Template-Type: ReDIF-Paper 1.0

Author-Name: Daniela Bragoli
Author-X-Name-First: Daniela
Author-X-Name-Last: Bragoli
Author-Email: daniela.bragoli@unicatt.it
Author-Workplace-Name: Università Cattolica del Sacro Cuore
Author-Workplace-Name: Dipartimento di Matematica per le Scienze economiche, finanziarie ed attuariali, Università Cattolica del Sacro Cuore

Author-Name: Camilla Ferretti
Author-X-Name-First: Camilla
Author-X-Name-Last: Ferretti

Author-Name: Piero Ganugi
Author-X-Name-First: Piero
Author-X-Name-Last: Ganugi

Author-Name: Giovanni Marseguerra
Author-X-Name-First: Giovanni
Author-X-Name-Last: Marseguerra
Author-Email: giovanni.marseguerra@unicatt.it
Author-Workplace-Name: Università Cattolica del Sacro Cuore
Author-Workplace-Name: Dipartimento di Matematica per le Scienze economiche, finanziarie ed attuariali, Università Cattolica del Sacro Cuore

Author-Name: Davide Mezzogori
Author-X-Name-First: Davide
Author-X-Name-Last: Mezzogori

Author-Name: Francesco Zammori
Author-X-Name-First: Francesco
Author-X-Name-Last: Zammori


Title: Machine Learning models for bankruptcy prediction in Italy:do industrial variables count?

Abstract: We aim to provide a predictive model, specifically designed for the Italian economy, 
which classifies solvent and insolvent firms one year in advance, using AIDA Bureau van Dijk dataset from 2007 to 2015.  
We apply a full battery of bankruptcy forecasting models, including both traditional and more sophisticated 
machine learning techniques, and add to the financial ratios used in the literature a set of industrial/regional variables.  
We find that XGBoost is the best performer and that industrial/regional variables are important. 
Moreover, belonging to a district,having a high markup and a greater market share diminish bankruptcy probability.

Length: 41

Creation-Date: 2019-03
File-URL: https://dipartimenti.unicatt.it/dime-dime19_03.pdf
File-Format: Application/pdf
File-Function: First version, 2019
Number: dime19_03
Classification-JEL: G33, C45, C52, R11, L23.
Keywords: Firm distress analysis, machine learning, logistic regression, industrial variables.
Handle: RePEc:ctc:sdimse:dime19_03