Evaluation of insolvency in mutual credit unions by application of the data mining using decision trees approach

Abstract

This study aimed to present an evaluation of the insolvency of mutual credit unions in the Paraná State (Brazil) by application of the data mining using decision trees approach. The information required to build the models were obtained from indicators applied to a sample of 62 mutual credit unions from which 31 are solvent and 31 are insolvent. The selection of indicators was made based on the PEARLS system, whose efficacy refers to the World Council of Credit Unions (WOCCU). The decision trees were built by training the J48, ADTree and LADTree algorithms. After the analysis of results, the best performance was observed for the ADTree algorithm. According to the Kappa statistics, its acceptance level was excellent. In addition to the evaluation of performance of the decision trees, the paths with the highest confidence levels for assessing insolvency was identified by the A3 indicator (Net Institutional and Transitory Capital + Non-Interest-bearing Liabilities/ Non-earning Assets) (> 0.052), this value indicate that the cooperative is solvent. The confidence level was set at 1.953 and the path is represented on the second node of the tree.