Dependence Modelling using GARCH, EGARCH, and Copula Models:

Copula become a popular tool to measure the dependency between financial data due to its ability to capture the non-normal distributions. Hence, this paper will inspect the impact of input models towards the parameter estimation of marginal and copula models for KLCI and FBMHS returns series by considering the ARMA-GARCH model and the ARMA-EGARCH model. This study also investigates the dependency of Islamic-conventional pair for Malaysia indices by using static copula and time-varying copula approach. The closing prices of Malaysia indices represented by KLCI (conventional) index and FBMHS (Islamic) index for the period of 21 May 2007 until 28 September 2018 are used as a sample data. The results show that KLCI-FBMHS pair is strongly correlated, different input models (ARMA-GARCH and ARMA-EGARCH) have identical dependence structure but slightly different value of parameter estimated, and the time-varying Gaussian copula is chosen as the best dependence model. Finding suggest that the diversification between Islamic-conventional pair is worthwhile during stable period.