Quantitative modeling of M&A success probability Predicting the Completion of Mergers and Acquisitions: An Econometric and Machine Learning Approach

M&A represent one of the most significant strategic decisions undertaken by firms. They involve substantial financial resources, complex negotiations, regulatory scrutiny, and high levels of uncertainty. Despite strategic intent, a significant number of these operations do not reach completion. Identifying the elements that influence whether a deal is successfully completed or fails is highly relevant for corporate finance, investors, and policymakers. This thesis explores the drivers of M&A deal completion by analysing a structured dataset of transactions enriched with financial and institutional variables extracted from Bloomberg platform. The empirical approach integrates conventional econometric models with ML techniques in order to identify the structural factors that affect the probability of completion. Particular attention is given to firm size, relative negotiating strength, financial performance, transaction characteristics, and contextual aspects such as industry and geographic proximity. Given the class imbalance between completed and non-completed deals, the study also emphasizes predictive performance and model selection, evaluating different methodologies to ensure robustness and interpretability. The objective is not only to explain completion outcomes, but also to evaluate how accurately completion can be anticipated using information available ex ante. Through this approach, the thesis contributes to the empirical literature on merger execution risk and provides practical insights into the structural features of successful M&A transactions.