المستخلص: |
The purpose of this study is to find the GARCH specification and innovations distribution combination which best models the returns volatility of four major Islamic equity indices DJIM, S&P500 SH, FTSE SWORLD.IS and MSCI ISWD. The conditionally heteroscedastic autoregressive models considered are GARCH, EGARCH, AGARCH, NARCH, NGARCH, GJR GARCH, APARCH and NGARCH whereas the distributions considered are the normal, student, cauchy, laplace, logistics and EVD distributions. The study of the statistical properties of the different return series confirms that GARCH models are the most suitable for modeling purposes. The results of the estimations suggest that the combinations offering the best volatility modeling are: NGARCH-Laplace for the DJIM, APGARCH-Laplace for the S&P500 SH, GJR GARCH-Logistics for the SWORLD.IS and GJR GARCH-Student for the MSCI ISWD.
|