FX Risk Sentinel: AI and Econometrics for Predictive Risk Analysis

Description

Our interdisciplinary team, with members from the School of Mathematics and Physics, the School of Advanced Technology, and the International Business School Suzhou, developed a robust FX risk signal detection system. By integrating technical, macroeconomic, and geopolitical risk indicators, we designed a mixed-frequency transformer-based model, deployed through both a web platform and an iOS application. These platforms generate daily risk signals, issue real-time alerts, and adjust hedging strategies to enhance returns while mitigating risk. The system has been validated by successfully capturing FX risk spikes between March 2024 and March 2025, demonstrating its reliability as a tool for crisis resilience.

Period	9 Jun 2025 → 28 Aug 2025