Maverick firms and merger policy

Abstract

A “maverick firm” is one that behaves in a manner that differs from the industry norm. As a result they are perceived to present a barrier to tacit collusion and for that reason competition authorities can seek to prevent mergers that might result in their removal from a market. However, the existing literature on mavericks is limited. No formal theory exists and there is no established empirical method for identifying such firms. This thesis seeks to address these gaps. First, we begin to formulate a formal theory of mavericks via a repeated games model in which firms have contrasting preferences and asymmetric capacities. Then we analyse the use of the concept in European Commission merger cases during 2000-2013. Finally, we explore three strategies for empirical maverick identification. The first consists of a replication of the only previous attempt, which used simple regressions and a ranking system. The second utilises cointegration and error correction techniques. The third explores asymmetric error correction, applying the nonlinear autoregressive distributed lag (NARDL) approach. These methods are used to identify the maverick in the market for bank and building society deposit accounts in the United Kingdom during 2000-2009. The theoretical model formally illustrates how a maverick can adversely impact non-mavericks’ payoffs. However, we also highlight scenarios where collusion could occur despite the presence of a maverick. The analysis of EC merger cases highlights the fact that application of the concept has been inconsistent with merger guidelines. In particular, the term was regularly used in conjunction with unilateral effects whereas guidelines only relate the concept to tacit coordination. Finally, each of the identification strategies produces contrasting results. Ultimately, the NARDL approach allows the identification of favourable behaviours (from a consumer perspective). As a result we are able to identify the most likely candidate mavericks in our market.