Return and Volatility Spillover between Financial Market Participants of Dhaka Stock Exchange Using Asymmetric GARCH Methods

Abstract—This paper is an investigation of the comovement in the form of return and volatility spillover across financial market participants of Bangladesh. This study uses daily price data of commercial banks, non-bank financial institutions (NBFI), and insurance companies traded in the Dhaka Stock Exchange (DSE) for the period spanning 2009 to 2016. Bayesian Vector Autoregressive (VAR) model has been used in the conditional mean equations of EGARCH and GJR-GARCH models have been used to test the return spillover effects whereas lagged squared residuals and lagged conditional variances have been used as variance regressors in conditional variance equations to test the spillover effects of historical volatility and innovations transmitting in the form of shock to other participants operating in the same market. Bayesian VAR output reveals a highly significant bi-directional return spillover between bank-insurance pair and also between NBFIs-insurance pair. However, return spillover between commercial banks and NBFIs is unidirectional; only bank returns are affecting returns from NBFIs. Conditional volatility of NBFIs exhibit a highly significant asymmetric effect implying that bad news increases volatility of NBFIs to a greater degree than good news. Both GJR-GARCH and EGARCH output reveal bidirectional volatility spillover in the form of historical volatility and innovations among commercial banks, NBFIs and insurance companies.