Volume : 3, Issue : 2, February - 2014

Bankruptcy Prediction Models: A Bet on Artificial Neural Networks

Ana Isabel Velasco Fernandez, Ricardo Jose Rejas Muslera, Juan Padilla Fernandez Vega, Maria Isabel Cepeda Gonzalez

Abstract :

The proper prediction of future bankruptcy of an organization is vital to all its stakeholders, from its owners/ shareholders to employees, passing of course through their creditors. This paper presents a quantitative approach using the discriminant analysis, analysis logit and a multilayer perceptron networks applied to economic data of one hundred forty-three companies to generate the models bankruptcy prediction. The results demonstrate that artificial neural networks provide analysis and predictions significantly superior than the statistical classical methods, resulting in an ideal tool for the financial management