Computational Methods in the Evaluation of Mergers and Acquisitions

Abstract

Computational antitrust is gaining popularity among competition authorities all around the globe. One of the areas of its application is merger review. Standard steps of merger review include: (1) selection of merger and acquisition cases for notification, (2) investigation phase, (3) court procedure, (4) final decision (with or without remedies), and (5) ex post control (if applicable). In this chapter, each sub-section addresses one of these steps and discusses the applicability and the existence of computational tools within the respective step. Furthermore, we focus on two research questions: (1) does the use of computational tools allow for a better selection of merger projects, and (2) are predictions of merger effects based on computational tools better than those based on typical types of evidence? The chapter is based on insights from computational research for merger review, modern economics, institutional economics, and political economics. Among other results, we conclude that each step of the merger review process offers space for computational tools, which, despite having considerable imperfections, can enhance merger control proceedings. However, some elements of the institutional framework of merger control require adaptations to specific characteristics of computational tools, which should contribute to their wider application.

Keywords: Computational antitrust, Merger review, Types of evidence, Algorithms, Artificial Intelligence Algorithms, Regulation

JEL Classification: K0, K21, K4, L0, L4, O3, G34, C63