This analysis compares forecasting accuracy using AR-GARCH, LSTM, and various machine learning models, suggesting hybrid models improve precision in financial markets.
Key Points
The hybrid model achieves superior accuracy in forecasting compared to traditional methods.
Root Mean Square Error metrics reveal that machine learning models like LSTM and CNN enhance predictions.
Assessment across indices including the S&P 500 highlights diverse strengths of individual models.
Integrating classical econometrics with machine learning may strengthen robustness in financial analysis.