The Next Frontier in Merger Review
There are two parallel, and seemingly contradictory, trends impacting the world of antitrust and competition in merger reviews.
The first is towards more sophisticated data analysis; competition Agencies on both sides of the Atlantic have developed strong internal Machine Learning and AI capabilities, so we expect to see “Big Tech-like” techniques deployed in merger reviews – even in traditional sectors, outside digital markets.
But the second trend is towards greater reliance on qualitative evidence and internal documents, potentially at the expense of quantitative analyses. How do these trends reconcile? Will data analysis find more or less space in merger reviews going forward?
In reality, these trends have a common root: increasing skepticism by Agencies towards some of the “more traditional” econometric techniques, which have often been found wanting when assessed ex post, and increasing interest in embedding a “business perspective” in merger reviews.
So, whether we will see more data analysis or less in merger review ultimately depends on the ability of economic advisors to produce more “realistic” analysis, based on state-of-the-art quantitative techniques which reflect how businesses look at their industry, and move away from some of the “more traditional” econometrics.
But what are these analyses?
In this article, Partick Bajari, Gianmarco Calanchi and Tega Akati-Udi examine:
- How to take the four key “traditional” merger analyses to the next level using ML and AI tools (they focus on entry analysis, demand estimation, merger simulation and synergy estimation).
- How to use novel ML and AI analyses currently used by Big Tech companies (but not only by them) in the ordinary course of business to learn causal effects.
These techniques can be deployed in traditional sectors, such as retail, as well as in digital markets. The Authors have deployed some of them in recent merger cases in traditional industries and found them very effective at addressing the typical limitations of the “more traditional” techniques.