Abstract
<jats:p>This study investigates how geopolitical risk shocks affect the Turkish stock market using monthly data for 2010:01–2025:11. Geopolitical risk is proxied by the news-based Geopolitical Risk Index (GPR), and BIST-100 monthly returns measure market performance. For robustness, we estimate four alternative models with gradually expanded control sets capturing global risk appetite (VIX), exchange-rate shocks (USD/TRY), commodity shocks (Brent oil and gold), and inflation changes (CPI). The empirical strategy combines dynamic regressions with HAC/Newey–West inference, Jordà (2005) local projections to trace horizon-specific responses, and a VAR(p=1) framework as a consistency check. Dynamic regressions indicate that contemporaneous GPR changes are negative but statistically insignificant, whereas VIX shocks have a large, highly significant negative impact on returns. Local projections show that the adverse effect of geopolitical risk increases with horizon length and becomes significantly negative at longer horizons, especially at 12 months. VAR-based impulse responses are imprecise and sensitive to orthogonalisation and are treated as complementary evidence. Overall, geopolitical risk matters mainly over longer horizons, whereas global risk appetite dominates short-run return dynamics</jats:p>