A Support Vector Machine model for classification of efficiency: An application to M&A

Abstract

One of the main issues in banking and finance sector is measuring the efficiency of mergers and acquisitions (M&A), due to a plethora of key performance indicators (KPI) and variables. In this study, the efficiency of 441 M&A deals is evaluated based on specific inputs and outputs, including the change of environmental and social governance (ESG) scores. Due to presence of negative data, two Data Envelopment Analysis (DEA) and second stage analyses have been applied. The first is a regression model, which examines the impact of control variables on the efficiency of DEA scores. The second is a Support Vector Machine (SVM) model, mapping efficiency based on gender diversity. Results indicate that the performance of M&A deals is positively affected by both gender diversity and relative size whereas is negatively affected by the deal value. The SVM model classification indicates which regions of efficiency and stability are reflected by good or bad representation of women on boards.