Flight to Safety in Volatility Forecasting
37 Pages Posted: 30 Jun 2018
Date Written: October 26, 2017
The benefits of using flight-to-safety (FTS) in volatility forecasting are assessed within a multivariate GARCH framework. In particular, we propose realized semi-covariance between falling equity and rising safe haven returns as a proxy of FTS and we use it to model the conditional distribution of asset returns. An empirical application based on US stock market, US bond and gold market data shows that the FTS measure is a statistically significant predictor of future equity volatility. Moreover, when comparing with a benchmark and controlling for data snooping, we find that the proposed model yields superior out-of-sample forecasts of the variance of the asset returns examined. However, the performance can be sensitive to the choice of the safe haven asset with the US Treasuries being the most effective in this regard.
Keywords: Flight to Safety, Realized Semi-Covariance, Volatility Forecasting
JEL Classification: C22, C53, C58, G17
Suggested Citation: Suggested Citation