Return to search

An empirical assessment of the key drivers of sovereign bond yields in South Africa: it’s not just about fundamentals

Thesis (M.Com. (Business Finance))--University of the Witwatersrand, Faculty of Commerce, Law and Management, School of Economic and Business Sciences, 2017 / The writer studies the short-run determinants of bond yield volatility in South Africa (SA) by
analyzing the impact that global factors –representing global funding conditions – have on the
changes to the rand denominated generic 10-year government bond yield (SAGB). This is
followed by a one-period forward forecast of this volatility. The explanatory variables tested
in this study are as follows: net bond purchases by foreign investors, Chicago Board Options
Volatility Index (VIX), JP Morgan Emerging Market Bond Index (JP EMBI) spread, the US
dollar to SA rand (USDZAR) exchange rate, the SA 5 year credit default swap (CDS) rate,
the 12 month interest rate expectation/9x12 forward rate agreement (FRA), dollar spot price
of gold and dollar spot price of oil. The study period ranges from January 2000 to December
2015. The GARCH modelling technique is used due to its ability to capture the volatility
clustering effects observed in time series return data. The writer used the Gaussian
distribution as the default model, however in order to control for the skewness and fat-tails in
financial market return data, the Student-T and Generalised Error distributions are also tested
to see if the non-normally distributed bond returns could be better captured by alternative
parametric assumptions. The results show that all the explanatory variables, with the
exception of the FRA, are statistically significant in explaining volatility in the local generic
10-year government bond. / GR2018

Identiferoai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:wits/oai:wiredspace.wits.ac.za:10539/24702
Date January 2017
CreatorsMpakama, Sinovuyo Lusanda
Source SetsSouth African National ETD Portal
LanguageEnglish
Detected LanguageEnglish
TypeThesis
FormatOnline resource (80 leaves), application/pdf

Page generated in 0.0024 seconds