O uso dos indicadores contábeis para análise de solvência das seguradoras brasileiras

• Ano de defesa: 2017
• Autor(a) principal: Silva, Camila Menezes da
• Orientador(a): Securato, José Roberto
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Pontifícia Universidade Católica de São Paulo
• Programa de Pós-Graduação: Programa de Estudos Pós-Graduados em Ciências Contábeis e Atuariais
• Departamento: Faculdade de Economia, Administração, Contábeis e Atuariais
• País: Brasil
• Palavras-chave em Inglês: Seguros - Contabilidade; Companhias de seguro; Regulação; Insurance accounting; Regulation; Insurance companies
• Área do conhecimento CNPq: CNPQ::CIENCIAS SOCIAIS APLICADAS::ADMINISTRACAO::CIENCIAS CONTABEIS
• Link de acesso: https://tede2.pucsp.br/handle/handle/19887

The main objective of this paper is to identify the indicator or the group of accounting indicators to classify the Brazilian insurance companies as solvent or insolvent. Insurance companies play an important role in the society: to restore the financial losses caused by uncertain events. In this context, the uncertainty is a fundamental problem for which insurance companies must be financially prepared not to become insolvent. In Brazil, the ruling entity is SUSEP and it is in charge of overseeing these institutions and establishing rules to help companies to not become insolvent. From the methodological perspective, this research used the discriminant analysis to develop a model that classifies the Brazilian insurance companies as solvent or insolvent. The sample consisted of 30 insurance companies divided into 2 groups: the first group with insurance companies that had some SUSEP intervention; the second one with insurance companies that had no intervention, considered solvent, between 2000 and 2015 years. The 15 indicators were selected to enter the model, after reviewing the prerequisite of the discriminant analysis technique, such normality and the absence of multicollinearity of variables; it was possible to obtain a classification model by the stepwise method. The finding of the developed model, statistically significant, has properly classified 90% of the observations