Forecasting Financial Failure of Firms via Genetic Algorithms

Authors:
Eduardo Acosta-González;Fernando Fernández-Rodríguez
Affiliations:
Faculty of Economics, Management and Tourism, University of Las Palmas de Gran Canaria, Canaria, Spain;Faculty of Economics, Management and Tourism, University of Las Palmas de Gran Canaria, Canaria, Spain
Venue:
Computational Economics
Year:
2014

Citing 3
Cited 0

Inducing rules for expert system development: an example using default and bankruptcy data

Management Science
Managerial applications of neural networks: the case of bank failure predictions

Management Science
Hybrid Classifiers for Financial Multicriteria Decision Making: TheCase of Bankruptcy Prediction

Computational Economics

Quantified Score

Hi-index	0.00

Visualization

Abstract

Given a wide amount of possible ratios available for constructing a LOGIT model for forecasting bankruptcy, this paper provides a computational search methodology, only guided by data, for selecting the financial ratios employed in the model. This procedure is based on genetic algorithms which are used to explore the universe of models made available by all possible existing financial ratios (with very redundant information). This search process of the correct model is guided by the Schwarz information criterion. As an empirical illustration, the methodology is applied to forecasting the failure of firms in the Spanish building industry using annual public accounting information.