Estudo de eventos e análise da rentabilidade ex post como metodologia de análise de fusões: o caso brasileiro

Abstract

The objective of this dissertation is to estimate the potential competitive effects on the relevant market of major mergers occurred in Brazil after the Real Plan, evaluating the effectiveness of methods used to estimate ex ante effects actually observed in the profitability of merged firms and rivals ex post. According to the analysis of the theoretical model proposed by Motta(2004), it is possible to use the result of rival firms in the market in which the merger occurred as an indicator of the presence of gains in efficiency or market power in the relevant market of the merged firms, where positive returns for rival firms indicate the presence of market power and, consequently, a decrease in consumer surplus and welfare in the economy. Similarly, negative returns for rival firms indicate the presence of efficiency gains and increased well-being. To estimate the effects of mergers on rivals and firms merged, is used in this work to study methodology in which events are calculated abnormal stock returns of rival firms as well as firms merged, resulting from the merger announcement. This comparison was made by means of a correlation analysis and regression. In this study, a sample of 42large horizontal mergers between firms with publicly traded shares listed on the BM&FBOVESPA and their respective rivals in the sector was used, and a breakdown for a specific sector, the Telecommunications. The empirical results obtained in the methodology of event studies show a higher presence of anticompetitive mergers in general (gain market power) and efficiency gains in the sector of Telecommunications effects. Furthermore, it was found for both the rival firms as for firms merged, positive and significant correlation between abnormal returns and the effect on ex post profitability, demonstrating the market's ability to anticipate the negative effects of mergers on the welfare. These results are also corroborated by regression analysis, fitting to emphasize the presence of the effect of mean reversion, thus demonstrating the usefulness of the event study methodology to capture the potential effects of ex ante mergers.