A Large-Scale Analysis of Mergers and Merger Simulations

Abstract

Antitrust agencies in the United States reviewed over two thousand prospective mergers and acquisitions per year in recent years. Their task is to trade off harms from market power with potential synergies that may be passed on to consumers—and to convince a court to either block mergers that they predict will cause significant consumer harm or to order that the merging parties adopt remedies.

These projects will study two aspects of antitrust enforcement in the US. First, the project will document the realized price and quantity effects of mergers that have been approved, studying a near-universe of sufficiently large deals in a number of industries. This quantification will help inform the effects of current policy and may point to avenues for future adjustments.

Second, the project will study the efficacy of one of the standard tools for predicted price effects of prospective mergers: the merger simulation. There will be systematic assessment of the predictive power of mergers simulations and characterize the reasons why predictions differ from realized observations. The results of the project will have important implications for antitrust policy, and will provide tools for conducting merger evaluations in the US for a variety of industries.

The project will first estimate the price and quantity effect of mergers—on both merging parties and non-merging parties—using data before and after the completion date of the merger and a variety of controls. The project will correlate these effects, both across mergers and within-mergers across markets, with measures of concentration, such as the Herfindahl-Hirschman Index (HHI) and the naïve change of the HHI.

There will be a documentation of the time path of the change in prices. Second, the project will estimate a structural model of demand for each of the mergers in the dataset using only data before the merger was consummated, choosing among a variety of demand specifications. The project will then predict post-merger prices using these estimates, the implied costs, and the assumption of static Bayesian-Nash equilibrium.

Finally, there will be a decomposition of the prediction error into that due to demand-side changes and supply-side changes (such as synergies and conduct) by estimating parameters of the structural model using post-merger data. This decomposition may provide guidance for how to adjust merger simulations in practice.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.