The reported 1.20x P/B sits between Itaú's 2.51x and Santander Brasil's 0.83x, and since the peer table shows Itaú converting 21.1% ROE into a two-turn book premium while Bradesco's roughly 15% annualized ROE (four times the 3.8% quarterly print) does not, the discount appears proportional to the return gap rather than evidence of mispricing. Investors should note that the tabular price-to-book series rising from 15.55 to 18.77 and the 25.8% dividend yield for 2026Q3 conflict with the 1.20x P/B and 5.5% yield in the snapshot, so the units and currency basis of these multiples warrant verification before anchoring any valuation conclusion. The 1.50x forward P/E alongside a 9.27x trailing figure looks internally inconsistent and, absent updated management guidance for the quarter reported 2026-10-22, should probably be treated as a data artifact rather than an implied earnings re-rating.

Because ROA has been pinned in a 0.2%–0.3% band for ten quarters while equity-to-assets has not moved off 0.08, the entire variance in ROE is coming from asset mix and provisioning rather than underlying yield generation, which suggests the DuPont quality is weaker than the headline return implies. The 41.5% fee contribution in 2026Q3, up from 25.9% in 2025Q1, does offset some of the spread weakness and is consistent with the stabilizing role of the insurance segment, though the wide quarter-to-quarter swing in that figure makes it difficult to conclude the fee base is fully recurring. With provisions reported at $8.5B in 2026Q3 versus $3.1B in 2024Q3, pre-provision profitability appears materially better than post-provision ROE, so the trajectory of credit costs, not pricing power, looks to be the swing factor in any ROE recovery case.

The 13.7-point efficiency improvement indicates the branch-network downsizing and IT modernization are beginning to bend the fixed-cost base, but the 111.6% efficiency print in 2025Q4 stands out as a distorted period that should be excluded before annualizing any run-rate improvement. NIM compression to 0.7% occurring alongside a high Selic environment is the more troubling signal, because it suggests the shift toward the $2.0T securities portfolio and slower loan-book growth is diluting asset yields faster than deposit costs can be repriced downward. Taken together, the two trends imply operating leverage is being manufactured by cost removal rather than revenue momentum, a form of improvement that typically has a finite runway once the restructuring backlog is exhausted.

CET1, Tier 1 and total capital adequacy ratios are not present in the supplied data, so regulatory headroom versus Brazilian minimums cannot be assessed here and should be drawn from the primary capital disclosures before any capacity-for-capital-return judgment is made. What can be observed is that equity grew about 20% over the period while total assets grew roughly 25%, a drift that suggests balance-sheet expansion is running slightly ahead of retained earnings and may help explain the lumpy, back-loaded dividend pattern with $9.6B paid in 2026Q3 after three quarters of zero distributions. With investment securities at $2.0T against $196.9B of equity, even a modest mark-to-market decline in the portfolio could consume a disproportionate share of book value, so investors should monitor unrealized-loss disclosures as closely as the headline equity ratio; the reported debt-to-equity of 4.46 also appears inconsistent with typical bank reporting and may require manual recalculation.

Only Itaú and Santander Brasil constitute a meaningful comparison set for Bradesco, since Ambev, Gerdau and Sid are non-financial issuers whose multiples, margins and capital structures are not comparable to a deposit-taking institution. On that trimmed peer set, Bradesco's efficiency ratio of 51.6% compares reasonably well against its own history, yet its ROE shortfall to Itaú appears to stem from asset yields and credit costs rather than the expense base, which would make the gap partly cyclical and tied to the retail and micro-SME credit cycle. The dividend yield comparison is less clean: the snapshot's 5.5% trails Itaú's 8.8%, but the 25.8% figure shown for 2026Q3 in the ratio table is implausible as a sustained payout and, if the snapshot is correct, would suggest payout capacity is currently constrained by the provisioning build rather than by capital limits.