The management of market risk is an essential determinant of the stability of a financial institution, and by extension, of the overall financial system. There are various variables which impact on the accuracy of a market risk management system. For various reasons which are discussed in this study, Value at Risk (VaR) is used as a measure of market risk. VaR has certain key features which make it adaptable to several types of scenarios in order to provide a measure of market risk. In order to assess these features of VaR, this study evaluates VaR using a range of techniques. This study analyses the performance of some of the most popular VaR models using the JSE ALSI's total daily returns. The VaR estimates were calculated for each model using varying parameters for confidence level, risk horizon, distributional assumptions and other variables. The study evaluates the relative accuracy of each model analysed, over specific subsets of the data set under consideration, and performs five different backtests to determine the accuracy of each model. The aim of this analysis is to identify the model most suited to predicting VaR in the South African environment. A key feature of this study is that it includes data during and after the financial crisis, and can, therefore, model the respective volatility characteristics of the data during this period. The results of the analysis indicate that the asymmetric GARCH models outperform the other models over both the full sample period and the crisis and post-crisis subperiods, and that the t distribution assumption produces more accurate forecasts. This implies that such models are better suited to capturing the effects of volatility for data with these characteristics.
Identifer | oai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:uct/oai:localhost:11427/18625 |
Date | January 2014 |
Creators | Chotee, Deepika |
Contributors | Toerien, Francois, Kruger, Ryan |
Publisher | University of Cape Town, Faculty of Commerce, Department of Finance and Tax |
Source Sets | South African National ETD Portal |
Language | English |
Detected Language | English |
Type | Master Thesis, Masters, MCom |
Format | application/pdf |
Page generated in 0.0169 seconds