Estimating Value at Risk for Sukuk Market Using Generalized Auto Regressive Conditional Heteroskedasticity Models

Abstract

In this paper, we compare the forecasting ability of different GARCH models to estimate value at risk in sukuk market. A wide extensive list of both symmetric and asymmetric GARCH models (including GARCH, EGARCH, GJR-GARCH, IGARCH and Asymmetric power GARCH) were considered in modeling the volatility in the sukuk market. All VaR estimations are carried out by " rugarch " package in " R " software. The performance of these models is compared by both in-sample and out-of-sample analysis. We found that the performance of asymmetric models in estimating value at risk are superior in both in-sample and out-of-sample evaluation. We also found that in most cases the student-t distribution is more preferable than normal or generalized error distribution (GED). This study is one of very few studies which have investigated the behavior of sukuk data in secondary market and describes characteristics of its statistical distribution function. Another contribution is comparing the estimation ability of various GARCH models in order to find superior model to estimate VAR in sukuk market.