Measuring Returns And Contagion Of Investor Behavior In Iraq Stock Exchange Using Multivariate GARCH Model

Abstract

This paper aims to use multivariate GARCH models (MVGARCH), including the dynamic conditional correlation (DCC-GARCH) model to reveal the nature of the correlations between the returns of Iraqi financial market, and then test the extent of transmission of fluctuations between the financial market sectors. The DCC-GARCH developed by Engle (2002) is used to detect changes occurred in the conditional correlations during the time. The model assumes that the statistical distribution of the time series is the normal with mean equal to zero and conditional variance.  The results of the statistical analysis showed that the correlation among banking, services, and industrial sectors is dynamic.  This result confirms that the occurrence of previous shocks of the banking sector index will lead to fluctuations in indices of services and industrial sectors of Iraqi Stock Exchange.  The results of the Jaque-Bera test confirm that the differences between all sectors are significant because P-value is less than 0.05.