The relationship between global risk aversion and returns from safe-haven assets

Abstract

We investigate the relationship between global risk aversion and safe-haven assets using the causality-in-quantiles test and the quantile-on-quantile regression method. Our empirical results show the predictability of global risk aversion on the returns of safe-haven assets. Furthermore, we find that several assets have consistent safe haven attributes regardless of the level of global risk aversion, while gold and Bitcoin cannot be considered consistent safe havens. Based on these findings, non-cash flow-induced shocks are not only an important predictor of asset returns but also their relevance cuts across general financial markets.