Analysis of mergers in bidding markets

Merger analysis --- the study of the determinants and effects of mergers on the merging firms and on the markets they operate --- has an impact on many areas of industrial organization research. Old questions find answers, new questions of firm behavior arise. Some ground-breaking research on merger analysis has already connected theoretical models with strategies for empirical identification. Empirical analysis of mergers simulation involving bidding markets has so far received little attention. The advanced level of auction theory and its empirical modelling, though, provide fertile ground for the application to merger analysis. With our research we break new ground and address the following topics: * The formulation of a merger simulation in an auction model. Based on a model of first-price auctions, we estimate and identify the effects of a merger, i.e., market power and efficiency effects. We evaluate the mergers in our data with respect to the non-merger benchmark as implied by the merger simulation model. * The study of entry behavior before and after the merger. In particular, the evaluation of the effects of the merger on post-merger entry of the merging firms and their competitors and the restraints that entry places on merging parties` market power. * The evaluation of the merger simulation model in comparison to the actual outcome of the observed mergers, with respect to (i) bidding behavior, (ii) cost efficiencies and (iii) entry behavior. * The application of a treatment effects model, based on the entry behavior of the bidders in different geographic markets and construction segments to establish a control group. * The evaluation of the treatment effects model in comparison to the predictions and performance of the structural merger estimation. We model mergers within a theoretical first-price auction model. Using data on construction procurement, we estimate bidders` construction costs within a private value auction model. Application of existing results on identification in auctions allows to recover the distribution of bidders` costs using the first order condition for optimal bidding behavior. Based on the marginal cost, we can evaluate the effect of the merger and observe the outcome for cost efficiencies, market power and entry behavior of merged entity and competitors. Ex-post evaluation of the merger can also be possible with a treatment approach. We identify the treatment effect by comparing separate geographic markets. Entry information identifies geographically separate markets. Our research advances the empirical analysis of auctions, extension of merger simulation models, informs on merits of different empirical approaches and furthers the understanding of firm behavior more generally.

The project led to three scientific papers. One of the earlier papers (Competition in the economic crisis: Analysis of Procurement auctions, European Economic Review) was the result of the project National Research Network S103, has already been published in a top economics journal and it has won the WU Best Paper Award 2016. The paper "Analysis of mergers in first-price auctions", which is the main paper of this FWF project, was published as a discussion paper at the Centre for Economic Policy Research (CEPR). Two other papers (Evaluation of bidding groups in first-price auctions and Effects of government spending on employment: Evidence from winners and runners-up in procurement auctions) are working papers and are in the submission process. The main questions of the papers do on the one hand differ substantially (from fiscal stimulus, labor market effects, effects of joint bidding groups to merger effects), however the papers have on the other hand a strong common methodological denominator. In the main paper of the project Analysis of mergers in first-price auctions, we analyze mergers in bidding markets. We utilize data from procurement auctions in the Austrian construction sector and estimate models of first-price sealed-bid auctions. Based on estimated costs and markups, we run merger simulations and disentangle the market power effect(s) from potential cost efficiencies. We also compare static and dynamic models of first price auctions. Finally, we compare the outcomes of the merger simulations with the actual effects of observed mergers. Our results show that market power and efficiency effects are present post-merger, leading to increased markups, but leaving the winning bid essentially unaffected by the merger. We find a good correspondence of predicted and actual effects for full mergers, but not for majority acquisitions. In a next version of the paper we will utilize the data that we have gathered in the project on geographical markets to generate more adequate control groups and use the knowledge that was built up to accommodate entry/exit decisions post-merger. The results of the project are important both from a scientific and a policy perspective. From a scientific perspective, we are the first to combine simulated and real world mergers accounting for entry/exit post-merger. From a policy perspective, to understand the competitive effects of mergers (and joint bidding groups) is of utmost importance.