Strategic M&A decisions powered by predictive insights : media and AI

Abstract

The big data era has revolutionized strategic decision-making by enabling organizations to leverage vast amounts of information for enhanced predictive accuracy. This dissertation explores how human-curated media data and machine-generated artificial intelligence (AI) predictions—two critical products of the big data age—support decision-making in the high-stakes context of mergers and acquisitions (M&A). Focusing on the key task of target selection in M&A, this research examines how these two data sources contribute to reducing uncertainty and improving decision outcomes. The first study investigates how media data influences ownership decisions in cross-border M&A. Assessing the value of a foreign target firm in cross-border acquisitions has historically been a challenge for acquirers because of information asymmetry. Although a limited number of M&A studies have suggested that signaling theory provides insights into mitigating information asymmetry, these studies have mainly focused on the signals sent by the acquirer and paid less attention to those released by foreign target firms. In this study, I suggest that the reputation of a foreign target firm represents a visible and credible signal released by that firm regarding its valuation. Using a sample of 5,647 cross-border acquisition deals between 2012 and 2017, and a sentiment analysis of 36,469 news articles pertaining to the target companies involved, I find that acquirers tend to increase their ownership level when the target company's reputation index is higher. I also argue that, despite the value of the reputational signal, its efficacy depends on its credibility, which can be violated by noises from both the foreign target firm and its host country. These noises will influence how much foreign acquirers rely on the reputational signal. The second study explores the interplay between artificial intelligence (AI) and human decision-making in mergers and acquisitions (M&A), examining whether AI predictions enhance or replace human judgment in shaping post-acquisition performance. Using a dataset of 2,906 M&A transactions from 2000 to 2023, I compare predictions generated by a neural network model with those of human analysts from the Institutional Brokers' Estimate System (IBES). My findings reveal that AI predictions consistently outperform human forecasts, with decisions aligned with AI evaluations leading to significantly higher post acquisition performance. By analyzing the alignment between M&A outcomes and predictions, this study provides empirical evidence on the relative effectiveness of AI-driven versus human-driven decision-making in M&A. These findings contribute to the growing discourse on AI in strategic decision-making and offer practical insights for managers navigating the evolving role of AI in corporate strategy. Together, these studies underscore the transformative potential of media data and AI in enhancing predictive precision and decision-making reliability in M&A. By integrating traditional strategic management theories with emerging technologies, this dissertation contributes to the growing discourse on leveraging big data for competitive advantage. It also raises critical ethical and practical questions about the integration of predictive tools into decision-making processes, advocating for hybrid frameworks that balance machine-driven insights with human creativity and judgment.