Are Prices Really Affected by Mergers?

Xavier BOUTIN et Lionel JANIN

During the 80s, several empirical studies have shown a positive correlation between concentration, prices and profits. It is well known that these estimates all suffer from simultaneity bias: market structure and prices are affected by common factors, some of which are not observable, which rules out any causal interpretation of cross-sectional correlations. Mergers are an interesting instrument to identify the (static) impact of concentration on prices, since they induce breaks in strategic interactions between actors. The few ex post studies on mergers that are currently available are difficult to generalize, because they pertain to specific markets. This study looks more systematically to selling prices in 63 sectors observed between 1989 and 2002. The approach that has been chosen is a difference in differences approach, applied to price movements around mergers. The rate of inflation in a sector where a merger has occurred is compared to a counterfactual. In a simple framework, in line with previous studies (McCabe 2002), this counterfactual would be built as the mean of inflation rates in other sectors. This paper focuses on more relevant estimates, provided by a factor model. This methodology allows tracking the profile of prices around mergers. We separate mergers between French firms and mergers between other European firms controlled by European authorities (and thus assumed to have affected the common market). We also distinguish mergers having led to an in-depth inquiry by competition authorities (« phase 2 ») and those benefiting from a shorter procedure (« phase 1 »). We observe an acceleration of price movements around the most important of French mergers, but not for the ones authorized under phase 1. We also observe a break in price movements for mergers between foreign firms examined by the European Commission, generally in the other direction.

Documents de travail N° G2008/08
No G2008/08
Paru le : 01/07/2008