When safe havens fail: regime-dependent spillover connectedness among African stock markets, precious metals, and cryptocurrencies

Abstract

Abstract This study examines the dynamic and quantile-dependent spillover connectedness among African stock markets, precious metals, and cryptocurrencies. Using daily data for seven major African stock markets, three precious metals, and three cryptocurrencies from July 2007 to July 2023, we employ a combined Time-Varying Parameter Vector Autoregression (TVP-VAR) and Quantile Vector Autoregression (QVAR) framework to capture both time-varying and tail-dependent spillovers. The findings reveal pronounced regime dependence, with connectedness intensifying substantially during bearish and bullish market states relative to normal conditions. Cryptocurrencies emerge as increasingly important transmitters of shocks, while African equity markets predominantly act as net shock receivers. The evidence shows that the traditional safe-haven role of precious metals is regime-dependent rather than universal, with these assets frequently transmitting, rather than absorbing, shocks during periods of market stress. Extensive robustness analyse, including wavelet coherence, minimum-spanning-tree networks, quantile-on-quantile estimation, chronological sub-sample analysis, and asymmetric volatility connectedness confirm that cross-market linkages are nonlinear, state-dependent, and structurally strengthened following 2020. The findings contribute to the literature on emerging market financial integration and provide implications for portfolio diversification, and digital asset regulation.