An Early Warning System for Systemic Risks
51 Pages Posted: 7 Mar 2018
Date Written: March 1, 2018
This paper formulates an Early Warning System (EWS) for systemic risks based on forecasts of Expected Shortfalls (ES) of real and financial indicators integrated with structural stress-tests via a structural VAR. The EWS delivers early warning signals as probabilities of tail risk realizations, as well as measures of the incidence of tail risks as ES forecasts conditional on configurations of structural shocks. Using monthly data of the G-7 economies for the period 1984:01-2016:12, the EWS is shown to have significant out-of-sample forecasting power, with its usefulness as an early warning system demonstrated through a pseudo-real-time application.
Keywords: Systemic Risks, VAR, Quantile Auto-regressions, Quantile curves
JEL Classification: C5, E3, G2
Suggested Citation: Suggested Citation