Revealing the Impacts of Rare Events on Financial Markets Using XAI

Rare events such as economic shocks or policy shifts can cause sudden volatility and significant losses in financial markets. Traditional risk models often fail to capture the complexity of these disruptions. This study explores the use of Explainable Artificial Intelligence (XAI) to improve the transparency and reliability of financial forecasting during such events. Developed an XAI-based framework using historical stock data, financial indicators, and an XGBoost regression model, enhanced with SHAP for interpretability. By applying this system to real-world events such as the 2019 corporate tax cut in India and analyzing market reactions across multiple sectors. The results showed that while the XAI models maintained strong predictive performance, they also offered clear insights into the influence of factors such as RSI and SMA on price behavior. Findings demonstrate that integrating XAI enhances both trust and accuracy in financial decision-making, offering valuable support for investors and regulators during periods of market stress.