Logistic Regression Analysis of Key Drivers in Mergers and Acquisitions

Abstract

The current study evaluates the predictive factors for mergers and acquisitions (M&A) using a logistic regression model, focusing on key financial variables such as Face Value (FV), Advertisement Expenses (AE), and Research & Development (RD). Model 1, with an AIC of 112.67 and an accuracy of 84.17%, performs best overall, providing strong predictive capability for M&A activity. The model reveals that companies with lower face value, higher advertisement expenses, and increased RD spending are more likely to engage in M&A. The ROC curve analysis indicates a robust model with an AUC of 0.9311, suggesting high classification accuracy. Despite its effectiveness, non-random residual patterns highlight areas for improvement, indicating potential non-linearity and outliers. Future improvements could involve refining the model through larger datasets, adding interaction terms, or exploring industry-specific models. These findings provide valuable insights for corporate strategists and investors in identifying potential M&A candidates.