Merger Simulation: A Simplified Approach with New Applications

Merger Simulation: A Simplified Approach with New Applications

Description

In recent years there have been significant developments in the use of empirical economic methods to study the likely competitive effects of mergers. These developments have been shaped by the increased use of unilateral effects analyses by the competition authorities, as is expressed in part in the 1997 Horizontal Merger Guidelines. Such analyses evaluate the ability of the post-transaction firm to raise the prices of some or all of its (often differentiated) products through unilateral decisions and without resort to overtly collusive activities. Unilateral effects analyses encompass a broad set of issues that arise when the differentiated brands produced by the merging firms constitute the first and second choices for some group of customers. Absent de novo entry or product repositioning, a unilateral price increase may become profitable as the result of a merger if a substantial number of customers who previously would have been lost to competitors can now be retained because the merged firm also offers the customers' second choice. If, however, this “1-2” customer group is relatively small, then at best only a minimal price increase will be profitable. In essence, the forgone profits from the lost sales to diverted customers would be roughly comparable to the incremental profits from price increases to customers that do not switch. The technique known as “merger simulation” has emerged as a promising framework for this analysis. Simulation uses economic models grounded in the theory of industrial organization to predict the effect of mergers on prices in relevant markets. There is a common theoretical core to all simulation approaches in use today, although the details of a given simulation will depend on data availability and on the mathematical characterization of the market or markets at issue. While merger simulation is not a panacea for all of the economic issues that arise in a difficult transaction, it nonetheless can offer assessments of competitive effects and remedies that are beyond the reach of other methods of inquiry. For example, simulation has been used to evaluate the likelihood that potential merger-specific efficiencies (associated with reductions in the marginal cost of production) are sufficiently great to offset predicted price increases. Simulation can also be used to analyze the competitive effects of product repositioning and de novo entry. Finally, simulation can help one to evaluate the adequacy of proposed divestitures. With time, we believe that simulation techniques will be better understood and more widely used by antitrust lawyers and economists. A variety of different economic models can be utilized as the basis for a simulation analysis. When sufficient data are available, demand models can be estimated econometrically. When these estimated-demand simulation models are not feasible, models requiring less data can be valuable if one is willing to make additional assumptions about the nature of demand. The logit demand model and “PCAIDS”—a new model to be introduced in this article-both fit into this calibrated-demand simulation model category. We will suggest that PCAIDS offers advantages over a number of other calibrated-demand models. We have undertaken this review and update of work on merger simulation with a number of goals in mind. First, we offer a relatively non-technical description of the principles of merger simulation-principles that are consistent with the methodologies currently in use by the competition authorities. Second, we describe PCAIDS, the new calibrated-demand merger simulation methodology. Third, we present examples that apply PCAIDS, including some applications that to our knowledge have not previously appeared in the literature on merger simulation. Fourth, we suggest how simulation analyses might be used to evaluate the safe harbors of the Merger Guidelines. Calibrated-demand models are relatively easy to implement and make detailed simulation feasible for nearly any transaction because they require neither scanner nor transaction-level data. The PCAIDS model, in particular, requires only information on market shares and reasonable estimates of two elasticities. Estimates of these elasticities often can be obtained from marketing information or, when appropriate, through demand estimation. As with any calibrated-demand simulation model, one can test the sensitivity of the PCAIDS results to changes in the values of the estimated elasticities and to other simulation parameters. We believe that calibrated-demand simulation models can offer valuable screening devices for “quick looks” by enforcement agencies and by merging firms. The models can be used to review the potential antitrust exposure resulting when unilateral effects issues are raised but sufficient information is not available to estimate reliably a full set of cross-price elasticities. The models also can offer a useful means of working out the implications of the range of qualitative judgments an analyst might make based on documentary and interview evidence, and to test the sensitivity of competitive effects predictions to plausible variations in those assumptions. The analyses may be particularly useful for weighing opposing forces, as when comparing the potential anticompetitive loss of localized competition to the procompetitive gain relating to merger-specific efficiencies and product repositioning. The balance of this article is organized as follows. Part II discusses the economic fundamentals of merger simulation. Because the pros and cons of merger simulation have been extensively debated elsewhere, we do not undertake such a treatment here. In Part III we introduce the PCAIDS approach to modeling demand. We explain how a key assumption about the relationship between market shares and the diversion of lost sales from price increases can be used to calibrate the PCAIDS model. Part IV offers some examples of merger simulation with PCAIDS that includes comparisons with other simulation models. In Part V we show how PCAIDS can be applied to the analysis of product repositioning and entry. Part VI presents an analysis of the Merger Guidelines's safe harbors using PCAIDS simulation, and Part VII contains some brief concluding remarks. We have relegated the more technical mathematical details to the Appendix.

Source Publication

The More Economic Approach to European Competition Law

Source Editors/Authors

Dieter Schmidtchen, Max Albert, Stefan Voigt

Publication Date

2007

Merger Simulation: A Simplified Approach with New Applications

Share

COinS