Machine learning in bank merger prediction: A text-based approach

Apostolos G. Katsafados, George N. Leledakis, Emmanouil G. Pyrgiotakis, Ion Androutsopoulos, Manos Fergadiotis

Abstract: Highlights•Textual features improve the classification ability of our models in bank merger prediction.•When we combine textual data with financial variables, we achieve the highest performance.•In predicting bidders, textual data are more important inputs that financial variables.•Our results highlight the importance of textual data in merger prediction tasks.