Research on Identifying Key Indicators for Financial Risk Warning in Electric Power Enterprises Based on XGBoost Model

Power enterprises operate in a complex environment characterized by multifaceted financial risks. This study constructs a financial risk early warning model for power enterprises, utilizing 33 financial and non-financial indicators and employing the XGBoost algorithm. The results demonstrate that the XGBoost model effectively predicts financial risks in power enterprises with high accuracy. Solvency indicators are identified as primary drivers of financial risk. Additionally, profitability metrics, operational efficiency indicators, and the Tobin's Q ratio significantly influence the financial risk profile of power enterprises.