Dynamic Weighting Methods in Portfolio Construction: A Hidden Markov Model Approach

Abstract

This paper explores the application of dynamic portfolio weighting strategies through the integration of Hidden Markov Models with risk parity and equal weighting techniques. Using the concept of market regimes, this study identifies distinct economic conditions based on inflation and growth metrics and develops an investment strategy that adjusts asset allocation conditioned on the macroeconomic climate. The paper contributes to the existing literature by creating a hybrid model that navigates through economic cycles with the goal of improving precision in asset allocation, portfolio diversification, and portfolio resilience. The results underline the Hidden Markov Model’s capability to identify market regime shifts based on macroeconomic data and yields positive outcomes in the application of estimated state probabilities as a weighting scheme in a dynamic regime-based investment strategy. The integration of a Hidden Markov Model in the Risk Parity framework, creating the papers hybrid model, improved the portfolio performance, resulted in a higher risk adjusted return and significantly lower drawdowns during the outof-sample period, compared to the standalone risk-parity portfolios and other portfolios used in the study.