What question did this study set out to answer?

The research aims to develop a hybrid forecasting framework for predicting stock indices by incorporating latent market states.

June 17, 2026ForecastingOpen Access

Regime-Aware Stock Index Forecasting Under Latent Market States: A Hybrid Statistical Learning Framework with Cross-Market Validation

View Full Paper

Ask AI

Bookmark

Authors

Hybrid framework predicts stock indices using Kalman filtering and recurrent learning, indicating strong regime awareness implications.

The research aims to develop a hybrid forecasting framework for predicting stock indices by incorporating latent market states.
Utilized Kalman Filtering (KF) for noise smoothing and Markov Switching (MS) to identify regimes.
Incorporated LSTM/GRU components for learning nonlinear temporal patterns.
Evaluated the framework with CSI 300, S&P 500, and Nikkei 225 indices through various accuracy measures and trading simulations.
Regime conditioning significantly improves forecasting performance and economic outcomes.
KF–MS–LSTM showed optimal results for the CSI 300; Standard MS excelled for the S&P 500; both KF–MS–LSTM and KF–MS–GRU were competitive for the Nikkei 225.
Models lacking regime information, like pure LSTM/GRU and standalone Transformers, demonstrated inferior forecasting capabilities.

Tian et al. (2026) studied this question.

View Full Paper

Ask AI

Bookmark