horizon.
tail risks density forecasts factor models quantile projections