May 9, 2024

AI: Using Natural Language Processing to Predict M&A Participants

This recent SSC/Intralinks blog discusses how natural language processing (NLP), an AI-powered machine learning technology, may be about to transform M&A. The blog notes NLP’s potential to assess corporate culture, unlock potential synergies, and, as this excerpt discusses, even help identify likely buyers and sellers:

Can NLP predict whether a firm will be acquired or be an acquirer? Based on an NLP analysis of a large set of Form 10-K filings, the answer is “yes.” A study by Carnegie Mellon used the frequencies of words and phrases, showing an improvement in predicting merger targets and acquirers. Further advancing the field, two academics at the University of Münster applied the RoBERTa algorithm, a sophisticated pre-trained transformer neural network capable of handling text cohesively. Their findings showed that textual disclosures in both the business description and the management discussion and analysis sections of Form 10-K filings can significantly enhance predictions of corporate acquisition targets, going beyond traditional financial metrics.

Like many people, I find a lot of this AI stuff a little unnerving, although I draw some comfort from the fact that the real-world AI applications I’ve seen so far look less like the HAL 9000 from 2001: A Space Odyssey and more like 1980s pop culture icon Max Headroom.

John Jenkins