A Comparison Between The Egarch Model and Multifactor Risk Modelin Predicting VaR: The Case of The Zimbabwe Stock Exchange.

Abstract

Following the latest global financial crisis and the ongoing sovereign debt crisis, accurate measuring of market losses has become a very current issue not only in developed markets but in less developing markets like Zimbabwe. One of the most popular risk measures is Value-at-Risk (VaR); the Basel committee has encouraged the use of VaR as a measure of credit risk and market risk (thus the breakthrough of Basel 111 after the Eurozone debt crisis). In this research two predictive models have been used for estimating VaR for the ZSE. The two models are compared in terms of their ability in giving an accurateVaR estimate which does not underestimate the market risk of the Zimbabwe Stock Exchange (ZSE). The EGARCH model used the actual stock returns of the ZSE indices and the Multifactor Risk model assumed stock returns to be a function of four risk factors in predicting VaR. Applying the Kupiec test and the Basel Traffic light approach EGARCH model was found to be more accurate in estimating VaR for the ZSE as compared to the Multifactor Risk model.Due to the fact that the EGARCH model was able to capture not only volatility clustering and leptokurtosisbut it was also able to capture leverage effects in stock returns. A recommendation was made for further studies that the Multifactor Risk Approach explained in this document can be combined with the EGARCH model in VaR estimation. To allow for the inclusion of the risk factors in estimating VaR at the same time taking into consideration the volatility clustering, leptokurtosis and leverage effects in stock returns. Basing on the research findings a recommendation was also made to the investors and the stakeholders of the stock market to use a predictive model that can capture volatility clustering, leverage effects and risk factors in predicting VaR for the stock market. Lastly a VaR program was designed to answer one of the research questions using PHP; a lower level programming language that allows for many more functions than other programming language, such as login functions and graphical displays. This program was named VaR-Estimator.