Avaliação econômico-financeira de empresas brasileiras de capital aberto no período de 2011 a 2018

Abstract

The analyzed companies live in a highly competitive environment, leading them to make certain decisions with different levels of uncertainties, either because of the difference in understanding, as well as the future behavior of the economy and, how this impacts their business or, still, due to the various risks that permeate its investment and market strategies. In this context, the present study aimed to assess the economic and financial situation (whether profitable or loss-making) of publicly traded Brazilian companies between 2011 and 2018. To achieve this objective, a backpropagation algorithm was built and estimated artificial neural network and a discriminant function, using a sample of 285 publicly traded Brazilian companies. In its main results, the work identified the backpropagation algorithm of the artificial neural network as the best method, verifying that the model that relates the situation of the company with its most recent past, proved to be more efficient in the classification of companies in profit or loss, with 83.80% assertiveness in the classification. In addition, it was identified that the discriminant analysis method presented the result of the test of statistical significance that invalidated its application. Finally, the equation with the highest degree of assertiveness showed that the important variables are𝐿iqG2018, 𝑀𝐿i𝑞2018, 𝐶ompEn𝑑2018, 𝑅O𝐼2018, 𝐺𝐴2018 and the most important variables of economic activity are 𝑃���𝐹���2018..