Identifying M&A Targets from Textual Disclosures: A Transformer Neural Network Approach

Abstract

Can textual information from firm disclosures help to identify M&A targets? We employ the state-of-the-art transformer neural network RoBERTa based on 113,000 annual financial reports of publicly listed US firms to estimate takeover likelihoods. We show that incorporating publicly available, highly standardized textual information can improve the predictability of corporate takeovers significantly in out-of-sample tests and that this information is not fully incorporated in stock prices. We use explainable artificial intelligence methods to examine the reasons for the improved predictions. Our analyses indicate that the machine learning algorithm is able to identify product offerings and firm-specific capabilities sought by acquirers.

Keywords: Mergers & Acquisitions, Takeover Probability, Machine Learning, Natural Language Processing, Deep Learning

JEL Classification: G34, C45