Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
July 8, 2026MathematicsOpen Access

Modelling South African Macroeconomic and Financial Time Series: A Comparative Analysis of Vector Autoregressive Moving Average and Asymmetric Generalised Autoregressive Conditional Heteroskedasticity Frameworks

View Full Paper
Ask AI
Bookmark
Share

Authors

TMThatoyaone Johannes ModiseJTJohannes Tshepiso TsokuTBTshegofatso Botlhoko

Discussion

Loading...

Member takes

Overview

Compares VAR and GARCH models to forecast macroeconomic indicators in South Africa, suggesting improved prediction techniques.

Key Points

  • The aim is to examine various models for forecasting macroeconomic and financial time series in South Africa.
  • Analyzed quarterly data from 1970 to 2024 for GDP growth, exchange rates, interest rates, and household consumption expenditure.
  • Employed VAR, VARMA, GARCH, EGARCH, and GJR-GARCH models to capture mean dynamics and volatility.
  • Used AIC, BIC, HQ, and ECCM for selecting optimal model specifications, including VAR (4) and VARMA (1,1).
  • VARMA (1,1) model showed improved forecasting performance compared to VAR (4).
  • GARCH models exhibited significant persistence and clustering in volatility, especially in exchange rates and interest rates.
  • Incorporating an ARMA (1,1) term into GARCH models enhanced model adequacy, reducing residual autocorrelation and heteroskedasticity.

Cite This Study

Modise et al. (2026) studied this question.

synapsesocial.com/papers/6a4de976d2ea289ef6283a79https://doi.org/10.3390/math14132427
View Full Paper
Ask AI
Bookmark
Share