Abstract This study investigates the impact of geopolitical conflict, specifically the Iran–Israel tension, on the volatility and extreme risk of stock markets in six indirectly involved countries: China, Russia, Turkey, the United States, the United Kingdom, and Germany. Using daily log returns from June 2023 to July 2025, the analysis begins with modelling the conditional mean using ARMA, followed by volatility modelling via the MS-GARCH(1,1) framework to capture regime changes between calm and turbulent market conditions. The presence of heavy tails and volatility clustering justifies the application of the Extreme Value Theory (EVT) using the Peaks Over Threshold (POT) method to estimate tail-related risks. The results reveal that stock markets in Iran-aligned countries (China, Russia, Turkey) generally remained in the calm regime, while Israel-aligned countries (USA, UK, Germany) exhibited more frequent transitions to high-volatility regimes. EVT parameters confirm the existence of substantial tail risks, especially in high quantiles. Value-at-Risk (VaR) estimation using the MS-GARCH–EVT(POT) model shows strong performance in capturing extreme losses. Backtesting using Kupiec’s tests validates the model’s accuracy in both frequency and timing of violations, with only minor exceptions. Overall, the combined MS-GARCH–EVT(POT) model proves to be an effective and statistically robust approach for quantifying extreme market risks during geopolitical uncertainty. These findings offer valuable insights for risk managers and policymakers in navigating financial markets under external shock scenarios.
Darmanto et al. (Sun,) studied this question.