What type of study is this?

Linear and nonlinear econometric models against machine learning models: realized volatility prediction

View Full Paper

Ask AI

Bookmark

Authors

This analysis compares realized volatility predictions from ML and econometric models, indicating regime-switching models excel in performance.

Regime-switching models like THAR and STHAR outperform ML and linear models in predicting realized volatility.
Using rolling forecasts since 2006, regime-switching models provide more accurate risk forecasts and optimal utility.
ML models show potential in capturing nonlinear patterns but lack an advantage over simpler econometric models.
Highlighting the importance of regime changes in risk management emphasizes the need for interpretable models.

Rehim Kılıç (2025) studied this question.

View Full Paper

Ask AI

Bookmark

This analysis compares realized volatility predictions from ML and econometric models, indicating regime-switching models excel in performance.