Consolidation, product quality and entry in the banking industry
This dissertation consists of three chapters, each of which is concerned with consolidation, product quality and entry in the banking industry. The three chapters are respectively titled Literature Review, Merger Simulation with Historical Merger Data, and Post-Merger Product Repositioning .
In Literature Review, I give a summary of relevant papers, and illustrate the position and contribution of this dissertation in the literature on antitrust analysis, banking industry, endogenous product quality and post-merger product repositioning respectively.
Merger analysis has typically focused on changes in price but not much on changes in product quality and entry. In the second chapter, Merger Simulation with Historical Merger Data, I present a methodology of merger simulation with historical merger data, which extends merger prediction from price to product quality and entry. Product quality in this paper is measured by both observed product quality, branch density, and unobserved product quality. I adopt a broad definition of entry: not only the creation of a new bank but also branch network extension of small banks. The historical bank merger data can be used to estimate the post-merger patterns of entry and changes in costs of merging banks, and help us select reasonable assumptions in merger simulation. I apply the methodology of merger simulation to eleven cases of within-market bank mergers. The predicted post-merger branch densities and market shares of merging banks are closer to the actual outcomes than a benchmark, which uses the sum of pre-merger branch densities and market shares of merging banks as their post-merger values, for eleven out of eleven times and nine out of eleven times respectively. The benchmark tends to overestimate post-merger branch densities and market shares.
Theory provides ambiguous predictions for post-merger product repositioning of merging and non-merging firms. Mergers can lead merging firms to spread similar products apart to reduce cannibalization, or to crowd products together to preempt entry. In the third chapter, Post-Merger Product Repositioning , I investigate how merging and non-merging banks reposition products, branches, using a database that includes bank mergers and detailed branch locations. If a branch of a target and a branch of an acquirer are in the same ZIP code pre merger, then the new merging bank is likely to close or sell one branch, and this branch usually belongs to the target. The exit probability of a branch of a non-merging bank is reduced if it is located in a ZIP code with within-ZIP mergers. A bank tends to build up new branches in ZIP codes with few within-ZIP mergers, high overall branch density, large population or high income. A merging bank prefers to build up new branches close to its old branches or its merger partner's old branches in the same market but not in the ZIP codes where it has overlap with its merger partner.