Risk Monitoring Systems in Real-time Based on Dynamic Factor Models

This work aims at prediction, prevention and mitigation of systemic risks providing an Early Warning System (EWS) based on forecasts of systemic real and financial risks. These risks are estimated using density forecasts of a set of unrestricted factor-augmented VARs, with factors extracted from a large number of macroeconomic series employed as predictors. Its forecasting power is assessed using quarterly time series of the G-7 economies in order to compare the major European economies with Japan, Canada and the United States.

Systemic financial risk is not new to researchers and policy-makers, who are still committed to the stability of financial systems1. More than 15 years ago, Group of Ten concluded that “[risk] interdependencies between large and complex banking organizations have increased over the last decade in the United States and Japan, and are beginning to do so in Europe”. In the early 2000s, several other studies documented the increased potential for systemic financial risk realizations in several advanced economies, consistent with the conclusion of the G-10 study.