Exploiting Fuzzy Regression Trees for M&A Sustainability Prediction

The increasing importance of sustainability in financial decision-making has driven the need for transparent and interpretable models for evaluating Mergers and Acquisitions (M&A) deals. Traditional Artificial Intelligence (AI) approaches, particularly deep learning models, often lack explainability, making their adoption in high-stakes financial contexts challenging. In this work, we propose a framework based on multiway Fuzzy Regression Trees to predict the sustainability of M&A deals. Fuzzy Regression Trees (FRTs) offer a high level of interpretability by leveraging linguistic if-then rules and fuzzy partitions, ensuring transparency in decision-making. Our approach integrates financial and Environmental Social and Governance (ESG)-related indicators to estimate the impact of an acquisition on the acquirer's sustainability score. Using a dataset of approximately 1,000 completed M&A deals sourced from LSEG Datastream, we conduct an extensive experimental analysis to assess the trade-offs between model complexity and predictive performance. The results demonstrate that Fuzzy Regression Tree provides a compelling balance between accuracy and interpretability, making it a valuable tool for decision-makers seeking to incorporate sustainability considerations into M&A strategies.