Modeling of stock indices using asymmetric GARCH models

Abstract

This study discusses GARCH modeling with a special focus on the fitting of GARCH models to financial return series. Most popular asymmetric GARCH models are considered by comparing the modeling performance of different conditional variance models. The data consist of daily closing levels of indices for the Nairobi Stock Exchange(NSE) running through 1996 to 2003 with reference to the equity of Uchumi Supermarket. The results suggest that improvement of the overall estimation are achieved when asymmetric GARCH are used and when fat-tailed densities are taken into account in the conditional variance.