Modeling and forecasting volatility of returns on the Jordanian stock index using garch models

Abstract EN

The purpose of this study is to examine and estimate the three Generalized Autoregressive Conditional Heteroskedasticity (GARCH) Models: one symmetric (GARCH) and two asymmetric types of GARCH models Exponential and Glosten, Jagannanthan, and Runkle (EGARCH and GJR).

This study were executed using the daily price data from the Jordanian Stock Index over a 16-year period beginning on January 4 1992 and ending on December 30, 2007.

The models used three different error distributions, Gaussian Normal, Student-t, General.

The results of estimations and forecasting performance revealed that the asymmetric modes are better than symmetric model, especially when fat-tailed was used.

The results also, revealed that the GJR model provide the best out-of sample forecast for the Jordanian Stock Index when the specified (16 years) applied for the next year.