A relação entre as expectativas macroeconômicas divulgadas no relatório focus e os parâmetros de Basileia nos bancos S1 do Brasil

Abstract

The systemic risk present in the financial system has been a source of concern and studies around the world for a long time, constituting one of the main motivations for the maturation of the Basel Accords in the last five decades. However, compliance by financial institutions occurs within each country, under the influence of different macroeconomic factors in each location. This dissertation explored the relationship between market expectations about Brazilian macroeconomic variables expressed in the Focus Report of the Central Bank and the main capital indicators of the largest banking institutions in Brazil, which are part of the prudential segment S1. The data used were collected in public databases made available by the Central Bank and included quarterly data from September 2008 to June 2021. Initially, the data collected were evaluated together, through regression of panel data, in order to assess in aggregate the relationship between expectations on macroeconomic indicators disclosed in the FOCUS report and the share of credit risk in risk-weighted assets (RWAc) of the banks in the sample. Then, the individual evaluation of the banks was carried out, through multiple linear regression, with the objective of evaluating the same relationship as before, but now in a unique way for each bank, comparing the results observed between the institutions. Finally, for each bank in the sample, the relationship between the expectations about the macroeconomic indicators disclosed in the Focus report and the Common Equity Tier 1 Capital (CP) was evaluated in order to compare the results observed between the institutions and to assess the difference in responses in RWAc ratio. The results indicate that, for the analysis of data in panels, the random effects models are the most adequate to explain the studied relationship, to the detriment of the pooled and fixed effects models. Among these, the model with the greatest explanatory capacity pointed out that the variables of Exchange, Current Account, Trade Balance, Public Sector Net Debt, Primary Result and Nominal Result are the most relevant to explain the RWAc. Regarding the individual analysis of the banks, it was observed for the models that sought to explain the RWAc that a single model typology was selected as the most robust for all banks, with a high explanatory capacity, above 90%. In these chosen models, the IPCA, Selic and Current Account variables were presented as relevant variables for all banks. Finally, for the individual assessment of banks to explain the CP, there was greater dispersion in the types of models that were more robust to explain the CP, the explanatory capacity of the models was lower, between 62.9% and 96,4%, and the explanatory variables were more dispersedly distributed, with none of them present in three or more models chosen for the banks.