Backtesting Global Growth-at-Risk
69 Pages Posted: 10 Oct 2019 Last revised: 2 Sep 2020
Date Written: September 29, 2019
We conduct an out-of-sample backtesting exercise of multivariate Growth-at-Risk (GaR) predictions for 24 OECD countries. We consider forecasts constructed from quantile regression (QR) and GARCH models. We find evidence of predictability up to one year ahead, and that GARCH outperforms QR. In particular, our results show that standard volatility models such as GARCH produce more accurate GaR forecasts than QR models based on predictors of downside risks to GDP.
Keywords: Growth-at-Risk, Backtesting, Quantile Regression, GARCH
JEL Classification: C22, C23, C52, C53, C58
Suggested Citation: Suggested Citation