VaR Prediction for Financial Asset Series Based on Stable Distribution

This paper performs a Value at Risk (VaR) prediction and analysis of the daily log-returns of the French CAC40 index utilizing the Stable distribution. To validate the effectiveness of the proposed method, we utilized daily closing data from January 1, 2000, to June 30, 2022, and converted these data into log returns. Through rolling window analysis, we evaluated the out-of-sample predictive performance of the Stable distribution VaR estimation method and compared it with the traditional RiskMetrics method. The experimental findings show that the Stable method provides more accurate and stable VaR estimates at the 5% confidence level, while the RiskMetrics method performs poorly under extreme market conditions. Additionally, descriptive statistical analysis reveals that the log returns of the CAC40 index exhibit significant heavy-tail characteristics and autocorrelation, further confirming the volatility clustering in financial markets. The findings of this study offer valuable insights for financial risk management and investment decision-making.