AUTHOR=Qiu Zihao , Chen Siyu , Feng Zixin , Luo Ruitong , Wang Zhiwei 
  
TITLE=Machine learning-based dynamic risk measurement for white sugar futures under geopolitical risks
  
JOURNAL=Frontiers in Physics
  
VOLUME=Volume 13 - 2025
  
YEAR=2025
  
URL=https://www.frontiersin.org/journals/physics/articles/10.3389/fphy.2025.1674717
  
DOI=10.3389/fphy.2025.1674717
  
ISSN=2296-424X
  
ABSTRACT=Futures, as significant financial derivatives, play a crucial role in financial markets by fulfilling price discovery functions and providing efficient risk hedging tools. Against the backdrop of geopolitical conflicts, market risk emerges not only from external shocks and random fluctuations but also from strategic interactions among diverse participants including hedgers, speculators, arbitrageurs, and regulators. This study integrates traditional VaR theory with machine learning methods to systematically examine risk characteristics and transmission mechanisms in the sugar futures market under geopolitical uncertainty. Utilizing sugar No. 5 futures trading data from the Zhengzhou Futures Exchange spanning 2015–2019 and 2024, we employ a Random Forest model for feature importance analysis and compare three risk measurement approaches: traditional parametric VaR, historical simulation methods, and machine learning-enhanced VaR models. We conduct empirical tests to validate the theoretical relationship √3 × VaRT(1,p) ≈ VaRT(3,p) and calculate epsilon values (relative deviation between actual and estimated tail risk occurrences) through return tests. Annual delta values range between 0.26 and 1.16, averaging approximately 35% below theoretical values. The machine learning-based Value at Risk (VaR) at 95% confidence level exhibits a violation rate of 5.00%, demonstrating superior accuracy compared to parametric VaR (26.67%) and traditional historical VaR (7.00%). Epsilon values show no statistically significant difference between 2024 (0.08) and the 2015–2019 average level (0.14), indicating stable risk transmission mechanisms despite geopolitical conflicts. The hybrid “machine learning-traditional theory” risk framework developed in this research provides a theoretical foundation and practical guidance for regulatory bodies to enhance risk prevention and control systems, as well as for market participants to optimize risk management strategies. Despite geopolitical impacts, the fundamental risk transmission mechanisms of the sugar futures market remain relatively stable, demonstrating market resilience.