Volatility and jumps in the Chinese Yuan using Gumbel distribution during the trade war and COVID-19 pandemic

This study investigates the dynamic structure of intraday volatility and jump behavior in the Chinese yuan and US dollar exchange market by employing a Gumbel-distribution-based threshold derived from extreme value theory. Using five-minute high-frequency data over 2,450 trading days, we construct a daily jump probability index and examine its responsiveness to major economic shocks, including the U.S.–China trade war and the COVID-19 pandemic. Regression and simulation analyses (5,000 replications) show that the Gumbel-based jump metric provides superior explanatory power in detecting large and irregular jumps. The Gumbel distribution offers a clear and theoretically grounded thresholding mechanism, making it particularly effective in identifying episodic volatility clusters during the 2010s. Tests by Lee and Hannig (2010) and Laurent and Shi (2020) also suggest that the Gumbel jump statistic is more appropriate for capturing the frequent and discontinuous jumps in foreign exchange volatility. Furthermore, after filtering for intraday periodicity, the estimated jump probabilities significantly decline, indicating the importance of periodicity adjustment. This study also confirms that jump probabilities were notably higher during the U.S.–China trade war and the COVID-19 pandemic than in the overall sample period. Sensitivity tests on volatility filtering and simulation parameters further demonstrate the robustness of the Gumbel-based jump distribution.