Volatility dynamics and the risk-return relationship in South Africa: A GARCH approach

  • Received January 6, 2021;
    Accepted April 5, 2021;
    Published May 7, 2021
  • Author(s)
  • DOI
    http://dx.doi.org/10.21511/imfi.18(2).2021.09
  • Article Info
    Volume 18 2021, Issue #2, pp. 106-117
  • TO CITE АНОТАЦІЯ
  • Cited by
    3 articles
  • 903 Views
  • 336 Downloads

Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 International License

This study is aimed at investigating the volatility dynamics and the risk-return relationship in the South African market, analyzing the FTSE/JSE All Share Index returns for an updated sample period of 2009–2019. The study employed several GARCH type models with different probability distributions governing the model’s innovations. Results have revealed strong persistent levels of volatility and a positive risk-return relationship in the South African market. Given the elaborate use of the GARCH approach of risk estimation in the existing finance literature, this study highlighted several weaknesses of the model. A noteworthy property of the GARCH approach was that the innovation distributions did not affect parameter estimation. Analyzing the GARCH type models, this theory was supported by the majority of the GARCH test results with respect to the volatility dynamics. On the contrary, it was strongly unsupported by the risk-return relationship. More specifically, it was found that while the innovations of the EGARCH (1, 1) model could account for the volatile nature of financial data, asymmetry remained uncaptured. As a result, misestimating of risks occurred, which could lead to inaccurate results. This study highlighted the significance of the innovation distribution of choice and recommended the exploration of different nonnormal innovation distributions to aid with capturing the asymmetry.

view full abstract hide full abstract
    • Table 1. Descriptive statistics of the ALSI returns
    • Table 2. Heteroscedasticity tests for the ALSI returns
    • Table 3. ML parameter estimates of the GARCH (1, 1) for different innovation distributions
    • Table 4. ML parameter estimates of GJR-GARCH for different innovation distributions
    • Table 5. ML parameter estimates of EGARCH with different innovation distributions
    • Table 6. ML parameter estimates of APARCH with different innovation distributions
    • Table 7. Preliminary test results for the innovations of the EGARCH
    • Table 8. ML parameter estimates for GARCH-M with different innovation distributions
    • Table 9. ML parameter estimates for EGARCH-M with different innovation distributions
    • Conceptualization
      Nitesha Dwarika, Peter Moores-Pitt
    • Data curation
      Nitesha Dwarika, Retius Chifurira
    • Formal Analysis
      Nitesha Dwarika, Peter Moores-Pitt
    • Investigation
      Nitesha Dwarika
    • Methodology
      Nitesha Dwarika, Retius Chifurira
    • Software
      Nitesha Dwarika
    • Validation
      Nitesha Dwarika, Peter Moores-Pitt, Retius Chifurira
    • Visualization
      Nitesha Dwarika, Peter Moores-Pitt, Retius Chifurira
    • Writing – original draft
      Nitesha Dwarika
    • Writing – review & editing
      Nitesha Dwarika, Peter Moores-Pitt, Retius Chifurira
    • Project administration
      Peter Moores-Pitt
    • Supervision
      Peter Moores-Pitt, Retius Chifurira