Ensemble Learning and NLP for M&A Target Selection: The CTT Case

The present study investigates the application of machine learning techniques to predict merger and acquisition (M&A) targets, using the logistics sector, and specifically the case of CTT's acquisition of Cacesa, as a real-world context. Although the decision-making process in M&A is frequently characterized by complexity and firm-specific factors, the increasing accessibility of structured financial data and unstructured textual information is creating new opportunities for systematic analysis. Drawing upon the Equity Research of CTT, this study proposes a methodology that combines quantitative indicators and text-based features to develop predictive models capable of identifying likely acquisition candidates. The application of a profit-oriented evaluation framework ensures alignment with practical decision-making criteria. The findings indicate that these models effectively enhance target screening accuracy, underscoring their relevance to M&A analysis. While acknowledging the constraints imposed by data limitations, particularly regarding private companies, the study highlights the promise of AI-driven methodologies in facilitating forward-looking strategic evaluations. This work provides actionable insights for industry practitioners and represents a significant contribution to the broader adoption of predictive analytics in corporate acquisition strategies.