Análise da relação de indicadores financeiros e a recuperação judicial : um estudo aplicado com redes neurais artificiais

Abstract

Companies aim to continue their activities and increase their profits, to do so, it is necessary to obtain information that expresses their economic and financial situations, in order to avoid situations of judicial recovery and, in more serious cases their early closure. This research was the subject of the analysis of some financial indicators for judicial recovery guidance, which took place in publicly traded companies listed on the CVM website. The research problem sought to investigate the predictive capacity of a model of artificial neural networks to explain judicial recovery through economic and financial indicators. As a way of answering the question raised, the general objective was to verify the existence of a relationship between judicial recovery and these indicators, which was possible through the specific objectives sought, through working capital, current liquidity and social capital, to build and validate the RNA model. The research was based on a typically classified descriptive study, case study, bibliographic, documentary and qualiquantitative. The theoretical framework was based on materials published by renowned authors in books, articles, magazines, dissertations and other sources. As a form of data collection, the accounting statements, for the year 2018, of publicly traded entities, arranged on the CVM website, were analyzed. The results obtained confirmed the hypothesis that artificial neural networks can express a satisfactory result from the prediction of their model, using social capital, working capital and current liquidity to demonstrate recovery Judicial. Because it has a very low error index, it is possible to continue this study, implementing more variables, to analyze and predict possible situations more safely.