Navigating Market Distress: Dynamic Portfolio Management Using Hidden Markov Models

Abstract

We develop a regime-switching model combining the approaches of Moreira and Muir and Daniel, Jagannathan, and Kim to construct regime-aware portfolios. Our model predicts and bypasses regimes of market distress, characterized by high volatility, enhancing riskadjusted returns for mean-variance investors. We document significant performance improvements across mean-variance efficient factor portfolios, including the momentum, market, size, value, profitability, investment, return on equity, and betting-against-beta factors. Specifically, our strategy reduces investment weights during high volatility and market downturns, significantly enhancing performance metrics such as Sharpe ratio, appraisal ratio, and utility for investors, the latter two by 0.87 and 66% respectively. This improvement is attributed to regime-switching models’ ability to capture market rebounds and mitigate downturns, outperforming traditional strategies.