Abstract
This study investigates tail-risk spillovers and portfolio resilience in the South African equity market using a quantile factor vector autoregression (QFVAR) framework. Focusing on the Johannesburg Stock Exchange (JSE) Top 20 equities from January 2019 to September 2025, the research examines how systemic interdependencies among major firms evolve across market regimes such as bearish, normal, or bullish, and how these dynamics influence optimal portfolio construction. By combining QFVAR estimation with Generalised Forecast Error Variance Decomposition (GFEVD), the study quantifies directional and total connectedness to identify key transmitters and receivers of systemic risk. The resulting spillover matrices are analysed through network topology eigenvector centrality, which reveals the structural hubs of contagion within the JSE. Furthermore, the study integrates these insights into portfolio optimisation frameworks, developing spillover-aware and network-augmented strategies that penalise assets with high systemic risk exposure. Empirical results show that connectedness is strongly state-dependent, intensifying in both the lower and upper quantiles, while financial and technology stocks emerge as the dominant transmitters of tail-risk. Incorporating network-informed spillover measures into portfolio design significantly improves risk-adjusted performance, enhancing resilience during extreme market conditions. The findings contribute to the literature on systemic risk and quantile connectedness in emerging markets, providing actionable insights for investors and regulators seeking to manage contagion and enhance financial stability in South Africa's equity market.