Five Things the Market Is Misreading in Bank of America's Q1 Numbers

Operating income went from $33.98 billion in 2021 to $30.97 billion in 2022 to $28.34 billion in 2023 to $29.25 billion in 2024. Then 2025: $37.70 billion. That is a 29% YoY jump and a break of the three-year decline.

The drivers are specific. Net interest margin expanded as the deposit beta cycle finally turned; fixed-income trading revenue printed its best year since 2022; and credit loss provisions normalised after the 2023-2024 over-reserving cycle. Each of these is, in isolation, cyclical. Together, they imply operating income closer to $35-38 billion is the new run-rate, not a peak.

That matters for the multiple. A bank earning $36 billion a year in operating income, paying 22% tax, and returning 80% of earnings to shareholders generates roughly $22 billion of annual distributable capital. At a $385 billion market cap, that is a 5.7% distribution yield before growth. You do not need much of a re-rating from here to generate double-digit total returns.

The single most important variable for a US money-centre bank in 2023-2024 was deposit beta, the rate at which deposit costs rose in response to Fed hikes. Bank of America was the most exposed to this dynamic because its consumer deposit base of roughly $1.9 trillion is disproportionately tilted toward interest-bearing accounts that reprice with policy rates.

That drag has inverted. With the Fed having cut materially off the 2023 highs and deposit costs now repricing lower faster than loan yields, net interest income has turned into a tailwind. BAC's net interest income in 2025 rose roughly 7% against a flat loan book; the same dynamic is visible at JPM and Citi but it is most pronounced at BAC because the deposit franchise is more rate-sensitive in both directions.

Historically, the window after a peak deposit beta and before a loan loss cycle has been the best earnings window for money-centre banks. The last clean version of this setup was 2015-2016. BAC returned 45% over that two-year window. The current setup is not identical, but the structure is familiar.

The pattern is clear.

Fixed-income, currencies and commodities (FICC) revenue has been volatile. That is the nature of the business. What has changed is the structural floor. In 2025, BAC's FICC revenue printed materially above the five-year median, and the Q4 commentary explicitly flagged the benefit of stronger secondary credit trading, mortgage trading, and rates volatility.

The point is not that trading will continue at this pace. The point is that the floor has lifted. The pre-2020 FICC run-rate was approximately $9 billion annually. The post-2020 run-rate appears to be in the $12-13 billion range, supported by rates volatility that looks structurally higher than the 2015-2019 period and by improved market-making share in credit. At current valuation, the market is still modelling the $9 billion floor.

If the $12 billion floor holds, that is roughly $3 billion of incremental pre-tax income that falls to the bottom line. On current share count, that is roughly $0.35 per share, or about 7% of consensus 2026 EPS, which is not in the bull case.

BAC's provision for credit losses in 2025 was notably benign. Net charge-offs are trending toward the long-run median of 50-55 basis points on the loan book, down from the peak of ~70 basis points in late 2023. The consumer credit card portfolio, which was the source of the 2023 over-provisioning, has stabilised.

There is a counterpoint here that the bears raise correctly. Credit costs at current levels are below normalised. A recession or even a garden-variety slowdown would push provisions higher by perhaps $2-3 billion, compressing net income by a commensurate amount. That is a real risk.

The offset is reserves. BAC's allowance for credit losses sits at roughly 1.3% of loans, above the pre-COVID baseline. The reserve coverage is adequate for a mild recession without requiring a top-up, and the 2023-2024 over-reserving cycle means the starting point is strong.

This is not a thesis that ignores the downside. It is a thesis that says the downside is manageable and the upside, at 12x forward earnings, is not priced.

Bank of America returned roughly $21 billion to shareholders in 2025, split between $8 billion of dividends and $13 billion of repurchases. At an average repurchase price of approximately $43, the buybacks were conducted at a 14% discount to the current share price. That is accretion that hits tangible book per share mechanically.

Management has signalled a 60-70% payout ratio as the ongoing target. At consensus 2026 earnings of roughly $32 billion, that is $19-22 billion of distribution, or 5.0-5.7% of current market cap. Combined with 3-4% EPS growth from the deposit beta tailwind, organic loan growth, and FICC floor elevation, the math points to a total return in the low double digits annually, not including any multiple re-rating.

If the multiple re-rates from 12.3x to 14x over the next 18 months (the historical trough to mid-cycle range), that is another 14% of price return on top of the distribution yield and the EPS growth. That is the listicle arithmetic: 5% distribution, 4% EPS growth, 14% re-rating, compounded. It is not complicated.