Rating de crédito não financeiro: análise do varejo alimentar brasileiro à luz de seus indicadores regionais e operacionais

Abstract

Contextualization: Accurately predicting bankruptcy and/or mitigating credit risk are relevant and current topics, their necessity ratified by the consolidation of demand for credit. Computational advancements, market liberalization, new financial agents, and the diversification of borrowers consolidate the evolution in this market, implying the need to re-examine the topic in light of new nuances. With the objective of analyzing the relevance of operational and regional metrics in the construction of a non-financial credit rating applicable to Brazilian food retail companies, this work followed the approach of the base articles (WU, MA, OLSON, 2022; CHAI et al., 2019), applying discriminant analysis and neural networks to a sample of 22 companies from 2015 to 2022 with a view to updating the prediction model in light of the risk categories of Altman's seminal model and constructing a model based on operational and regional variables, without dependence on structured financial indicators. Results: Due to accounting differences compared to the original Altman model, the variables calculated for updating the prediction model resulted in a new configuration, while preserving its evaluation categories. The results indicated an accuracy of 98% using Discriminant Analysis and 95% when processed by neural networks and applied for prediction one year in advance. In the second empirical test, although with a lower accuracy rate (85% via Discriminant Analysis and 95% with Neural Networks), the model outperformed the traditional Z-score and achieved the benefit of constructing this prediction without the need to collect structured economic-financial variables. Research alignment with the PPGAdm (Regionality and Management) area of ​​concentration and with the research line: The construction of a tool that aligns operational decisions and indicators with the impacts on the financial evaluation of companies, using regional variables, is aligned with the research line, especially due to this last factor. Impact and innovative character in intellectual production: The innovation proposed in this work is the method of assessing the risk of bankruptcy by including operational and regional variables instead of financial ones, focusing on the Brazilian food retail sector (unlisted companies whose information is not fully available) and replicating a hybrid method. It reinforces the need to include regionality in bankruptcy prediction and overcomes the gap created by the common absence of structured financial information. Economic, social, and regional impact: The economic impact of the research involves the potential for better and more adequate sizing of the cost and access to credit, given that companies, even without structured financial statements and economic-financial indicators, can be evaluated. From a social perspective, it highlights the breadth of operational decisions in the company's evaluation. From a regional perspective, the contribution was to include variables in the predictive models, overcoming the regional particularities of the predictors. Regional implications: The research is important for the region, given its geographical characteristics; importance and logistical advantage, with wholesalers, financial institutions, individual agents, and especially various retailers located here, which, in addition to being included in the analysis, can benefit from the results obtained. Sustainable Development Goals addressed in the research: Promote sustained, inclusive and sustainable economic growth, full and productive employment and decent work for all (Goal 8), contributing to strengthening the capacity of institutions to encourage the expansion of access to their services for all.