Optimal multi-period dynamic portfolios for sustainable investing

Sustainable investing involves balancing traditional investment strategies with environmental, social, and governance-related (ESG) objectives to improve long-term outcomes. We develop a stochastic control framework for constructing optimal multi-period dynamic ESG portfolios, in which investors form dynamic portfolios by maximizing utility functions that account for both profitability and the average ESG scores of their portfolios in the near and distant future. We provide an explicit solution to the investors' utility maximization problem, and show that the optimal weight is a linear combination of three components: a financial performance component, an ESG component, and a component reflecting the weight from the previous period. Using the U.S. stock market data, we find that the multi-period dynamic ESG portfolio outperforms single-period portfolios in terms of Sharpe ratios, alphas, and ESG scores. Our article extends the dynamic portfolio theory in the context of sustainable investing, and provides guidance to practitioners on tailoring ESG portfolios to their specific preferences and to market conditions related to ESG.