This report applies a balance-sheet approach to the Brazilian banking system over 2019–2025, drawing on 7,000 institution-year observations for 1,094 banking institutions from the IF.data dataset of the Central Bank of Brazil. Return on equity is treated as an endpoint rather than a premise, and is decomposed through the identity ROE = Rb + (Rb − Cd) · δ, where Rb is the broad asset return, Cd the broad structural cost of liabilities, and δ the debt-to-equity ratio. Sufficiency is assessed against the criterion ROE ≥ max(Cd, kE), operationalized as the Structural Sufficiency Index, SSI = min(DCI, ECI); the weighted average cost of capital is shown to be an inadequate benchmark for banks, since it never exceeds the greater of the two costs. The reference return on equity (kE) is set, for each year, equal to the realized equity-weighted ROE of the large commercial and universal banks (group b1), and is therefore an ex-post peer benchmark; the sufficiency verdicts are robust to this choice. The evidence indicates an unlevered spread that is persistently thin, an inverted regime in 2023 in which the cost of liabilities exceeded the required return on equity and the binding constraint shifted to the debt side, and a system operating close to its sufficiency thresholds despite reported accounting profitability. Concentration conditions the aggregates without substituting for balance-sheet analysis: the dominance of the largest banks rests on lower structural cost and higher leverage rather than on a superior asset return.
Fernando Barros (Mon,) studied this question.