Original Research
Value-at-risk for the USD/ZAR exchange rate: The Variance-Gamma model
South African Journal of Economic and Management Sciences | Vol 18, No 4 | a966 |
DOI: https://doi.org/10.4102/sajems.v18i4.966
| © 2015 Lionel Establet Kemda, Chun-Kai Huang, Knowledge Chinhamu
| This work is licensed under CC Attribution 4.0
Submitted: 17 March 2014 | Published: 27 November 2015
Submitted: 17 March 2014 | Published: 27 November 2015
About the author(s)
Lionel Establet Kemda, University of KwaZulu-Natal, South AfricaChun-Kai Huang, University of Cape Town, South Africa
Knowledge Chinhamu, University of KwaZulu-Natal, South Africa
Full Text:
PDF (4MB)Abstract
A country’s level of exchange risk is closely linked to its financial stability, on a macro-economic scale. South African exchange rates, in particular, have a significant impact on imports, inflation, consumer prices and monetary policies. Consequently, it is imperative for economists and investors to assess accurately the associated exchange risks. Exchange rates, like most financial time series, are leptokurtic and contradict the classical Gaussian assumption. We therefore introduce subclasses of the generalised hyperbolic distribution as alternative models and contrast these with the normal distribution. We conclude that the variance-gamma model is the most robust for describing the log-returns of daily USD/ZAR exchange rates and their related Value-at-Risk (VaR) estimates. The model selection methodologies utilised in our analyses include the robust Kolmogorov-Smirnov test and the Akaike information criterion. Backtesting on the adequacy of VaR estimates is also performed using the Kupiec likelihood ratio test.
Keywords
No related keywords in the metadata.
Metrics
Total abstract views: 3432Total article views: 1824
Crossref Citations
1. A new lossy compression algorithm for wireless sensor networks using Bayesian predictive coding
Chen Chen, Limao Zhang, Robert Lee Kong Tiong
Wireless Networks vol: 26 issue: 8 first page: 5981 year: 2020
doi: 10.1007/s11276-020-02425-w