Parâmetros Ambientais, Sociais e Governança (ASG) para criação de valor: implementação de modelos de previsão de desempenho para empresas brasileiras.

Abstract

As foundations, governments and society become interested in questions about sustainable social, environmental and economic financing, interest in ESG issues is growing. Considering only the financial report, without considering sustainability measures, today it no longer meets informational needs. Studies seek to investigate the relationship between ESG parameters and financial performance, but evidence on the topic is still inconsistent and inconclusive. One of the big challenges is that ESG reporting, for the most part, is voluntary and unregulated in many countries. As a result, concerns arise about the comparability, accuracy and usefulness of ESG disclosure. In this context, the problem question arises: How can we relate corporate financial performance to the implementation of ESG requirements in companies in a reliable and standardized way, with a high level of analysis efficiency? In Brazil, research is still embryonic in its conclusions. With a methodology that seeks to develop a prediction algorithm for the relationship between ESG parameters and financial performance. This study is justified by contributing to the debate on ESG initiatives and maximizing the financial performance of corporations, which does not yet present robust lessons on a relationship between variables representing the socio-environmental and financial fields. The objective was achieved through empirical study, where classification tree techniques proved to be more efficient than the model based on logistic regression when comparing accuracies and ROC curves. It is worth noting that both the profit variable and appreciation had a direct relationship with some of the ESG variables.