Ceris-Cnr, W.P. N° 10/2006 

Models for Default Risk Analysis:
Focus on Artificial Neural Networks, Model Comparisons, Hybrid Frameworks

Greta Falavigna
(Ceris-CNR and University of Bergamo)
 National Research Council of Italy, Ceris-Cnr
Via Real Collegio, 30
10024 Moncalieri (To) – Italy
Tel.: +39.011.6824.937; Fax: +39.011.6824.966; Email: g.falavigna@ceris.cnr.it

Abstract:

During the last three decades various models have been proposed by the literature to predict the risk of bankruptcy and of firm insolvency.

In this work there is a survey on the methodologies used by the author for the analysis of default risk, taking into account several approaches suggested by the literature.

The focus is to analyse the Artificial Neural Networks as a tool for the study of this problem and to verify the ability of classification of these models.

Finally, an analysis of variables introduced in the Artificial Neural Network models and some considerations about these.

Keywords: Artificial Neural Networks, Hybrid neural network models Expert Systems, Default, Bankruptcy, Rating Systems, Credit scoring models
 

SCARICA QUESTO WP