This study compares the market-specific safe-haven performance of gold, the US dollar, and Bitcoin across advanced and emerging equity markets and examines whether daily changes in geopolitical risk are associated with conditional volatility in equity markets.
Using daily data from 5th January, 2016 to 9th March, 2026, the study applies market-specific exponential generalized autoregressive conditional heteroskedasticity models with exogenous variables (EGARCH-X) and dynamic conditional correlation generalized autoregressive conditional heteroskedasticity (DCC-GARCH) models, as well as downside safe-haven regressions, across eight equity markets. The downside models evaluate asset performance under bottom-10% and bottom-5% equity-market conditions.
Standardized daily changes in geopolitical risk are statistically insignificant at the 5% level across all EGARCH-X variance specifications. The DCC-GARCH estimates reveal predominantly negative correlations between the US dollar and equity markets, correlations around zero for gold, and mostly positive correlations for Bitcoin. Under bottom-5% conditions, the US dollar provides the broadest protection, servin g as a strong safe haven in two markets and a weak one in three. Gold provides weak protection in one market, whereas Bitcoin exhibits limited safe-haven performance.
The study distinguishes time-varying diversification characteristics from actual downside protection, demonstrating that predominantly negative dynamic correlations do not necessarily translate into safe-haven effectiveness during severe equity-market declines.
