The Dynamic Conditional Correlation and Volatility Linkages between Green and Conventional Bonds: Empirical Evidence on a Global level.

Abstract

This study aims at examining the dynamic conditional correlations and the volatility linkages between the green bonds and the conventional bonds market on a global level. The paper chooses the Bloomberg Barclays MSCI Global Green Bond Index (GB) and the Bloomberg Barclays Global Aggregate Total Return Index (CB) to represent the green and conventional bonds markets on a global level, respectively. The paper gathers their weekly data over a period of six years from 17th October 2014 to 18th September 2020 from Bloomberg. It adopts Engle (2002) two-steps DCC multivariate GARCH model to carry out the analysis. In the first step, this paper finds the best fitting univariate GARCH model is ARMA (8,8)-GARCH (1,2) and finds evidence that GB is more sensitive and has higher reaction to market events than CB does. In addition, GB exhibit less persistency in its conditional volatility than CB does. In the second step, using the DCC-MGARCH (1,2), this paper finds short-term volatility spillover between GB and CB but the persistency of a shock in both markets relative to the other is low and fades away quickly. This paper concludes that a time-varying, positive, and strong conditional correlation exists between GB and CB. Also, it finds evidence of strong positive volatility linkages between GB and CB. Lastly, the paper identifies a structural break in March 2020 caused by the COVID-19 pandemic. The implications of this paper are important to investors, portfolio managers, and policymakers as they aid in making educated decisions related to portfolio diversification. Based on the results, this paper does not find evidence of gaining diversification benefits and, hence, does not recommend placing both types of bonds in the same portfolio.