ESG-Integrated machine learning portfolio opimization strategies: performance analysis and applications

Abstract

This master thesis, through the integration of deep learning models and the inclusion of ESG preferences, presents a comprehensive approach to portfolio management that can serve as a reference for sustainable investing in the future. By providing insights into the complexities and trade-offs involved in portfolio optimization and ESG integration, this research facilitates more informed decision-making for portfolio managers, enabling them to balance performance goals with ethical considerations effectively. The integration of these key components reinforces the thesis's primary goal and substantiates its significant contribution to the fields of portfolio management and sustainable investing.

Building upon these findings, the research discovered the Multilayer Perceptrons (MLP) model to be an especially powerful tool in asset allocation. Its robust performance promises to increase the efficiency of portfolio management, effectively managing the risk-return trade-off. However, it is of the utmost importance for asset managers to align the choice of regression models with specific investor profiles and market conditions, highlighting the need for flexibility in strategy implementation.

In relation to ESG integration, our study underscored the importance of balancing ethical considerations with financial performance. The trade-off observed between ESG performance and portfolio returns prompts asset managers to approach ESG integration with caution. However, the research also demonstrates that the application of strategies implied by machine learning models like the MLP can potentially mitigate the return cost associated with ESG integration, paving the way for more sustainable investing while limiting the compromise on returns.

This research not only offers practical insights for portfolio managers but also opens up exciting paths for future exploration. With the ongoing evolution of financial markets and sustainable investing, this study underscores the importance of continual review and innovation in portfolio optimization techniques and ESG integration. As such, this master thesis contributes to the current discourse in the field of portfolio management and sustainable investing, serving as a key starting point for more advanced and comprehensive approaches to sustainable portfolio management in the future.