Incorporating Market Regimes into Large-Scale Stock Portfolios: A Hidden Markov Model Approach

We propose a portfolio construction method that accounts for the regime-dependent behavior of stocks, thereby impacting their expected returns. Using a hidden Markov model (HMM) and a regime-weighted least-squares approach, we estimate forward-looking regime-conditional factors. These factors help build large-scale stock portfolios for systematic investment management, considering financial market regimes. In historical simulations, our framework achieves superior risk-adjusted performance compared to passive portfolios in both relative and absolute management settings.