Toward green finance: applying Bayesian machine learning in environmental portfolio management

Abstract In recent years, the importance of climate change, environmental sustainability, and climate finance has witnessed a significant surge in recognition and relevance. The pressing global need to address environmental challenges and promote sustainable financial practices has become more pronounced than ever before. Traditional portfolio optimization often overlooks environmental considerations, resulting in sub-optimal investment decisions. In this paper, we propose the E-Sharpe Ratio, a metric tailored for evaluating environmental risk-adjusted returns. By combining this ratio with Bayesian machine learning, our methodology provides a comprehensive framework for assessing stocks and portfolios, accounting for both financial and environmental performance metrics. Our research contributes to the field of environmental and climate finance by bridging financial and environmental considerations, enabling investors to make environmentally-aware decisions, and enhancing the stock selection process underscoring the importance of integrating environmental criteria into modern investment strategies, paving the way for a more sustainable financial future.