Escolhendo ações de bancos brasileiros: um modelo de análise fundamentalista para melhor performance

Abstract

Researchers and analysts are constantly searching for the most efficient technique in making stock investment decisions. However, most of the usual methods have limitations such as complexity and the need for a vast amount of historical information to make the necessary predictions. Furthermore, although banks have peculiar characteristics such as the nature of the operation, high regulation, high risks and different leverage from other sectors, there is still no specific indicator to evaluate them. Therefore, the objective of this thesis was to search for a fundamentalist model formed by specific accounting information from banking institutions and verify how it behaves to identify investment opportunities among shares of Brazilian banks. The study sample is made up of Brazilian banking institutions and the analysis period covers the years 2010 to 2021 using annual data. A model was adopted, Bank_SCORE, which is made up of four large groups of variables: 1) general profitability; 2) profitability components; 3) risk; and 4) growth opportunities. Then, macroeconomic variables were also tested. The statistical techniques used were parametric and non-parametric mean tests and multiple linear regressions. The results showed that Bank_SCORE was less efficient than ROE and book-to-market in predicting future returns. Furthermore, macroeconomic variables were relevant in explaining the returns and performance of Brazilian banking shares. Thus, this study contributed to reinforcing the gap in knowledge of ways in which specific indicators of bank financial statement accounts can be used by investors in general.