Cluster analysis and segmentation of Global M&A transactions

Abstract

The present thesis is the analysis a dataset of mergers and acquisitions (M&A) through a segmentation process by cluster analysis, to better understand combined explanatory variables and characteristics of global M&A transactions. Past researched has strongly focused on (A) whether or not M&A creates wealth for investors or (B) which factors and variables help explain value this wealth (des)creation. The present thesis is rather an attempt to reach a third leg of research which is that, by segmenting and understanding these “natural” groupings we may develop a richer understanding of this form of corporate transactions. The paper comprises a study-event dataset from global completed M&A since 1994 with high disclosure filters, a factor analysis that selected 7 out of 13 variables from previous literature review, preceded by a the cluster analysis for variable selection. The end result indicated a connection between several explanatory variables and the formation of clusters with economical meaning. Six clusters were formed under a two-step clustering process. The paper has three relevant highlights: (1) the application of cluster analysis in a M&A setting; (2) the selection of surrogate variables from the factor analysis, providing better economic representation and (3) a clustering method that automatically captures the natural grouping the dataset.