Modelos de previsão de insolvência utilizando a análise por envoltória de dados: aplicação a empresas brasileiras
Main Article Content
Abstract
This article aimed to develop an insolvency forecasting model through a mathematical technique that was based on Operational Research: Data Envelopment Analysis (DEA). The research methodology involved the following steps: (1) we selected ten companies that faced a bankruptcy concordat process between 1995 and 2003; (2) for each insolvent company, we chose by lot five companies among those that performed best in the sector and had a similar size, summing up a sample of 50 health companies; (3) the sample companies' accounting information for a three-year period before the bankruptcy/concordat were obtained from Fipecafi/Exame's Melhores e Maiores database, and submitted to statistical analyses; (4) we developed a DEA model and calculated efficiency indicators for the sample companies; (5) we determined the cut-off point to rank companies as insolvent or solvent; (6) the classifications obtained through the DEA indicators were confronted with the company's actual situation. Preliminary results have shown that the developed DEA model was capable of distinguishing between solvent and insolvent companies with a reasonable level of exactness: it correctly classified 90% of the insolvent companies.
Downloads
Download data is not yet available.
Download data is not yet available.
Article Details
How to Cite
Onusic, L. M., Casa Nova, S. P. de C., & Almeida, F. C. de. (1). Modelos de previsão de insolvência utilizando a análise por envoltória de dados: aplicação a empresas brasileiras. Journal of Contemporary Administration, 11(spe2), 77-97. https://doi.org/10.1590/S1415-65552007000600005
Section
Articles
This journal remains the copyright holder of articles published. In order to be published, authors must sign the Transfer of Copyrights Document, which is sent to the authors by e-mail, thus granting rights, including on translation, to the Journal of Contemporary Administration. The journal grants third parties the right to use, reproduce, and share the article according to the Creative Commons license agreement (CC-BY 4.0), as stated in the article’s PDF documents.