What is the Deal?: Predicting M&A Outcomes with Machine Learning

Abstract

We examine whether machine learning algorithms that incorporate accounting fundamentals, deal characteristics, and macroeconomic indicators can predict which merger and acquisition (M&A) deals will be value creative versus value destructive, and offer three main results. First, non-linear machine learning models predict two-year post-deal announcement returns relatively well. A trading strategy that buys the stock of the deals with the highest predicted score quintile and sells short the stock of the deals with the lowest predicted score quintile generates market adjusted returns around 11.9%, while a linear prediction model does not provide significant returns. Second, the link between the machine learning prediction score and post-acquisition accounting earnings is not direct under machine learning or linear prediction models. This suggests that the M&A deals the market views as successful are not directly linked to earnings in the immediate years after the deal closes. Finally, we identify the specific macroeconomic, firm, and deal variables that explain why non-linear machine learning models succeed in predicting post-merger return success. Overall, we shed light on the M&A outcome puzzle that has persisted in the literature for decades, suggesting that prior papers may have been unable to clearly predict future M&A success due to being limited to linear models, excluding macroeconomic variables, and, in some cases, linking deal success to accounting earnings or cash flows in the first three years post-acquisition.

Keywords: Mergers and Acquisitions, Fundamental Analysis, Financial Accounting

JEL Classification: G11, G14