What is Driving Connectedness? Stylized Facts from Mean and Volatility Dynamics

Abstract

We investigate connectedness within a network of environmental indices and crude oil by utilizing state-of-the-art empirical methods. Focusing on the relevance of both mean and volatility dynamics, we find that mean dynamics persistently account for less than 5% of overall dynamics implying that, volatility is the main driver of developments in the network. In turn, we propose a novel mixed bivariate DCC-GARCH connectedness model whereby dynamics adjust faster and more accurately compared to all presented alternatives.

Keywords: Dynamic connectedness, VAR, DCC-GARCH

JEL Classification: C50, F65, G15